Training in progress, epoch 4, checkpoint

checkpoint-16932/README.md

@@ -0,0 +1,34 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: False
+- load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: bfloat16
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- load_in_8bit: False
+- load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: bfloat16
+### Framework versions
+
+- PEFT 0.4.0
+
+- PEFT 0.4.0

checkpoint-16932/adapter_config.json

@@ -0,0 +1,22 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "mistralai/Mistral-7B-v0.1",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "revision": null,
+  "target_modules": [
+    "q_proj",
+    "k_proj",
+    "v_projo_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

checkpoint-16932/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad25a7eb9cb72fbbe9d3edf00e78584870bb9d937630c0da28792dbbbba6ab7a
+size 13677261

checkpoint-16932/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d15b53270b9b62da1821794aae28d51abba7588b4682e862f81c75a8307c93ae
+size 13648432

checkpoint-16932/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:353ed746cb298288d34d9195e72babee57facb5b7767a4b38d57063bfb4a4a07
+size 27370181

checkpoint-16932/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e5c2cdbf773ee990ad0d518e03aaae819c64bb772e74572b22503fbd7aec7ff5
+size 14575

checkpoint-16932/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce9e342ddf59a6fb9d1a6d60ac34229635882ae60c7b7312aa42575f54c44b0e
+size 627

checkpoint-16932/special_tokens_map.json

@@ -0,0 +1,11 @@
+{
+  "additional_special_tokens": [
+    "<unk>",
+    "<s>",
+    "</s>"
+  ],
+  "bos_token": "<s>",
+  "eos_token": "</s>",
+  "pad_token": "</s>",
+  "unk_token": "<unk>"
+}

checkpoint-16932/tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<unk>",
+    "<s>",
+    "</s>"
+  ],
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "legacy": true,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "</s>",
+  "sp_model_kwargs": {},
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "LlamaTokenizer",
+  "unk_token": "<unk>",
+  "use_default_system_prompt": true
+}

checkpoint-16932/trainer_state.json

@@ -0,0 +1,257 @@
+{
+  "best_metric": 0.8134331107139587,
+  "best_model_checkpoint": "./Zeroshot/01-12-23-mistralai-Mistral-7B-v0.1_multilang-dataset-3.0.3-portuguese-2_epochs-10_batch_3/checkpoints/checkpoint-16932",
+  "epoch": 4.999852354938727,
+  "eval_steps": 500,
+  "global_step": 16932,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.835794447725931e-05,
+      "loss": 1.4468,
+      "step": 500
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00011742468989958655,
+      "loss": 0.9754,
+      "step": 1000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00017649143532191377,
+      "loss": 0.9429,
+      "step": 1500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.000235558180744241,
+      "loss": 0.9147,
+      "step": 2000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00029462492616656825,
+      "loss": 0.9067,
+      "step": 2500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00035369167158889544,
+      "loss": 0.8978,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.8805813789367676,
+      "eval_runtime": 88.2945,
+      "eval_samples_per_second": 17.057,
+      "eval_steps_per_second": 2.141,
+      "step": 3386
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.00039998760393503537,
+      "loss": 0.895,
+      "step": 3500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0003996072594095129,
+      "loss": 0.8687,
+      "step": 4000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.00039869668890858337,
+      "loss": 0.8884,
+      "step": 4500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00039725831122269285,
+      "loss": 0.8715,
+      "step": 5000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00039529594718087214,
+      "loss": 0.8645,
+      "step": 5500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0003928148095012922,
+      "loss": 0.8666,
+      "step": 6000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0003898214889444803,
+      "loss": 0.8719,
+      "step": 6500
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.8552550673484802,
+      "eval_runtime": 88.3232,
+      "eval_samples_per_second": 17.051,
+      "eval_steps_per_second": 2.14,
+      "step": 6773
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00038632393680597854,
+      "loss": 0.8438,
+      "step": 7000
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.0003823314437949511,
+      "loss": 0.8308,
+      "step": 7500
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00037785461535484375,
+      "loss": 0.8259,
+      "step": 8000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0003729053434916558,
+      "loss": 0.8324,
+      "step": 8500
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.0003674967751846552,
+      "loss": 0.8413,
+      "step": 9000
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0003616554183563445,
+      "loss": 0.8322,
+      "step": 9500
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00035537338261496887,
+      "loss": 0.8368,
+      "step": 10000
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 0.8443310260772705,
+      "eval_runtime": 88.3102,
+      "eval_samples_per_second": 17.054,
+      "eval_steps_per_second": 2.14,
+      "step": 10159
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.0003486786213865893,
+      "loss": 0.8088,
+      "step": 10500
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.0003415889182744321,
+      "loss": 0.8003,
+      "step": 11000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.0003341231059840768,
+      "loss": 0.805,
+      "step": 11500
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.0003263010162972709,
+      "loss": 0.8061,
+      "step": 12000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.00031814342739185336,
+      "loss": 0.8008,
+      "step": 12500
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 0.000309672008647721,
+      "loss": 0.8029,
+      "step": 13000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00030090926308545536,
+      "loss": 0.8056,
+      "step": 13500
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.8322489857673645,
+      "eval_runtime": 88.3294,
+      "eval_samples_per_second": 17.05,
+      "eval_steps_per_second": 2.14,
+      "step": 13546
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 0.00029187846759051,
+      "loss": 0.7649,
+      "step": 14000
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 0.00028260361108174584,
+      "loss": 0.7674,
+      "step": 14500
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 0.0002731093307885585,
+      "loss": 0.7635,
+      "step": 15000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 0.0002634208468058692,
+      "loss": 0.7759,
+      "step": 15500
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.000253563895100822,
+      "loss": 0.7669,
+      "step": 16000
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 0.0002435847820221107,
+      "loss": 0.7708,
+      "step": 16500
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 0.8134331107139587,
+      "eval_runtime": 88.3002,
+      "eval_samples_per_second": 17.055,
+      "eval_steps_per_second": 2.14,
+      "step": 16932
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 33860,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 8.503491256748114e+17,
+  "trial_name": null,
+  "trial_params": null
+}

checkpoint-16932/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57ba9992e8be82ca13275abd7c4d38c76e2e922e71387553242d20e094340bc1
+size 4347