Upload folder using huggingface_hub
Browse files- config.json +29 -0
- generation_config.json +6 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state_0.pth +3 -0
- rng_state_1.pth +3 -0
- rng_state_2.pth +3 -0
- rng_state_3.pth +3 -0
- rng_state_4.pth +3 -0
- rng_state_5.pth +3 -0
- rng_state_6.pth +3 -0
- rng_state_7.pth +3 -0
- rng_state_8.pth +3 -0
- rng_state_9.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +43 -0
- tokenizer.json +0 -0
- tokenizer_config.json +169 -0
- trainer_state.json +2882 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Trelis/TrelisLM-100M-layer-hidden-pruned",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 0,
|
| 9 |
+
"eos_token_id": 0,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 660,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 1760,
|
| 14 |
+
"max_position_embeddings": 2048,
|
| 15 |
+
"mlp_bias": false,
|
| 16 |
+
"model_type": "llama",
|
| 17 |
+
"num_attention_heads": 15,
|
| 18 |
+
"num_hidden_layers": 16,
|
| 19 |
+
"num_key_value_heads": 5,
|
| 20 |
+
"pretraining_tp": 1,
|
| 21 |
+
"rms_norm_eps": 1e-05,
|
| 22 |
+
"rope_scaling": null,
|
| 23 |
+
"rope_theta": 10000.0,
|
| 24 |
+
"tie_word_embeddings": true,
|
| 25 |
+
"torch_dtype": "bfloat16",
|
| 26 |
+
"transformers_version": "4.44.2",
|
| 27 |
+
"use_cache": true,
|
| 28 |
+
"vocab_size": 49152
|
| 29 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 0,
|
| 5 |
+
"transformers_version": "4.44.2"
|
| 6 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c9b2c4db3a2d556ee001eb8fdf128d644b789b40b2f7cf64684b6fe78989053b
|
| 3 |
+
size 213625344
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:129003e0eae631ea59961c42baa93d6ece566b523e5b57d228356535ea34946d
|
| 3 |
+
size 427334458
|
rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:593cfa780b4b09ba583a139eb81c2eae72992c19fc5f8f38c81bd37ea47dbe04
|
| 3 |
+
size 16433
|
rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9560f8c460a856a55828494146d2d52ecf0d95a3dec5919d8f29a972450cec34
|
| 3 |
+
size 16433
|
rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e7b7bce88125710e1c78de933cf62b48d6ec5a97b36fc43a09d7f70aebd0307
|
| 3 |
+
size 16433
|
rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4cb51f987fa17a879db4d368b0564564ae49379d5c5ce803d79d24b4b5a43c13
|
| 3 |
+
size 16433
|
rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33ed1a64882ec192e7bdbd0b9dda7c3dd977bc8ef889d26ddca3e2380d9debae
|
| 3 |
+
size 16433
|
rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4dd0f015bb5215ff40f1555f5c47c89a7bd89b00e7ef4568ca045dc1c2b5514a
|
| 3 |
+
size 16433
|
rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac5bc7ba9f4a9e405864d41e902bc7509a5b6fa554a6cf09f24491e00dac06fb
|
| 3 |
+
size 16433
|
rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:77e384968f192bdc497600d0108b82695a12247413143c0c9bd4e09fbb718212
|
| 3 |
+
size 16433
|
rng_state_8.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc46c9518ec829f507eb5833115c977024d13a12bc4e0ecff2238d818e6eb6dc
|
| 3 |
+
size 16433
|
rng_state_9.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56cb5d3bf2f9602568c76013ebf1c626061418df7f27c779c448921d362d5232
|
| 3 |
+
size 16433
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83a9c85c7b2c29125f99f000e54e900b05be0859260af7e4a0abf634beb2c469
|
| 3 |
+
size 1064
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|endoftext|>",
|
| 4 |
+
"<|im_start|>",
|
| 5 |
+
"<|im_end|>",
|
| 6 |
+
"<repo_name>",
|
| 7 |
+
"<reponame>",
|
| 8 |
+
"<file_sep>",
|
| 9 |
+
"<filename>",
|
| 10 |
+
"<gh_stars>",
|
| 11 |
+
"<issue_start>",
|
| 12 |
+
"<issue_comment>",
|
| 13 |
+
"<issue_closed>",
|
| 14 |
+
"<jupyter_start>",
|
| 15 |
+
"<jupyter_text>",
|
| 16 |
+
"<jupyter_code>",
|
| 17 |
+
"<jupyter_output>",
|
| 18 |
+
"<jupyter_script>",
|
| 19 |
+
"<empty_output>"
|
| 20 |
+
],
|
| 21 |
+
"bos_token": {
|
| 22 |
+
"content": "<|endoftext|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false
|
| 27 |
+
},
|
| 28 |
+
"eos_token": {
|
| 29 |
+
"content": "<|endoftext|>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false
|
| 34 |
+
},
|
| 35 |
+
"pad_token": "<|endoftext|>",
|
| 36 |
+
"unk_token": {
|
| 37 |
+
"content": "<|endoftext|>",
|
| 38 |
+
"lstrip": false,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false
|
| 42 |
+
}
|
| 43 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,169 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_eos_token": true,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"0": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"1": {
|
| 14 |
+
"content": "<|im_start|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"2": {
|
| 22 |
+
"content": "<|im_end|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"3": {
|
| 30 |
+
"content": "<repo_name>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"4": {
|
| 38 |
+
"content": "<reponame>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
},
|
| 45 |
+
"5": {
|
| 46 |
+
"content": "<file_sep>",
|
| 47 |
+
"lstrip": false,
|
| 48 |
+
"normalized": false,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false,
|
| 51 |
+
"special": true
|
| 52 |
+
},
|
| 53 |
+
"6": {
|
| 54 |
+
"content": "<filename>",
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"normalized": false,
|
| 57 |
+
"rstrip": false,
|
| 58 |
+
"single_word": false,
|
| 59 |
+
"special": true
|
| 60 |
+
},
|
| 61 |
+
"7": {
|
| 62 |
+
"content": "<gh_stars>",
|
| 63 |
+
"lstrip": false,
|
| 64 |
+
"normalized": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"single_word": false,
|
| 67 |
+
"special": true
|
| 68 |
+
},
|
| 69 |
+
"8": {
|
| 70 |
+
"content": "<issue_start>",
|
| 71 |
+
"lstrip": false,
|
| 72 |
+
"normalized": false,
|
| 73 |
+
"rstrip": false,
|
| 74 |
+
"single_word": false,
|
| 75 |
+
"special": true
|
| 76 |
+
},
|
| 77 |
+
"9": {
|
| 78 |
+
"content": "<issue_comment>",
|
| 79 |
+
"lstrip": false,
|
| 80 |
+
"normalized": false,
|
| 81 |
+
"rstrip": false,
|
| 82 |
+
"single_word": false,
|
| 83 |
+
"special": true
|
| 84 |
+
},
|
| 85 |
+
"10": {
|
| 86 |
+
"content": "<issue_closed>",
|
| 87 |
+
"lstrip": false,
|
| 88 |
+
"normalized": false,
|
| 89 |
+
"rstrip": false,
|
| 90 |
+
"single_word": false,
|
| 91 |
+
"special": true
|
| 92 |
+
},
|
| 93 |
+
"11": {
|
| 94 |
+
"content": "<jupyter_start>",
|
| 95 |
+
"lstrip": false,
|
| 96 |
+
"normalized": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"single_word": false,
|
| 99 |
+
"special": true
|
| 100 |
+
},
|
| 101 |
+
"12": {
|
| 102 |
+
"content": "<jupyter_text>",
|
| 103 |
+
"lstrip": false,
|
| 104 |
+
"normalized": false,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false,
|
| 107 |
+
"special": true
|
| 108 |
+
},
|
| 109 |
+
"13": {
|
| 110 |
+
"content": "<jupyter_code>",
|
| 111 |
+
"lstrip": false,
|
| 112 |
+
"normalized": false,
|
| 113 |
+
"rstrip": false,
|
| 114 |
+
"single_word": false,
|
| 115 |
+
"special": true
|
| 116 |
+
},
|
| 117 |
+
"14": {
|
| 118 |
+
"content": "<jupyter_output>",
|
| 119 |
+
"lstrip": false,
|
| 120 |
+
"normalized": false,
|
| 121 |
+
"rstrip": false,
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"special": true
|
| 124 |
+
},
|
| 125 |
+
"15": {
|
| 126 |
+
"content": "<jupyter_script>",
|
| 127 |
+
"lstrip": false,
|
| 128 |
+
"normalized": false,
|
| 129 |
+
"rstrip": false,
|
| 130 |
+
"single_word": false,
|
| 131 |
+
"special": true
|
| 132 |
+
},
|
| 133 |
+
"16": {
|
| 134 |
+
"content": "<empty_output>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": false,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false,
|
| 139 |
+
"special": true
|
| 140 |
+
}
|
| 141 |
+
},
|
| 142 |
+
"additional_special_tokens": [
|
| 143 |
+
"<|endoftext|>",
|
| 144 |
+
"<|im_start|>",
|
| 145 |
+
"<|im_end|>",
|
| 146 |
+
"<repo_name>",
|
| 147 |
+
"<reponame>",
|
| 148 |
+
"<file_sep>",
|
| 149 |
+
"<filename>",
|
| 150 |
+
"<gh_stars>",
|
| 151 |
+
"<issue_start>",
|
| 152 |
+
"<issue_comment>",
|
| 153 |
+
"<issue_closed>",
|
| 154 |
+
"<jupyter_start>",
|
| 155 |
+
"<jupyter_text>",
|
| 156 |
+
"<jupyter_code>",
|
| 157 |
+
"<jupyter_output>",
|
| 158 |
+
"<jupyter_script>",
|
| 159 |
+
"<empty_output>"
|
| 160 |
+
],
|
| 161 |
+
"bos_token": "<|endoftext|>",
|
| 162 |
+
"clean_up_tokenization_spaces": false,
|
| 163 |
+
"eos_token": "<|endoftext|>",
|
| 164 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 165 |
+
"pad_token": "<|endoftext|>",
|
| 166 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 167 |
+
"unk_token": "<|endoftext|>",
|
| 168 |
+
"vocab_size": 49152
|
| 169 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2882 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.3998792426790874,
|
| 5 |
+
"eval_steps": 2318,
|
| 6 |
+
"global_step": 9272,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0009919351360676242,
|
| 13 |
+
"grad_norm": 0.91796875,
|
| 14 |
+
"learning_rate": 0.001,
|
| 15 |
+
"loss": 9.0124,
|
| 16 |
+
"step": 23
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0019838702721352484,
|
| 20 |
+
"grad_norm": 0.1376953125,
|
| 21 |
+
"learning_rate": 0.0009990070799516491,
|
| 22 |
+
"loss": 4.9678,
|
| 23 |
+
"step": 46
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0029758054082028723,
|
| 27 |
+
"grad_norm": 0.10400390625,
|
| 28 |
+
"learning_rate": 0.0009980141599032983,
|
| 29 |
+
"loss": 4.8546,
|
| 30 |
+
"step": 69
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.003967740544270497,
|
| 34 |
+
"grad_norm": 0.2109375,
|
| 35 |
+
"learning_rate": 0.0009970212398549474,
|
| 36 |
+
"loss": 4.7284,
|
| 37 |
+
"step": 92
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.00495967568033812,
|
| 41 |
+
"grad_norm": 0.201171875,
|
| 42 |
+
"learning_rate": 0.0009960283198065965,
|
| 43 |
+
"loss": 4.5063,
|
| 44 |
+
"step": 115
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.005951610816405745,
|
| 48 |
+
"grad_norm": 0.302734375,
|
| 49 |
+
"learning_rate": 0.0009950353997582456,
|
| 50 |
+
"loss": 4.2107,
|
| 51 |
+
"step": 138
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.006943545952473369,
|
| 55 |
+
"grad_norm": 0.2314453125,
|
| 56 |
+
"learning_rate": 0.0009940424797098947,
|
| 57 |
+
"loss": 4.0923,
|
| 58 |
+
"step": 161
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.007935481088540993,
|
| 62 |
+
"grad_norm": 0.310546875,
|
| 63 |
+
"learning_rate": 0.0009930495596615439,
|
| 64 |
+
"loss": 3.7695,
|
| 65 |
+
"step": 184
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.008927416224608617,
|
| 69 |
+
"grad_norm": 0.30078125,
|
| 70 |
+
"learning_rate": 0.000992056639613193,
|
| 71 |
+
"loss": 3.6712,
|
| 72 |
+
"step": 207
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.00991935136067624,
|
| 76 |
+
"grad_norm": 0.322265625,
|
| 77 |
+
"learning_rate": 0.0009910637195648419,
|
| 78 |
+
"loss": 3.5194,
|
| 79 |
+
"step": 230
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.010911286496743866,
|
| 83 |
+
"grad_norm": 0.314453125,
|
| 84 |
+
"learning_rate": 0.0009900707995164912,
|
| 85 |
+
"loss": 3.4792,
|
| 86 |
+
"step": 253
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.01190322163281149,
|
| 90 |
+
"grad_norm": 0.296875,
|
| 91 |
+
"learning_rate": 0.0009890778794681401,
|
| 92 |
+
"loss": 3.3372,
|
| 93 |
+
"step": 276
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.012895156768879113,
|
| 97 |
+
"grad_norm": 0.310546875,
|
| 98 |
+
"learning_rate": 0.0009880849594197895,
|
| 99 |
+
"loss": 3.2479,
|
| 100 |
+
"step": 299
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.013887091904946738,
|
| 104 |
+
"grad_norm": 0.34765625,
|
| 105 |
+
"learning_rate": 0.0009870920393714384,
|
| 106 |
+
"loss": 3.2265,
|
| 107 |
+
"step": 322
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.014879027041014362,
|
| 111 |
+
"grad_norm": 0.3515625,
|
| 112 |
+
"learning_rate": 0.0009860991193230877,
|
| 113 |
+
"loss": 3.1277,
|
| 114 |
+
"step": 345
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.015870962177081987,
|
| 118 |
+
"grad_norm": 0.2578125,
|
| 119 |
+
"learning_rate": 0.0009851061992747366,
|
| 120 |
+
"loss": 3.0156,
|
| 121 |
+
"step": 368
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.01686289731314961,
|
| 125 |
+
"grad_norm": 0.2578125,
|
| 126 |
+
"learning_rate": 0.0009841132792263857,
|
| 127 |
+
"loss": 2.9304,
|
| 128 |
+
"step": 391
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.017854832449217234,
|
| 132 |
+
"grad_norm": 0.486328125,
|
| 133 |
+
"learning_rate": 0.0009831203591780348,
|
| 134 |
+
"loss": 2.9364,
|
| 135 |
+
"step": 414
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.01884676758528486,
|
| 139 |
+
"grad_norm": 0.37890625,
|
| 140 |
+
"learning_rate": 0.000982127439129684,
|
| 141 |
+
"loss": 2.8848,
|
| 142 |
+
"step": 437
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.01983870272135248,
|
| 146 |
+
"grad_norm": 0.28125,
|
| 147 |
+
"learning_rate": 0.000981134519081333,
|
| 148 |
+
"loss": 2.8368,
|
| 149 |
+
"step": 460
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.020830637857420106,
|
| 153 |
+
"grad_norm": 0.310546875,
|
| 154 |
+
"learning_rate": 0.0009801415990329822,
|
| 155 |
+
"loss": 2.8024,
|
| 156 |
+
"step": 483
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.02182257299348773,
|
| 160 |
+
"grad_norm": 0.498046875,
|
| 161 |
+
"learning_rate": 0.0009791486789846313,
|
| 162 |
+
"loss": 2.7152,
|
| 163 |
+
"step": 506
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.022814508129555353,
|
| 167 |
+
"grad_norm": 0.357421875,
|
| 168 |
+
"learning_rate": 0.0009781557589362804,
|
| 169 |
+
"loss": 2.6818,
|
| 170 |
+
"step": 529
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.02380644326562298,
|
| 174 |
+
"grad_norm": 0.353515625,
|
| 175 |
+
"learning_rate": 0.0009771628388879296,
|
| 176 |
+
"loss": 2.6625,
|
| 177 |
+
"step": 552
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.024798378401690604,
|
| 181 |
+
"grad_norm": 0.390625,
|
| 182 |
+
"learning_rate": 0.0009761699188395787,
|
| 183 |
+
"loss": 2.5967,
|
| 184 |
+
"step": 575
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.025790313537758226,
|
| 188 |
+
"grad_norm": 0.427734375,
|
| 189 |
+
"learning_rate": 0.0009751769987912278,
|
| 190 |
+
"loss": 2.541,
|
| 191 |
+
"step": 598
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.02678224867382585,
|
| 195 |
+
"grad_norm": 0.314453125,
|
| 196 |
+
"learning_rate": 0.0009741840787428768,
|
| 197 |
+
"loss": 2.5455,
|
| 198 |
+
"step": 621
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.027774183809893476,
|
| 202 |
+
"grad_norm": 0.302734375,
|
| 203 |
+
"learning_rate": 0.000973191158694526,
|
| 204 |
+
"loss": 2.5566,
|
| 205 |
+
"step": 644
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.028766118945961098,
|
| 209 |
+
"grad_norm": 0.357421875,
|
| 210 |
+
"learning_rate": 0.000972198238646175,
|
| 211 |
+
"loss": 2.4318,
|
| 212 |
+
"step": 667
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.029758054082028723,
|
| 216 |
+
"grad_norm": 0.3125,
|
| 217 |
+
"learning_rate": 0.0009712053185978243,
|
| 218 |
+
"loss": 2.4449,
|
| 219 |
+
"step": 690
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.03074998921809635,
|
| 223 |
+
"grad_norm": 0.302734375,
|
| 224 |
+
"learning_rate": 0.0009702123985494733,
|
| 225 |
+
"loss": 2.4232,
|
| 226 |
+
"step": 713
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.031741924354163974,
|
| 230 |
+
"grad_norm": 0.53125,
|
| 231 |
+
"learning_rate": 0.0009692194785011224,
|
| 232 |
+
"loss": 2.3352,
|
| 233 |
+
"step": 736
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.032733859490231595,
|
| 237 |
+
"grad_norm": 0.30859375,
|
| 238 |
+
"learning_rate": 0.0009682265584527715,
|
| 239 |
+
"loss": 2.332,
|
| 240 |
+
"step": 759
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.03372579462629922,
|
| 244 |
+
"grad_norm": 0.42578125,
|
| 245 |
+
"learning_rate": 0.0009672336384044207,
|
| 246 |
+
"loss": 2.3185,
|
| 247 |
+
"step": 782
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.034717729762366846,
|
| 251 |
+
"grad_norm": 0.314453125,
|
| 252 |
+
"learning_rate": 0.0009662407183560698,
|
| 253 |
+
"loss": 2.2945,
|
| 254 |
+
"step": 805
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.03570966489843447,
|
| 258 |
+
"grad_norm": 0.3203125,
|
| 259 |
+
"learning_rate": 0.0009652477983077189,
|
| 260 |
+
"loss": 2.2821,
|
| 261 |
+
"step": 828
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.03670160003450209,
|
| 265 |
+
"grad_norm": 0.265625,
|
| 266 |
+
"learning_rate": 0.000964254878259368,
|
| 267 |
+
"loss": 2.2184,
|
| 268 |
+
"step": 851
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.03769353517056972,
|
| 272 |
+
"grad_norm": 0.40234375,
|
| 273 |
+
"learning_rate": 0.0009632619582110171,
|
| 274 |
+
"loss": 2.2383,
|
| 275 |
+
"step": 874
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.03868547030663734,
|
| 279 |
+
"grad_norm": 0.30859375,
|
| 280 |
+
"learning_rate": 0.0009622690381626662,
|
| 281 |
+
"loss": 2.1755,
|
| 282 |
+
"step": 897
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.03967740544270496,
|
| 286 |
+
"grad_norm": 0.3203125,
|
| 287 |
+
"learning_rate": 0.0009612761181143154,
|
| 288 |
+
"loss": 2.1569,
|
| 289 |
+
"step": 920
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.04066934057877259,
|
| 293 |
+
"grad_norm": 0.314453125,
|
| 294 |
+
"learning_rate": 0.0009602831980659644,
|
| 295 |
+
"loss": 2.1272,
|
| 296 |
+
"step": 943
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.04166127571484021,
|
| 300 |
+
"grad_norm": 0.54296875,
|
| 301 |
+
"learning_rate": 0.0009592902780176136,
|
| 302 |
+
"loss": 2.1358,
|
| 303 |
+
"step": 966
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.042653210850907834,
|
| 307 |
+
"grad_norm": 0.279296875,
|
| 308 |
+
"learning_rate": 0.0009582973579692626,
|
| 309 |
+
"loss": 2.1388,
|
| 310 |
+
"step": 989
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.04364514598697546,
|
| 314 |
+
"grad_norm": 0.376953125,
|
| 315 |
+
"learning_rate": 0.0009573044379209118,
|
| 316 |
+
"loss": 2.0937,
|
| 317 |
+
"step": 1012
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.044637081123043085,
|
| 321 |
+
"grad_norm": 0.4140625,
|
| 322 |
+
"learning_rate": 0.0009563115178725609,
|
| 323 |
+
"loss": 2.0818,
|
| 324 |
+
"step": 1035
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.045629016259110707,
|
| 328 |
+
"grad_norm": 0.384765625,
|
| 329 |
+
"learning_rate": 0.0009553185978242101,
|
| 330 |
+
"loss": 2.0259,
|
| 331 |
+
"step": 1058
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.046620951395178335,
|
| 335 |
+
"grad_norm": 0.3359375,
|
| 336 |
+
"learning_rate": 0.0009543256777758591,
|
| 337 |
+
"loss": 2.0098,
|
| 338 |
+
"step": 1081
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.04761288653124596,
|
| 342 |
+
"grad_norm": 0.30078125,
|
| 343 |
+
"learning_rate": 0.0009533327577275082,
|
| 344 |
+
"loss": 2.0276,
|
| 345 |
+
"step": 1104
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.04860482166731358,
|
| 349 |
+
"grad_norm": 0.3046875,
|
| 350 |
+
"learning_rate": 0.0009523398376791573,
|
| 351 |
+
"loss": 2.0285,
|
| 352 |
+
"step": 1127
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.04959675680338121,
|
| 356 |
+
"grad_norm": 0.306640625,
|
| 357 |
+
"learning_rate": 0.0009513469176308065,
|
| 358 |
+
"loss": 1.9488,
|
| 359 |
+
"step": 1150
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.05058869193944883,
|
| 363 |
+
"grad_norm": 0.32421875,
|
| 364 |
+
"learning_rate": 0.0009503539975824556,
|
| 365 |
+
"loss": 1.9922,
|
| 366 |
+
"step": 1173
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.05158062707551645,
|
| 370 |
+
"grad_norm": 0.267578125,
|
| 371 |
+
"learning_rate": 0.0009493610775341046,
|
| 372 |
+
"loss": 1.9759,
|
| 373 |
+
"step": 1196
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.05257256221158408,
|
| 377 |
+
"grad_norm": 0.41015625,
|
| 378 |
+
"learning_rate": 0.0009483681574857538,
|
| 379 |
+
"loss": 1.93,
|
| 380 |
+
"step": 1219
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.0535644973476517,
|
| 384 |
+
"grad_norm": 0.36328125,
|
| 385 |
+
"learning_rate": 0.0009473752374374028,
|
| 386 |
+
"loss": 1.9264,
|
| 387 |
+
"step": 1242
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.05455643248371932,
|
| 391 |
+
"grad_norm": 0.25390625,
|
| 392 |
+
"learning_rate": 0.0009463823173890521,
|
| 393 |
+
"loss": 1.8782,
|
| 394 |
+
"step": 1265
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.05554836761978695,
|
| 398 |
+
"grad_norm": 0.51171875,
|
| 399 |
+
"learning_rate": 0.0009453893973407011,
|
| 400 |
+
"loss": 1.9079,
|
| 401 |
+
"step": 1288
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.056540302755854574,
|
| 405 |
+
"grad_norm": 0.27734375,
|
| 406 |
+
"learning_rate": 0.0009443964772923502,
|
| 407 |
+
"loss": 1.9279,
|
| 408 |
+
"step": 1311
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.057532237891922196,
|
| 412 |
+
"grad_norm": 0.291015625,
|
| 413 |
+
"learning_rate": 0.0009434035572439993,
|
| 414 |
+
"loss": 1.8501,
|
| 415 |
+
"step": 1334
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.058524173027989825,
|
| 419 |
+
"grad_norm": 0.484375,
|
| 420 |
+
"learning_rate": 0.0009424106371956484,
|
| 421 |
+
"loss": 1.8732,
|
| 422 |
+
"step": 1357
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.059516108164057446,
|
| 426 |
+
"grad_norm": 0.27734375,
|
| 427 |
+
"learning_rate": 0.0009414177171472975,
|
| 428 |
+
"loss": 1.8397,
|
| 429 |
+
"step": 1380
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.06050804330012507,
|
| 433 |
+
"grad_norm": 0.361328125,
|
| 434 |
+
"learning_rate": 0.0009404247970989467,
|
| 435 |
+
"loss": 1.8288,
|
| 436 |
+
"step": 1403
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.0614999784361927,
|
| 440 |
+
"grad_norm": 0.298828125,
|
| 441 |
+
"learning_rate": 0.0009394318770505958,
|
| 442 |
+
"loss": 1.8162,
|
| 443 |
+
"step": 1426
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.06249191357226032,
|
| 447 |
+
"grad_norm": 0.27734375,
|
| 448 |
+
"learning_rate": 0.0009384389570022449,
|
| 449 |
+
"loss": 1.8204,
|
| 450 |
+
"step": 1449
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.06348384870832795,
|
| 454 |
+
"grad_norm": 0.248046875,
|
| 455 |
+
"learning_rate": 0.0009374460369538939,
|
| 456 |
+
"loss": 1.8123,
|
| 457 |
+
"step": 1472
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.06447578384439556,
|
| 461 |
+
"grad_norm": 0.275390625,
|
| 462 |
+
"learning_rate": 0.0009364531169055431,
|
| 463 |
+
"loss": 1.8225,
|
| 464 |
+
"step": 1495
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.06546771898046319,
|
| 468 |
+
"grad_norm": 0.349609375,
|
| 469 |
+
"learning_rate": 0.0009354601968571922,
|
| 470 |
+
"loss": 1.7787,
|
| 471 |
+
"step": 1518
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.06645965411653082,
|
| 475 |
+
"grad_norm": 0.3671875,
|
| 476 |
+
"learning_rate": 0.0009344672768088414,
|
| 477 |
+
"loss": 1.7703,
|
| 478 |
+
"step": 1541
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.06745158925259843,
|
| 482 |
+
"grad_norm": 0.2578125,
|
| 483 |
+
"learning_rate": 0.0009334743567604904,
|
| 484 |
+
"loss": 1.7955,
|
| 485 |
+
"step": 1564
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.06844352438866606,
|
| 489 |
+
"grad_norm": 0.392578125,
|
| 490 |
+
"learning_rate": 0.0009324814367121396,
|
| 491 |
+
"loss": 1.7675,
|
| 492 |
+
"step": 1587
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.06943545952473369,
|
| 496 |
+
"grad_norm": 0.31640625,
|
| 497 |
+
"learning_rate": 0.0009314885166637886,
|
| 498 |
+
"loss": 1.7548,
|
| 499 |
+
"step": 1610
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.0704273946608013,
|
| 503 |
+
"grad_norm": 0.255859375,
|
| 504 |
+
"learning_rate": 0.0009304955966154378,
|
| 505 |
+
"loss": 1.7549,
|
| 506 |
+
"step": 1633
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.07141932979686894,
|
| 510 |
+
"grad_norm": 0.2578125,
|
| 511 |
+
"learning_rate": 0.0009295026765670869,
|
| 512 |
+
"loss": 1.7057,
|
| 513 |
+
"step": 1656
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.07241126493293656,
|
| 517 |
+
"grad_norm": 0.3125,
|
| 518 |
+
"learning_rate": 0.000928509756518736,
|
| 519 |
+
"loss": 1.7128,
|
| 520 |
+
"step": 1679
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.07340320006900418,
|
| 524 |
+
"grad_norm": 0.267578125,
|
| 525 |
+
"learning_rate": 0.0009275168364703851,
|
| 526 |
+
"loss": 1.7606,
|
| 527 |
+
"step": 1702
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.07439513520507181,
|
| 531 |
+
"grad_norm": 0.271484375,
|
| 532 |
+
"learning_rate": 0.0009265239164220342,
|
| 533 |
+
"loss": 1.6989,
|
| 534 |
+
"step": 1725
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.07538707034113944,
|
| 538 |
+
"grad_norm": 0.302734375,
|
| 539 |
+
"learning_rate": 0.0009255309963736834,
|
| 540 |
+
"loss": 1.7052,
|
| 541 |
+
"step": 1748
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.07637900547720705,
|
| 545 |
+
"grad_norm": 0.28515625,
|
| 546 |
+
"learning_rate": 0.0009245380763253324,
|
| 547 |
+
"loss": 1.7021,
|
| 548 |
+
"step": 1771
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.07737094061327468,
|
| 552 |
+
"grad_norm": 0.2314453125,
|
| 553 |
+
"learning_rate": 0.0009235451562769816,
|
| 554 |
+
"loss": 1.7096,
|
| 555 |
+
"step": 1794
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.07836287574934231,
|
| 559 |
+
"grad_norm": 0.29296875,
|
| 560 |
+
"learning_rate": 0.0009225522362286306,
|
| 561 |
+
"loss": 1.6537,
|
| 562 |
+
"step": 1817
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.07935481088540992,
|
| 566 |
+
"grad_norm": 0.26953125,
|
| 567 |
+
"learning_rate": 0.0009215593161802797,
|
| 568 |
+
"loss": 1.6359,
|
| 569 |
+
"step": 1840
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.08034674602147755,
|
| 573 |
+
"grad_norm": 0.298828125,
|
| 574 |
+
"learning_rate": 0.0009205663961319288,
|
| 575 |
+
"loss": 1.6107,
|
| 576 |
+
"step": 1863
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.08133868115754518,
|
| 580 |
+
"grad_norm": 0.271484375,
|
| 581 |
+
"learning_rate": 0.000919573476083578,
|
| 582 |
+
"loss": 1.6137,
|
| 583 |
+
"step": 1886
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.0823306162936128,
|
| 587 |
+
"grad_norm": 0.3359375,
|
| 588 |
+
"learning_rate": 0.0009185805560352271,
|
| 589 |
+
"loss": 1.6428,
|
| 590 |
+
"step": 1909
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.08332255142968042,
|
| 594 |
+
"grad_norm": 0.421875,
|
| 595 |
+
"learning_rate": 0.0009175876359868762,
|
| 596 |
+
"loss": 1.6146,
|
| 597 |
+
"step": 1932
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.08431448656574805,
|
| 601 |
+
"grad_norm": 0.283203125,
|
| 602 |
+
"learning_rate": 0.0009165947159385253,
|
| 603 |
+
"loss": 1.5996,
|
| 604 |
+
"step": 1955
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.08530642170181567,
|
| 608 |
+
"grad_norm": 0.3203125,
|
| 609 |
+
"learning_rate": 0.0009156017958901744,
|
| 610 |
+
"loss": 1.6153,
|
| 611 |
+
"step": 1978
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.0862983568378833,
|
| 615 |
+
"grad_norm": 0.283203125,
|
| 616 |
+
"learning_rate": 0.0009146088758418235,
|
| 617 |
+
"loss": 1.6178,
|
| 618 |
+
"step": 2001
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.08729029197395093,
|
| 622 |
+
"grad_norm": 0.3671875,
|
| 623 |
+
"learning_rate": 0.0009136159557934727,
|
| 624 |
+
"loss": 1.5988,
|
| 625 |
+
"step": 2024
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.08828222711001854,
|
| 629 |
+
"grad_norm": 0.279296875,
|
| 630 |
+
"learning_rate": 0.0009126230357451217,
|
| 631 |
+
"loss": 1.578,
|
| 632 |
+
"step": 2047
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.08927416224608617,
|
| 636 |
+
"grad_norm": 0.30859375,
|
| 637 |
+
"learning_rate": 0.0009116301156967709,
|
| 638 |
+
"loss": 1.5755,
|
| 639 |
+
"step": 2070
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.0902660973821538,
|
| 643 |
+
"grad_norm": 0.24609375,
|
| 644 |
+
"learning_rate": 0.0009106371956484199,
|
| 645 |
+
"loss": 1.5772,
|
| 646 |
+
"step": 2093
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.09125803251822141,
|
| 650 |
+
"grad_norm": 0.2734375,
|
| 651 |
+
"learning_rate": 0.0009096442756000692,
|
| 652 |
+
"loss": 1.5405,
|
| 653 |
+
"step": 2116
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.09224996765428904,
|
| 657 |
+
"grad_norm": 0.333984375,
|
| 658 |
+
"learning_rate": 0.0009086513555517182,
|
| 659 |
+
"loss": 1.5458,
|
| 660 |
+
"step": 2139
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.09324190279035667,
|
| 664 |
+
"grad_norm": 0.36328125,
|
| 665 |
+
"learning_rate": 0.0009076584355033673,
|
| 666 |
+
"loss": 1.5783,
|
| 667 |
+
"step": 2162
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.09423383792642429,
|
| 671 |
+
"grad_norm": 0.314453125,
|
| 672 |
+
"learning_rate": 0.0009066655154550164,
|
| 673 |
+
"loss": 1.5237,
|
| 674 |
+
"step": 2185
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.09522577306249191,
|
| 678 |
+
"grad_norm": 0.302734375,
|
| 679 |
+
"learning_rate": 0.0009056725954066655,
|
| 680 |
+
"loss": 1.5361,
|
| 681 |
+
"step": 2208
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.09621770819855954,
|
| 685 |
+
"grad_norm": 0.28125,
|
| 686 |
+
"learning_rate": 0.0009046796753583147,
|
| 687 |
+
"loss": 1.5493,
|
| 688 |
+
"step": 2231
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.09720964333462716,
|
| 692 |
+
"grad_norm": 0.369140625,
|
| 693 |
+
"learning_rate": 0.0009036867553099638,
|
| 694 |
+
"loss": 1.5243,
|
| 695 |
+
"step": 2254
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.09820157847069479,
|
| 699 |
+
"grad_norm": 0.41796875,
|
| 700 |
+
"learning_rate": 0.0009026938352616129,
|
| 701 |
+
"loss": 1.5061,
|
| 702 |
+
"step": 2277
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.09919351360676242,
|
| 706 |
+
"grad_norm": 0.3203125,
|
| 707 |
+
"learning_rate": 0.000901700915213262,
|
| 708 |
+
"loss": 1.5141,
|
| 709 |
+
"step": 2300
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.09996981066977186,
|
| 713 |
+
"eval_runtime": 163.8773,
|
| 714 |
+
"eval_samples_per_second": 610.212,
|
| 715 |
+
"eval_steps_per_second": 7.628,
|
| 716 |
+
"step": 2318
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.10018544874283003,
|
| 720 |
+
"grad_norm": 0.27734375,
|
| 721 |
+
"learning_rate": 0.0009007079951649111,
|
| 722 |
+
"loss": 1.5069,
|
| 723 |
+
"step": 2323
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.10117738387889766,
|
| 727 |
+
"grad_norm": 0.310546875,
|
| 728 |
+
"learning_rate": 0.0008997150751165601,
|
| 729 |
+
"loss": 1.513,
|
| 730 |
+
"step": 2346
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.10216931901496529,
|
| 734 |
+
"grad_norm": 0.33203125,
|
| 735 |
+
"learning_rate": 0.0008987221550682093,
|
| 736 |
+
"loss": 1.5015,
|
| 737 |
+
"step": 2369
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.1031612541510329,
|
| 741 |
+
"grad_norm": 0.3359375,
|
| 742 |
+
"learning_rate": 0.0008977292350198584,
|
| 743 |
+
"loss": 1.4961,
|
| 744 |
+
"step": 2392
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.10415318928710053,
|
| 748 |
+
"grad_norm": 0.271484375,
|
| 749 |
+
"learning_rate": 0.0008967363149715075,
|
| 750 |
+
"loss": 1.4802,
|
| 751 |
+
"step": 2415
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.10514512442316816,
|
| 755 |
+
"grad_norm": 0.369140625,
|
| 756 |
+
"learning_rate": 0.0008957433949231566,
|
| 757 |
+
"loss": 1.4725,
|
| 758 |
+
"step": 2438
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.10613705955923577,
|
| 762 |
+
"grad_norm": 0.23828125,
|
| 763 |
+
"learning_rate": 0.0008947504748748057,
|
| 764 |
+
"loss": 1.4698,
|
| 765 |
+
"step": 2461
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.1071289946953034,
|
| 769 |
+
"grad_norm": 0.357421875,
|
| 770 |
+
"learning_rate": 0.0008937575548264549,
|
| 771 |
+
"loss": 1.4954,
|
| 772 |
+
"step": 2484
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.10812092983137103,
|
| 776 |
+
"grad_norm": 0.43359375,
|
| 777 |
+
"learning_rate": 0.000892764634778104,
|
| 778 |
+
"loss": 1.4221,
|
| 779 |
+
"step": 2507
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.10911286496743865,
|
| 783 |
+
"grad_norm": 0.376953125,
|
| 784 |
+
"learning_rate": 0.000891771714729753,
|
| 785 |
+
"loss": 1.4364,
|
| 786 |
+
"step": 2530
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.11010480010350628,
|
| 790 |
+
"grad_norm": 0.29296875,
|
| 791 |
+
"learning_rate": 0.0008907787946814022,
|
| 792 |
+
"loss": 1.4479,
|
| 793 |
+
"step": 2553
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.1110967352395739,
|
| 797 |
+
"grad_norm": 0.341796875,
|
| 798 |
+
"learning_rate": 0.0008897858746330512,
|
| 799 |
+
"loss": 1.4662,
|
| 800 |
+
"step": 2576
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.11208867037564152,
|
| 804 |
+
"grad_norm": 0.31640625,
|
| 805 |
+
"learning_rate": 0.0008887929545847005,
|
| 806 |
+
"loss": 1.4463,
|
| 807 |
+
"step": 2599
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.11308060551170915,
|
| 811 |
+
"grad_norm": 0.322265625,
|
| 812 |
+
"learning_rate": 0.0008878000345363495,
|
| 813 |
+
"loss": 1.4342,
|
| 814 |
+
"step": 2622
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.11407254064777678,
|
| 818 |
+
"grad_norm": 0.283203125,
|
| 819 |
+
"learning_rate": 0.0008868071144879987,
|
| 820 |
+
"loss": 1.4457,
|
| 821 |
+
"step": 2645
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.11506447578384439,
|
| 825 |
+
"grad_norm": 0.34765625,
|
| 826 |
+
"learning_rate": 0.0008858141944396477,
|
| 827 |
+
"loss": 1.4413,
|
| 828 |
+
"step": 2668
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.11605641091991202,
|
| 832 |
+
"grad_norm": 0.35546875,
|
| 833 |
+
"learning_rate": 0.000884821274391297,
|
| 834 |
+
"loss": 1.4272,
|
| 835 |
+
"step": 2691
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.11704834605597965,
|
| 839 |
+
"grad_norm": 0.31640625,
|
| 840 |
+
"learning_rate": 0.000883828354342946,
|
| 841 |
+
"loss": 1.4108,
|
| 842 |
+
"step": 2714
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.11804028119204726,
|
| 846 |
+
"grad_norm": 0.3359375,
|
| 847 |
+
"learning_rate": 0.0008828354342945951,
|
| 848 |
+
"loss": 1.4226,
|
| 849 |
+
"step": 2737
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.11903221632811489,
|
| 853 |
+
"grad_norm": 0.287109375,
|
| 854 |
+
"learning_rate": 0.0008818425142462442,
|
| 855 |
+
"loss": 1.4096,
|
| 856 |
+
"step": 2760
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.12002415146418252,
|
| 860 |
+
"grad_norm": 0.3125,
|
| 861 |
+
"learning_rate": 0.0008808495941978933,
|
| 862 |
+
"loss": 1.3829,
|
| 863 |
+
"step": 2783
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.12101608660025014,
|
| 867 |
+
"grad_norm": 0.396484375,
|
| 868 |
+
"learning_rate": 0.0008798566741495424,
|
| 869 |
+
"loss": 1.3841,
|
| 870 |
+
"step": 2806
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.12200802173631776,
|
| 874 |
+
"grad_norm": 0.27734375,
|
| 875 |
+
"learning_rate": 0.0008788637541011916,
|
| 876 |
+
"loss": 1.4068,
|
| 877 |
+
"step": 2829
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.1229999568723854,
|
| 881 |
+
"grad_norm": 0.380859375,
|
| 882 |
+
"learning_rate": 0.0008778708340528407,
|
| 883 |
+
"loss": 1.3872,
|
| 884 |
+
"step": 2852
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.12399189200845301,
|
| 888 |
+
"grad_norm": 0.2890625,
|
| 889 |
+
"learning_rate": 0.0008768779140044898,
|
| 890 |
+
"loss": 1.3999,
|
| 891 |
+
"step": 2875
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.12498382714452064,
|
| 895 |
+
"grad_norm": 0.3359375,
|
| 896 |
+
"learning_rate": 0.0008758849939561388,
|
| 897 |
+
"loss": 1.3889,
|
| 898 |
+
"step": 2898
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.12597576228058827,
|
| 902 |
+
"grad_norm": 0.271484375,
|
| 903 |
+
"learning_rate": 0.0008748920739077879,
|
| 904 |
+
"loss": 1.3763,
|
| 905 |
+
"step": 2921
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.1269676974166559,
|
| 909 |
+
"grad_norm": 0.328125,
|
| 910 |
+
"learning_rate": 0.000873899153859437,
|
| 911 |
+
"loss": 1.39,
|
| 912 |
+
"step": 2944
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.1279596325527235,
|
| 916 |
+
"grad_norm": 0.3203125,
|
| 917 |
+
"learning_rate": 0.0008729062338110862,
|
| 918 |
+
"loss": 1.3858,
|
| 919 |
+
"step": 2967
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.12895156768879112,
|
| 923 |
+
"grad_norm": 0.2734375,
|
| 924 |
+
"learning_rate": 0.0008719133137627353,
|
| 925 |
+
"loss": 1.3674,
|
| 926 |
+
"step": 2990
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.12994350282485875,
|
| 930 |
+
"grad_norm": 0.30078125,
|
| 931 |
+
"learning_rate": 0.0008709203937143844,
|
| 932 |
+
"loss": 1.3458,
|
| 933 |
+
"step": 3013
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.13093543796092638,
|
| 937 |
+
"grad_norm": 0.349609375,
|
| 938 |
+
"learning_rate": 0.0008699274736660335,
|
| 939 |
+
"loss": 1.3549,
|
| 940 |
+
"step": 3036
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.131927373096994,
|
| 944 |
+
"grad_norm": 0.2890625,
|
| 945 |
+
"learning_rate": 0.0008689345536176825,
|
| 946 |
+
"loss": 1.3683,
|
| 947 |
+
"step": 3059
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.13291930823306164,
|
| 951 |
+
"grad_norm": 0.298828125,
|
| 952 |
+
"learning_rate": 0.0008679416335693318,
|
| 953 |
+
"loss": 1.3625,
|
| 954 |
+
"step": 3082
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.13391124336912924,
|
| 958 |
+
"grad_norm": 0.318359375,
|
| 959 |
+
"learning_rate": 0.0008669487135209808,
|
| 960 |
+
"loss": 1.3462,
|
| 961 |
+
"step": 3105
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.13490317850519687,
|
| 965 |
+
"grad_norm": 0.34765625,
|
| 966 |
+
"learning_rate": 0.00086595579347263,
|
| 967 |
+
"loss": 1.345,
|
| 968 |
+
"step": 3128
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.1358951136412645,
|
| 972 |
+
"grad_norm": 0.259765625,
|
| 973 |
+
"learning_rate": 0.000864962873424279,
|
| 974 |
+
"loss": 1.3552,
|
| 975 |
+
"step": 3151
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.13688704877733213,
|
| 979 |
+
"grad_norm": 0.240234375,
|
| 980 |
+
"learning_rate": 0.0008639699533759282,
|
| 981 |
+
"loss": 1.3356,
|
| 982 |
+
"step": 3174
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.13787898391339976,
|
| 986 |
+
"grad_norm": 0.255859375,
|
| 987 |
+
"learning_rate": 0.0008629770333275773,
|
| 988 |
+
"loss": 1.3373,
|
| 989 |
+
"step": 3197
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.13887091904946738,
|
| 993 |
+
"grad_norm": 0.35546875,
|
| 994 |
+
"learning_rate": 0.0008619841132792265,
|
| 995 |
+
"loss": 1.3514,
|
| 996 |
+
"step": 3220
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.13986285418553498,
|
| 1000 |
+
"grad_norm": 0.310546875,
|
| 1001 |
+
"learning_rate": 0.0008609911932308755,
|
| 1002 |
+
"loss": 1.3291,
|
| 1003 |
+
"step": 3243
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.1408547893216026,
|
| 1007 |
+
"grad_norm": 0.36328125,
|
| 1008 |
+
"learning_rate": 0.0008599982731825246,
|
| 1009 |
+
"loss": 1.3233,
|
| 1010 |
+
"step": 3266
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.14184672445767024,
|
| 1014 |
+
"grad_norm": 0.38671875,
|
| 1015 |
+
"learning_rate": 0.0008590053531341737,
|
| 1016 |
+
"loss": 1.3174,
|
| 1017 |
+
"step": 3289
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.14283865959373787,
|
| 1021 |
+
"grad_norm": 0.25,
|
| 1022 |
+
"learning_rate": 0.0008580124330858229,
|
| 1023 |
+
"loss": 1.3304,
|
| 1024 |
+
"step": 3312
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.1438305947298055,
|
| 1028 |
+
"grad_norm": 0.3125,
|
| 1029 |
+
"learning_rate": 0.000857019513037472,
|
| 1030 |
+
"loss": 1.3287,
|
| 1031 |
+
"step": 3335
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.14482252986587313,
|
| 1035 |
+
"grad_norm": 0.34375,
|
| 1036 |
+
"learning_rate": 0.0008560265929891211,
|
| 1037 |
+
"loss": 1.3201,
|
| 1038 |
+
"step": 3358
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.14581446500194073,
|
| 1042 |
+
"grad_norm": 0.330078125,
|
| 1043 |
+
"learning_rate": 0.0008550336729407702,
|
| 1044 |
+
"loss": 1.3027,
|
| 1045 |
+
"step": 3381
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.14680640013800836,
|
| 1049 |
+
"grad_norm": 0.34765625,
|
| 1050 |
+
"learning_rate": 0.0008540407528924193,
|
| 1051 |
+
"loss": 1.3128,
|
| 1052 |
+
"step": 3404
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.147798335274076,
|
| 1056 |
+
"grad_norm": 0.3828125,
|
| 1057 |
+
"learning_rate": 0.0008530478328440683,
|
| 1058 |
+
"loss": 1.3182,
|
| 1059 |
+
"step": 3427
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.14879027041014362,
|
| 1063 |
+
"grad_norm": 0.287109375,
|
| 1064 |
+
"learning_rate": 0.0008520549127957176,
|
| 1065 |
+
"loss": 1.2906,
|
| 1066 |
+
"step": 3450
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.14978220554621124,
|
| 1070 |
+
"grad_norm": 0.271484375,
|
| 1071 |
+
"learning_rate": 0.0008510619927473666,
|
| 1072 |
+
"loss": 1.2837,
|
| 1073 |
+
"step": 3473
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.15077414068227887,
|
| 1077 |
+
"grad_norm": 0.365234375,
|
| 1078 |
+
"learning_rate": 0.0008500690726990157,
|
| 1079 |
+
"loss": 1.3114,
|
| 1080 |
+
"step": 3496
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.15176607581834647,
|
| 1084 |
+
"grad_norm": 0.365234375,
|
| 1085 |
+
"learning_rate": 0.0008490761526506648,
|
| 1086 |
+
"loss": 1.3047,
|
| 1087 |
+
"step": 3519
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.1527580109544141,
|
| 1091 |
+
"grad_norm": 0.322265625,
|
| 1092 |
+
"learning_rate": 0.000848083232602314,
|
| 1093 |
+
"loss": 1.2654,
|
| 1094 |
+
"step": 3542
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.15374994609048173,
|
| 1098 |
+
"grad_norm": 0.302734375,
|
| 1099 |
+
"learning_rate": 0.0008470903125539631,
|
| 1100 |
+
"loss": 1.2675,
|
| 1101 |
+
"step": 3565
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.15474188122654936,
|
| 1105 |
+
"grad_norm": 0.3671875,
|
| 1106 |
+
"learning_rate": 0.0008460973925056121,
|
| 1107 |
+
"loss": 1.2671,
|
| 1108 |
+
"step": 3588
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.155733816362617,
|
| 1112 |
+
"grad_norm": 0.26953125,
|
| 1113 |
+
"learning_rate": 0.0008451044724572613,
|
| 1114 |
+
"loss": 1.2719,
|
| 1115 |
+
"step": 3611
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.15672575149868462,
|
| 1119 |
+
"grad_norm": 0.29296875,
|
| 1120 |
+
"learning_rate": 0.0008441115524089103,
|
| 1121 |
+
"loss": 1.2755,
|
| 1122 |
+
"step": 3634
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.15771768663475222,
|
| 1126 |
+
"grad_norm": 0.318359375,
|
| 1127 |
+
"learning_rate": 0.0008431186323605595,
|
| 1128 |
+
"loss": 1.2876,
|
| 1129 |
+
"step": 3657
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.15870962177081985,
|
| 1133 |
+
"grad_norm": 0.35546875,
|
| 1134 |
+
"learning_rate": 0.0008421257123122086,
|
| 1135 |
+
"loss": 1.2928,
|
| 1136 |
+
"step": 3680
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.15970155690688748,
|
| 1140 |
+
"grad_norm": 0.3046875,
|
| 1141 |
+
"learning_rate": 0.0008411327922638578,
|
| 1142 |
+
"loss": 1.2855,
|
| 1143 |
+
"step": 3703
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.1606934920429551,
|
| 1147 |
+
"grad_norm": 0.341796875,
|
| 1148 |
+
"learning_rate": 0.0008401398722155068,
|
| 1149 |
+
"loss": 1.2558,
|
| 1150 |
+
"step": 3726
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.16168542717902273,
|
| 1154 |
+
"grad_norm": 0.34375,
|
| 1155 |
+
"learning_rate": 0.000839146952167156,
|
| 1156 |
+
"loss": 1.2561,
|
| 1157 |
+
"step": 3749
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.16267736231509036,
|
| 1161 |
+
"grad_norm": 0.361328125,
|
| 1162 |
+
"learning_rate": 0.000838154032118805,
|
| 1163 |
+
"loss": 1.2851,
|
| 1164 |
+
"step": 3772
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.16366929745115796,
|
| 1168 |
+
"grad_norm": 0.3359375,
|
| 1169 |
+
"learning_rate": 0.0008371611120704542,
|
| 1170 |
+
"loss": 1.2682,
|
| 1171 |
+
"step": 3795
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.1646612325872256,
|
| 1175 |
+
"grad_norm": 0.33203125,
|
| 1176 |
+
"learning_rate": 0.0008361681920221033,
|
| 1177 |
+
"loss": 1.2709,
|
| 1178 |
+
"step": 3818
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.16565316772329322,
|
| 1182 |
+
"grad_norm": 0.302734375,
|
| 1183 |
+
"learning_rate": 0.0008351752719737524,
|
| 1184 |
+
"loss": 1.2459,
|
| 1185 |
+
"step": 3841
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.16664510285936085,
|
| 1189 |
+
"grad_norm": 0.3203125,
|
| 1190 |
+
"learning_rate": 0.0008341823519254015,
|
| 1191 |
+
"loss": 1.2134,
|
| 1192 |
+
"step": 3864
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.16763703799542848,
|
| 1196 |
+
"grad_norm": 0.365234375,
|
| 1197 |
+
"learning_rate": 0.0008331894318770506,
|
| 1198 |
+
"loss": 1.2445,
|
| 1199 |
+
"step": 3887
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.1686289731314961,
|
| 1203 |
+
"grad_norm": 0.306640625,
|
| 1204 |
+
"learning_rate": 0.0008321965118286998,
|
| 1205 |
+
"loss": 1.2522,
|
| 1206 |
+
"step": 3910
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.1696209082675637,
|
| 1210 |
+
"grad_norm": 0.314453125,
|
| 1211 |
+
"learning_rate": 0.0008312035917803489,
|
| 1212 |
+
"loss": 1.2296,
|
| 1213 |
+
"step": 3933
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.17061284340363134,
|
| 1217 |
+
"grad_norm": 0.33203125,
|
| 1218 |
+
"learning_rate": 0.0008302106717319979,
|
| 1219 |
+
"loss": 1.252,
|
| 1220 |
+
"step": 3956
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.17160477853969897,
|
| 1224 |
+
"grad_norm": 0.29296875,
|
| 1225 |
+
"learning_rate": 0.0008292177516836471,
|
| 1226 |
+
"loss": 1.2501,
|
| 1227 |
+
"step": 3979
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.1725967136757666,
|
| 1231 |
+
"grad_norm": 0.337890625,
|
| 1232 |
+
"learning_rate": 0.0008282248316352961,
|
| 1233 |
+
"loss": 1.2328,
|
| 1234 |
+
"step": 4002
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.17358864881183422,
|
| 1238 |
+
"grad_norm": 0.306640625,
|
| 1239 |
+
"learning_rate": 0.0008272319115869454,
|
| 1240 |
+
"loss": 1.2182,
|
| 1241 |
+
"step": 4025
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.17458058394790185,
|
| 1245 |
+
"grad_norm": 0.400390625,
|
| 1246 |
+
"learning_rate": 0.0008262389915385944,
|
| 1247 |
+
"loss": 1.2515,
|
| 1248 |
+
"step": 4048
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.17557251908396945,
|
| 1252 |
+
"grad_norm": 0.32421875,
|
| 1253 |
+
"learning_rate": 0.0008252460714902435,
|
| 1254 |
+
"loss": 1.2076,
|
| 1255 |
+
"step": 4071
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.17656445422003708,
|
| 1259 |
+
"grad_norm": 0.322265625,
|
| 1260 |
+
"learning_rate": 0.0008242531514418926,
|
| 1261 |
+
"loss": 1.2219,
|
| 1262 |
+
"step": 4094
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.1775563893561047,
|
| 1266 |
+
"grad_norm": 0.2734375,
|
| 1267 |
+
"learning_rate": 0.0008232602313935416,
|
| 1268 |
+
"loss": 1.2307,
|
| 1269 |
+
"step": 4117
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.17854832449217234,
|
| 1273 |
+
"grad_norm": 0.41015625,
|
| 1274 |
+
"learning_rate": 0.0008222673113451908,
|
| 1275 |
+
"loss": 1.2368,
|
| 1276 |
+
"step": 4140
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.17954025962823997,
|
| 1280 |
+
"grad_norm": 0.322265625,
|
| 1281 |
+
"learning_rate": 0.0008212743912968399,
|
| 1282 |
+
"loss": 1.2332,
|
| 1283 |
+
"step": 4163
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.1805321947643076,
|
| 1287 |
+
"grad_norm": 0.2734375,
|
| 1288 |
+
"learning_rate": 0.0008202814712484891,
|
| 1289 |
+
"loss": 1.2086,
|
| 1290 |
+
"step": 4186
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.1815241299003752,
|
| 1294 |
+
"grad_norm": 0.4375,
|
| 1295 |
+
"learning_rate": 0.0008192885512001381,
|
| 1296 |
+
"loss": 1.2204,
|
| 1297 |
+
"step": 4209
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.18251606503644283,
|
| 1301 |
+
"grad_norm": 0.365234375,
|
| 1302 |
+
"learning_rate": 0.0008182956311517873,
|
| 1303 |
+
"loss": 1.2379,
|
| 1304 |
+
"step": 4232
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.18350800017251045,
|
| 1308 |
+
"grad_norm": 0.294921875,
|
| 1309 |
+
"learning_rate": 0.0008173027111034363,
|
| 1310 |
+
"loss": 1.2387,
|
| 1311 |
+
"step": 4255
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.18449993530857808,
|
| 1315 |
+
"grad_norm": 0.28125,
|
| 1316 |
+
"learning_rate": 0.0008163097910550856,
|
| 1317 |
+
"loss": 1.2008,
|
| 1318 |
+
"step": 4278
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.1854918704446457,
|
| 1322 |
+
"grad_norm": 0.28515625,
|
| 1323 |
+
"learning_rate": 0.0008153168710067346,
|
| 1324 |
+
"loss": 1.2191,
|
| 1325 |
+
"step": 4301
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.18648380558071334,
|
| 1329 |
+
"grad_norm": 0.30859375,
|
| 1330 |
+
"learning_rate": 0.0008143239509583837,
|
| 1331 |
+
"loss": 1.2153,
|
| 1332 |
+
"step": 4324
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.18747574071678094,
|
| 1336 |
+
"grad_norm": 0.404296875,
|
| 1337 |
+
"learning_rate": 0.0008133310309100328,
|
| 1338 |
+
"loss": 1.2171,
|
| 1339 |
+
"step": 4347
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.18846767585284857,
|
| 1343 |
+
"grad_norm": 0.298828125,
|
| 1344 |
+
"learning_rate": 0.0008123381108616819,
|
| 1345 |
+
"loss": 1.2154,
|
| 1346 |
+
"step": 4370
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.1894596109889162,
|
| 1350 |
+
"grad_norm": 0.353515625,
|
| 1351 |
+
"learning_rate": 0.0008113451908133311,
|
| 1352 |
+
"loss": 1.2055,
|
| 1353 |
+
"step": 4393
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.19045154612498383,
|
| 1357 |
+
"grad_norm": 0.337890625,
|
| 1358 |
+
"learning_rate": 0.0008103522707649802,
|
| 1359 |
+
"loss": 1.2215,
|
| 1360 |
+
"step": 4416
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.19144348126105146,
|
| 1364 |
+
"grad_norm": 0.3203125,
|
| 1365 |
+
"learning_rate": 0.0008093593507166293,
|
| 1366 |
+
"loss": 1.2035,
|
| 1367 |
+
"step": 4439
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.19243541639711909,
|
| 1371 |
+
"grad_norm": 0.326171875,
|
| 1372 |
+
"learning_rate": 0.0008083664306682784,
|
| 1373 |
+
"loss": 1.183,
|
| 1374 |
+
"step": 4462
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.1934273515331867,
|
| 1378 |
+
"grad_norm": 0.34765625,
|
| 1379 |
+
"learning_rate": 0.0008073735106199274,
|
| 1380 |
+
"loss": 1.1892,
|
| 1381 |
+
"step": 4485
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.19441928666925432,
|
| 1385 |
+
"grad_norm": 0.337890625,
|
| 1386 |
+
"learning_rate": 0.0008063805905715767,
|
| 1387 |
+
"loss": 1.2006,
|
| 1388 |
+
"step": 4508
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.19541122180532194,
|
| 1392 |
+
"grad_norm": 0.41796875,
|
| 1393 |
+
"learning_rate": 0.0008053876705232257,
|
| 1394 |
+
"loss": 1.2081,
|
| 1395 |
+
"step": 4531
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.19640315694138957,
|
| 1399 |
+
"grad_norm": 0.357421875,
|
| 1400 |
+
"learning_rate": 0.0008043947504748749,
|
| 1401 |
+
"loss": 1.1878,
|
| 1402 |
+
"step": 4554
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.1973950920774572,
|
| 1406 |
+
"grad_norm": 0.3125,
|
| 1407 |
+
"learning_rate": 0.0008034018304265239,
|
| 1408 |
+
"loss": 1.1701,
|
| 1409 |
+
"step": 4577
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.19838702721352483,
|
| 1413 |
+
"grad_norm": 0.337890625,
|
| 1414 |
+
"learning_rate": 0.0008024089103781731,
|
| 1415 |
+
"loss": 1.179,
|
| 1416 |
+
"step": 4600
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.19937896234959243,
|
| 1420 |
+
"grad_norm": 0.361328125,
|
| 1421 |
+
"learning_rate": 0.0008014159903298221,
|
| 1422 |
+
"loss": 1.1794,
|
| 1423 |
+
"step": 4623
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.1999396213395437,
|
| 1427 |
+
"eval_runtime": 164.1241,
|
| 1428 |
+
"eval_samples_per_second": 609.295,
|
| 1429 |
+
"eval_steps_per_second": 7.616,
|
| 1430 |
+
"step": 4636
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.20037089748566006,
|
| 1434 |
+
"grad_norm": 0.357421875,
|
| 1435 |
+
"learning_rate": 0.0008004230702814712,
|
| 1436 |
+
"loss": 1.181,
|
| 1437 |
+
"step": 4646
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.2013628326217277,
|
| 1441 |
+
"grad_norm": 0.369140625,
|
| 1442 |
+
"learning_rate": 0.0007994301502331204,
|
| 1443 |
+
"loss": 1.1706,
|
| 1444 |
+
"step": 4669
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.20235476775779532,
|
| 1448 |
+
"grad_norm": 0.30078125,
|
| 1449 |
+
"learning_rate": 0.0007984372301847694,
|
| 1450 |
+
"loss": 1.2005,
|
| 1451 |
+
"step": 4692
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.20334670289386295,
|
| 1455 |
+
"grad_norm": 0.34375,
|
| 1456 |
+
"learning_rate": 0.0007974443101364186,
|
| 1457 |
+
"loss": 1.1784,
|
| 1458 |
+
"step": 4715
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.20433863802993057,
|
| 1462 |
+
"grad_norm": 0.310546875,
|
| 1463 |
+
"learning_rate": 0.0007964513900880676,
|
| 1464 |
+
"loss": 1.1774,
|
| 1465 |
+
"step": 4738
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.20533057316599818,
|
| 1469 |
+
"grad_norm": 0.26953125,
|
| 1470 |
+
"learning_rate": 0.0007954584700397169,
|
| 1471 |
+
"loss": 1.1589,
|
| 1472 |
+
"step": 4761
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.2063225083020658,
|
| 1476 |
+
"grad_norm": 0.345703125,
|
| 1477 |
+
"learning_rate": 0.0007944655499913659,
|
| 1478 |
+
"loss": 1.1821,
|
| 1479 |
+
"step": 4784
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.20731444343813343,
|
| 1483 |
+
"grad_norm": 0.287109375,
|
| 1484 |
+
"learning_rate": 0.0007934726299430151,
|
| 1485 |
+
"loss": 1.1496,
|
| 1486 |
+
"step": 4807
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.20830637857420106,
|
| 1490 |
+
"grad_norm": 0.3671875,
|
| 1491 |
+
"learning_rate": 0.0007924797098946641,
|
| 1492 |
+
"loss": 1.1666,
|
| 1493 |
+
"step": 4830
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.2092983137102687,
|
| 1497 |
+
"grad_norm": 0.3046875,
|
| 1498 |
+
"learning_rate": 0.0007914867898463132,
|
| 1499 |
+
"loss": 1.1546,
|
| 1500 |
+
"step": 4853
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.21029024884633632,
|
| 1504 |
+
"grad_norm": 0.361328125,
|
| 1505 |
+
"learning_rate": 0.0007904938697979624,
|
| 1506 |
+
"loss": 1.2045,
|
| 1507 |
+
"step": 4876
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.21128218398240392,
|
| 1511 |
+
"grad_norm": 0.56640625,
|
| 1512 |
+
"learning_rate": 0.0007895009497496115,
|
| 1513 |
+
"loss": 1.1757,
|
| 1514 |
+
"step": 4899
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.21227411911847155,
|
| 1518 |
+
"grad_norm": 0.36328125,
|
| 1519 |
+
"learning_rate": 0.0007885080297012606,
|
| 1520 |
+
"loss": 1.1894,
|
| 1521 |
+
"step": 4922
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.21326605425453918,
|
| 1525 |
+
"grad_norm": 0.3046875,
|
| 1526 |
+
"learning_rate": 0.0007875151096529097,
|
| 1527 |
+
"loss": 1.174,
|
| 1528 |
+
"step": 4945
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.2142579893906068,
|
| 1532 |
+
"grad_norm": 0.322265625,
|
| 1533 |
+
"learning_rate": 0.0007865221896045588,
|
| 1534 |
+
"loss": 1.1423,
|
| 1535 |
+
"step": 4968
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.21524992452667444,
|
| 1539 |
+
"grad_norm": 0.421875,
|
| 1540 |
+
"learning_rate": 0.000785529269556208,
|
| 1541 |
+
"loss": 1.1532,
|
| 1542 |
+
"step": 4991
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.21624185966274206,
|
| 1546 |
+
"grad_norm": 0.3203125,
|
| 1547 |
+
"learning_rate": 0.000784536349507857,
|
| 1548 |
+
"loss": 1.1688,
|
| 1549 |
+
"step": 5014
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.21723379479880967,
|
| 1553 |
+
"grad_norm": 0.2734375,
|
| 1554 |
+
"learning_rate": 0.0007835434294595062,
|
| 1555 |
+
"loss": 1.1446,
|
| 1556 |
+
"step": 5037
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.2182257299348773,
|
| 1560 |
+
"grad_norm": 0.365234375,
|
| 1561 |
+
"learning_rate": 0.0007825505094111552,
|
| 1562 |
+
"loss": 1.1792,
|
| 1563 |
+
"step": 5060
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.21921766507094492,
|
| 1567 |
+
"grad_norm": 0.33203125,
|
| 1568 |
+
"learning_rate": 0.0007815575893628044,
|
| 1569 |
+
"loss": 1.1516,
|
| 1570 |
+
"step": 5083
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.22020960020701255,
|
| 1574 |
+
"grad_norm": 0.296875,
|
| 1575 |
+
"learning_rate": 0.0007805646693144534,
|
| 1576 |
+
"loss": 1.1712,
|
| 1577 |
+
"step": 5106
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.22120153534308018,
|
| 1581 |
+
"grad_norm": 0.341796875,
|
| 1582 |
+
"learning_rate": 0.0007795717492661027,
|
| 1583 |
+
"loss": 1.1562,
|
| 1584 |
+
"step": 5129
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.2221934704791478,
|
| 1588 |
+
"grad_norm": 0.470703125,
|
| 1589 |
+
"learning_rate": 0.0007785788292177517,
|
| 1590 |
+
"loss": 1.1641,
|
| 1591 |
+
"step": 5152
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.2231854056152154,
|
| 1595 |
+
"grad_norm": 0.4765625,
|
| 1596 |
+
"learning_rate": 0.0007775859091694009,
|
| 1597 |
+
"loss": 1.1649,
|
| 1598 |
+
"step": 5175
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.22417734075128304,
|
| 1602 |
+
"grad_norm": 0.333984375,
|
| 1603 |
+
"learning_rate": 0.0007765929891210499,
|
| 1604 |
+
"loss": 1.1684,
|
| 1605 |
+
"step": 5198
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.22516927588735067,
|
| 1609 |
+
"grad_norm": 0.33984375,
|
| 1610 |
+
"learning_rate": 0.0007756000690726989,
|
| 1611 |
+
"loss": 1.1659,
|
| 1612 |
+
"step": 5221
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.2261612110234183,
|
| 1616 |
+
"grad_norm": 0.419921875,
|
| 1617 |
+
"learning_rate": 0.0007746071490243482,
|
| 1618 |
+
"loss": 1.1683,
|
| 1619 |
+
"step": 5244
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.22715314615948592,
|
| 1623 |
+
"grad_norm": 0.33203125,
|
| 1624 |
+
"learning_rate": 0.0007736142289759972,
|
| 1625 |
+
"loss": 1.1536,
|
| 1626 |
+
"step": 5267
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.22814508129555355,
|
| 1630 |
+
"grad_norm": 0.3203125,
|
| 1631 |
+
"learning_rate": 0.0007726213089276464,
|
| 1632 |
+
"loss": 1.1451,
|
| 1633 |
+
"step": 5290
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.22913701643162115,
|
| 1637 |
+
"grad_norm": 0.35546875,
|
| 1638 |
+
"learning_rate": 0.0007716283888792954,
|
| 1639 |
+
"loss": 1.1453,
|
| 1640 |
+
"step": 5313
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.23012895156768878,
|
| 1644 |
+
"grad_norm": 0.298828125,
|
| 1645 |
+
"learning_rate": 0.0007706354688309446,
|
| 1646 |
+
"loss": 1.1585,
|
| 1647 |
+
"step": 5336
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.2311208867037564,
|
| 1651 |
+
"grad_norm": 0.3828125,
|
| 1652 |
+
"learning_rate": 0.0007696425487825937,
|
| 1653 |
+
"loss": 1.1585,
|
| 1654 |
+
"step": 5359
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.23211282183982404,
|
| 1658 |
+
"grad_norm": 0.298828125,
|
| 1659 |
+
"learning_rate": 0.0007686496287342428,
|
| 1660 |
+
"loss": 1.1396,
|
| 1661 |
+
"step": 5382
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.23310475697589167,
|
| 1665 |
+
"grad_norm": 0.35546875,
|
| 1666 |
+
"learning_rate": 0.0007676567086858919,
|
| 1667 |
+
"loss": 1.1527,
|
| 1668 |
+
"step": 5405
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.2340966921119593,
|
| 1672 |
+
"grad_norm": 0.310546875,
|
| 1673 |
+
"learning_rate": 0.000766663788637541,
|
| 1674 |
+
"loss": 1.1157,
|
| 1675 |
+
"step": 5428
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.2350886272480269,
|
| 1679 |
+
"grad_norm": 0.31640625,
|
| 1680 |
+
"learning_rate": 0.0007656708685891901,
|
| 1681 |
+
"loss": 1.1232,
|
| 1682 |
+
"step": 5451
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.23608056238409453,
|
| 1686 |
+
"grad_norm": 0.3359375,
|
| 1687 |
+
"learning_rate": 0.0007646779485408393,
|
| 1688 |
+
"loss": 1.1411,
|
| 1689 |
+
"step": 5474
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.23707249752016216,
|
| 1693 |
+
"grad_norm": 0.314453125,
|
| 1694 |
+
"learning_rate": 0.0007636850284924884,
|
| 1695 |
+
"loss": 1.1329,
|
| 1696 |
+
"step": 5497
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.23806443265622979,
|
| 1700 |
+
"grad_norm": 0.314453125,
|
| 1701 |
+
"learning_rate": 0.0007626921084441375,
|
| 1702 |
+
"loss": 1.1262,
|
| 1703 |
+
"step": 5520
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.23905636779229741,
|
| 1707 |
+
"grad_norm": 0.3359375,
|
| 1708 |
+
"learning_rate": 0.0007616991883957865,
|
| 1709 |
+
"loss": 1.1346,
|
| 1710 |
+
"step": 5543
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.24004830292836504,
|
| 1714 |
+
"grad_norm": 0.306640625,
|
| 1715 |
+
"learning_rate": 0.0007607062683474357,
|
| 1716 |
+
"loss": 1.139,
|
| 1717 |
+
"step": 5566
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.24104023806443264,
|
| 1721 |
+
"grad_norm": 0.396484375,
|
| 1722 |
+
"learning_rate": 0.0007597133482990847,
|
| 1723 |
+
"loss": 1.1318,
|
| 1724 |
+
"step": 5589
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.24203217320050027,
|
| 1728 |
+
"grad_norm": 0.35546875,
|
| 1729 |
+
"learning_rate": 0.000758720428250734,
|
| 1730 |
+
"loss": 1.1178,
|
| 1731 |
+
"step": 5612
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.2430241083365679,
|
| 1735 |
+
"grad_norm": 0.357421875,
|
| 1736 |
+
"learning_rate": 0.000757727508202383,
|
| 1737 |
+
"loss": 1.1312,
|
| 1738 |
+
"step": 5635
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.24401604347263553,
|
| 1742 |
+
"grad_norm": 0.373046875,
|
| 1743 |
+
"learning_rate": 0.0007567345881540322,
|
| 1744 |
+
"loss": 1.1263,
|
| 1745 |
+
"step": 5658
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.24500797860870316,
|
| 1749 |
+
"grad_norm": 0.306640625,
|
| 1750 |
+
"learning_rate": 0.0007557416681056812,
|
| 1751 |
+
"loss": 1.1303,
|
| 1752 |
+
"step": 5681
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.2459999137447708,
|
| 1756 |
+
"grad_norm": 0.412109375,
|
| 1757 |
+
"learning_rate": 0.0007547487480573305,
|
| 1758 |
+
"loss": 1.1365,
|
| 1759 |
+
"step": 5704
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.2469918488808384,
|
| 1763 |
+
"grad_norm": 0.375,
|
| 1764 |
+
"learning_rate": 0.0007537558280089795,
|
| 1765 |
+
"loss": 1.1239,
|
| 1766 |
+
"step": 5727
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.24798378401690602,
|
| 1770 |
+
"grad_norm": 0.33984375,
|
| 1771 |
+
"learning_rate": 0.0007527629079606286,
|
| 1772 |
+
"loss": 1.1212,
|
| 1773 |
+
"step": 5750
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.24897571915297365,
|
| 1777 |
+
"grad_norm": 0.396484375,
|
| 1778 |
+
"learning_rate": 0.0007517699879122777,
|
| 1779 |
+
"loss": 1.1128,
|
| 1780 |
+
"step": 5773
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.24996765428904127,
|
| 1784 |
+
"grad_norm": 0.345703125,
|
| 1785 |
+
"learning_rate": 0.0007507770678639267,
|
| 1786 |
+
"loss": 1.1153,
|
| 1787 |
+
"step": 5796
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.2509595894251089,
|
| 1791 |
+
"grad_norm": 0.337890625,
|
| 1792 |
+
"learning_rate": 0.0007497841478155759,
|
| 1793 |
+
"loss": 1.1191,
|
| 1794 |
+
"step": 5819
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.25195152456117653,
|
| 1798 |
+
"grad_norm": 0.357421875,
|
| 1799 |
+
"learning_rate": 0.000748791227767225,
|
| 1800 |
+
"loss": 1.1115,
|
| 1801 |
+
"step": 5842
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.25294345969724413,
|
| 1805 |
+
"grad_norm": 0.484375,
|
| 1806 |
+
"learning_rate": 0.0007477983077188742,
|
| 1807 |
+
"loss": 1.1377,
|
| 1808 |
+
"step": 5865
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.2539353948333118,
|
| 1812 |
+
"grad_norm": 0.3671875,
|
| 1813 |
+
"learning_rate": 0.0007468053876705232,
|
| 1814 |
+
"loss": 1.1345,
|
| 1815 |
+
"step": 5888
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.2549273299693794,
|
| 1819 |
+
"grad_norm": 0.3359375,
|
| 1820 |
+
"learning_rate": 0.0007458124676221723,
|
| 1821 |
+
"loss": 1.1226,
|
| 1822 |
+
"step": 5911
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.255919265105447,
|
| 1826 |
+
"grad_norm": 0.361328125,
|
| 1827 |
+
"learning_rate": 0.0007448195475738214,
|
| 1828 |
+
"loss": 1.1163,
|
| 1829 |
+
"step": 5934
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.25691120024151465,
|
| 1833 |
+
"grad_norm": 0.361328125,
|
| 1834 |
+
"learning_rate": 0.0007438266275254706,
|
| 1835 |
+
"loss": 1.1255,
|
| 1836 |
+
"step": 5957
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.25790313537758225,
|
| 1840 |
+
"grad_norm": 0.322265625,
|
| 1841 |
+
"learning_rate": 0.0007428337074771197,
|
| 1842 |
+
"loss": 1.1049,
|
| 1843 |
+
"step": 5980
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.2588950705136499,
|
| 1847 |
+
"grad_norm": 0.427734375,
|
| 1848 |
+
"learning_rate": 0.0007418407874287688,
|
| 1849 |
+
"loss": 1.1294,
|
| 1850 |
+
"step": 6003
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.2598870056497175,
|
| 1854 |
+
"grad_norm": 0.34765625,
|
| 1855 |
+
"learning_rate": 0.0007408478673804179,
|
| 1856 |
+
"loss": 1.1167,
|
| 1857 |
+
"step": 6026
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.26087894078578516,
|
| 1861 |
+
"grad_norm": 0.388671875,
|
| 1862 |
+
"learning_rate": 0.000739854947332067,
|
| 1863 |
+
"loss": 1.1041,
|
| 1864 |
+
"step": 6049
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.26187087592185276,
|
| 1868 |
+
"grad_norm": 0.302734375,
|
| 1869 |
+
"learning_rate": 0.000738862027283716,
|
| 1870 |
+
"loss": 1.1146,
|
| 1871 |
+
"step": 6072
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.26286281105792036,
|
| 1875 |
+
"grad_norm": 0.341796875,
|
| 1876 |
+
"learning_rate": 0.0007378691072353653,
|
| 1877 |
+
"loss": 1.104,
|
| 1878 |
+
"step": 6095
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.263854746193988,
|
| 1882 |
+
"grad_norm": 0.416015625,
|
| 1883 |
+
"learning_rate": 0.0007368761871870143,
|
| 1884 |
+
"loss": 1.1091,
|
| 1885 |
+
"step": 6118
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.2648466813300556,
|
| 1889 |
+
"grad_norm": 0.326171875,
|
| 1890 |
+
"learning_rate": 0.0007358832671386635,
|
| 1891 |
+
"loss": 1.1295,
|
| 1892 |
+
"step": 6141
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.2658386164661233,
|
| 1896 |
+
"grad_norm": 0.341796875,
|
| 1897 |
+
"learning_rate": 0.0007348903470903125,
|
| 1898 |
+
"loss": 1.0913,
|
| 1899 |
+
"step": 6164
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.2668305516021909,
|
| 1903 |
+
"grad_norm": 0.32421875,
|
| 1904 |
+
"learning_rate": 0.0007338974270419618,
|
| 1905 |
+
"loss": 1.0806,
|
| 1906 |
+
"step": 6187
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.2678224867382585,
|
| 1910 |
+
"grad_norm": 0.380859375,
|
| 1911 |
+
"learning_rate": 0.0007329045069936108,
|
| 1912 |
+
"loss": 1.099,
|
| 1913 |
+
"step": 6210
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.26881442187432614,
|
| 1917 |
+
"grad_norm": 0.453125,
|
| 1918 |
+
"learning_rate": 0.00073191158694526,
|
| 1919 |
+
"loss": 1.0982,
|
| 1920 |
+
"step": 6233
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.26980635701039374,
|
| 1924 |
+
"grad_norm": 0.3515625,
|
| 1925 |
+
"learning_rate": 0.000730918666896909,
|
| 1926 |
+
"loss": 1.093,
|
| 1927 |
+
"step": 6256
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.2707982921464614,
|
| 1931 |
+
"grad_norm": 0.3359375,
|
| 1932 |
+
"learning_rate": 0.0007299257468485581,
|
| 1933 |
+
"loss": 1.0773,
|
| 1934 |
+
"step": 6279
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.271790227282529,
|
| 1938 |
+
"grad_norm": 0.33984375,
|
| 1939 |
+
"learning_rate": 0.0007289328268002072,
|
| 1940 |
+
"loss": 1.101,
|
| 1941 |
+
"step": 6302
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.27278216241859665,
|
| 1945 |
+
"grad_norm": 0.328125,
|
| 1946 |
+
"learning_rate": 0.0007279399067518564,
|
| 1947 |
+
"loss": 1.1084,
|
| 1948 |
+
"step": 6325
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.27377409755466425,
|
| 1952 |
+
"grad_norm": 0.359375,
|
| 1953 |
+
"learning_rate": 0.0007269469867035055,
|
| 1954 |
+
"loss": 1.0869,
|
| 1955 |
+
"step": 6348
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.27476603269073185,
|
| 1959 |
+
"grad_norm": 0.3203125,
|
| 1960 |
+
"learning_rate": 0.0007259540666551545,
|
| 1961 |
+
"loss": 1.088,
|
| 1962 |
+
"step": 6371
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.2757579678267995,
|
| 1966 |
+
"grad_norm": 0.3984375,
|
| 1967 |
+
"learning_rate": 0.0007249611466068037,
|
| 1968 |
+
"loss": 1.1002,
|
| 1969 |
+
"step": 6394
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.2767499029628671,
|
| 1973 |
+
"grad_norm": 0.30859375,
|
| 1974 |
+
"learning_rate": 0.0007239682265584527,
|
| 1975 |
+
"loss": 1.083,
|
| 1976 |
+
"step": 6417
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.27774183809893477,
|
| 1980 |
+
"grad_norm": 0.306640625,
|
| 1981 |
+
"learning_rate": 0.0007229753065101019,
|
| 1982 |
+
"loss": 1.0884,
|
| 1983 |
+
"step": 6440
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.27873377323500237,
|
| 1987 |
+
"grad_norm": 0.41015625,
|
| 1988 |
+
"learning_rate": 0.000721982386461751,
|
| 1989 |
+
"loss": 1.092,
|
| 1990 |
+
"step": 6463
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.27972570837106997,
|
| 1994 |
+
"grad_norm": 0.3359375,
|
| 1995 |
+
"learning_rate": 0.0007209894664134001,
|
| 1996 |
+
"loss": 1.0736,
|
| 1997 |
+
"step": 6486
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.2807176435071376,
|
| 2001 |
+
"grad_norm": 0.353515625,
|
| 2002 |
+
"learning_rate": 0.0007199965463650492,
|
| 2003 |
+
"loss": 1.091,
|
| 2004 |
+
"step": 6509
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.2817095786432052,
|
| 2008 |
+
"grad_norm": 0.34765625,
|
| 2009 |
+
"learning_rate": 0.0007190036263166983,
|
| 2010 |
+
"loss": 1.0893,
|
| 2011 |
+
"step": 6532
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.2827015137792729,
|
| 2015 |
+
"grad_norm": 0.41015625,
|
| 2016 |
+
"learning_rate": 0.0007180107062683475,
|
| 2017 |
+
"loss": 1.0936,
|
| 2018 |
+
"step": 6555
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.2836934489153405,
|
| 2022 |
+
"grad_norm": 0.423828125,
|
| 2023 |
+
"learning_rate": 0.0007170177862199966,
|
| 2024 |
+
"loss": 1.0955,
|
| 2025 |
+
"step": 6578
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.28468538405140814,
|
| 2029 |
+
"grad_norm": 0.34765625,
|
| 2030 |
+
"learning_rate": 0.0007160248661716456,
|
| 2031 |
+
"loss": 1.1016,
|
| 2032 |
+
"step": 6601
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.28567731918747574,
|
| 2036 |
+
"grad_norm": 0.34765625,
|
| 2037 |
+
"learning_rate": 0.0007150319461232948,
|
| 2038 |
+
"loss": 1.102,
|
| 2039 |
+
"step": 6624
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.28666925432354334,
|
| 2043 |
+
"grad_norm": 0.404296875,
|
| 2044 |
+
"learning_rate": 0.0007140390260749438,
|
| 2045 |
+
"loss": 1.08,
|
| 2046 |
+
"step": 6647
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.287661189459611,
|
| 2050 |
+
"grad_norm": 0.44140625,
|
| 2051 |
+
"learning_rate": 0.000713046106026593,
|
| 2052 |
+
"loss": 1.0785,
|
| 2053 |
+
"step": 6670
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.2886531245956786,
|
| 2057 |
+
"grad_norm": 0.392578125,
|
| 2058 |
+
"learning_rate": 0.0007120531859782421,
|
| 2059 |
+
"loss": 1.0944,
|
| 2060 |
+
"step": 6693
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.28964505973174626,
|
| 2064 |
+
"grad_norm": 0.34375,
|
| 2065 |
+
"learning_rate": 0.0007110602659298913,
|
| 2066 |
+
"loss": 1.0851,
|
| 2067 |
+
"step": 6716
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.29063699486781386,
|
| 2071 |
+
"grad_norm": 0.40234375,
|
| 2072 |
+
"learning_rate": 0.0007100673458815403,
|
| 2073 |
+
"loss": 1.0942,
|
| 2074 |
+
"step": 6739
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.29162893000388146,
|
| 2078 |
+
"grad_norm": 0.42578125,
|
| 2079 |
+
"learning_rate": 0.0007090744258331895,
|
| 2080 |
+
"loss": 1.0831,
|
| 2081 |
+
"step": 6762
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.2926208651399491,
|
| 2085 |
+
"grad_norm": 0.408203125,
|
| 2086 |
+
"learning_rate": 0.0007080815057848385,
|
| 2087 |
+
"loss": 1.0724,
|
| 2088 |
+
"step": 6785
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.2936128002760167,
|
| 2092 |
+
"grad_norm": 0.31640625,
|
| 2093 |
+
"learning_rate": 0.0007070885857364877,
|
| 2094 |
+
"loss": 1.0875,
|
| 2095 |
+
"step": 6808
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.2946047354120844,
|
| 2099 |
+
"grad_norm": 0.466796875,
|
| 2100 |
+
"learning_rate": 0.0007060956656881368,
|
| 2101 |
+
"loss": 1.0896,
|
| 2102 |
+
"step": 6831
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.295596670548152,
|
| 2106 |
+
"grad_norm": 0.3984375,
|
| 2107 |
+
"learning_rate": 0.0007051027456397859,
|
| 2108 |
+
"loss": 1.0702,
|
| 2109 |
+
"step": 6854
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.29658860568421963,
|
| 2113 |
+
"grad_norm": 0.390625,
|
| 2114 |
+
"learning_rate": 0.000704109825591435,
|
| 2115 |
+
"loss": 1.0887,
|
| 2116 |
+
"step": 6877
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.29758054082028723,
|
| 2120 |
+
"grad_norm": 0.341796875,
|
| 2121 |
+
"learning_rate": 0.0007031169055430841,
|
| 2122 |
+
"loss": 1.0708,
|
| 2123 |
+
"step": 6900
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.29857247595635483,
|
| 2127 |
+
"grad_norm": 0.41796875,
|
| 2128 |
+
"learning_rate": 0.0007021239854947333,
|
| 2129 |
+
"loss": 1.0828,
|
| 2130 |
+
"step": 6923
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.2995644110924225,
|
| 2134 |
+
"grad_norm": 0.33984375,
|
| 2135 |
+
"learning_rate": 0.0007011310654463823,
|
| 2136 |
+
"loss": 1.061,
|
| 2137 |
+
"step": 6946
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.2999094320093156,
|
| 2141 |
+
"eval_runtime": 164.053,
|
| 2142 |
+
"eval_samples_per_second": 609.559,
|
| 2143 |
+
"eval_steps_per_second": 7.619,
|
| 2144 |
+
"step": 6954
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.3005563462284901,
|
| 2148 |
+
"grad_norm": 0.345703125,
|
| 2149 |
+
"learning_rate": 0.0007001381453980314,
|
| 2150 |
+
"loss": 1.0634,
|
| 2151 |
+
"step": 6969
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.30154828136455775,
|
| 2155 |
+
"grad_norm": 0.359375,
|
| 2156 |
+
"learning_rate": 0.0006991452253496805,
|
| 2157 |
+
"loss": 1.071,
|
| 2158 |
+
"step": 6992
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.30254021650062535,
|
| 2162 |
+
"grad_norm": 0.3828125,
|
| 2163 |
+
"learning_rate": 0.0006981523053013296,
|
| 2164 |
+
"loss": 1.0822,
|
| 2165 |
+
"step": 7015
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.30353215163669295,
|
| 2169 |
+
"grad_norm": 0.4140625,
|
| 2170 |
+
"learning_rate": 0.0006971593852529788,
|
| 2171 |
+
"loss": 1.0618,
|
| 2172 |
+
"step": 7038
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.3045240867727606,
|
| 2176 |
+
"grad_norm": 0.322265625,
|
| 2177 |
+
"learning_rate": 0.0006961664652046279,
|
| 2178 |
+
"loss": 1.0722,
|
| 2179 |
+
"step": 7061
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.3055160219088282,
|
| 2183 |
+
"grad_norm": 0.3359375,
|
| 2184 |
+
"learning_rate": 0.000695173545156277,
|
| 2185 |
+
"loss": 1.0651,
|
| 2186 |
+
"step": 7084
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.30650795704489586,
|
| 2190 |
+
"grad_norm": 0.34765625,
|
| 2191 |
+
"learning_rate": 0.0006941806251079261,
|
| 2192 |
+
"loss": 1.0648,
|
| 2193 |
+
"step": 7107
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.30749989218096346,
|
| 2197 |
+
"grad_norm": 0.359375,
|
| 2198 |
+
"learning_rate": 0.0006931877050595752,
|
| 2199 |
+
"loss": 1.0615,
|
| 2200 |
+
"step": 7130
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.3084918273170311,
|
| 2204 |
+
"grad_norm": 0.349609375,
|
| 2205 |
+
"learning_rate": 0.0006921947850112244,
|
| 2206 |
+
"loss": 1.0673,
|
| 2207 |
+
"step": 7153
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.3094837624530987,
|
| 2211 |
+
"grad_norm": 0.38671875,
|
| 2212 |
+
"learning_rate": 0.0006912018649628734,
|
| 2213 |
+
"loss": 1.0596,
|
| 2214 |
+
"step": 7176
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.3104756975891663,
|
| 2218 |
+
"grad_norm": 0.337890625,
|
| 2219 |
+
"learning_rate": 0.0006902089449145226,
|
| 2220 |
+
"loss": 1.0665,
|
| 2221 |
+
"step": 7199
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.311467632725234,
|
| 2225 |
+
"grad_norm": 0.388671875,
|
| 2226 |
+
"learning_rate": 0.0006892160248661716,
|
| 2227 |
+
"loss": 1.0784,
|
| 2228 |
+
"step": 7222
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.3124595678613016,
|
| 2232 |
+
"grad_norm": 0.33203125,
|
| 2233 |
+
"learning_rate": 0.0006882231048178208,
|
| 2234 |
+
"loss": 1.0737,
|
| 2235 |
+
"step": 7245
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.31345150299736924,
|
| 2239 |
+
"grad_norm": 0.326171875,
|
| 2240 |
+
"learning_rate": 0.0006872301847694698,
|
| 2241 |
+
"loss": 1.0647,
|
| 2242 |
+
"step": 7268
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.31444343813343684,
|
| 2246 |
+
"grad_norm": 0.33984375,
|
| 2247 |
+
"learning_rate": 0.0006862372647211191,
|
| 2248 |
+
"loss": 1.0859,
|
| 2249 |
+
"step": 7291
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.31543537326950444,
|
| 2253 |
+
"grad_norm": 0.40625,
|
| 2254 |
+
"learning_rate": 0.0006852443446727681,
|
| 2255 |
+
"loss": 1.0703,
|
| 2256 |
+
"step": 7314
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.3164273084055721,
|
| 2260 |
+
"grad_norm": 0.37890625,
|
| 2261 |
+
"learning_rate": 0.0006842514246244172,
|
| 2262 |
+
"loss": 1.0581,
|
| 2263 |
+
"step": 7337
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.3174192435416397,
|
| 2267 |
+
"grad_norm": 0.408203125,
|
| 2268 |
+
"learning_rate": 0.0006832585045760663,
|
| 2269 |
+
"loss": 1.0557,
|
| 2270 |
+
"step": 7360
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.31841117867770735,
|
| 2274 |
+
"grad_norm": 0.3203125,
|
| 2275 |
+
"learning_rate": 0.0006822655845277154,
|
| 2276 |
+
"loss": 1.0853,
|
| 2277 |
+
"step": 7383
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.31940311381377495,
|
| 2281 |
+
"grad_norm": 0.4296875,
|
| 2282 |
+
"learning_rate": 0.0006812726644793646,
|
| 2283 |
+
"loss": 1.0602,
|
| 2284 |
+
"step": 7406
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.3203950489498426,
|
| 2288 |
+
"grad_norm": 0.41015625,
|
| 2289 |
+
"learning_rate": 0.0006802797444310137,
|
| 2290 |
+
"loss": 1.0611,
|
| 2291 |
+
"step": 7429
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.3213869840859102,
|
| 2295 |
+
"grad_norm": 0.3203125,
|
| 2296 |
+
"learning_rate": 0.0006792868243826628,
|
| 2297 |
+
"loss": 1.0703,
|
| 2298 |
+
"step": 7452
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.3223789192219778,
|
| 2302 |
+
"grad_norm": 0.44140625,
|
| 2303 |
+
"learning_rate": 0.0006782939043343119,
|
| 2304 |
+
"loss": 1.0357,
|
| 2305 |
+
"step": 7475
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.32337085435804547,
|
| 2309 |
+
"grad_norm": 0.345703125,
|
| 2310 |
+
"learning_rate": 0.0006773009842859609,
|
| 2311 |
+
"loss": 1.0556,
|
| 2312 |
+
"step": 7498
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.32436278949411307,
|
| 2316 |
+
"grad_norm": 0.37109375,
|
| 2317 |
+
"learning_rate": 0.00067630806423761,
|
| 2318 |
+
"loss": 1.0561,
|
| 2319 |
+
"step": 7521
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.3253547246301807,
|
| 2323 |
+
"grad_norm": 0.341796875,
|
| 2324 |
+
"learning_rate": 0.0006753151441892592,
|
| 2325 |
+
"loss": 1.0693,
|
| 2326 |
+
"step": 7544
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.3263466597662483,
|
| 2330 |
+
"grad_norm": 0.48046875,
|
| 2331 |
+
"learning_rate": 0.0006743222241409083,
|
| 2332 |
+
"loss": 1.0668,
|
| 2333 |
+
"step": 7567
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.3273385949023159,
|
| 2337 |
+
"grad_norm": 0.33984375,
|
| 2338 |
+
"learning_rate": 0.0006733293040925574,
|
| 2339 |
+
"loss": 1.038,
|
| 2340 |
+
"step": 7590
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.3283305300383836,
|
| 2344 |
+
"grad_norm": 0.34765625,
|
| 2345 |
+
"learning_rate": 0.0006723363840442065,
|
| 2346 |
+
"loss": 1.056,
|
| 2347 |
+
"step": 7613
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.3293224651744512,
|
| 2351 |
+
"grad_norm": 0.376953125,
|
| 2352 |
+
"learning_rate": 0.0006713434639958557,
|
| 2353 |
+
"loss": 1.0444,
|
| 2354 |
+
"step": 7636
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.33031440031051884,
|
| 2358 |
+
"grad_norm": 0.34375,
|
| 2359 |
+
"learning_rate": 0.0006703505439475048,
|
| 2360 |
+
"loss": 1.0534,
|
| 2361 |
+
"step": 7659
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.33130633544658644,
|
| 2365 |
+
"grad_norm": 0.341796875,
|
| 2366 |
+
"learning_rate": 0.0006693576238991539,
|
| 2367 |
+
"loss": 1.0597,
|
| 2368 |
+
"step": 7682
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.3322982705826541,
|
| 2372 |
+
"grad_norm": 0.41796875,
|
| 2373 |
+
"learning_rate": 0.0006683647038508029,
|
| 2374 |
+
"loss": 1.0305,
|
| 2375 |
+
"step": 7705
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.3332902057187217,
|
| 2379 |
+
"grad_norm": 0.3515625,
|
| 2380 |
+
"learning_rate": 0.0006673717838024521,
|
| 2381 |
+
"loss": 1.0632,
|
| 2382 |
+
"step": 7728
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.3342821408547893,
|
| 2386 |
+
"grad_norm": 0.341796875,
|
| 2387 |
+
"learning_rate": 0.0006663788637541011,
|
| 2388 |
+
"loss": 1.0462,
|
| 2389 |
+
"step": 7751
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.33527407599085696,
|
| 2393 |
+
"grad_norm": 0.423828125,
|
| 2394 |
+
"learning_rate": 0.0006653859437057504,
|
| 2395 |
+
"loss": 1.0534,
|
| 2396 |
+
"step": 7774
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.33626601112692456,
|
| 2400 |
+
"grad_norm": 0.37109375,
|
| 2401 |
+
"learning_rate": 0.0006643930236573994,
|
| 2402 |
+
"loss": 1.0468,
|
| 2403 |
+
"step": 7797
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.3372579462629922,
|
| 2407 |
+
"grad_norm": 0.37890625,
|
| 2408 |
+
"learning_rate": 0.0006634001036090486,
|
| 2409 |
+
"loss": 1.0473,
|
| 2410 |
+
"step": 7820
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 0.3382498813990598,
|
| 2414 |
+
"grad_norm": 0.349609375,
|
| 2415 |
+
"learning_rate": 0.0006624071835606976,
|
| 2416 |
+
"loss": 1.0402,
|
| 2417 |
+
"step": 7843
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 0.3392418165351274,
|
| 2421 |
+
"grad_norm": 0.384765625,
|
| 2422 |
+
"learning_rate": 0.0006614142635123467,
|
| 2423 |
+
"loss": 1.0713,
|
| 2424 |
+
"step": 7866
|
| 2425 |
+
},
|
| 2426 |
+
{
|
| 2427 |
+
"epoch": 0.3402337516711951,
|
| 2428 |
+
"grad_norm": 0.388671875,
|
| 2429 |
+
"learning_rate": 0.0006604213434639959,
|
| 2430 |
+
"loss": 1.0549,
|
| 2431 |
+
"step": 7889
|
| 2432 |
+
},
|
| 2433 |
+
{
|
| 2434 |
+
"epoch": 0.3412256868072627,
|
| 2435 |
+
"grad_norm": 0.3359375,
|
| 2436 |
+
"learning_rate": 0.000659428423415645,
|
| 2437 |
+
"loss": 1.0527,
|
| 2438 |
+
"step": 7912
|
| 2439 |
+
},
|
| 2440 |
+
{
|
| 2441 |
+
"epoch": 0.34221762194333033,
|
| 2442 |
+
"grad_norm": 0.421875,
|
| 2443 |
+
"learning_rate": 0.0006584355033672941,
|
| 2444 |
+
"loss": 1.0599,
|
| 2445 |
+
"step": 7935
|
| 2446 |
+
},
|
| 2447 |
+
{
|
| 2448 |
+
"epoch": 0.34320955707939793,
|
| 2449 |
+
"grad_norm": 0.322265625,
|
| 2450 |
+
"learning_rate": 0.0006574425833189432,
|
| 2451 |
+
"loss": 1.0503,
|
| 2452 |
+
"step": 7958
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 0.3442014922154656,
|
| 2456 |
+
"grad_norm": 0.388671875,
|
| 2457 |
+
"learning_rate": 0.0006564496632705923,
|
| 2458 |
+
"loss": 1.0465,
|
| 2459 |
+
"step": 7981
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 0.3451934273515332,
|
| 2463 |
+
"grad_norm": 0.314453125,
|
| 2464 |
+
"learning_rate": 0.0006554567432222415,
|
| 2465 |
+
"loss": 1.0246,
|
| 2466 |
+
"step": 8004
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 0.3461853624876008,
|
| 2470 |
+
"grad_norm": 0.36328125,
|
| 2471 |
+
"learning_rate": 0.0006544638231738905,
|
| 2472 |
+
"loss": 1.0463,
|
| 2473 |
+
"step": 8027
|
| 2474 |
+
},
|
| 2475 |
+
{
|
| 2476 |
+
"epoch": 0.34717729762366845,
|
| 2477 |
+
"grad_norm": 0.42578125,
|
| 2478 |
+
"learning_rate": 0.0006534709031255397,
|
| 2479 |
+
"loss": 1.0471,
|
| 2480 |
+
"step": 8050
|
| 2481 |
+
},
|
| 2482 |
+
{
|
| 2483 |
+
"epoch": 0.34816923275973605,
|
| 2484 |
+
"grad_norm": 0.470703125,
|
| 2485 |
+
"learning_rate": 0.0006524779830771887,
|
| 2486 |
+
"loss": 1.0414,
|
| 2487 |
+
"step": 8073
|
| 2488 |
+
},
|
| 2489 |
+
{
|
| 2490 |
+
"epoch": 0.3491611678958037,
|
| 2491 |
+
"grad_norm": 0.515625,
|
| 2492 |
+
"learning_rate": 0.0006514850630288378,
|
| 2493 |
+
"loss": 1.0431,
|
| 2494 |
+
"step": 8096
|
| 2495 |
+
},
|
| 2496 |
+
{
|
| 2497 |
+
"epoch": 0.3501531030318713,
|
| 2498 |
+
"grad_norm": 0.37109375,
|
| 2499 |
+
"learning_rate": 0.000650492142980487,
|
| 2500 |
+
"loss": 1.0382,
|
| 2501 |
+
"step": 8119
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 0.3511450381679389,
|
| 2505 |
+
"grad_norm": 0.39453125,
|
| 2506 |
+
"learning_rate": 0.0006494992229321361,
|
| 2507 |
+
"loss": 1.0557,
|
| 2508 |
+
"step": 8142
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 0.35213697330400656,
|
| 2512 |
+
"grad_norm": 0.376953125,
|
| 2513 |
+
"learning_rate": 0.0006485063028837852,
|
| 2514 |
+
"loss": 1.0414,
|
| 2515 |
+
"step": 8165
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.35312890844007416,
|
| 2519 |
+
"grad_norm": 0.365234375,
|
| 2520 |
+
"learning_rate": 0.0006475133828354343,
|
| 2521 |
+
"loss": 1.0221,
|
| 2522 |
+
"step": 8188
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 0.3541208435761418,
|
| 2526 |
+
"grad_norm": 0.384765625,
|
| 2527 |
+
"learning_rate": 0.0006465204627870834,
|
| 2528 |
+
"loss": 1.0207,
|
| 2529 |
+
"step": 8211
|
| 2530 |
+
},
|
| 2531 |
+
{
|
| 2532 |
+
"epoch": 0.3551127787122094,
|
| 2533 |
+
"grad_norm": 0.486328125,
|
| 2534 |
+
"learning_rate": 0.0006455275427387324,
|
| 2535 |
+
"loss": 1.0399,
|
| 2536 |
+
"step": 8234
|
| 2537 |
+
},
|
| 2538 |
+
{
|
| 2539 |
+
"epoch": 0.3561047138482771,
|
| 2540 |
+
"grad_norm": 0.43359375,
|
| 2541 |
+
"learning_rate": 0.0006445346226903817,
|
| 2542 |
+
"loss": 1.0256,
|
| 2543 |
+
"step": 8257
|
| 2544 |
+
},
|
| 2545 |
+
{
|
| 2546 |
+
"epoch": 0.3570966489843447,
|
| 2547 |
+
"grad_norm": 0.30859375,
|
| 2548 |
+
"learning_rate": 0.0006435417026420307,
|
| 2549 |
+
"loss": 1.0452,
|
| 2550 |
+
"step": 8280
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 0.3580885841204123,
|
| 2554 |
+
"grad_norm": 0.412109375,
|
| 2555 |
+
"learning_rate": 0.0006425487825936799,
|
| 2556 |
+
"loss": 1.0388,
|
| 2557 |
+
"step": 8303
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.35908051925647994,
|
| 2561 |
+
"grad_norm": 0.3828125,
|
| 2562 |
+
"learning_rate": 0.0006415558625453289,
|
| 2563 |
+
"loss": 1.0401,
|
| 2564 |
+
"step": 8326
|
| 2565 |
+
},
|
| 2566 |
+
{
|
| 2567 |
+
"epoch": 0.36007245439254754,
|
| 2568 |
+
"grad_norm": 0.474609375,
|
| 2569 |
+
"learning_rate": 0.0006405629424969782,
|
| 2570 |
+
"loss": 1.0345,
|
| 2571 |
+
"step": 8349
|
| 2572 |
+
},
|
| 2573 |
+
{
|
| 2574 |
+
"epoch": 0.3610643895286152,
|
| 2575 |
+
"grad_norm": 0.375,
|
| 2576 |
+
"learning_rate": 0.0006395700224486272,
|
| 2577 |
+
"loss": 1.0362,
|
| 2578 |
+
"step": 8372
|
| 2579 |
+
},
|
| 2580 |
+
{
|
| 2581 |
+
"epoch": 0.3620563246646828,
|
| 2582 |
+
"grad_norm": 0.3984375,
|
| 2583 |
+
"learning_rate": 0.0006385771024002763,
|
| 2584 |
+
"loss": 1.045,
|
| 2585 |
+
"step": 8395
|
| 2586 |
+
},
|
| 2587 |
+
{
|
| 2588 |
+
"epoch": 0.3630482598007504,
|
| 2589 |
+
"grad_norm": 0.357421875,
|
| 2590 |
+
"learning_rate": 0.0006375841823519254,
|
| 2591 |
+
"loss": 1.0417,
|
| 2592 |
+
"step": 8418
|
| 2593 |
+
},
|
| 2594 |
+
{
|
| 2595 |
+
"epoch": 0.36404019493681805,
|
| 2596 |
+
"grad_norm": 0.306640625,
|
| 2597 |
+
"learning_rate": 0.0006365912623035745,
|
| 2598 |
+
"loss": 1.0388,
|
| 2599 |
+
"step": 8441
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.36503213007288565,
|
| 2603 |
+
"grad_norm": 0.34765625,
|
| 2604 |
+
"learning_rate": 0.0006355983422552236,
|
| 2605 |
+
"loss": 1.0303,
|
| 2606 |
+
"step": 8464
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 0.3660240652089533,
|
| 2610 |
+
"grad_norm": 0.3359375,
|
| 2611 |
+
"learning_rate": 0.0006346054222068728,
|
| 2612 |
+
"loss": 1.017,
|
| 2613 |
+
"step": 8487
|
| 2614 |
+
},
|
| 2615 |
+
{
|
| 2616 |
+
"epoch": 0.3670160003450209,
|
| 2617 |
+
"grad_norm": 0.384765625,
|
| 2618 |
+
"learning_rate": 0.0006336125021585219,
|
| 2619 |
+
"loss": 1.0274,
|
| 2620 |
+
"step": 8510
|
| 2621 |
+
},
|
| 2622 |
+
{
|
| 2623 |
+
"epoch": 0.36800793548108857,
|
| 2624 |
+
"grad_norm": 0.357421875,
|
| 2625 |
+
"learning_rate": 0.000632619582110171,
|
| 2626 |
+
"loss": 1.0257,
|
| 2627 |
+
"step": 8533
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 0.36899987061715617,
|
| 2631 |
+
"grad_norm": 0.33203125,
|
| 2632 |
+
"learning_rate": 0.0006316266620618201,
|
| 2633 |
+
"loss": 1.0389,
|
| 2634 |
+
"step": 8556
|
| 2635 |
+
},
|
| 2636 |
+
{
|
| 2637 |
+
"epoch": 0.36999180575322377,
|
| 2638 |
+
"grad_norm": 0.369140625,
|
| 2639 |
+
"learning_rate": 0.0006306337420134692,
|
| 2640 |
+
"loss": 1.0295,
|
| 2641 |
+
"step": 8579
|
| 2642 |
+
},
|
| 2643 |
+
{
|
| 2644 |
+
"epoch": 0.3709837408892914,
|
| 2645 |
+
"grad_norm": 0.4140625,
|
| 2646 |
+
"learning_rate": 0.0006296408219651183,
|
| 2647 |
+
"loss": 1.045,
|
| 2648 |
+
"step": 8602
|
| 2649 |
+
},
|
| 2650 |
+
{
|
| 2651 |
+
"epoch": 0.371975676025359,
|
| 2652 |
+
"grad_norm": 0.392578125,
|
| 2653 |
+
"learning_rate": 0.0006286479019167675,
|
| 2654 |
+
"loss": 1.0581,
|
| 2655 |
+
"step": 8625
|
| 2656 |
+
},
|
| 2657 |
+
{
|
| 2658 |
+
"epoch": 0.3729676111614267,
|
| 2659 |
+
"grad_norm": 0.44140625,
|
| 2660 |
+
"learning_rate": 0.0006276549818684165,
|
| 2661 |
+
"loss": 1.0521,
|
| 2662 |
+
"step": 8648
|
| 2663 |
+
},
|
| 2664 |
+
{
|
| 2665 |
+
"epoch": 0.3739595462974943,
|
| 2666 |
+
"grad_norm": 0.36328125,
|
| 2667 |
+
"learning_rate": 0.0006266620618200657,
|
| 2668 |
+
"loss": 1.0355,
|
| 2669 |
+
"step": 8671
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.3749514814335619,
|
| 2673 |
+
"grad_norm": 0.349609375,
|
| 2674 |
+
"learning_rate": 0.0006256691417717147,
|
| 2675 |
+
"loss": 1.0366,
|
| 2676 |
+
"step": 8694
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.37594341656962954,
|
| 2680 |
+
"grad_norm": 0.349609375,
|
| 2681 |
+
"learning_rate": 0.0006246762217233639,
|
| 2682 |
+
"loss": 1.0148,
|
| 2683 |
+
"step": 8717
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.37693535170569714,
|
| 2687 |
+
"grad_norm": 0.421875,
|
| 2688 |
+
"learning_rate": 0.000623683301675013,
|
| 2689 |
+
"loss": 1.0401,
|
| 2690 |
+
"step": 8740
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.3779272868417648,
|
| 2694 |
+
"grad_norm": 0.3203125,
|
| 2695 |
+
"learning_rate": 0.000622690381626662,
|
| 2696 |
+
"loss": 1.0346,
|
| 2697 |
+
"step": 8763
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.3789192219778324,
|
| 2701 |
+
"grad_norm": 0.34375,
|
| 2702 |
+
"learning_rate": 0.0006216974615783112,
|
| 2703 |
+
"loss": 1.0218,
|
| 2704 |
+
"step": 8786
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.37991115711390006,
|
| 2708 |
+
"grad_norm": 0.427734375,
|
| 2709 |
+
"learning_rate": 0.0006207045415299602,
|
| 2710 |
+
"loss": 1.0405,
|
| 2711 |
+
"step": 8809
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.38090309224996766,
|
| 2715 |
+
"grad_norm": 0.392578125,
|
| 2716 |
+
"learning_rate": 0.0006197116214816095,
|
| 2717 |
+
"loss": 1.0253,
|
| 2718 |
+
"step": 8832
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.38189502738603526,
|
| 2722 |
+
"grad_norm": 0.345703125,
|
| 2723 |
+
"learning_rate": 0.0006187187014332585,
|
| 2724 |
+
"loss": 1.0409,
|
| 2725 |
+
"step": 8855
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.3828869625221029,
|
| 2729 |
+
"grad_norm": 0.349609375,
|
| 2730 |
+
"learning_rate": 0.0006177257813849077,
|
| 2731 |
+
"loss": 1.0349,
|
| 2732 |
+
"step": 8878
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.3838788976581705,
|
| 2736 |
+
"grad_norm": 0.49609375,
|
| 2737 |
+
"learning_rate": 0.0006167328613365567,
|
| 2738 |
+
"loss": 1.029,
|
| 2739 |
+
"step": 8901
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.38487083279423817,
|
| 2743 |
+
"grad_norm": 0.447265625,
|
| 2744 |
+
"learning_rate": 0.0006157399412882058,
|
| 2745 |
+
"loss": 1.0338,
|
| 2746 |
+
"step": 8924
|
| 2747 |
+
},
|
| 2748 |
+
{
|
| 2749 |
+
"epoch": 0.3858627679303058,
|
| 2750 |
+
"grad_norm": 0.416015625,
|
| 2751 |
+
"learning_rate": 0.0006147470212398549,
|
| 2752 |
+
"loss": 1.0335,
|
| 2753 |
+
"step": 8947
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 0.3868547030663734,
|
| 2757 |
+
"grad_norm": 0.40234375,
|
| 2758 |
+
"learning_rate": 0.0006137541011915041,
|
| 2759 |
+
"loss": 1.0512,
|
| 2760 |
+
"step": 8970
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.38784663820244103,
|
| 2764 |
+
"grad_norm": 0.388671875,
|
| 2765 |
+
"learning_rate": 0.0006127611811431532,
|
| 2766 |
+
"loss": 1.0187,
|
| 2767 |
+
"step": 8993
|
| 2768 |
+
},
|
| 2769 |
+
{
|
| 2770 |
+
"epoch": 0.38883857333850863,
|
| 2771 |
+
"grad_norm": 0.298828125,
|
| 2772 |
+
"learning_rate": 0.0006117682610948023,
|
| 2773 |
+
"loss": 1.0321,
|
| 2774 |
+
"step": 9016
|
| 2775 |
+
},
|
| 2776 |
+
{
|
| 2777 |
+
"epoch": 0.3898305084745763,
|
| 2778 |
+
"grad_norm": 0.353515625,
|
| 2779 |
+
"learning_rate": 0.0006107753410464514,
|
| 2780 |
+
"loss": 1.0314,
|
| 2781 |
+
"step": 9039
|
| 2782 |
+
},
|
| 2783 |
+
{
|
| 2784 |
+
"epoch": 0.3908224436106439,
|
| 2785 |
+
"grad_norm": 0.3984375,
|
| 2786 |
+
"learning_rate": 0.0006097824209981005,
|
| 2787 |
+
"loss": 1.011,
|
| 2788 |
+
"step": 9062
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 0.39181437874671154,
|
| 2792 |
+
"grad_norm": 0.37890625,
|
| 2793 |
+
"learning_rate": 0.0006087895009497497,
|
| 2794 |
+
"loss": 1.0127,
|
| 2795 |
+
"step": 9085
|
| 2796 |
+
},
|
| 2797 |
+
{
|
| 2798 |
+
"epoch": 0.39280631388277915,
|
| 2799 |
+
"grad_norm": 0.40625,
|
| 2800 |
+
"learning_rate": 0.0006077965809013988,
|
| 2801 |
+
"loss": 1.0045,
|
| 2802 |
+
"step": 9108
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.39379824901884675,
|
| 2806 |
+
"grad_norm": 0.349609375,
|
| 2807 |
+
"learning_rate": 0.0006068036608530478,
|
| 2808 |
+
"loss": 0.9966,
|
| 2809 |
+
"step": 9131
|
| 2810 |
+
},
|
| 2811 |
+
{
|
| 2812 |
+
"epoch": 0.3947901841549144,
|
| 2813 |
+
"grad_norm": 0.318359375,
|
| 2814 |
+
"learning_rate": 0.000605810740804697,
|
| 2815 |
+
"loss": 1.0235,
|
| 2816 |
+
"step": 9154
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 0.395782119290982,
|
| 2820 |
+
"grad_norm": 0.427734375,
|
| 2821 |
+
"learning_rate": 0.000604817820756346,
|
| 2822 |
+
"loss": 1.0448,
|
| 2823 |
+
"step": 9177
|
| 2824 |
+
},
|
| 2825 |
+
{
|
| 2826 |
+
"epoch": 0.39677405442704966,
|
| 2827 |
+
"grad_norm": 0.337890625,
|
| 2828 |
+
"learning_rate": 0.0006038249007079953,
|
| 2829 |
+
"loss": 1.0197,
|
| 2830 |
+
"step": 9200
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 0.39776598956311726,
|
| 2834 |
+
"grad_norm": 0.447265625,
|
| 2835 |
+
"learning_rate": 0.0006028319806596443,
|
| 2836 |
+
"loss": 1.0093,
|
| 2837 |
+
"step": 9223
|
| 2838 |
+
},
|
| 2839 |
+
{
|
| 2840 |
+
"epoch": 0.39875792469918486,
|
| 2841 |
+
"grad_norm": 0.412109375,
|
| 2842 |
+
"learning_rate": 0.0006018390606112935,
|
| 2843 |
+
"loss": 1.014,
|
| 2844 |
+
"step": 9246
|
| 2845 |
+
},
|
| 2846 |
+
{
|
| 2847 |
+
"epoch": 0.3997498598352525,
|
| 2848 |
+
"grad_norm": 0.462890625,
|
| 2849 |
+
"learning_rate": 0.0006008461405629425,
|
| 2850 |
+
"loss": 1.0269,
|
| 2851 |
+
"step": 9269
|
| 2852 |
+
},
|
| 2853 |
+
{
|
| 2854 |
+
"epoch": 0.3998792426790874,
|
| 2855 |
+
"eval_runtime": 163.9225,
|
| 2856 |
+
"eval_samples_per_second": 610.045,
|
| 2857 |
+
"eval_steps_per_second": 7.626,
|
| 2858 |
+
"step": 9272
|
| 2859 |
+
}
|
| 2860 |
+
],
|
| 2861 |
+
"logging_steps": 23,
|
| 2862 |
+
"max_steps": 23187,
|
| 2863 |
+
"num_input_tokens_seen": 0,
|
| 2864 |
+
"num_train_epochs": 9223372036854775807,
|
| 2865 |
+
"save_steps": 2318,
|
| 2866 |
+
"stateful_callbacks": {
|
| 2867 |
+
"TrainerControl": {
|
| 2868 |
+
"args": {
|
| 2869 |
+
"should_epoch_stop": false,
|
| 2870 |
+
"should_evaluate": false,
|
| 2871 |
+
"should_log": false,
|
| 2872 |
+
"should_save": true,
|
| 2873 |
+
"should_training_stop": false
|
| 2874 |
+
},
|
| 2875 |
+
"attributes": {}
|
| 2876 |
+
}
|
| 2877 |
+
},
|
| 2878 |
+
"total_flos": 6.778106242599485e+17,
|
| 2879 |
+
"train_batch_size": 8,
|
| 2880 |
+
"trial_name": null,
|
| 2881 |
+
"trial_params": null
|
| 2882 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1ce2265cfdc88c58cc9e3f8a83920eb0394e14800c5a5bf7370808a2299aee92
|
| 3 |
+
size 5496
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|