Hugging Face's logo

HueyNemud
/
das22-10-camembert_pretrained

Fill-Mask
Transformers
PyTorch
camembert
generated_from_trainer
Inference Endpoints
Model card Files Community
1
HueyNemud commited on
Commit
9e621d7
•
1 Parent(s): 756e3cb

News version trained on 845k raw entries

Browse files
Files changed (9) hide show
  1. README.md +10 -20
  2. all_results.json +11 -11
  3. config.json +1 -1
  4. eval_results.json +6 -6
  5. pytorch_model.bin +1 -1
  6. tokenizer_config.json +1 -1
  7. train_results.json +5 -5
  8. trainer_state.json +3590 -136
  9. training_args.bin +2 -2
README.md CHANGED
@@ -2,18 +2,18 @@
2
  tags:
3
  - generated_from_trainer
4
  model-index:
5
- - name: berties-pretrained
6
  results: []
7
  ---
8
 
9
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
10
  should probably proofread and complete it, then remove this comment. -->
11
 
12
- # berties-pretrained
13
 
14
  This model is a fine-tuned version of [Jean-Baptiste/camembert-ner](https://huggingface.co/Jean-Baptiste/camembert-ner) on an unknown dataset.
15
  It achieves the following results on the evaluation set:
16
- - Loss: 3.3922
17
 
18
  ## Model description
19
 
@@ -42,26 +42,16 @@ The following hyperparameters were used during training:
42
 
43
  ### Training results
44
 
45
- | Training Loss | Epoch | Step | Validation Loss |
46
- |:-------------:|:-----:|:----:|:---------------:|
47
- | 5.7047 | 0.21 | 500 | 5.0371 |
48
- | 4.4116 | 0.43 | 1000 | 4.5143 |
49
- | 4.0711 | 0.64 | 1500 | 4.3272 |
50
- | 3.8657 | 0.86 | 2000 | 4.1704 |
51
- | 3.6971 | 1.07 | 2500 | 3.9951 |
52
- | 3.5093 | 1.29 | 3000 | 3.7682 |
53
- | 3.3598 | 1.5 | 3500 | 3.6554 |
54
- | 3.2389 | 1.72 | 4000 | 3.6167 |
55
- | 3.1984 | 1.93 | 4500 | 3.4445 |
56
- | 3.1063 | 2.15 | 5000 | 3.4206 |
57
- | 3.0063 | 2.36 | 5500 | 3.4164 |
58
- | 3.0069 | 2.57 | 6000 | 3.4115 |
59
- | 2.9954 | 2.79 | 6500 | 3.3463 |
60
 
61
 
62
  ### Framework versions
63
 
64
- - Transformers 4.15.0
65
- - Pytorch 1.10.0+cu111
66
  - Datasets 1.17.0
67
  - Tokenizers 0.10.3
 
2
  tags:
3
  - generated_from_trainer
4
  model-index:
5
+ - name: model_pretrained
6
  results: []
7
  ---
8
 
9
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
10
  should probably proofread and complete it, then remove this comment. -->
11
 
12
+ # model_pretrained
13
 
14
  This model is a fine-tuned version of [Jean-Baptiste/camembert-ner](https://huggingface.co/Jean-Baptiste/camembert-ner) on an unknown dataset.
15
  It achieves the following results on the evaluation set:
16
+ - Loss: 1.5619
17
 
18
  ## Model description
19
 
 
42
 
43
  ### Training results
44
 
45
+ | Training Loss | Epoch | Step | Validation Loss |
46
+ |:-------------:|:-----:|:------:|:---------------:|
47
+ | 1.9603 | 1.0 | 100346 | 1.8005 |
48
+ | 1.7032 | 2.0 | 200692 | 1.6460 |
49
+ | 1.5879 | 3.0 | 301038 | 1.5570 |
 
 
 
 
 
 
 
 
 
 
50
 
51
 
52
  ### Framework versions
53
 
54
+ - Transformers 4.16.0.dev0
55
+ - Pytorch 1.10.1+cu102
56
  - Datasets 1.17.0
57
  - Tokenizers 0.10.3
all_results.json CHANGED
@@ -1,14 +1,14 @@
1
  {
2
  "epoch": 3.0,
3
- "eval_loss": 3.392239809036255,
4
- "eval_runtime": 8.4705,
5
- "eval_samples": 981,
6
- "eval_samples_per_second": 115.814,
7
- "eval_steps_per_second": 14.521,
8
- "perplexity": 29.732472807173135,
9
- "train_loss": 3.5777804869601746,
10
- "train_runtime": 2142.1553,
11
- "train_samples": 18647,
12
- "train_samples_per_second": 26.114,
13
- "train_steps_per_second": 3.264
14
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "eval_loss": 1.5619168281555176,
4
+ "eval_runtime": 63.0861,
5
+ "eval_samples": 42251,
6
+ "eval_samples_per_second": 669.735,
7
+ "eval_steps_per_second": 83.727,
8
+ "perplexity": 4.76795183577856,
9
+ "train_loss": 1.8846093698359827,
10
+ "train_runtime": 20632.0794,
11
+ "train_samples": 802763,
12
+ "train_samples_per_second": 116.725,
13
+ "train_steps_per_second": 14.591
14
  }
config.json CHANGED
@@ -36,7 +36,7 @@
36
  "pad_token_id": 1,
37
  "position_embedding_type": "absolute",
38
  "torch_dtype": "float32",
39
- "transformers_version": "4.15.0",
40
  "type_vocab_size": 1,
41
  "use_cache": true,
42
  "vocab_size": 32005
 
36
  "pad_token_id": 1,
37
  "position_embedding_type": "absolute",
38
  "torch_dtype": "float32",
39
+ "transformers_version": "4.16.0.dev0",
40
  "type_vocab_size": 1,
41
  "use_cache": true,
42
  "vocab_size": 32005
eval_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "epoch": 3.0,
3
- "eval_loss": 3.392239809036255,
4
- "eval_runtime": 8.4705,
5
- "eval_samples": 981,
6
- "eval_samples_per_second": 115.814,
7
- "eval_steps_per_second": 14.521,
8
- "perplexity": 29.732472807173135
9
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "eval_loss": 1.5619168281555176,
4
+ "eval_runtime": 63.0861,
5
+ "eval_samples": 42251,
6
+ "eval_samples_per_second": 669.735,
7
+ "eval_steps_per_second": 83.727,
8
+ "perplexity": 4.76795183577856
9
  }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3006eea8290de245cf656d4dfa4d2d8246ee705987ad6e513087e30a8167da22
3
  size 442707627
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1bb51a694376014a1a3bc6d41f41039249527833b088ca87ea248412242e135b
3
  size 442707627
tokenizer_config.json CHANGED
@@ -1 +1 @@
1
- {"bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "additional_special_tokens": ["<s>NOTUSED", "</s>NOTUSED"], "model_max_length": 512, "name_or_path": "Jean-Baptiste/camembert-ner", "special_tokens_map_file": "/root/.cache/huggingface/transformers/fe0e213c44079a9ee091098f81fff2941484006e9ba3001a9bf1ee9f87537599.cb3ec3a6c1200d181228d8825ae9767572abca54efa1bbb37fd83d721b2ef323", "sp_model_kwargs": {}, "tokenizer_class": "CamembertTokenizer"}
 
1
+ {"bos_token": "<s>", "eos_token": "</s>", "sep_token": "</s>", "cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "additional_special_tokens": ["<s>NOTUSED", "</s>NOTUSED"], "model_max_length": 512, "name_or_path": "Jean-Baptiste/camembert-ner", "special_tokens_map_file": "/lrde/home2/jchazalo/.cache/huggingface/transformers/fe0e213c44079a9ee091098f81fff2941484006e9ba3001a9bf1ee9f87537599.cb3ec3a6c1200d181228d8825ae9767572abca54efa1bbb37fd83d721b2ef323", "sp_model_kwargs": {}, "tokenizer_class": "CamembertTokenizer"}
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "epoch": 3.0,
3
- "train_loss": 3.5777804869601746,
4
- "train_runtime": 2142.1553,
5
- "train_samples": 18647,
6
- "train_samples_per_second": 26.114,
7
- "train_steps_per_second": 3.264
8
  }
 
1
  {
2
  "epoch": 3.0,
3
+ "train_loss": 1.8846093698359827,
4
+ "train_runtime": 20632.0794,
5
+ "train_samples": 802763,
6
+ "train_samples_per_second": 116.725,
7
+ "train_steps_per_second": 14.591
8
  }
trainer_state.json CHANGED
@@ -1,207 +1,3661 @@
1
  {
2
- "best_metric": 3.346250057220459,
3
- "best_model_checkpoint": "/tmp/berties-pretrained/checkpoint-6500",
4
  "epoch": 3.0,
5
- "global_step": 6993,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
- "epoch": 0.21,
12
- "learning_rate": 4.642499642499642e-05,
13
- "loss": 5.7047,
14
  "step": 500
15
  },
16
  {
17
- "epoch": 0.21,
18
- "eval_loss": 5.037093639373779,
19
- "eval_runtime": 8.5329,
20
- "eval_samples_per_second": 114.967,
21
- "eval_steps_per_second": 14.415,
22
- "step": 500
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
23
  },
24
  {
25
- "epoch": 0.43,
26
- "learning_rate": 4.2849992849992856e-05,
27
- "loss": 4.4116,
28
- "step": 1000
29
  },
30
  {
31
- "epoch": 0.43,
32
- "eval_loss": 4.514286041259766,
33
- "eval_runtime": 8.497,
34
- "eval_samples_per_second": 115.452,
35
- "eval_steps_per_second": 14.476,
36
- "step": 1000
37
  },
38
  {
39
- "epoch": 0.64,
40
- "learning_rate": 3.9274989274989276e-05,
41
- "loss": 4.0711,
42
- "step": 1500
43
  },
44
  {
45
- "epoch": 0.64,
46
- "eval_loss": 4.327243804931641,
47
- "eval_runtime": 8.5359,
48
- "eval_samples_per_second": 114.926,
49
- "eval_steps_per_second": 14.41,
50
- "step": 1500
51
  },
52
  {
53
- "epoch": 0.86,
54
- "learning_rate": 3.56999856999857e-05,
55
- "loss": 3.8657,
56
- "step": 2000
57
  },
58
  {
59
- "epoch": 0.86,
60
- "eval_loss": 4.17035436630249,
61
- "eval_runtime": 8.541,
62
- "eval_samples_per_second": 114.858,
63
- "eval_steps_per_second": 14.401,
64
- "step": 2000
65
  },
66
  {
67
- "epoch": 1.07,
68
- "learning_rate": 3.212498212498213e-05,
69
- "loss": 3.6971,
70
- "step": 2500
71
  },
72
  {
73
- "epoch": 1.07,
74
- "eval_loss": 3.9950718879699707,
75
- "eval_runtime": 8.4686,
76
- "eval_samples_per_second": 115.84,
77
- "eval_steps_per_second": 14.524,
78
- "step": 2500
79
  },
80
  {
81
- "epoch": 1.29,
82
- "learning_rate": 2.854997854997855e-05,
83
- "loss": 3.5093,
84
- "step": 3000
85
  },
86
  {
87
- "epoch": 1.29,
88
- "eval_loss": 3.768172264099121,
89
- "eval_runtime": 8.4968,
90
- "eval_samples_per_second": 115.455,
91
- "eval_steps_per_second": 14.476,
92
- "step": 3000
93
  },
94
  {
95
- "epoch": 1.5,
96
- "learning_rate": 2.4974974974974976e-05,
97
- "loss": 3.3598,
98
- "step": 3500
99
  },
100
  {
101
- "epoch": 1.5,
102
- "eval_loss": 3.655388832092285,
103
- "eval_runtime": 8.4823,
104
- "eval_samples_per_second": 115.652,
105
- "eval_steps_per_second": 14.501,
106
- "step": 3500
107
  },
108
  {
109
- "epoch": 1.72,
110
- "learning_rate": 2.13999713999714e-05,
111
- "loss": 3.2389,
112
- "step": 4000
113
  },
114
  {
115
- "epoch": 1.72,
116
- "eval_loss": 3.6167075634002686,
117
- "eval_runtime": 8.4937,
118
- "eval_samples_per_second": 115.497,
119
- "eval_steps_per_second": 14.481,
120
- "step": 4000
121
  },
122
  {
123
- "epoch": 1.93,
124
- "learning_rate": 1.7824967824967826e-05,
125
- "loss": 3.1984,
126
- "step": 4500
127
  },
128
  {
129
- "epoch": 1.93,
130
- "eval_loss": 3.4444828033447266,
131
- "eval_runtime": 8.5242,
132
- "eval_samples_per_second": 115.084,
133
- "eval_steps_per_second": 14.429,
134
- "step": 4500
135
  },
136
  {
137
- "epoch": 2.15,
138
- "learning_rate": 1.4249964249964251e-05,
139
- "loss": 3.1063,
140
- "step": 5000
141
  },
142
  {
143
- "epoch": 2.15,
144
- "eval_loss": 3.4205729961395264,
145
- "eval_runtime": 8.5145,
146
- "eval_samples_per_second": 115.215,
147
- "eval_steps_per_second": 14.446,
148
- "step": 5000
149
  },
150
  {
151
- "epoch": 2.36,
152
- "learning_rate": 1.0674960674960675e-05,
153
- "loss": 3.0063,
154
- "step": 5500
155
  },
156
  {
157
- "epoch": 2.36,
158
- "eval_loss": 3.416414976119995,
159
- "eval_runtime": 8.5224,
160
- "eval_samples_per_second": 115.108,
161
- "eval_steps_per_second": 14.432,
162
- "step": 5500
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
163
  },
164
  {
165
  "epoch": 2.57,
166
- "learning_rate": 7.0999570999571e-06,
167
- "loss": 3.0069,
168
- "step": 6000
169
  },
170
  {
171
  "epoch": 2.57,
172
- "eval_loss": 3.4115302562713623,
173
- "eval_runtime": 8.4853,
174
- "eval_samples_per_second": 115.611,
175
- "eval_steps_per_second": 14.496,
176
- "step": 6000
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
177
  },
178
  {
179
  "epoch": 2.79,
180
- "learning_rate": 3.5249535249535254e-06,
181
- "loss": 2.9954,
182
- "step": 6500
183
  },
184
  {
185
  "epoch": 2.79,
186
- "eval_loss": 3.346250057220459,
187
- "eval_runtime": 8.4716,
188
- "eval_samples_per_second": 115.799,
189
- "eval_steps_per_second": 14.519,
190
- "step": 6500
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
191
  },
192
  {
193
  "epoch": 3.0,
194
- "step": 6993,
195
- "total_flos": 1415406837275046.0,
196
- "train_loss": 3.5777804869601746,
197
- "train_runtime": 2142.1553,
198
- "train_samples_per_second": 26.114,
199
- "train_steps_per_second": 3.264
200
  }
201
  ],
202
- "max_steps": 6993,
203
  "num_train_epochs": 3,
204
- "total_flos": 1415406837275046.0,
205
  "trial_name": null,
206
  "trial_params": null
207
  }
 
1
  {
2
+ "best_metric": 1.5570189952850342,
3
+ "best_model_checkpoint": "/lrde/home2/jchazalo/soduco/article_das_2022/model_pretrained/checkpoint-301038",
4
  "epoch": 3.0,
5
+ "global_step": 301038,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.991695400580658e-05,
13
+ "loss": 5.5377,
14
  "step": 500
15
  },
16
  {
17
+ "epoch": 0.01,
18
+ "learning_rate": 4.983390801161315e-05,
19
+ "loss": 4.3409,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.975086201741973e-05,
25
+ "loss": 4.0691,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 4.9667816023226304e-05,
31
+ "loss": 3.8186,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 4.9584770029032886e-05,
37
+ "loss": 3.6468,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 4.950172403483946e-05,
43
+ "loss": 3.4779,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 4.941867804064603e-05,
49
+ "loss": 3.4288,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.04,
54
+ "learning_rate": 4.933563204645261e-05,
55
+ "loss": 3.2539,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.04,
60
+ "learning_rate": 4.925258605225919e-05,
61
+ "loss": 3.1983,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.05,
66
+ "learning_rate": 4.916954005806576e-05,
67
+ "loss": 3.0558,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.05,
72
+ "learning_rate": 4.908649406387234e-05,
73
+ "loss": 3.0145,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.06,
78
+ "learning_rate": 4.900344806967891e-05,
79
+ "loss": 2.9689,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.06,
84
+ "learning_rate": 4.892040207548549e-05,
85
+ "loss": 2.9319,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.07,
90
+ "learning_rate": 4.883735608129206e-05,
91
+ "loss": 2.9474,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.07,
96
+ "learning_rate": 4.8754310087098645e-05,
97
+ "loss": 2.8579,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.08,
102
+ "learning_rate": 4.8671264092905214e-05,
103
+ "loss": 2.8567,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.08,
108
+ "learning_rate": 4.858821809871179e-05,
109
+ "loss": 2.8157,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.09,
114
+ "learning_rate": 4.850517210451837e-05,
115
+ "loss": 2.7768,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.09,
120
+ "learning_rate": 4.8422126110324947e-05,
121
+ "loss": 2.7482,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.1,
126
+ "learning_rate": 4.833908011613152e-05,
127
+ "loss": 2.7077,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.1,
132
+ "learning_rate": 4.82560341219381e-05,
133
+ "loss": 2.6328,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.11,
138
+ "learning_rate": 4.817298812774467e-05,
139
+ "loss": 2.6814,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.11,
144
+ "learning_rate": 4.808994213355125e-05,
145
+ "loss": 2.612,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.12,
150
+ "learning_rate": 4.800689613935783e-05,
151
+ "loss": 2.533,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.12,
156
+ "learning_rate": 4.79238501451644e-05,
157
+ "loss": 2.6266,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.13,
162
+ "learning_rate": 4.7840804150970974e-05,
163
+ "loss": 2.5415,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.13,
168
+ "learning_rate": 4.7757758156777556e-05,
169
+ "loss": 2.6189,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.14,
174
+ "learning_rate": 4.767471216258413e-05,
175
+ "loss": 2.521,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.14,
180
+ "learning_rate": 4.75916661683907e-05,
181
+ "loss": 2.5118,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.15,
186
+ "learning_rate": 4.7508620174197275e-05,
187
+ "loss": 2.5452,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.15,
192
+ "learning_rate": 4.742557418000386e-05,
193
+ "loss": 2.4781,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.16,
198
+ "learning_rate": 4.734252818581043e-05,
199
+ "loss": 2.4724,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.16,
204
+ "learning_rate": 4.725948219161701e-05,
205
+ "loss": 2.4601,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.17,
210
+ "learning_rate": 4.717643619742358e-05,
211
+ "loss": 2.4624,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.17,
216
+ "learning_rate": 4.709339020323016e-05,
217
+ "loss": 2.4557,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.18,
222
+ "learning_rate": 4.701034420903673e-05,
223
+ "loss": 2.4614,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.18,
228
+ "learning_rate": 4.6927298214843315e-05,
229
+ "loss": 2.4857,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.19,
234
+ "learning_rate": 4.6844252220649884e-05,
235
+ "loss": 2.4286,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.19,
240
+ "learning_rate": 4.676120622645646e-05,
241
+ "loss": 2.4001,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.2,
246
+ "learning_rate": 4.667816023226304e-05,
247
+ "loss": 2.3912,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.2,
252
+ "learning_rate": 4.659511423806962e-05,
253
+ "loss": 2.4437,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.21,
258
+ "learning_rate": 4.651206824387619e-05,
259
+ "loss": 2.3163,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.21,
264
+ "learning_rate": 4.642902224968277e-05,
265
+ "loss": 2.3766,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.22,
270
+ "learning_rate": 4.634597625548934e-05,
271
+ "loss": 2.4041,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.22,
276
+ "learning_rate": 4.626293026129592e-05,
277
+ "loss": 2.3604,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.23,
282
+ "learning_rate": 4.61798842671025e-05,
283
+ "loss": 2.2922,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.23,
288
+ "learning_rate": 4.609683827290907e-05,
289
+ "loss": 2.3668,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.24,
294
+ "learning_rate": 4.6013792278715644e-05,
295
+ "loss": 2.2715,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.24,
300
+ "learning_rate": 4.5930746284522226e-05,
301
+ "loss": 2.3434,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.25,
306
+ "learning_rate": 4.58477002903288e-05,
307
+ "loss": 2.3345,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.25,
312
+ "learning_rate": 4.5764654296135376e-05,
313
+ "loss": 2.3421,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.26,
318
+ "learning_rate": 4.5681608301941945e-05,
319
+ "loss": 2.2897,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.26,
324
+ "learning_rate": 4.559856230774853e-05,
325
+ "loss": 2.3008,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.27,
330
+ "learning_rate": 4.55155163135551e-05,
331
+ "loss": 2.2646,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.27,
336
+ "learning_rate": 4.543247031936168e-05,
337
+ "loss": 2.3077,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.28,
342
+ "learning_rate": 4.534942432516825e-05,
343
+ "loss": 2.271,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.28,
348
+ "learning_rate": 4.526637833097483e-05,
349
+ "loss": 2.3123,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.29,
354
+ "learning_rate": 4.5183332336781403e-05,
355
+ "loss": 2.2825,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.29,
360
+ "learning_rate": 4.5100286342587985e-05,
361
+ "loss": 2.233,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.3,
366
+ "learning_rate": 4.5017240348394554e-05,
367
+ "loss": 2.2651,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.3,
372
+ "learning_rate": 4.493419435420113e-05,
373
+ "loss": 2.2706,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.31,
378
+ "learning_rate": 4.485114836000771e-05,
379
+ "loss": 2.2816,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.31,
384
+ "learning_rate": 4.476810236581429e-05,
385
+ "loss": 2.2128,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.32,
390
+ "learning_rate": 4.468505637162086e-05,
391
+ "loss": 2.2582,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.32,
396
+ "learning_rate": 4.460201037742744e-05,
397
+ "loss": 2.2227,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.33,
402
+ "learning_rate": 4.451896438323401e-05,
403
+ "loss": 2.2747,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.33,
408
+ "learning_rate": 4.443591838904059e-05,
409
+ "loss": 2.2617,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.34,
414
+ "learning_rate": 4.435287239484717e-05,
415
+ "loss": 2.2421,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.34,
420
+ "learning_rate": 4.426982640065374e-05,
421
+ "loss": 2.2419,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.35,
426
+ "learning_rate": 4.4186780406460314e-05,
427
+ "loss": 2.2061,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.35,
432
+ "learning_rate": 4.410373441226689e-05,
433
+ "loss": 2.2277,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.36,
438
+ "learning_rate": 4.402068841807347e-05,
439
+ "loss": 2.2249,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.36,
444
+ "learning_rate": 4.3937642423880046e-05,
445
+ "loss": 2.2255,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.37,
450
+ "learning_rate": 4.3854596429686615e-05,
451
+ "loss": 2.1634,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.37,
456
+ "learning_rate": 4.37715504354932e-05,
457
+ "loss": 2.2847,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.38,
462
+ "learning_rate": 4.368850444129977e-05,
463
+ "loss": 2.1552,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.38,
468
+ "learning_rate": 4.360545844710635e-05,
469
+ "loss": 2.0838,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.39,
474
+ "learning_rate": 4.352241245291292e-05,
475
+ "loss": 2.158,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.39,
480
+ "learning_rate": 4.34393664587195e-05,
481
+ "loss": 2.1425,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.4,
486
+ "learning_rate": 4.3356320464526073e-05,
487
+ "loss": 2.2127,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.4,
492
+ "learning_rate": 4.3273274470332656e-05,
493
+ "loss": 2.1698,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.41,
498
+ "learning_rate": 4.3190228476139224e-05,
499
+ "loss": 2.117,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.41,
504
+ "learning_rate": 4.31071824819458e-05,
505
+ "loss": 2.2138,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.42,
510
+ "learning_rate": 4.302413648775238e-05,
511
+ "loss": 2.156,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.42,
516
+ "learning_rate": 4.294109049355896e-05,
517
+ "loss": 2.1724,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.43,
522
+ "learning_rate": 4.285804449936553e-05,
523
+ "loss": 2.2016,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.43,
528
+ "learning_rate": 4.277499850517211e-05,
529
+ "loss": 2.1283,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.44,
534
+ "learning_rate": 4.269195251097868e-05,
535
+ "loss": 2.1544,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.44,
540
+ "learning_rate": 4.260890651678526e-05,
541
+ "loss": 2.155,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.45,
546
+ "learning_rate": 4.252586052259183e-05,
547
+ "loss": 2.1064,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.45,
552
+ "learning_rate": 4.244281452839841e-05,
553
+ "loss": 2.1419,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.46,
558
+ "learning_rate": 4.2359768534204984e-05,
559
+ "loss": 2.1219,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.46,
564
+ "learning_rate": 4.227672254001156e-05,
565
+ "loss": 2.1187,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.47,
570
+ "learning_rate": 4.219367654581814e-05,
571
+ "loss": 2.1089,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.47,
576
+ "learning_rate": 4.2110630551624716e-05,
577
+ "loss": 2.0924,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.48,
582
+ "learning_rate": 4.2027584557431285e-05,
583
+ "loss": 2.1622,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.48,
588
+ "learning_rate": 4.194453856323787e-05,
589
+ "loss": 2.1085,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.49,
594
+ "learning_rate": 4.186149256904444e-05,
595
+ "loss": 2.1393,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.49,
600
+ "learning_rate": 4.177844657485102e-05,
601
+ "loss": 2.1106,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.5,
606
+ "learning_rate": 4.169540058065759e-05,
607
+ "loss": 2.0899,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.5,
612
+ "learning_rate": 4.161235458646417e-05,
613
+ "loss": 2.0945,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.51,
618
+ "learning_rate": 4.1529308592270744e-05,
619
+ "loss": 2.1392,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.51,
624
+ "learning_rate": 4.1446262598077326e-05,
625
+ "loss": 2.0147,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.52,
630
+ "learning_rate": 4.13632166038839e-05,
631
+ "loss": 2.1004,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.52,
636
+ "learning_rate": 4.128017060969047e-05,
637
+ "loss": 2.082,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.53,
642
+ "learning_rate": 4.119712461549705e-05,
643
+ "loss": 2.0517,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.53,
648
+ "learning_rate": 4.111407862130363e-05,
649
+ "loss": 2.0961,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.54,
654
+ "learning_rate": 4.10310326271102e-05,
655
+ "loss": 2.0637,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.54,
660
+ "learning_rate": 4.094798663291677e-05,
661
+ "loss": 2.0417,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.55,
666
+ "learning_rate": 4.086494063872335e-05,
667
+ "loss": 2.1015,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.55,
672
+ "learning_rate": 4.078189464452993e-05,
673
+ "loss": 2.0383,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.56,
678
+ "learning_rate": 4.06988486503365e-05,
679
+ "loss": 2.1315,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.56,
684
+ "learning_rate": 4.061580265614308e-05,
685
+ "loss": 2.1347,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.57,
690
+ "learning_rate": 4.0532756661949654e-05,
691
+ "loss": 2.0428,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.57,
696
+ "learning_rate": 4.044971066775623e-05,
697
+ "loss": 2.0839,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.58,
702
+ "learning_rate": 4.036666467356281e-05,
703
+ "loss": 2.0674,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.58,
708
+ "learning_rate": 4.0283618679369387e-05,
709
+ "loss": 2.0504,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.59,
714
+ "learning_rate": 4.0200572685175955e-05,
715
+ "loss": 2.0351,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.59,
720
+ "learning_rate": 4.011752669098254e-05,
721
+ "loss": 2.0675,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.6,
726
+ "learning_rate": 4.003448069678911e-05,
727
+ "loss": 1.9957,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.6,
732
+ "learning_rate": 3.995143470259569e-05,
733
+ "loss": 2.0214,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.61,
738
+ "learning_rate": 3.986838870840226e-05,
739
+ "loss": 2.0477,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.61,
744
+ "learning_rate": 3.978534271420884e-05,
745
+ "loss": 2.0456,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.62,
750
+ "learning_rate": 3.9702296720015414e-05,
751
+ "loss": 2.0229,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.62,
756
+ "learning_rate": 3.9619250725821996e-05,
757
+ "loss": 2.0165,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.63,
762
+ "learning_rate": 3.953620473162857e-05,
763
+ "loss": 1.9758,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.63,
768
+ "learning_rate": 3.945315873743514e-05,
769
+ "loss": 2.0707,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.64,
774
+ "learning_rate": 3.937011274324172e-05,
775
+ "loss": 2.0189,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 0.64,
780
+ "learning_rate": 3.92870667490483e-05,
781
+ "loss": 1.9979,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 0.65,
786
+ "learning_rate": 3.920402075485487e-05,
787
+ "loss": 2.0263,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 0.65,
792
+ "learning_rate": 3.912097476066145e-05,
793
+ "loss": 2.0242,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 0.66,
798
+ "learning_rate": 3.903792876646802e-05,
799
+ "loss": 2.0036,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 0.66,
804
+ "learning_rate": 3.89548827722746e-05,
805
+ "loss": 1.9882,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 0.67,
810
+ "learning_rate": 3.887183677808117e-05,
811
+ "loss": 1.9732,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 0.67,
816
+ "learning_rate": 3.878879078388775e-05,
817
+ "loss": 1.9732,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 0.68,
822
+ "learning_rate": 3.8705744789694324e-05,
823
+ "loss": 2.0815,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 0.68,
828
+ "learning_rate": 3.86226987955009e-05,
829
+ "loss": 2.0008,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 0.69,
834
+ "learning_rate": 3.853965280130748e-05,
835
+ "loss": 1.9707,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 0.69,
840
+ "learning_rate": 3.8456606807114057e-05,
841
+ "loss": 1.9812,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 0.7,
846
+ "learning_rate": 3.8373560812920625e-05,
847
+ "loss": 1.9863,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 0.7,
852
+ "learning_rate": 3.829051481872721e-05,
853
+ "loss": 1.9619,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 0.71,
858
+ "learning_rate": 3.820746882453378e-05,
859
+ "loss": 1.9504,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 0.71,
864
+ "learning_rate": 3.812442283034036e-05,
865
+ "loss": 2.015,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 0.72,
870
+ "learning_rate": 3.804137683614693e-05,
871
+ "loss": 2.0166,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 0.72,
876
+ "learning_rate": 3.795833084195351e-05,
877
+ "loss": 1.9393,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 0.73,
882
+ "learning_rate": 3.7875284847760084e-05,
883
+ "loss": 1.9721,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 0.73,
888
+ "learning_rate": 3.7792238853566666e-05,
889
+ "loss": 1.9453,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 0.74,
894
+ "learning_rate": 3.770919285937324e-05,
895
+ "loss": 1.9656,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 0.74,
900
+ "learning_rate": 3.762614686517981e-05,
901
+ "loss": 1.9776,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 0.75,
906
+ "learning_rate": 3.7543100870986385e-05,
907
+ "loss": 1.9679,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 0.75,
912
+ "learning_rate": 3.746005487679297e-05,
913
+ "loss": 2.0017,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 0.76,
918
+ "learning_rate": 3.737700888259954e-05,
919
+ "loss": 1.9355,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 0.76,
924
+ "learning_rate": 3.729396288840612e-05,
925
+ "loss": 2.0124,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 0.77,
930
+ "learning_rate": 3.721091689421269e-05,
931
+ "loss": 1.9671,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 0.77,
936
+ "learning_rate": 3.712787090001927e-05,
937
+ "loss": 1.9514,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 0.78,
942
+ "learning_rate": 3.704482490582584e-05,
943
+ "loss": 1.9842,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 0.78,
948
+ "learning_rate": 3.696177891163242e-05,
949
+ "loss": 1.9971,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 0.79,
954
+ "learning_rate": 3.6878732917438994e-05,
955
+ "loss": 1.9237,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 0.79,
960
+ "learning_rate": 3.679568692324557e-05,
961
+ "loss": 1.9427,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 0.8,
966
+ "learning_rate": 3.671264092905215e-05,
967
+ "loss": 1.9638,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 0.8,
972
+ "learning_rate": 3.662959493485873e-05,
973
+ "loss": 1.9808,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 0.81,
978
+ "learning_rate": 3.6546548940665295e-05,
979
+ "loss": 1.9247,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 0.81,
984
+ "learning_rate": 3.646350294647188e-05,
985
+ "loss": 1.9813,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 0.82,
990
+ "learning_rate": 3.638045695227845e-05,
991
+ "loss": 1.8976,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 0.82,
996
+ "learning_rate": 3.629741095808503e-05,
997
+ "loss": 1.9592,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 0.83,
1002
+ "learning_rate": 3.62143649638916e-05,
1003
+ "loss": 1.9644,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 0.83,
1008
+ "learning_rate": 3.613131896969818e-05,
1009
+ "loss": 1.9495,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 0.84,
1014
+ "learning_rate": 3.6048272975504754e-05,
1015
+ "loss": 1.9123,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 0.84,
1020
+ "learning_rate": 3.596522698131133e-05,
1021
+ "loss": 1.9777,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 0.85,
1026
+ "learning_rate": 3.588218098711791e-05,
1027
+ "loss": 1.9372,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 0.85,
1032
+ "learning_rate": 3.579913499292448e-05,
1033
+ "loss": 1.9425,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 0.86,
1038
+ "learning_rate": 3.5716088998731055e-05,
1039
+ "loss": 1.8948,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 0.86,
1044
+ "learning_rate": 3.563304300453764e-05,
1045
+ "loss": 1.9407,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 0.87,
1050
+ "learning_rate": 3.554999701034421e-05,
1051
+ "loss": 1.9079,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 0.87,
1056
+ "learning_rate": 3.546695101615079e-05,
1057
+ "loss": 1.9738,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 0.88,
1062
+ "learning_rate": 3.538390502195736e-05,
1063
+ "loss": 1.9636,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 0.88,
1068
+ "learning_rate": 3.530085902776394e-05,
1069
+ "loss": 1.9252,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 0.89,
1074
+ "learning_rate": 3.5217813033570513e-05,
1075
+ "loss": 1.9185,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 0.89,
1080
+ "learning_rate": 3.5134767039377096e-05,
1081
+ "loss": 1.9472,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 0.9,
1086
+ "learning_rate": 3.5051721045183664e-05,
1087
+ "loss": 1.9579,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 0.9,
1092
+ "learning_rate": 3.496867505099024e-05,
1093
+ "loss": 1.9113,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 0.91,
1098
+ "learning_rate": 3.488562905679682e-05,
1099
+ "loss": 1.949,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 0.91,
1104
+ "learning_rate": 3.48025830626034e-05,
1105
+ "loss": 1.9589,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 0.92,
1110
+ "learning_rate": 3.4719537068409965e-05,
1111
+ "loss": 1.8764,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 0.92,
1116
+ "learning_rate": 3.463649107421655e-05,
1117
+ "loss": 1.9191,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 0.93,
1122
+ "learning_rate": 3.455344508002312e-05,
1123
+ "loss": 1.8962,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 0.93,
1128
+ "learning_rate": 3.44703990858297e-05,
1129
+ "loss": 1.9213,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 0.94,
1134
+ "learning_rate": 3.438735309163627e-05,
1135
+ "loss": 1.8769,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 0.94,
1140
+ "learning_rate": 3.430430709744285e-05,
1141
+ "loss": 1.9639,
1142
+ "step": 94500
1143
+ },
1144
+ {
1145
+ "epoch": 0.95,
1146
+ "learning_rate": 3.4221261103249424e-05,
1147
+ "loss": 1.919,
1148
+ "step": 95000
1149
+ },
1150
+ {
1151
+ "epoch": 0.95,
1152
+ "learning_rate": 3.4138215109056e-05,
1153
+ "loss": 1.911,
1154
+ "step": 95500
1155
+ },
1156
+ {
1157
+ "epoch": 0.96,
1158
+ "learning_rate": 3.405516911486258e-05,
1159
+ "loss": 1.9024,
1160
+ "step": 96000
1161
+ },
1162
+ {
1163
+ "epoch": 0.96,
1164
+ "learning_rate": 3.397212312066915e-05,
1165
+ "loss": 1.871,
1166
+ "step": 96500
1167
+ },
1168
+ {
1169
+ "epoch": 0.97,
1170
+ "learning_rate": 3.3889077126475725e-05,
1171
+ "loss": 1.9182,
1172
+ "step": 97000
1173
+ },
1174
+ {
1175
+ "epoch": 0.97,
1176
+ "learning_rate": 3.380603113228231e-05,
1177
+ "loss": 1.919,
1178
+ "step": 97500
1179
+ },
1180
+ {
1181
+ "epoch": 0.98,
1182
+ "learning_rate": 3.372298513808888e-05,
1183
+ "loss": 1.9493,
1184
+ "step": 98000
1185
+ },
1186
+ {
1187
+ "epoch": 0.98,
1188
+ "learning_rate": 3.363993914389546e-05,
1189
+ "loss": 1.8757,
1190
+ "step": 98500
1191
+ },
1192
+ {
1193
+ "epoch": 0.99,
1194
+ "learning_rate": 3.355689314970203e-05,
1195
+ "loss": 1.9248,
1196
+ "step": 99000
1197
+ },
1198
+ {
1199
+ "epoch": 0.99,
1200
+ "learning_rate": 3.347384715550861e-05,
1201
+ "loss": 1.8484,
1202
+ "step": 99500
1203
+ },
1204
+ {
1205
+ "epoch": 1.0,
1206
+ "learning_rate": 3.3390801161315183e-05,
1207
+ "loss": 1.9603,
1208
+ "step": 100000
1209
+ },
1210
+ {
1211
+ "epoch": 1.0,
1212
+ "eval_loss": 1.8004786968231201,
1213
+ "eval_runtime": 62.0584,
1214
+ "eval_samples_per_second": 680.826,
1215
+ "eval_steps_per_second": 85.113,
1216
+ "step": 100346
1217
+ },
1218
+ {
1219
+ "epoch": 1.0,
1220
+ "learning_rate": 3.3307755167121766e-05,
1221
+ "loss": 1.9152,
1222
+ "step": 100500
1223
+ },
1224
+ {
1225
+ "epoch": 1.01,
1226
+ "learning_rate": 3.3224709172928334e-05,
1227
+ "loss": 1.8557,
1228
+ "step": 101000
1229
+ },
1230
+ {
1231
+ "epoch": 1.01,
1232
+ "learning_rate": 3.314166317873491e-05,
1233
+ "loss": 1.9144,
1234
+ "step": 101500
1235
+ },
1236
+ {
1237
+ "epoch": 1.02,
1238
+ "learning_rate": 3.305861718454149e-05,
1239
+ "loss": 1.8657,
1240
+ "step": 102000
1241
+ },
1242
+ {
1243
+ "epoch": 1.02,
1244
+ "learning_rate": 3.297557119034807e-05,
1245
+ "loss": 1.8971,
1246
+ "step": 102500
1247
+ },
1248
+ {
1249
+ "epoch": 1.03,
1250
+ "learning_rate": 3.289252519615464e-05,
1251
+ "loss": 1.7939,
1252
+ "step": 103000
1253
+ },
1254
+ {
1255
+ "epoch": 1.03,
1256
+ "learning_rate": 3.280947920196122e-05,
1257
+ "loss": 1.8882,
1258
+ "step": 103500
1259
+ },
1260
+ {
1261
+ "epoch": 1.04,
1262
+ "learning_rate": 3.272643320776779e-05,
1263
+ "loss": 1.8827,
1264
+ "step": 104000
1265
+ },
1266
+ {
1267
+ "epoch": 1.04,
1268
+ "learning_rate": 3.264338721357437e-05,
1269
+ "loss": 1.8452,
1270
+ "step": 104500
1271
+ },
1272
+ {
1273
+ "epoch": 1.05,
1274
+ "learning_rate": 3.256034121938094e-05,
1275
+ "loss": 1.8755,
1276
+ "step": 105000
1277
+ },
1278
+ {
1279
+ "epoch": 1.05,
1280
+ "learning_rate": 3.247729522518752e-05,
1281
+ "loss": 1.8606,
1282
+ "step": 105500
1283
+ },
1284
+ {
1285
+ "epoch": 1.06,
1286
+ "learning_rate": 3.2394249230994094e-05,
1287
+ "loss": 1.8403,
1288
+ "step": 106000
1289
+ },
1290
+ {
1291
+ "epoch": 1.06,
1292
+ "learning_rate": 3.231120323680067e-05,
1293
+ "loss": 1.8451,
1294
+ "step": 106500
1295
+ },
1296
+ {
1297
+ "epoch": 1.07,
1298
+ "learning_rate": 3.222815724260725e-05,
1299
+ "loss": 1.8768,
1300
+ "step": 107000
1301
+ },
1302
+ {
1303
+ "epoch": 1.07,
1304
+ "learning_rate": 3.214511124841382e-05,
1305
+ "loss": 1.8505,
1306
+ "step": 107500
1307
+ },
1308
+ {
1309
+ "epoch": 1.08,
1310
+ "learning_rate": 3.2062065254220395e-05,
1311
+ "loss": 1.8806,
1312
+ "step": 108000
1313
+ },
1314
+ {
1315
+ "epoch": 1.08,
1316
+ "learning_rate": 3.197901926002698e-05,
1317
+ "loss": 1.8913,
1318
+ "step": 108500
1319
+ },
1320
+ {
1321
+ "epoch": 1.09,
1322
+ "learning_rate": 3.189597326583355e-05,
1323
+ "loss": 1.8635,
1324
+ "step": 109000
1325
+ },
1326
+ {
1327
+ "epoch": 1.09,
1328
+ "learning_rate": 3.181292727164013e-05,
1329
+ "loss": 1.8529,
1330
+ "step": 109500
1331
+ },
1332
+ {
1333
+ "epoch": 1.1,
1334
+ "learning_rate": 3.17298812774467e-05,
1335
+ "loss": 1.7873,
1336
+ "step": 110000
1337
+ },
1338
+ {
1339
+ "epoch": 1.1,
1340
+ "learning_rate": 3.164683528325328e-05,
1341
+ "loss": 1.8758,
1342
+ "step": 110500
1343
+ },
1344
+ {
1345
+ "epoch": 1.11,
1346
+ "learning_rate": 3.1563789289059854e-05,
1347
+ "loss": 1.8157,
1348
+ "step": 111000
1349
+ },
1350
+ {
1351
+ "epoch": 1.11,
1352
+ "learning_rate": 3.1480743294866436e-05,
1353
+ "loss": 1.8983,
1354
+ "step": 111500
1355
+ },
1356
+ {
1357
+ "epoch": 1.12,
1358
+ "learning_rate": 3.1397697300673004e-05,
1359
+ "loss": 1.8393,
1360
+ "step": 112000
1361
+ },
1362
+ {
1363
+ "epoch": 1.12,
1364
+ "learning_rate": 3.131465130647958e-05,
1365
+ "loss": 1.8297,
1366
+ "step": 112500
1367
+ },
1368
+ {
1369
+ "epoch": 1.13,
1370
+ "learning_rate": 3.123160531228616e-05,
1371
+ "loss": 1.8603,
1372
+ "step": 113000
1373
+ },
1374
+ {
1375
+ "epoch": 1.13,
1376
+ "learning_rate": 3.114855931809274e-05,
1377
+ "loss": 1.8201,
1378
+ "step": 113500
1379
+ },
1380
+ {
1381
+ "epoch": 1.14,
1382
+ "learning_rate": 3.106551332389931e-05,
1383
+ "loss": 1.823,
1384
+ "step": 114000
1385
+ },
1386
+ {
1387
+ "epoch": 1.14,
1388
+ "learning_rate": 3.098246732970588e-05,
1389
+ "loss": 1.8348,
1390
+ "step": 114500
1391
+ },
1392
+ {
1393
+ "epoch": 1.15,
1394
+ "learning_rate": 3.089942133551246e-05,
1395
+ "loss": 1.7525,
1396
+ "step": 115000
1397
+ },
1398
+ {
1399
+ "epoch": 1.15,
1400
+ "learning_rate": 3.081637534131904e-05,
1401
+ "loss": 1.9044,
1402
+ "step": 115500
1403
+ },
1404
+ {
1405
+ "epoch": 1.16,
1406
+ "learning_rate": 3.073332934712561e-05,
1407
+ "loss": 1.8235,
1408
+ "step": 116000
1409
+ },
1410
+ {
1411
+ "epoch": 1.16,
1412
+ "learning_rate": 3.065028335293219e-05,
1413
+ "loss": 1.8582,
1414
+ "step": 116500
1415
+ },
1416
+ {
1417
+ "epoch": 1.17,
1418
+ "learning_rate": 3.0567237358738764e-05,
1419
+ "loss": 1.8067,
1420
+ "step": 117000
1421
+ },
1422
+ {
1423
+ "epoch": 1.17,
1424
+ "learning_rate": 3.048419136454534e-05,
1425
+ "loss": 1.8657,
1426
+ "step": 117500
1427
+ },
1428
+ {
1429
+ "epoch": 1.18,
1430
+ "learning_rate": 3.0401145370351918e-05,
1431
+ "loss": 1.8455,
1432
+ "step": 118000
1433
+ },
1434
+ {
1435
+ "epoch": 1.18,
1436
+ "learning_rate": 3.0318099376158493e-05,
1437
+ "loss": 1.8431,
1438
+ "step": 118500
1439
+ },
1440
+ {
1441
+ "epoch": 1.19,
1442
+ "learning_rate": 3.023505338196507e-05,
1443
+ "loss": 1.8478,
1444
+ "step": 119000
1445
+ },
1446
+ {
1447
+ "epoch": 1.19,
1448
+ "learning_rate": 3.0152007387771647e-05,
1449
+ "loss": 1.8247,
1450
+ "step": 119500
1451
+ },
1452
+ {
1453
+ "epoch": 1.2,
1454
+ "learning_rate": 3.0068961393578222e-05,
1455
+ "loss": 1.846,
1456
+ "step": 120000
1457
+ },
1458
+ {
1459
+ "epoch": 1.2,
1460
+ "learning_rate": 2.9985915399384794e-05,
1461
+ "loss": 1.8054,
1462
+ "step": 120500
1463
+ },
1464
+ {
1465
+ "epoch": 1.21,
1466
+ "learning_rate": 2.9902869405191376e-05,
1467
+ "loss": 1.8511,
1468
+ "step": 121000
1469
+ },
1470
+ {
1471
+ "epoch": 1.21,
1472
+ "learning_rate": 2.981982341099795e-05,
1473
+ "loss": 1.826,
1474
+ "step": 121500
1475
+ },
1476
+ {
1477
+ "epoch": 1.22,
1478
+ "learning_rate": 2.9736777416804524e-05,
1479
+ "loss": 1.807,
1480
+ "step": 122000
1481
+ },
1482
+ {
1483
+ "epoch": 1.22,
1484
+ "learning_rate": 2.9653731422611102e-05,
1485
+ "loss": 1.812,
1486
+ "step": 122500
1487
+ },
1488
+ {
1489
+ "epoch": 1.23,
1490
+ "learning_rate": 2.9570685428417678e-05,
1491
+ "loss": 1.8308,
1492
+ "step": 123000
1493
+ },
1494
+ {
1495
+ "epoch": 1.23,
1496
+ "learning_rate": 2.9487639434224253e-05,
1497
+ "loss": 1.7933,
1498
+ "step": 123500
1499
+ },
1500
+ {
1501
+ "epoch": 1.24,
1502
+ "learning_rate": 2.9404593440030825e-05,
1503
+ "loss": 1.8197,
1504
+ "step": 124000
1505
+ },
1506
+ {
1507
+ "epoch": 1.24,
1508
+ "learning_rate": 2.9321547445837403e-05,
1509
+ "loss": 1.8258,
1510
+ "step": 124500
1511
+ },
1512
+ {
1513
+ "epoch": 1.25,
1514
+ "learning_rate": 2.923850145164398e-05,
1515
+ "loss": 1.8341,
1516
+ "step": 125000
1517
+ },
1518
+ {
1519
+ "epoch": 1.25,
1520
+ "learning_rate": 2.9155455457450554e-05,
1521
+ "loss": 1.8089,
1522
+ "step": 125500
1523
+ },
1524
+ {
1525
+ "epoch": 1.26,
1526
+ "learning_rate": 2.9072409463257133e-05,
1527
+ "loss": 1.8254,
1528
+ "step": 126000
1529
+ },
1530
+ {
1531
+ "epoch": 1.26,
1532
+ "learning_rate": 2.8989363469063708e-05,
1533
+ "loss": 1.7734,
1534
+ "step": 126500
1535
+ },
1536
+ {
1537
+ "epoch": 1.27,
1538
+ "learning_rate": 2.890631747487028e-05,
1539
+ "loss": 1.811,
1540
+ "step": 127000
1541
+ },
1542
+ {
1543
+ "epoch": 1.27,
1544
+ "learning_rate": 2.8823271480676862e-05,
1545
+ "loss": 1.775,
1546
+ "step": 127500
1547
+ },
1548
+ {
1549
+ "epoch": 1.28,
1550
+ "learning_rate": 2.8740225486483434e-05,
1551
+ "loss": 1.863,
1552
+ "step": 128000
1553
+ },
1554
+ {
1555
+ "epoch": 1.28,
1556
+ "learning_rate": 2.865717949229001e-05,
1557
+ "loss": 1.8061,
1558
+ "step": 128500
1559
+ },
1560
+ {
1561
+ "epoch": 1.29,
1562
+ "learning_rate": 2.8574133498096588e-05,
1563
+ "loss": 1.7966,
1564
+ "step": 129000
1565
+ },
1566
+ {
1567
+ "epoch": 1.29,
1568
+ "learning_rate": 2.8491087503903163e-05,
1569
+ "loss": 1.8331,
1570
+ "step": 129500
1571
+ },
1572
+ {
1573
+ "epoch": 1.3,
1574
+ "learning_rate": 2.840804150970974e-05,
1575
+ "loss": 1.7619,
1576
+ "step": 130000
1577
+ },
1578
+ {
1579
+ "epoch": 1.3,
1580
+ "learning_rate": 2.8324995515516317e-05,
1581
+ "loss": 1.8034,
1582
+ "step": 130500
1583
+ },
1584
+ {
1585
+ "epoch": 1.31,
1586
+ "learning_rate": 2.8241949521322892e-05,
1587
+ "loss": 1.8151,
1588
+ "step": 131000
1589
+ },
1590
+ {
1591
+ "epoch": 1.31,
1592
+ "learning_rate": 2.8158903527129464e-05,
1593
+ "loss": 1.817,
1594
+ "step": 131500
1595
+ },
1596
+ {
1597
+ "epoch": 1.32,
1598
+ "learning_rate": 2.8075857532936046e-05,
1599
+ "loss": 1.8221,
1600
+ "step": 132000
1601
+ },
1602
+ {
1603
+ "epoch": 1.32,
1604
+ "learning_rate": 2.799281153874262e-05,
1605
+ "loss": 1.8163,
1606
+ "step": 132500
1607
+ },
1608
+ {
1609
+ "epoch": 1.33,
1610
+ "learning_rate": 2.7909765544549194e-05,
1611
+ "loss": 1.8357,
1612
+ "step": 133000
1613
+ },
1614
+ {
1615
+ "epoch": 1.33,
1616
+ "learning_rate": 2.782671955035577e-05,
1617
+ "loss": 1.7666,
1618
+ "step": 133500
1619
+ },
1620
+ {
1621
+ "epoch": 1.34,
1622
+ "learning_rate": 2.7743673556162348e-05,
1623
+ "loss": 1.8299,
1624
+ "step": 134000
1625
+ },
1626
+ {
1627
+ "epoch": 1.34,
1628
+ "learning_rate": 2.7660627561968923e-05,
1629
+ "loss": 1.7589,
1630
+ "step": 134500
1631
+ },
1632
+ {
1633
+ "epoch": 1.35,
1634
+ "learning_rate": 2.7577581567775495e-05,
1635
+ "loss": 1.7555,
1636
+ "step": 135000
1637
+ },
1638
+ {
1639
+ "epoch": 1.35,
1640
+ "learning_rate": 2.7494535573582077e-05,
1641
+ "loss": 1.7597,
1642
+ "step": 135500
1643
+ },
1644
+ {
1645
+ "epoch": 1.36,
1646
+ "learning_rate": 2.741148957938865e-05,
1647
+ "loss": 1.8411,
1648
+ "step": 136000
1649
+ },
1650
+ {
1651
+ "epoch": 1.36,
1652
+ "learning_rate": 2.7328443585195224e-05,
1653
+ "loss": 1.8237,
1654
+ "step": 136500
1655
+ },
1656
+ {
1657
+ "epoch": 1.37,
1658
+ "learning_rate": 2.7245397591001803e-05,
1659
+ "loss": 1.8343,
1660
+ "step": 137000
1661
+ },
1662
+ {
1663
+ "epoch": 1.37,
1664
+ "learning_rate": 2.7162351596808378e-05,
1665
+ "loss": 1.8008,
1666
+ "step": 137500
1667
+ },
1668
+ {
1669
+ "epoch": 1.38,
1670
+ "learning_rate": 2.707930560261495e-05,
1671
+ "loss": 1.7594,
1672
+ "step": 138000
1673
+ },
1674
+ {
1675
+ "epoch": 1.38,
1676
+ "learning_rate": 2.6996259608421532e-05,
1677
+ "loss": 1.8164,
1678
+ "step": 138500
1679
+ },
1680
+ {
1681
+ "epoch": 1.39,
1682
+ "learning_rate": 2.6913213614228104e-05,
1683
+ "loss": 1.8277,
1684
+ "step": 139000
1685
+ },
1686
+ {
1687
+ "epoch": 1.39,
1688
+ "learning_rate": 2.683016762003468e-05,
1689
+ "loss": 1.7493,
1690
+ "step": 139500
1691
+ },
1692
+ {
1693
+ "epoch": 1.4,
1694
+ "learning_rate": 2.6747121625841258e-05,
1695
+ "loss": 1.8044,
1696
+ "step": 140000
1697
+ },
1698
+ {
1699
+ "epoch": 1.4,
1700
+ "learning_rate": 2.6664075631647833e-05,
1701
+ "loss": 1.8229,
1702
+ "step": 140500
1703
+ },
1704
+ {
1705
+ "epoch": 1.41,
1706
+ "learning_rate": 2.658102963745441e-05,
1707
+ "loss": 1.8213,
1708
+ "step": 141000
1709
+ },
1710
+ {
1711
+ "epoch": 1.41,
1712
+ "learning_rate": 2.6497983643260987e-05,
1713
+ "loss": 1.8313,
1714
+ "step": 141500
1715
+ },
1716
+ {
1717
+ "epoch": 1.42,
1718
+ "learning_rate": 2.6414937649067563e-05,
1719
+ "loss": 1.7616,
1720
+ "step": 142000
1721
+ },
1722
+ {
1723
+ "epoch": 1.42,
1724
+ "learning_rate": 2.6331891654874134e-05,
1725
+ "loss": 1.7895,
1726
+ "step": 142500
1727
+ },
1728
+ {
1729
+ "epoch": 1.43,
1730
+ "learning_rate": 2.6248845660680717e-05,
1731
+ "loss": 1.8175,
1732
+ "step": 143000
1733
+ },
1734
+ {
1735
+ "epoch": 1.43,
1736
+ "learning_rate": 2.616579966648729e-05,
1737
+ "loss": 1.7599,
1738
+ "step": 143500
1739
+ },
1740
+ {
1741
+ "epoch": 1.44,
1742
+ "learning_rate": 2.6082753672293864e-05,
1743
+ "loss": 1.789,
1744
+ "step": 144000
1745
+ },
1746
+ {
1747
+ "epoch": 1.44,
1748
+ "learning_rate": 2.599970767810044e-05,
1749
+ "loss": 1.7954,
1750
+ "step": 144500
1751
+ },
1752
+ {
1753
+ "epoch": 1.45,
1754
+ "learning_rate": 2.5916661683907018e-05,
1755
+ "loss": 1.7553,
1756
+ "step": 145000
1757
+ },
1758
+ {
1759
+ "epoch": 1.45,
1760
+ "learning_rate": 2.5833615689713593e-05,
1761
+ "loss": 1.8179,
1762
+ "step": 145500
1763
+ },
1764
+ {
1765
+ "epoch": 1.45,
1766
+ "learning_rate": 2.5750569695520165e-05,
1767
+ "loss": 1.794,
1768
+ "step": 146000
1769
+ },
1770
+ {
1771
+ "epoch": 1.46,
1772
+ "learning_rate": 2.5667523701326747e-05,
1773
+ "loss": 1.7899,
1774
+ "step": 146500
1775
+ },
1776
+ {
1777
+ "epoch": 1.46,
1778
+ "learning_rate": 2.558447770713332e-05,
1779
+ "loss": 1.7393,
1780
+ "step": 147000
1781
+ },
1782
+ {
1783
+ "epoch": 1.47,
1784
+ "learning_rate": 2.5501431712939894e-05,
1785
+ "loss": 1.7655,
1786
+ "step": 147500
1787
+ },
1788
+ {
1789
+ "epoch": 1.47,
1790
+ "learning_rate": 2.5418385718746473e-05,
1791
+ "loss": 1.8001,
1792
+ "step": 148000
1793
+ },
1794
+ {
1795
+ "epoch": 1.48,
1796
+ "learning_rate": 2.5335339724553048e-05,
1797
+ "loss": 1.7506,
1798
+ "step": 148500
1799
+ },
1800
+ {
1801
+ "epoch": 1.48,
1802
+ "learning_rate": 2.5252293730359623e-05,
1803
+ "loss": 1.7704,
1804
+ "step": 149000
1805
+ },
1806
+ {
1807
+ "epoch": 1.49,
1808
+ "learning_rate": 2.5169247736166202e-05,
1809
+ "loss": 1.7745,
1810
+ "step": 149500
1811
+ },
1812
+ {
1813
+ "epoch": 1.49,
1814
+ "learning_rate": 2.5086201741972774e-05,
1815
+ "loss": 1.7585,
1816
+ "step": 150000
1817
+ },
1818
+ {
1819
+ "epoch": 1.5,
1820
+ "learning_rate": 2.500315574777935e-05,
1821
+ "loss": 1.7453,
1822
+ "step": 150500
1823
+ },
1824
+ {
1825
+ "epoch": 1.5,
1826
+ "learning_rate": 2.4920109753585928e-05,
1827
+ "loss": 1.7873,
1828
+ "step": 151000
1829
+ },
1830
+ {
1831
+ "epoch": 1.51,
1832
+ "learning_rate": 2.4837063759392503e-05,
1833
+ "loss": 1.7854,
1834
+ "step": 151500
1835
+ },
1836
+ {
1837
+ "epoch": 1.51,
1838
+ "learning_rate": 2.475401776519908e-05,
1839
+ "loss": 1.7683,
1840
+ "step": 152000
1841
+ },
1842
+ {
1843
+ "epoch": 1.52,
1844
+ "learning_rate": 2.4670971771005654e-05,
1845
+ "loss": 1.8352,
1846
+ "step": 152500
1847
+ },
1848
+ {
1849
+ "epoch": 1.52,
1850
+ "learning_rate": 2.4587925776812233e-05,
1851
+ "loss": 1.734,
1852
+ "step": 153000
1853
+ },
1854
+ {
1855
+ "epoch": 1.53,
1856
+ "learning_rate": 2.4504879782618805e-05,
1857
+ "loss": 1.7883,
1858
+ "step": 153500
1859
+ },
1860
+ {
1861
+ "epoch": 1.53,
1862
+ "learning_rate": 2.4421833788425383e-05,
1863
+ "loss": 1.7501,
1864
+ "step": 154000
1865
+ },
1866
+ {
1867
+ "epoch": 1.54,
1868
+ "learning_rate": 2.433878779423196e-05,
1869
+ "loss": 1.8167,
1870
+ "step": 154500
1871
+ },
1872
+ {
1873
+ "epoch": 1.54,
1874
+ "learning_rate": 2.4255741800038534e-05,
1875
+ "loss": 1.7432,
1876
+ "step": 155000
1877
+ },
1878
+ {
1879
+ "epoch": 1.55,
1880
+ "learning_rate": 2.4172695805845112e-05,
1881
+ "loss": 1.7148,
1882
+ "step": 155500
1883
+ },
1884
+ {
1885
+ "epoch": 1.55,
1886
+ "learning_rate": 2.4089649811651684e-05,
1887
+ "loss": 1.7508,
1888
+ "step": 156000
1889
+ },
1890
+ {
1891
+ "epoch": 1.56,
1892
+ "learning_rate": 2.4006603817458263e-05,
1893
+ "loss": 1.7503,
1894
+ "step": 156500
1895
+ },
1896
+ {
1897
+ "epoch": 1.56,
1898
+ "learning_rate": 2.392355782326484e-05,
1899
+ "loss": 1.7762,
1900
+ "step": 157000
1901
+ },
1902
+ {
1903
+ "epoch": 1.57,
1904
+ "learning_rate": 2.3840511829071414e-05,
1905
+ "loss": 1.7546,
1906
+ "step": 157500
1907
+ },
1908
+ {
1909
+ "epoch": 1.57,
1910
+ "learning_rate": 2.375746583487799e-05,
1911
+ "loss": 1.7867,
1912
+ "step": 158000
1913
+ },
1914
+ {
1915
+ "epoch": 1.58,
1916
+ "learning_rate": 2.3674419840684568e-05,
1917
+ "loss": 1.7745,
1918
+ "step": 158500
1919
+ },
1920
+ {
1921
+ "epoch": 1.58,
1922
+ "learning_rate": 2.359137384649114e-05,
1923
+ "loss": 1.7915,
1924
+ "step": 159000
1925
+ },
1926
+ {
1927
+ "epoch": 1.59,
1928
+ "learning_rate": 2.3508327852297718e-05,
1929
+ "loss": 1.6905,
1930
+ "step": 159500
1931
+ },
1932
+ {
1933
+ "epoch": 1.59,
1934
+ "learning_rate": 2.3425281858104294e-05,
1935
+ "loss": 1.7627,
1936
+ "step": 160000
1937
+ },
1938
+ {
1939
+ "epoch": 1.6,
1940
+ "learning_rate": 2.334223586391087e-05,
1941
+ "loss": 1.7578,
1942
+ "step": 160500
1943
+ },
1944
+ {
1945
+ "epoch": 1.6,
1946
+ "learning_rate": 2.3259189869717447e-05,
1947
+ "loss": 1.7387,
1948
+ "step": 161000
1949
+ },
1950
+ {
1951
+ "epoch": 1.61,
1952
+ "learning_rate": 2.317614387552402e-05,
1953
+ "loss": 1.767,
1954
+ "step": 161500
1955
+ },
1956
+ {
1957
+ "epoch": 1.61,
1958
+ "learning_rate": 2.3093097881330598e-05,
1959
+ "loss": 1.7562,
1960
+ "step": 162000
1961
+ },
1962
+ {
1963
+ "epoch": 1.62,
1964
+ "learning_rate": 2.3010051887137173e-05,
1965
+ "loss": 1.7326,
1966
+ "step": 162500
1967
+ },
1968
+ {
1969
+ "epoch": 1.62,
1970
+ "learning_rate": 2.292700589294375e-05,
1971
+ "loss": 1.7177,
1972
+ "step": 163000
1973
+ },
1974
+ {
1975
+ "epoch": 1.63,
1976
+ "learning_rate": 2.2843959898750324e-05,
1977
+ "loss": 1.8186,
1978
+ "step": 163500
1979
+ },
1980
+ {
1981
+ "epoch": 1.63,
1982
+ "learning_rate": 2.2760913904556903e-05,
1983
+ "loss": 1.756,
1984
+ "step": 164000
1985
+ },
1986
+ {
1987
+ "epoch": 1.64,
1988
+ "learning_rate": 2.2677867910363475e-05,
1989
+ "loss": 1.7344,
1990
+ "step": 164500
1991
+ },
1992
+ {
1993
+ "epoch": 1.64,
1994
+ "learning_rate": 2.2594821916170053e-05,
1995
+ "loss": 1.8005,
1996
+ "step": 165000
1997
+ },
1998
+ {
1999
+ "epoch": 1.65,
2000
+ "learning_rate": 2.251177592197663e-05,
2001
+ "loss": 1.7913,
2002
+ "step": 165500
2003
+ },
2004
+ {
2005
+ "epoch": 1.65,
2006
+ "learning_rate": 2.2428729927783204e-05,
2007
+ "loss": 1.7065,
2008
+ "step": 166000
2009
+ },
2010
+ {
2011
+ "epoch": 1.66,
2012
+ "learning_rate": 2.2345683933589783e-05,
2013
+ "loss": 1.7689,
2014
+ "step": 166500
2015
+ },
2016
+ {
2017
+ "epoch": 1.66,
2018
+ "learning_rate": 2.2262637939396354e-05,
2019
+ "loss": 1.7677,
2020
+ "step": 167000
2021
+ },
2022
+ {
2023
+ "epoch": 1.67,
2024
+ "learning_rate": 2.2179591945202933e-05,
2025
+ "loss": 1.7108,
2026
+ "step": 167500
2027
+ },
2028
+ {
2029
+ "epoch": 1.67,
2030
+ "learning_rate": 2.209654595100951e-05,
2031
+ "loss": 1.7906,
2032
+ "step": 168000
2033
+ },
2034
+ {
2035
+ "epoch": 1.68,
2036
+ "learning_rate": 2.2013499956816084e-05,
2037
+ "loss": 1.7315,
2038
+ "step": 168500
2039
+ },
2040
+ {
2041
+ "epoch": 1.68,
2042
+ "learning_rate": 2.193045396262266e-05,
2043
+ "loss": 1.7523,
2044
+ "step": 169000
2045
+ },
2046
+ {
2047
+ "epoch": 1.69,
2048
+ "learning_rate": 2.1847407968429238e-05,
2049
+ "loss": 1.723,
2050
+ "step": 169500
2051
+ },
2052
+ {
2053
+ "epoch": 1.69,
2054
+ "learning_rate": 2.176436197423581e-05,
2055
+ "loss": 1.7672,
2056
+ "step": 170000
2057
+ },
2058
+ {
2059
+ "epoch": 1.7,
2060
+ "learning_rate": 2.1681315980042388e-05,
2061
+ "loss": 1.7129,
2062
+ "step": 170500
2063
+ },
2064
+ {
2065
+ "epoch": 1.7,
2066
+ "learning_rate": 2.1598269985848964e-05,
2067
+ "loss": 1.6996,
2068
+ "step": 171000
2069
+ },
2070
+ {
2071
+ "epoch": 1.71,
2072
+ "learning_rate": 2.151522399165554e-05,
2073
+ "loss": 1.802,
2074
+ "step": 171500
2075
+ },
2076
+ {
2077
+ "epoch": 1.71,
2078
+ "learning_rate": 2.1432177997462118e-05,
2079
+ "loss": 1.7453,
2080
+ "step": 172000
2081
+ },
2082
+ {
2083
+ "epoch": 1.72,
2084
+ "learning_rate": 2.134913200326869e-05,
2085
+ "loss": 1.784,
2086
+ "step": 172500
2087
+ },
2088
+ {
2089
+ "epoch": 1.72,
2090
+ "learning_rate": 2.1266086009075268e-05,
2091
+ "loss": 1.7123,
2092
+ "step": 173000
2093
+ },
2094
+ {
2095
+ "epoch": 1.73,
2096
+ "learning_rate": 2.1183040014881843e-05,
2097
+ "loss": 1.7359,
2098
+ "step": 173500
2099
+ },
2100
+ {
2101
+ "epoch": 1.73,
2102
+ "learning_rate": 2.109999402068842e-05,
2103
+ "loss": 1.7265,
2104
+ "step": 174000
2105
+ },
2106
+ {
2107
+ "epoch": 1.74,
2108
+ "learning_rate": 2.1016948026494994e-05,
2109
+ "loss": 1.735,
2110
+ "step": 174500
2111
+ },
2112
+ {
2113
+ "epoch": 1.74,
2114
+ "learning_rate": 2.0933902032301573e-05,
2115
+ "loss": 1.7794,
2116
+ "step": 175000
2117
+ },
2118
+ {
2119
+ "epoch": 1.75,
2120
+ "learning_rate": 2.0850856038108148e-05,
2121
+ "loss": 1.7365,
2122
+ "step": 175500
2123
+ },
2124
+ {
2125
+ "epoch": 1.75,
2126
+ "learning_rate": 2.0767810043914723e-05,
2127
+ "loss": 1.703,
2128
+ "step": 176000
2129
+ },
2130
+ {
2131
+ "epoch": 1.76,
2132
+ "learning_rate": 2.06847640497213e-05,
2133
+ "loss": 1.7378,
2134
+ "step": 176500
2135
+ },
2136
+ {
2137
+ "epoch": 1.76,
2138
+ "learning_rate": 2.0601718055527874e-05,
2139
+ "loss": 1.7197,
2140
+ "step": 177000
2141
+ },
2142
+ {
2143
+ "epoch": 1.77,
2144
+ "learning_rate": 2.0518672061334453e-05,
2145
+ "loss": 1.7206,
2146
+ "step": 177500
2147
+ },
2148
+ {
2149
+ "epoch": 1.77,
2150
+ "learning_rate": 2.0435626067141024e-05,
2151
+ "loss": 1.692,
2152
+ "step": 178000
2153
+ },
2154
+ {
2155
+ "epoch": 1.78,
2156
+ "learning_rate": 2.0352580072947603e-05,
2157
+ "loss": 1.7309,
2158
+ "step": 178500
2159
+ },
2160
+ {
2161
+ "epoch": 1.78,
2162
+ "learning_rate": 2.026953407875418e-05,
2163
+ "loss": 1.7436,
2164
+ "step": 179000
2165
+ },
2166
+ {
2167
+ "epoch": 1.79,
2168
+ "learning_rate": 2.0186488084560754e-05,
2169
+ "loss": 1.7035,
2170
+ "step": 179500
2171
+ },
2172
+ {
2173
+ "epoch": 1.79,
2174
+ "learning_rate": 2.010344209036733e-05,
2175
+ "loss": 1.714,
2176
+ "step": 180000
2177
+ },
2178
+ {
2179
+ "epoch": 1.8,
2180
+ "learning_rate": 2.0020396096173904e-05,
2181
+ "loss": 1.7232,
2182
+ "step": 180500
2183
+ },
2184
+ {
2185
+ "epoch": 1.8,
2186
+ "learning_rate": 1.9937350101980483e-05,
2187
+ "loss": 1.7051,
2188
+ "step": 181000
2189
+ },
2190
+ {
2191
+ "epoch": 1.81,
2192
+ "learning_rate": 1.985430410778706e-05,
2193
+ "loss": 1.7289,
2194
+ "step": 181500
2195
+ },
2196
+ {
2197
+ "epoch": 1.81,
2198
+ "learning_rate": 1.9771258113593634e-05,
2199
+ "loss": 1.7324,
2200
+ "step": 182000
2201
+ },
2202
+ {
2203
+ "epoch": 1.82,
2204
+ "learning_rate": 1.968821211940021e-05,
2205
+ "loss": 1.7407,
2206
+ "step": 182500
2207
+ },
2208
+ {
2209
+ "epoch": 1.82,
2210
+ "learning_rate": 1.9605166125206788e-05,
2211
+ "loss": 1.7414,
2212
+ "step": 183000
2213
+ },
2214
+ {
2215
+ "epoch": 1.83,
2216
+ "learning_rate": 1.952212013101336e-05,
2217
+ "loss": 1.7127,
2218
+ "step": 183500
2219
+ },
2220
+ {
2221
+ "epoch": 1.83,
2222
+ "learning_rate": 1.9439074136819938e-05,
2223
+ "loss": 1.7178,
2224
+ "step": 184000
2225
+ },
2226
+ {
2227
+ "epoch": 1.84,
2228
+ "learning_rate": 1.9356028142626513e-05,
2229
+ "loss": 1.7367,
2230
+ "step": 184500
2231
+ },
2232
+ {
2233
+ "epoch": 1.84,
2234
+ "learning_rate": 1.927298214843309e-05,
2235
+ "loss": 1.7537,
2236
+ "step": 185000
2237
+ },
2238
+ {
2239
+ "epoch": 1.85,
2240
+ "learning_rate": 1.9189936154239664e-05,
2241
+ "loss": 1.7151,
2242
+ "step": 185500
2243
+ },
2244
+ {
2245
+ "epoch": 1.85,
2246
+ "learning_rate": 1.910689016004624e-05,
2247
+ "loss": 1.6802,
2248
+ "step": 186000
2249
+ },
2250
+ {
2251
+ "epoch": 1.86,
2252
+ "learning_rate": 1.9023844165852818e-05,
2253
+ "loss": 1.6828,
2254
+ "step": 186500
2255
+ },
2256
+ {
2257
+ "epoch": 1.86,
2258
+ "learning_rate": 1.8940798171659393e-05,
2259
+ "loss": 1.6664,
2260
+ "step": 187000
2261
+ },
2262
+ {
2263
+ "epoch": 1.87,
2264
+ "learning_rate": 1.885775217746597e-05,
2265
+ "loss": 1.6923,
2266
+ "step": 187500
2267
+ },
2268
+ {
2269
+ "epoch": 1.87,
2270
+ "learning_rate": 1.8774706183272544e-05,
2271
+ "loss": 1.7051,
2272
+ "step": 188000
2273
+ },
2274
+ {
2275
+ "epoch": 1.88,
2276
+ "learning_rate": 1.8691660189079123e-05,
2277
+ "loss": 1.7262,
2278
+ "step": 188500
2279
+ },
2280
+ {
2281
+ "epoch": 1.88,
2282
+ "learning_rate": 1.8608614194885695e-05,
2283
+ "loss": 1.7232,
2284
+ "step": 189000
2285
+ },
2286
+ {
2287
+ "epoch": 1.89,
2288
+ "learning_rate": 1.8525568200692273e-05,
2289
+ "loss": 1.6761,
2290
+ "step": 189500
2291
+ },
2292
+ {
2293
+ "epoch": 1.89,
2294
+ "learning_rate": 1.844252220649885e-05,
2295
+ "loss": 1.7318,
2296
+ "step": 190000
2297
+ },
2298
+ {
2299
+ "epoch": 1.9,
2300
+ "learning_rate": 1.8359476212305424e-05,
2301
+ "loss": 1.7338,
2302
+ "step": 190500
2303
+ },
2304
+ {
2305
+ "epoch": 1.9,
2306
+ "learning_rate": 1.8276430218112e-05,
2307
+ "loss": 1.725,
2308
+ "step": 191000
2309
+ },
2310
+ {
2311
+ "epoch": 1.91,
2312
+ "learning_rate": 1.8193384223918574e-05,
2313
+ "loss": 1.6966,
2314
+ "step": 191500
2315
+ },
2316
+ {
2317
+ "epoch": 1.91,
2318
+ "learning_rate": 1.8110338229725153e-05,
2319
+ "loss": 1.7098,
2320
+ "step": 192000
2321
+ },
2322
+ {
2323
+ "epoch": 1.92,
2324
+ "learning_rate": 1.802729223553173e-05,
2325
+ "loss": 1.6887,
2326
+ "step": 192500
2327
+ },
2328
+ {
2329
+ "epoch": 1.92,
2330
+ "learning_rate": 1.7944246241338304e-05,
2331
+ "loss": 1.6948,
2332
+ "step": 193000
2333
+ },
2334
+ {
2335
+ "epoch": 1.93,
2336
+ "learning_rate": 1.786120024714488e-05,
2337
+ "loss": 1.7347,
2338
+ "step": 193500
2339
+ },
2340
+ {
2341
+ "epoch": 1.93,
2342
+ "learning_rate": 1.7778154252951458e-05,
2343
+ "loss": 1.6812,
2344
+ "step": 194000
2345
+ },
2346
+ {
2347
+ "epoch": 1.94,
2348
+ "learning_rate": 1.769510825875803e-05,
2349
+ "loss": 1.7496,
2350
+ "step": 194500
2351
+ },
2352
+ {
2353
+ "epoch": 1.94,
2354
+ "learning_rate": 1.7612062264564608e-05,
2355
+ "loss": 1.7149,
2356
+ "step": 195000
2357
+ },
2358
+ {
2359
+ "epoch": 1.95,
2360
+ "learning_rate": 1.7529016270371184e-05,
2361
+ "loss": 1.7519,
2362
+ "step": 195500
2363
+ },
2364
+ {
2365
+ "epoch": 1.95,
2366
+ "learning_rate": 1.744597027617776e-05,
2367
+ "loss": 1.6595,
2368
+ "step": 196000
2369
+ },
2370
+ {
2371
+ "epoch": 1.96,
2372
+ "learning_rate": 1.7362924281984334e-05,
2373
+ "loss": 1.7009,
2374
+ "step": 196500
2375
+ },
2376
+ {
2377
+ "epoch": 1.96,
2378
+ "learning_rate": 1.727987828779091e-05,
2379
+ "loss": 1.7369,
2380
+ "step": 197000
2381
+ },
2382
+ {
2383
+ "epoch": 1.97,
2384
+ "learning_rate": 1.7196832293597488e-05,
2385
+ "loss": 1.6914,
2386
+ "step": 197500
2387
+ },
2388
+ {
2389
+ "epoch": 1.97,
2390
+ "learning_rate": 1.7113786299404063e-05,
2391
+ "loss": 1.74,
2392
+ "step": 198000
2393
+ },
2394
+ {
2395
+ "epoch": 1.98,
2396
+ "learning_rate": 1.703074030521064e-05,
2397
+ "loss": 1.7348,
2398
+ "step": 198500
2399
+ },
2400
+ {
2401
+ "epoch": 1.98,
2402
+ "learning_rate": 1.6947694311017214e-05,
2403
+ "loss": 1.6941,
2404
+ "step": 199000
2405
+ },
2406
+ {
2407
+ "epoch": 1.99,
2408
+ "learning_rate": 1.6864648316823793e-05,
2409
+ "loss": 1.6831,
2410
+ "step": 199500
2411
+ },
2412
+ {
2413
+ "epoch": 1.99,
2414
+ "learning_rate": 1.6781602322630365e-05,
2415
+ "loss": 1.6855,
2416
+ "step": 200000
2417
+ },
2418
+ {
2419
+ "epoch": 2.0,
2420
+ "learning_rate": 1.6698556328436943e-05,
2421
+ "loss": 1.7032,
2422
+ "step": 200500
2423
+ },
2424
+ {
2425
+ "epoch": 2.0,
2426
+ "eval_loss": 1.6460107564926147,
2427
+ "eval_runtime": 62.456,
2428
+ "eval_samples_per_second": 676.493,
2429
+ "eval_steps_per_second": 84.572,
2430
+ "step": 200692
2431
+ },
2432
+ {
2433
+ "epoch": 2.0,
2434
+ "learning_rate": 1.661551033424352e-05,
2435
+ "loss": 1.6659,
2436
+ "step": 201000
2437
+ },
2438
+ {
2439
+ "epoch": 2.01,
2440
+ "learning_rate": 1.6532464340050094e-05,
2441
+ "loss": 1.7258,
2442
+ "step": 201500
2443
+ },
2444
+ {
2445
+ "epoch": 2.01,
2446
+ "learning_rate": 1.644941834585667e-05,
2447
+ "loss": 1.6603,
2448
+ "step": 202000
2449
+ },
2450
+ {
2451
+ "epoch": 2.02,
2452
+ "learning_rate": 1.6366372351663244e-05,
2453
+ "loss": 1.6643,
2454
+ "step": 202500
2455
+ },
2456
+ {
2457
+ "epoch": 2.02,
2458
+ "learning_rate": 1.6283326357469823e-05,
2459
+ "loss": 1.67,
2460
+ "step": 203000
2461
+ },
2462
+ {
2463
+ "epoch": 2.03,
2464
+ "learning_rate": 1.62002803632764e-05,
2465
+ "loss": 1.7112,
2466
+ "step": 203500
2467
+ },
2468
+ {
2469
+ "epoch": 2.03,
2470
+ "learning_rate": 1.6117234369082974e-05,
2471
+ "loss": 1.672,
2472
+ "step": 204000
2473
+ },
2474
+ {
2475
+ "epoch": 2.04,
2476
+ "learning_rate": 1.603418837488955e-05,
2477
+ "loss": 1.6758,
2478
+ "step": 204500
2479
+ },
2480
+ {
2481
+ "epoch": 2.04,
2482
+ "learning_rate": 1.5951142380696128e-05,
2483
+ "loss": 1.7029,
2484
+ "step": 205000
2485
+ },
2486
+ {
2487
+ "epoch": 2.05,
2488
+ "learning_rate": 1.58680963865027e-05,
2489
+ "loss": 1.673,
2490
+ "step": 205500
2491
+ },
2492
+ {
2493
+ "epoch": 2.05,
2494
+ "learning_rate": 1.578505039230928e-05,
2495
+ "loss": 1.684,
2496
+ "step": 206000
2497
+ },
2498
+ {
2499
+ "epoch": 2.06,
2500
+ "learning_rate": 1.5702004398115854e-05,
2501
+ "loss": 1.6506,
2502
+ "step": 206500
2503
+ },
2504
+ {
2505
+ "epoch": 2.06,
2506
+ "learning_rate": 1.561895840392243e-05,
2507
+ "loss": 1.6968,
2508
+ "step": 207000
2509
+ },
2510
+ {
2511
+ "epoch": 2.07,
2512
+ "learning_rate": 1.5535912409729008e-05,
2513
+ "loss": 1.7132,
2514
+ "step": 207500
2515
+ },
2516
+ {
2517
+ "epoch": 2.07,
2518
+ "learning_rate": 1.545286641553558e-05,
2519
+ "loss": 1.7283,
2520
+ "step": 208000
2521
+ },
2522
+ {
2523
+ "epoch": 2.08,
2524
+ "learning_rate": 1.5369820421342158e-05,
2525
+ "loss": 1.6659,
2526
+ "step": 208500
2527
+ },
2528
+ {
2529
+ "epoch": 2.08,
2530
+ "learning_rate": 1.5286774427148733e-05,
2531
+ "loss": 1.653,
2532
+ "step": 209000
2533
+ },
2534
+ {
2535
+ "epoch": 2.09,
2536
+ "learning_rate": 1.5203728432955309e-05,
2537
+ "loss": 1.6335,
2538
+ "step": 209500
2539
+ },
2540
+ {
2541
+ "epoch": 2.09,
2542
+ "learning_rate": 1.5120682438761886e-05,
2543
+ "loss": 1.661,
2544
+ "step": 210000
2545
+ },
2546
+ {
2547
+ "epoch": 2.1,
2548
+ "learning_rate": 1.503763644456846e-05,
2549
+ "loss": 1.6556,
2550
+ "step": 210500
2551
+ },
2552
+ {
2553
+ "epoch": 2.1,
2554
+ "learning_rate": 1.4954590450375036e-05,
2555
+ "loss": 1.7251,
2556
+ "step": 211000
2557
+ },
2558
+ {
2559
+ "epoch": 2.11,
2560
+ "learning_rate": 1.4871544456181613e-05,
2561
+ "loss": 1.6816,
2562
+ "step": 211500
2563
+ },
2564
+ {
2565
+ "epoch": 2.11,
2566
+ "learning_rate": 1.4788498461988187e-05,
2567
+ "loss": 1.6591,
2568
+ "step": 212000
2569
+ },
2570
+ {
2571
+ "epoch": 2.12,
2572
+ "learning_rate": 1.4705452467794764e-05,
2573
+ "loss": 1.7129,
2574
+ "step": 212500
2575
+ },
2576
+ {
2577
+ "epoch": 2.12,
2578
+ "learning_rate": 1.4622406473601341e-05,
2579
+ "loss": 1.6808,
2580
+ "step": 213000
2581
+ },
2582
+ {
2583
+ "epoch": 2.13,
2584
+ "learning_rate": 1.4539360479407915e-05,
2585
+ "loss": 1.6958,
2586
+ "step": 213500
2587
+ },
2588
+ {
2589
+ "epoch": 2.13,
2590
+ "learning_rate": 1.4456314485214492e-05,
2591
+ "loss": 1.6892,
2592
+ "step": 214000
2593
+ },
2594
+ {
2595
+ "epoch": 2.14,
2596
+ "learning_rate": 1.4373268491021069e-05,
2597
+ "loss": 1.7093,
2598
+ "step": 214500
2599
+ },
2600
+ {
2601
+ "epoch": 2.14,
2602
+ "learning_rate": 1.4290222496827644e-05,
2603
+ "loss": 1.669,
2604
+ "step": 215000
2605
+ },
2606
+ {
2607
+ "epoch": 2.15,
2608
+ "learning_rate": 1.420717650263422e-05,
2609
+ "loss": 1.6376,
2610
+ "step": 215500
2611
+ },
2612
+ {
2613
+ "epoch": 2.15,
2614
+ "learning_rate": 1.4124130508440794e-05,
2615
+ "loss": 1.6988,
2616
+ "step": 216000
2617
+ },
2618
+ {
2619
+ "epoch": 2.16,
2620
+ "learning_rate": 1.4041084514247371e-05,
2621
+ "loss": 1.699,
2622
+ "step": 216500
2623
+ },
2624
+ {
2625
+ "epoch": 2.16,
2626
+ "learning_rate": 1.3958038520053948e-05,
2627
+ "loss": 1.7098,
2628
+ "step": 217000
2629
+ },
2630
+ {
2631
+ "epoch": 2.17,
2632
+ "learning_rate": 1.3874992525860522e-05,
2633
+ "loss": 1.6469,
2634
+ "step": 217500
2635
+ },
2636
+ {
2637
+ "epoch": 2.17,
2638
+ "learning_rate": 1.3791946531667099e-05,
2639
+ "loss": 1.6594,
2640
+ "step": 218000
2641
+ },
2642
+ {
2643
+ "epoch": 2.18,
2644
+ "learning_rate": 1.3708900537473676e-05,
2645
+ "loss": 1.6656,
2646
+ "step": 218500
2647
+ },
2648
+ {
2649
+ "epoch": 2.18,
2650
+ "learning_rate": 1.362585454328025e-05,
2651
+ "loss": 1.6813,
2652
+ "step": 219000
2653
+ },
2654
+ {
2655
+ "epoch": 2.19,
2656
+ "learning_rate": 1.3542808549086827e-05,
2657
+ "loss": 1.6124,
2658
+ "step": 219500
2659
+ },
2660
+ {
2661
+ "epoch": 2.19,
2662
+ "learning_rate": 1.3459762554893402e-05,
2663
+ "loss": 1.6395,
2664
+ "step": 220000
2665
+ },
2666
+ {
2667
+ "epoch": 2.2,
2668
+ "learning_rate": 1.3376716560699979e-05,
2669
+ "loss": 1.7241,
2670
+ "step": 220500
2671
+ },
2672
+ {
2673
+ "epoch": 2.2,
2674
+ "learning_rate": 1.3293670566506556e-05,
2675
+ "loss": 1.6459,
2676
+ "step": 221000
2677
+ },
2678
+ {
2679
+ "epoch": 2.21,
2680
+ "learning_rate": 1.321062457231313e-05,
2681
+ "loss": 1.6462,
2682
+ "step": 221500
2683
+ },
2684
+ {
2685
+ "epoch": 2.21,
2686
+ "learning_rate": 1.3127578578119706e-05,
2687
+ "loss": 1.65,
2688
+ "step": 222000
2689
+ },
2690
+ {
2691
+ "epoch": 2.22,
2692
+ "learning_rate": 1.3044532583926283e-05,
2693
+ "loss": 1.6135,
2694
+ "step": 222500
2695
+ },
2696
+ {
2697
+ "epoch": 2.22,
2698
+ "learning_rate": 1.2961486589732857e-05,
2699
+ "loss": 1.6306,
2700
+ "step": 223000
2701
+ },
2702
+ {
2703
+ "epoch": 2.23,
2704
+ "learning_rate": 1.2878440595539434e-05,
2705
+ "loss": 1.6445,
2706
+ "step": 223500
2707
+ },
2708
+ {
2709
+ "epoch": 2.23,
2710
+ "learning_rate": 1.2795394601346011e-05,
2711
+ "loss": 1.6473,
2712
+ "step": 224000
2713
+ },
2714
+ {
2715
+ "epoch": 2.24,
2716
+ "learning_rate": 1.2712348607152586e-05,
2717
+ "loss": 1.7171,
2718
+ "step": 224500
2719
+ },
2720
+ {
2721
+ "epoch": 2.24,
2722
+ "learning_rate": 1.2629302612959162e-05,
2723
+ "loss": 1.7322,
2724
+ "step": 225000
2725
+ },
2726
+ {
2727
+ "epoch": 2.25,
2728
+ "learning_rate": 1.2546256618765737e-05,
2729
+ "loss": 1.6733,
2730
+ "step": 225500
2731
+ },
2732
+ {
2733
+ "epoch": 2.25,
2734
+ "learning_rate": 1.2463210624572314e-05,
2735
+ "loss": 1.6405,
2736
+ "step": 226000
2737
+ },
2738
+ {
2739
+ "epoch": 2.26,
2740
+ "learning_rate": 1.2380164630378889e-05,
2741
+ "loss": 1.6611,
2742
+ "step": 226500
2743
+ },
2744
+ {
2745
+ "epoch": 2.26,
2746
+ "learning_rate": 1.2297118636185466e-05,
2747
+ "loss": 1.6675,
2748
+ "step": 227000
2749
+ },
2750
+ {
2751
+ "epoch": 2.27,
2752
+ "learning_rate": 1.2214072641992041e-05,
2753
+ "loss": 1.6807,
2754
+ "step": 227500
2755
+ },
2756
+ {
2757
+ "epoch": 2.27,
2758
+ "learning_rate": 1.2131026647798617e-05,
2759
+ "loss": 1.6737,
2760
+ "step": 228000
2761
+ },
2762
+ {
2763
+ "epoch": 2.28,
2764
+ "learning_rate": 1.2047980653605194e-05,
2765
+ "loss": 1.6794,
2766
+ "step": 228500
2767
+ },
2768
+ {
2769
+ "epoch": 2.28,
2770
+ "learning_rate": 1.1964934659411769e-05,
2771
+ "loss": 1.6674,
2772
+ "step": 229000
2773
+ },
2774
+ {
2775
+ "epoch": 2.29,
2776
+ "learning_rate": 1.1881888665218344e-05,
2777
+ "loss": 1.648,
2778
+ "step": 229500
2779
+ },
2780
+ {
2781
+ "epoch": 2.29,
2782
+ "learning_rate": 1.1798842671024921e-05,
2783
+ "loss": 1.6825,
2784
+ "step": 230000
2785
+ },
2786
+ {
2787
+ "epoch": 2.3,
2788
+ "learning_rate": 1.1715796676831498e-05,
2789
+ "loss": 1.6233,
2790
+ "step": 230500
2791
+ },
2792
+ {
2793
+ "epoch": 2.3,
2794
+ "learning_rate": 1.1632750682638074e-05,
2795
+ "loss": 1.669,
2796
+ "step": 231000
2797
+ },
2798
+ {
2799
+ "epoch": 2.31,
2800
+ "learning_rate": 1.1549704688444649e-05,
2801
+ "loss": 1.7087,
2802
+ "step": 231500
2803
+ },
2804
+ {
2805
+ "epoch": 2.31,
2806
+ "learning_rate": 1.1466658694251224e-05,
2807
+ "loss": 1.6436,
2808
+ "step": 232000
2809
+ },
2810
+ {
2811
+ "epoch": 2.32,
2812
+ "learning_rate": 1.1383612700057801e-05,
2813
+ "loss": 1.6433,
2814
+ "step": 232500
2815
+ },
2816
+ {
2817
+ "epoch": 2.32,
2818
+ "learning_rate": 1.1300566705864376e-05,
2819
+ "loss": 1.611,
2820
+ "step": 233000
2821
+ },
2822
+ {
2823
+ "epoch": 2.33,
2824
+ "learning_rate": 1.1217520711670952e-05,
2825
+ "loss": 1.6508,
2826
+ "step": 233500
2827
+ },
2828
+ {
2829
+ "epoch": 2.33,
2830
+ "learning_rate": 1.1134474717477527e-05,
2831
+ "loss": 1.634,
2832
+ "step": 234000
2833
+ },
2834
+ {
2835
+ "epoch": 2.34,
2836
+ "learning_rate": 1.1051428723284104e-05,
2837
+ "loss": 1.6744,
2838
+ "step": 234500
2839
+ },
2840
+ {
2841
+ "epoch": 2.34,
2842
+ "learning_rate": 1.096838272909068e-05,
2843
+ "loss": 1.6358,
2844
+ "step": 235000
2845
+ },
2846
+ {
2847
+ "epoch": 2.35,
2848
+ "learning_rate": 1.0885336734897256e-05,
2849
+ "loss": 1.6318,
2850
+ "step": 235500
2851
+ },
2852
+ {
2853
+ "epoch": 2.35,
2854
+ "learning_rate": 1.0802290740703832e-05,
2855
+ "loss": 1.6344,
2856
+ "step": 236000
2857
+ },
2858
+ {
2859
+ "epoch": 2.36,
2860
+ "learning_rate": 1.0719244746510409e-05,
2861
+ "loss": 1.6591,
2862
+ "step": 236500
2863
+ },
2864
+ {
2865
+ "epoch": 2.36,
2866
+ "learning_rate": 1.0636198752316984e-05,
2867
+ "loss": 1.6757,
2868
+ "step": 237000
2869
+ },
2870
+ {
2871
+ "epoch": 2.37,
2872
+ "learning_rate": 1.055315275812356e-05,
2873
+ "loss": 1.6628,
2874
+ "step": 237500
2875
  },
2876
  {
2877
+ "epoch": 2.37,
2878
+ "learning_rate": 1.0470106763930136e-05,
2879
+ "loss": 1.6095,
2880
+ "step": 238000
2881
  },
2882
  {
2883
+ "epoch": 2.38,
2884
+ "learning_rate": 1.0387060769736711e-05,
2885
+ "loss": 1.6678,
2886
+ "step": 238500
 
 
2887
  },
2888
  {
2889
+ "epoch": 2.38,
2890
+ "learning_rate": 1.0304014775543287e-05,
2891
+ "loss": 1.6727,
2892
+ "step": 239000
2893
  },
2894
  {
2895
+ "epoch": 2.39,
2896
+ "learning_rate": 1.0220968781349862e-05,
2897
+ "loss": 1.6702,
2898
+ "step": 239500
 
 
2899
  },
2900
  {
2901
+ "epoch": 2.39,
2902
+ "learning_rate": 1.0137922787156439e-05,
2903
+ "loss": 1.7455,
2904
+ "step": 240000
2905
  },
2906
  {
2907
+ "epoch": 2.4,
2908
+ "learning_rate": 1.0054876792963016e-05,
2909
+ "loss": 1.604,
2910
+ "step": 240500
 
 
2911
  },
2912
  {
2913
+ "epoch": 2.4,
2914
+ "learning_rate": 9.971830798769591e-06,
2915
+ "loss": 1.6657,
2916
+ "step": 241000
2917
  },
2918
  {
2919
+ "epoch": 2.41,
2920
+ "learning_rate": 9.888784804576167e-06,
2921
+ "loss": 1.6505,
2922
+ "step": 241500
 
 
2923
  },
2924
  {
2925
+ "epoch": 2.41,
2926
+ "learning_rate": 9.805738810382744e-06,
2927
+ "loss": 1.6435,
2928
+ "step": 242000
2929
  },
2930
  {
2931
+ "epoch": 2.42,
2932
+ "learning_rate": 9.722692816189319e-06,
2933
+ "loss": 1.6541,
2934
+ "step": 242500
 
 
2935
  },
2936
  {
2937
+ "epoch": 2.42,
2938
+ "learning_rate": 9.639646821995894e-06,
2939
+ "loss": 1.6212,
2940
+ "step": 243000
2941
  },
2942
  {
2943
+ "epoch": 2.43,
2944
+ "learning_rate": 9.55660082780247e-06,
2945
+ "loss": 1.6384,
2946
+ "step": 243500
 
 
2947
  },
2948
  {
2949
+ "epoch": 2.43,
2950
+ "learning_rate": 9.473554833609047e-06,
2951
+ "loss": 1.5894,
2952
+ "step": 244000
2953
  },
2954
  {
2955
+ "epoch": 2.44,
2956
+ "learning_rate": 9.390508839415622e-06,
2957
+ "loss": 1.6558,
2958
+ "step": 244500
 
 
2959
  },
2960
  {
2961
+ "epoch": 2.44,
2962
+ "learning_rate": 9.307462845222197e-06,
2963
+ "loss": 1.6277,
2964
+ "step": 245000
2965
  },
2966
  {
2967
+ "epoch": 2.45,
2968
+ "learning_rate": 9.224416851028774e-06,
2969
+ "loss": 1.6418,
2970
+ "step": 245500
 
 
2971
  },
2972
  {
2973
+ "epoch": 2.45,
2974
+ "learning_rate": 9.141370856835351e-06,
2975
+ "loss": 1.5842,
2976
+ "step": 246000
2977
  },
2978
  {
2979
+ "epoch": 2.46,
2980
+ "learning_rate": 9.058324862641926e-06,
2981
+ "loss": 1.6372,
2982
+ "step": 246500
 
 
2983
  },
2984
  {
2985
+ "epoch": 2.46,
2986
+ "learning_rate": 8.975278868448502e-06,
2987
+ "loss": 1.6669,
2988
+ "step": 247000
2989
  },
2990
  {
2991
+ "epoch": 2.47,
2992
+ "learning_rate": 8.892232874255079e-06,
2993
+ "loss": 1.617,
2994
+ "step": 247500
2995
+ },
2996
+ {
2997
+ "epoch": 2.47,
2998
+ "learning_rate": 8.809186880061654e-06,
2999
+ "loss": 1.6377,
3000
+ "step": 248000
3001
+ },
3002
+ {
3003
+ "epoch": 2.48,
3004
+ "learning_rate": 8.72614088586823e-06,
3005
+ "loss": 1.6603,
3006
+ "step": 248500
3007
+ },
3008
+ {
3009
+ "epoch": 2.48,
3010
+ "learning_rate": 8.643094891674805e-06,
3011
+ "loss": 1.6087,
3012
+ "step": 249000
3013
+ },
3014
+ {
3015
+ "epoch": 2.49,
3016
+ "learning_rate": 8.560048897481382e-06,
3017
+ "loss": 1.6408,
3018
+ "step": 249500
3019
+ },
3020
+ {
3021
+ "epoch": 2.49,
3022
+ "learning_rate": 8.477002903287957e-06,
3023
+ "loss": 1.6131,
3024
+ "step": 250000
3025
+ },
3026
+ {
3027
+ "epoch": 2.5,
3028
+ "learning_rate": 8.393956909094534e-06,
3029
+ "loss": 1.662,
3030
+ "step": 250500
3031
+ },
3032
+ {
3033
+ "epoch": 2.5,
3034
+ "learning_rate": 8.310910914901109e-06,
3035
+ "loss": 1.635,
3036
+ "step": 251000
3037
+ },
3038
+ {
3039
+ "epoch": 2.51,
3040
+ "learning_rate": 8.227864920707686e-06,
3041
+ "loss": 1.655,
3042
+ "step": 251500
3043
+ },
3044
+ {
3045
+ "epoch": 2.51,
3046
+ "learning_rate": 8.144818926514261e-06,
3047
+ "loss": 1.609,
3048
+ "step": 252000
3049
+ },
3050
+ {
3051
+ "epoch": 2.52,
3052
+ "learning_rate": 8.061772932320837e-06,
3053
+ "loss": 1.6342,
3054
+ "step": 252500
3055
+ },
3056
+ {
3057
+ "epoch": 2.52,
3058
+ "learning_rate": 7.978726938127414e-06,
3059
+ "loss": 1.6073,
3060
+ "step": 253000
3061
+ },
3062
+ {
3063
+ "epoch": 2.53,
3064
+ "learning_rate": 7.895680943933989e-06,
3065
+ "loss": 1.6242,
3066
+ "step": 253500
3067
+ },
3068
+ {
3069
+ "epoch": 2.53,
3070
+ "learning_rate": 7.812634949740564e-06,
3071
+ "loss": 1.5926,
3072
+ "step": 254000
3073
+ },
3074
+ {
3075
+ "epoch": 2.54,
3076
+ "learning_rate": 7.72958895554714e-06,
3077
+ "loss": 1.5771,
3078
+ "step": 254500
3079
+ },
3080
+ {
3081
+ "epoch": 2.54,
3082
+ "learning_rate": 7.646542961353717e-06,
3083
+ "loss": 1.6155,
3084
+ "step": 255000
3085
+ },
3086
+ {
3087
+ "epoch": 2.55,
3088
+ "learning_rate": 7.563496967160293e-06,
3089
+ "loss": 1.6359,
3090
+ "step": 255500
3091
+ },
3092
+ {
3093
+ "epoch": 2.55,
3094
+ "learning_rate": 7.480450972966868e-06,
3095
+ "loss": 1.6789,
3096
+ "step": 256000
3097
+ },
3098
+ {
3099
+ "epoch": 2.56,
3100
+ "learning_rate": 7.397404978773443e-06,
3101
+ "loss": 1.6543,
3102
+ "step": 256500
3103
+ },
3104
+ {
3105
+ "epoch": 2.56,
3106
+ "learning_rate": 7.31435898458002e-06,
3107
+ "loss": 1.6069,
3108
+ "step": 257000
3109
  },
3110
  {
3111
  "epoch": 2.57,
3112
+ "learning_rate": 7.2313129903865964e-06,
3113
+ "loss": 1.633,
3114
+ "step": 257500
3115
  },
3116
  {
3117
  "epoch": 2.57,
3118
+ "learning_rate": 7.148266996193172e-06,
3119
+ "loss": 1.5973,
3120
+ "step": 258000
3121
+ },
3122
+ {
3123
+ "epoch": 2.58,
3124
+ "learning_rate": 7.065221001999747e-06,
3125
+ "loss": 1.6708,
3126
+ "step": 258500
3127
+ },
3128
+ {
3129
+ "epoch": 2.58,
3130
+ "learning_rate": 6.982175007806324e-06,
3131
+ "loss": 1.5986,
3132
+ "step": 259000
3133
+ },
3134
+ {
3135
+ "epoch": 2.59,
3136
+ "learning_rate": 6.899129013612899e-06,
3137
+ "loss": 1.6398,
3138
+ "step": 259500
3139
+ },
3140
+ {
3141
+ "epoch": 2.59,
3142
+ "learning_rate": 6.8160830194194755e-06,
3143
+ "loss": 1.6438,
3144
+ "step": 260000
3145
+ },
3146
+ {
3147
+ "epoch": 2.6,
3148
+ "learning_rate": 6.7330370252260525e-06,
3149
+ "loss": 1.6719,
3150
+ "step": 260500
3151
+ },
3152
+ {
3153
+ "epoch": 2.6,
3154
+ "learning_rate": 6.649991031032628e-06,
3155
+ "loss": 1.6032,
3156
+ "step": 261000
3157
+ },
3158
+ {
3159
+ "epoch": 2.61,
3160
+ "learning_rate": 6.566945036839203e-06,
3161
+ "loss": 1.6564,
3162
+ "step": 261500
3163
+ },
3164
+ {
3165
+ "epoch": 2.61,
3166
+ "learning_rate": 6.483899042645779e-06,
3167
+ "loss": 1.5988,
3168
+ "step": 262000
3169
+ },
3170
+ {
3171
+ "epoch": 2.62,
3172
+ "learning_rate": 6.400853048452355e-06,
3173
+ "loss": 1.5879,
3174
+ "step": 262500
3175
+ },
3176
+ {
3177
+ "epoch": 2.62,
3178
+ "learning_rate": 6.3178070542589315e-06,
3179
+ "loss": 1.6442,
3180
+ "step": 263000
3181
+ },
3182
+ {
3183
+ "epoch": 2.63,
3184
+ "learning_rate": 6.234761060065507e-06,
3185
+ "loss": 1.664,
3186
+ "step": 263500
3187
+ },
3188
+ {
3189
+ "epoch": 2.63,
3190
+ "learning_rate": 6.151715065872083e-06,
3191
+ "loss": 1.622,
3192
+ "step": 264000
3193
+ },
3194
+ {
3195
+ "epoch": 2.64,
3196
+ "learning_rate": 6.068669071678658e-06,
3197
+ "loss": 1.6038,
3198
+ "step": 264500
3199
+ },
3200
+ {
3201
+ "epoch": 2.64,
3202
+ "learning_rate": 5.985623077485235e-06,
3203
+ "loss": 1.6131,
3204
+ "step": 265000
3205
+ },
3206
+ {
3207
+ "epoch": 2.65,
3208
+ "learning_rate": 5.9025770832918105e-06,
3209
+ "loss": 1.6065,
3210
+ "step": 265500
3211
+ },
3212
+ {
3213
+ "epoch": 2.65,
3214
+ "learning_rate": 5.819531089098387e-06,
3215
+ "loss": 1.6532,
3216
+ "step": 266000
3217
+ },
3218
+ {
3219
+ "epoch": 2.66,
3220
+ "learning_rate": 5.736485094904962e-06,
3221
+ "loss": 1.6297,
3222
+ "step": 266500
3223
+ },
3224
+ {
3225
+ "epoch": 2.66,
3226
+ "learning_rate": 5.653439100711538e-06,
3227
+ "loss": 1.6374,
3228
+ "step": 267000
3229
+ },
3230
+ {
3231
+ "epoch": 2.67,
3232
+ "learning_rate": 5.570393106518114e-06,
3233
+ "loss": 1.6533,
3234
+ "step": 267500
3235
+ },
3236
+ {
3237
+ "epoch": 2.67,
3238
+ "learning_rate": 5.48734711232469e-06,
3239
+ "loss": 1.6394,
3240
+ "step": 268000
3241
+ },
3242
+ {
3243
+ "epoch": 2.68,
3244
+ "learning_rate": 5.4043011181312665e-06,
3245
+ "loss": 1.6146,
3246
+ "step": 268500
3247
+ },
3248
+ {
3249
+ "epoch": 2.68,
3250
+ "learning_rate": 5.321255123937842e-06,
3251
+ "loss": 1.6221,
3252
+ "step": 269000
3253
+ },
3254
+ {
3255
+ "epoch": 2.69,
3256
+ "learning_rate": 5.238209129744418e-06,
3257
+ "loss": 1.5852,
3258
+ "step": 269500
3259
+ },
3260
+ {
3261
+ "epoch": 2.69,
3262
+ "learning_rate": 5.155163135550994e-06,
3263
+ "loss": 1.5987,
3264
+ "step": 270000
3265
+ },
3266
+ {
3267
+ "epoch": 2.7,
3268
+ "learning_rate": 5.07211714135757e-06,
3269
+ "loss": 1.622,
3270
+ "step": 270500
3271
+ },
3272
+ {
3273
+ "epoch": 2.7,
3274
+ "learning_rate": 4.9890711471641455e-06,
3275
+ "loss": 1.6285,
3276
+ "step": 271000
3277
+ },
3278
+ {
3279
+ "epoch": 2.71,
3280
+ "learning_rate": 4.906025152970722e-06,
3281
+ "loss": 1.5902,
3282
+ "step": 271500
3283
+ },
3284
+ {
3285
+ "epoch": 2.71,
3286
+ "learning_rate": 4.822979158777297e-06,
3287
+ "loss": 1.6129,
3288
+ "step": 272000
3289
+ },
3290
+ {
3291
+ "epoch": 2.72,
3292
+ "learning_rate": 4.739933164583873e-06,
3293
+ "loss": 1.6248,
3294
+ "step": 272500
3295
+ },
3296
+ {
3297
+ "epoch": 2.72,
3298
+ "learning_rate": 4.656887170390449e-06,
3299
+ "loss": 1.5923,
3300
+ "step": 273000
3301
+ },
3302
+ {
3303
+ "epoch": 2.73,
3304
+ "learning_rate": 4.573841176197025e-06,
3305
+ "loss": 1.5975,
3306
+ "step": 273500
3307
+ },
3308
+ {
3309
+ "epoch": 2.73,
3310
+ "learning_rate": 4.490795182003601e-06,
3311
+ "loss": 1.5937,
3312
+ "step": 274000
3313
+ },
3314
+ {
3315
+ "epoch": 2.74,
3316
+ "learning_rate": 4.407749187810177e-06,
3317
+ "loss": 1.5886,
3318
+ "step": 274500
3319
+ },
3320
+ {
3321
+ "epoch": 2.74,
3322
+ "learning_rate": 4.324703193616753e-06,
3323
+ "loss": 1.6767,
3324
+ "step": 275000
3325
+ },
3326
+ {
3327
+ "epoch": 2.75,
3328
+ "learning_rate": 4.241657199423329e-06,
3329
+ "loss": 1.5599,
3330
+ "step": 275500
3331
+ },
3332
+ {
3333
+ "epoch": 2.75,
3334
+ "learning_rate": 4.158611205229905e-06,
3335
+ "loss": 1.6116,
3336
+ "step": 276000
3337
+ },
3338
+ {
3339
+ "epoch": 2.76,
3340
+ "learning_rate": 4.0755652110364805e-06,
3341
+ "loss": 1.6349,
3342
+ "step": 276500
3343
+ },
3344
+ {
3345
+ "epoch": 2.76,
3346
+ "learning_rate": 3.992519216843057e-06,
3347
+ "loss": 1.6263,
3348
+ "step": 277000
3349
+ },
3350
+ {
3351
+ "epoch": 2.77,
3352
+ "learning_rate": 3.909473222649632e-06,
3353
+ "loss": 1.5866,
3354
+ "step": 277500
3355
+ },
3356
+ {
3357
+ "epoch": 2.77,
3358
+ "learning_rate": 3.826427228456209e-06,
3359
+ "loss": 1.5412,
3360
+ "step": 278000
3361
+ },
3362
+ {
3363
+ "epoch": 2.78,
3364
+ "learning_rate": 3.7433812342627843e-06,
3365
+ "loss": 1.5899,
3366
+ "step": 278500
3367
+ },
3368
+ {
3369
+ "epoch": 2.78,
3370
+ "learning_rate": 3.6603352400693604e-06,
3371
+ "loss": 1.5569,
3372
+ "step": 279000
3373
  },
3374
  {
3375
  "epoch": 2.79,
3376
+ "learning_rate": 3.5772892458759357e-06,
3377
+ "loss": 1.5831,
3378
+ "step": 279500
3379
  },
3380
  {
3381
  "epoch": 2.79,
3382
+ "learning_rate": 3.4942432516825123e-06,
3383
+ "loss": 1.6188,
3384
+ "step": 280000
3385
+ },
3386
+ {
3387
+ "epoch": 2.8,
3388
+ "learning_rate": 3.4111972574890876e-06,
3389
+ "loss": 1.6345,
3390
+ "step": 280500
3391
+ },
3392
+ {
3393
+ "epoch": 2.8,
3394
+ "learning_rate": 3.328151263295664e-06,
3395
+ "loss": 1.6034,
3396
+ "step": 281000
3397
+ },
3398
+ {
3399
+ "epoch": 2.81,
3400
+ "learning_rate": 3.2451052691022394e-06,
3401
+ "loss": 1.6262,
3402
+ "step": 281500
3403
+ },
3404
+ {
3405
+ "epoch": 2.81,
3406
+ "learning_rate": 3.1620592749088156e-06,
3407
+ "loss": 1.6173,
3408
+ "step": 282000
3409
+ },
3410
+ {
3411
+ "epoch": 2.82,
3412
+ "learning_rate": 3.0790132807153917e-06,
3413
+ "loss": 1.6379,
3414
+ "step": 282500
3415
+ },
3416
+ {
3417
+ "epoch": 2.82,
3418
+ "learning_rate": 2.9959672865219674e-06,
3419
+ "loss": 1.5932,
3420
+ "step": 283000
3421
+ },
3422
+ {
3423
+ "epoch": 2.83,
3424
+ "learning_rate": 2.912921292328543e-06,
3425
+ "loss": 1.599,
3426
+ "step": 283500
3427
+ },
3428
+ {
3429
+ "epoch": 2.83,
3430
+ "learning_rate": 2.8298752981351193e-06,
3431
+ "loss": 1.6222,
3432
+ "step": 284000
3433
+ },
3434
+ {
3435
+ "epoch": 2.84,
3436
+ "learning_rate": 2.746829303941695e-06,
3437
+ "loss": 1.6023,
3438
+ "step": 284500
3439
+ },
3440
+ {
3441
+ "epoch": 2.84,
3442
+ "learning_rate": 2.663783309748271e-06,
3443
+ "loss": 1.6039,
3444
+ "step": 285000
3445
+ },
3446
+ {
3447
+ "epoch": 2.85,
3448
+ "learning_rate": 2.580737315554847e-06,
3449
+ "loss": 1.5626,
3450
+ "step": 285500
3451
+ },
3452
+ {
3453
+ "epoch": 2.85,
3454
+ "learning_rate": 2.4976913213614226e-06,
3455
+ "loss": 1.6007,
3456
+ "step": 286000
3457
+ },
3458
+ {
3459
+ "epoch": 2.86,
3460
+ "learning_rate": 2.4146453271679987e-06,
3461
+ "loss": 1.6075,
3462
+ "step": 286500
3463
+ },
3464
+ {
3465
+ "epoch": 2.86,
3466
+ "learning_rate": 2.3315993329745745e-06,
3467
+ "loss": 1.6126,
3468
+ "step": 287000
3469
+ },
3470
+ {
3471
+ "epoch": 2.87,
3472
+ "learning_rate": 2.248553338781151e-06,
3473
+ "loss": 1.5963,
3474
+ "step": 287500
3475
+ },
3476
+ {
3477
+ "epoch": 2.87,
3478
+ "learning_rate": 2.1655073445877267e-06,
3479
+ "loss": 1.5668,
3480
+ "step": 288000
3481
+ },
3482
+ {
3483
+ "epoch": 2.88,
3484
+ "learning_rate": 2.0824613503943025e-06,
3485
+ "loss": 1.564,
3486
+ "step": 288500
3487
+ },
3488
+ {
3489
+ "epoch": 2.88,
3490
+ "learning_rate": 1.9994153562008786e-06,
3491
+ "loss": 1.6204,
3492
+ "step": 289000
3493
+ },
3494
+ {
3495
+ "epoch": 2.89,
3496
+ "learning_rate": 1.9163693620074543e-06,
3497
+ "loss": 1.5741,
3498
+ "step": 289500
3499
+ },
3500
+ {
3501
+ "epoch": 2.89,
3502
+ "learning_rate": 1.8333233678140303e-06,
3503
+ "loss": 1.608,
3504
+ "step": 290000
3505
+ },
3506
+ {
3507
+ "epoch": 2.89,
3508
+ "learning_rate": 1.7502773736206062e-06,
3509
+ "loss": 1.5894,
3510
+ "step": 290500
3511
+ },
3512
+ {
3513
+ "epoch": 2.9,
3514
+ "learning_rate": 1.6672313794271821e-06,
3515
+ "loss": 1.5644,
3516
+ "step": 291000
3517
+ },
3518
+ {
3519
+ "epoch": 2.9,
3520
+ "learning_rate": 1.5841853852337578e-06,
3521
+ "loss": 1.5474,
3522
+ "step": 291500
3523
+ },
3524
+ {
3525
+ "epoch": 2.91,
3526
+ "learning_rate": 1.5011393910403338e-06,
3527
+ "loss": 1.6177,
3528
+ "step": 292000
3529
+ },
3530
+ {
3531
+ "epoch": 2.91,
3532
+ "learning_rate": 1.4180933968469097e-06,
3533
+ "loss": 1.579,
3534
+ "step": 292500
3535
+ },
3536
+ {
3537
+ "epoch": 2.92,
3538
+ "learning_rate": 1.3350474026534856e-06,
3539
+ "loss": 1.5752,
3540
+ "step": 293000
3541
+ },
3542
+ {
3543
+ "epoch": 2.92,
3544
+ "learning_rate": 1.2520014084600616e-06,
3545
+ "loss": 1.5702,
3546
+ "step": 293500
3547
+ },
3548
+ {
3549
+ "epoch": 2.93,
3550
+ "learning_rate": 1.1689554142666375e-06,
3551
+ "loss": 1.5727,
3552
+ "step": 294000
3553
+ },
3554
+ {
3555
+ "epoch": 2.93,
3556
+ "learning_rate": 1.0859094200732134e-06,
3557
+ "loss": 1.695,
3558
+ "step": 294500
3559
+ },
3560
+ {
3561
+ "epoch": 2.94,
3562
+ "learning_rate": 1.0028634258797894e-06,
3563
+ "loss": 1.608,
3564
+ "step": 295000
3565
+ },
3566
+ {
3567
+ "epoch": 2.94,
3568
+ "learning_rate": 9.198174316863652e-07,
3569
+ "loss": 1.6115,
3570
+ "step": 295500
3571
+ },
3572
+ {
3573
+ "epoch": 2.95,
3574
+ "learning_rate": 8.367714374929411e-07,
3575
+ "loss": 1.6197,
3576
+ "step": 296000
3577
+ },
3578
+ {
3579
+ "epoch": 2.95,
3580
+ "learning_rate": 7.53725443299517e-07,
3581
+ "loss": 1.5447,
3582
+ "step": 296500
3583
+ },
3584
+ {
3585
+ "epoch": 2.96,
3586
+ "learning_rate": 6.70679449106093e-07,
3587
+ "loss": 1.6058,
3588
+ "step": 297000
3589
+ },
3590
+ {
3591
+ "epoch": 2.96,
3592
+ "learning_rate": 5.876334549126688e-07,
3593
+ "loss": 1.6225,
3594
+ "step": 297500
3595
+ },
3596
+ {
3597
+ "epoch": 2.97,
3598
+ "learning_rate": 5.045874607192447e-07,
3599
+ "loss": 1.5389,
3600
+ "step": 298000
3601
+ },
3602
+ {
3603
+ "epoch": 2.97,
3604
+ "learning_rate": 4.215414665258207e-07,
3605
+ "loss": 1.5693,
3606
+ "step": 298500
3607
+ },
3608
+ {
3609
+ "epoch": 2.98,
3610
+ "learning_rate": 3.384954723323966e-07,
3611
+ "loss": 1.5898,
3612
+ "step": 299000
3613
+ },
3614
+ {
3615
+ "epoch": 2.98,
3616
+ "learning_rate": 2.554494781389725e-07,
3617
+ "loss": 1.6425,
3618
+ "step": 299500
3619
+ },
3620
+ {
3621
+ "epoch": 2.99,
3622
+ "learning_rate": 1.724034839455484e-07,
3623
+ "loss": 1.6377,
3624
+ "step": 300000
3625
+ },
3626
+ {
3627
+ "epoch": 2.99,
3628
+ "learning_rate": 8.935748975212432e-08,
3629
+ "loss": 1.5687,
3630
+ "step": 300500
3631
+ },
3632
+ {
3633
+ "epoch": 3.0,
3634
+ "learning_rate": 6.3114955587002305e-09,
3635
+ "loss": 1.5879,
3636
+ "step": 301000
3637
+ },
3638
+ {
3639
+ "epoch": 3.0,
3640
+ "eval_loss": 1.5570189952850342,
3641
+ "eval_runtime": 63.1399,
3642
+ "eval_samples_per_second": 669.165,
3643
+ "eval_steps_per_second": 83.655,
3644
+ "step": 301038
3645
  },
3646
  {
3647
  "epoch": 3.0,
3648
+ "step": 301038,
3649
+ "total_flos": 5.871271452031616e+16,
3650
+ "train_loss": 1.8846093698359827,
3651
+ "train_runtime": 20632.0794,
3652
+ "train_samples_per_second": 116.725,
3653
+ "train_steps_per_second": 14.591
3654
  }
3655
  ],
3656
+ "max_steps": 301038,
3657
  "num_train_epochs": 3,
3658
+ "total_flos": 5.871271452031616e+16,
3659
  "trial_name": null,
3660
  "trial_params": null
3661
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9afce8d3d3462ca4f935b775e19a9239db8230d9e97923bdf6e43d2df813e9fb
3
- size 2927
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:57164c54e9b323c74564e3a969a7302c9a9ab39fbcce19ed9a34e82b4058c79b
3
+ size 3119