Hugging Face's logo

craa
/
exceptions

Safetensors
Model card Files
xet
Community
craa commited on
Commit
706b3e4
·
verified ·
1 Parent(s): 3fa1a9f

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. cost_to_hit_frequency_40817/README.md +129 -0
  2. cost_to_hit_frequency_40817/all_results.json +16 -0
  3. cost_to_hit_frequency_40817/checkpoint-20000/config.json +31 -0
  4. cost_to_hit_frequency_40817/checkpoint-20000/generation_config.json +6 -0
  5. cost_to_hit_frequency_40817/checkpoint-20000/merges.txt +0 -0
  6. cost_to_hit_frequency_40817/checkpoint-20000/model.safetensors +3 -0
  7. cost_to_hit_frequency_40817/checkpoint-20000/optimizer.pt +3 -0
  8. cost_to_hit_frequency_40817/checkpoint-20000/rng_state.pth +3 -0
  9. cost_to_hit_frequency_40817/checkpoint-20000/scaler.pt +3 -0
  10. cost_to_hit_frequency_40817/checkpoint-20000/scheduler.pt +3 -0
  11. cost_to_hit_frequency_40817/checkpoint-20000/special_tokens_map.json +5 -0
  12. cost_to_hit_frequency_40817/checkpoint-20000/tokenizer.json +0 -0
  13. cost_to_hit_frequency_40817/checkpoint-20000/tokenizer_config.json +20 -0
  14. cost_to_hit_frequency_40817/checkpoint-20000/trainer_state.json +3023 -0
  15. cost_to_hit_frequency_40817/checkpoint-20000/training_args.bin +3 -0
  16. cost_to_hit_frequency_40817/checkpoint-20000/vocab.json +0 -0
  17. cost_to_hit_frequency_40817/checkpoint-40000/config.json +31 -0
  18. cost_to_hit_frequency_40817/checkpoint-40000/generation_config.json +6 -0
  19. cost_to_hit_frequency_40817/checkpoint-40000/merges.txt +0 -0
  20. cost_to_hit_frequency_40817/checkpoint-40000/model.safetensors +3 -0
  21. cost_to_hit_frequency_40817/checkpoint-40000/optimizer.pt +3 -0
  22. cost_to_hit_frequency_40817/checkpoint-40000/rng_state.pth +3 -0
  23. cost_to_hit_frequency_40817/checkpoint-40000/scaler.pt +3 -0
  24. cost_to_hit_frequency_40817/checkpoint-40000/scheduler.pt +3 -0
  25. cost_to_hit_frequency_40817/checkpoint-40000/special_tokens_map.json +5 -0
  26. cost_to_hit_frequency_40817/checkpoint-40000/tokenizer.json +0 -0
  27. cost_to_hit_frequency_40817/checkpoint-40000/tokenizer_config.json +20 -0
  28. cost_to_hit_frequency_40817/checkpoint-40000/trainer_state.json +0 -0
  29. cost_to_hit_frequency_40817/checkpoint-40000/training_args.bin +3 -0
  30. cost_to_hit_frequency_40817/checkpoint-40000/vocab.json +0 -0
  31. cost_to_hit_frequency_40817/checkpoint-50000/config.json +31 -0
  32. cost_to_hit_frequency_40817/checkpoint-50000/generation_config.json +6 -0
  33. cost_to_hit_frequency_40817/checkpoint-50000/merges.txt +0 -0
  34. cost_to_hit_frequency_40817/checkpoint-50000/model.safetensors +3 -0
  35. cost_to_hit_frequency_40817/checkpoint-50000/optimizer.pt +3 -0
  36. cost_to_hit_frequency_40817/checkpoint-50000/rng_state.pth +3 -0
  37. cost_to_hit_frequency_40817/checkpoint-50000/scaler.pt +3 -0
  38. cost_to_hit_frequency_40817/checkpoint-50000/scheduler.pt +3 -0
  39. cost_to_hit_frequency_40817/checkpoint-50000/special_tokens_map.json +5 -0
  40. cost_to_hit_frequency_40817/checkpoint-50000/tokenizer.json +0 -0
  41. cost_to_hit_frequency_40817/checkpoint-50000/tokenizer_config.json +20 -0
  42. cost_to_hit_frequency_40817/checkpoint-50000/trainer_state.json +0 -0
  43. cost_to_hit_frequency_40817/checkpoint-50000/training_args.bin +3 -0
  44. cost_to_hit_frequency_40817/checkpoint-50000/vocab.json +0 -0
  45. cost_to_hit_frequency_40817/checkpoint-60000/config.json +31 -0
  46. cost_to_hit_frequency_40817/checkpoint-60000/generation_config.json +6 -0
  47. cost_to_hit_frequency_40817/checkpoint-60000/merges.txt +0 -0
  48. cost_to_hit_frequency_40817/checkpoint-60000/model.safetensors +3 -0
  49. cost_to_hit_frequency_40817/checkpoint-60000/optimizer.pt +3 -0
  50. cost_to_hit_frequency_40817/checkpoint-60000/rng_state.pth +3 -0
cost_to_hit_frequency_40817/README.md ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - accuracy
7
+ model-index:
8
+ - name: exceptions_exp2_cost_to_hit_frequency_40817
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/craaaa/exceptions_exp2/runs/kdm13v9q)
16
+ # exceptions_exp2_cost_to_hit_frequency_40817
17
+
18
+ This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
19
+ It achieves the following results on the evaluation set:
20
+ - Loss: 3.6227
21
+ - Accuracy: 0.3621
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 0.0006
41
+ - train_batch_size: 16
42
+ - eval_batch_size: 16
43
+ - seed: 40817
44
+ - gradient_accumulation_steps: 5
45
+ - total_train_batch_size: 80
46
+ - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
47
+ - lr_scheduler_type: linear
48
+ - lr_scheduler_warmup_steps: 100
49
+ - num_epochs: 20.0
50
+ - mixed_precision_training: Native AMP
51
+
52
+ ### Training results
53
+
54
+ | Training Loss | Epoch | Step | Validation Loss | Accuracy |
55
+ |:-------------:|:-------:|:-----:|:---------------:|:--------:|
56
+ | 4.8382 | 0.3024 | 1000 | 4.7672 | 0.2525 |
57
+ | 4.3297 | 0.6047 | 2000 | 4.2839 | 0.2995 |
58
+ | 4.1437 | 0.9071 | 3000 | 4.0920 | 0.3157 |
59
+ | 3.9999 | 1.2092 | 4000 | 3.9848 | 0.3256 |
60
+ | 3.9251 | 1.5116 | 5000 | 3.9090 | 0.3324 |
61
+ | 3.8666 | 1.8140 | 6000 | 3.8535 | 0.3376 |
62
+ | 3.748 | 2.1161 | 7000 | 3.8154 | 0.3413 |
63
+ | 3.7526 | 2.4185 | 8000 | 3.7824 | 0.3448 |
64
+ | 3.7241 | 2.7209 | 9000 | 3.7498 | 0.3476 |
65
+ | 3.5917 | 3.0230 | 10000 | 3.7303 | 0.3500 |
66
+ | 3.6298 | 3.3254 | 11000 | 3.7141 | 0.3517 |
67
+ | 3.6253 | 3.6277 | 12000 | 3.6941 | 0.3541 |
68
+ | 3.6323 | 3.9301 | 13000 | 3.6762 | 0.3555 |
69
+ | 3.5399 | 4.2322 | 14000 | 3.6729 | 0.3564 |
70
+ | 3.5538 | 4.5346 | 15000 | 3.6593 | 0.3578 |
71
+ | 3.5512 | 4.8370 | 16000 | 3.6441 | 0.3590 |
72
+ | 3.4722 | 5.1391 | 17000 | 3.6467 | 0.3597 |
73
+ | 3.5072 | 5.4415 | 18000 | 3.6349 | 0.3605 |
74
+ | 3.501 | 5.7438 | 19000 | 3.6229 | 0.3618 |
75
+ | 3.411 | 6.0460 | 20000 | 3.6227 | 0.3621 |
76
+ | 3.4363 | 6.3483 | 21000 | 3.6190 | 0.3628 |
77
+ | 3.4492 | 6.6507 | 22000 | 3.6093 | 0.3639 |
78
+ | 3.4632 | 6.9531 | 23000 | 3.5975 | 0.3648 |
79
+ | 3.3846 | 7.2552 | 24000 | 3.6061 | 0.3648 |
80
+ | 3.4205 | 7.5576 | 25000 | 3.5999 | 0.3655 |
81
+ | 3.4231 | 7.8599 | 26000 | 3.5885 | 0.3661 |
82
+ | 3.349 | 8.1621 | 27000 | 3.5982 | 0.3660 |
83
+ | 3.3654 | 8.4644 | 28000 | 3.5887 | 0.3667 |
84
+ | 3.3779 | 8.7668 | 29000 | 3.5791 | 0.3674 |
85
+ | 3.2906 | 9.0689 | 30000 | 3.5892 | 0.3672 |
86
+ | 3.3281 | 9.3713 | 31000 | 3.5846 | 0.3679 |
87
+ | 3.3594 | 9.6737 | 32000 | 3.5767 | 0.3686 |
88
+ | 3.3589 | 9.9761 | 33000 | 3.5681 | 0.3691 |
89
+ | 3.2886 | 10.2782 | 34000 | 3.5810 | 0.3684 |
90
+ | 3.3164 | 10.5806 | 35000 | 3.5735 | 0.3693 |
91
+ | 3.333 | 10.8829 | 36000 | 3.5634 | 0.3698 |
92
+ | 3.2511 | 11.1851 | 37000 | 3.5760 | 0.3694 |
93
+ | 3.2867 | 11.4874 | 38000 | 3.5710 | 0.3700 |
94
+ | 3.2981 | 11.7898 | 39000 | 3.5618 | 0.3704 |
95
+ | 3.2113 | 12.0919 | 40000 | 3.5728 | 0.3702 |
96
+ | 3.2584 | 12.3943 | 41000 | 3.5681 | 0.3705 |
97
+ | 3.2717 | 12.6967 | 42000 | 3.5603 | 0.3713 |
98
+ | 3.2805 | 12.9990 | 43000 | 3.5555 | 0.3715 |
99
+ | 3.2124 | 13.3012 | 44000 | 3.5677 | 0.3712 |
100
+ | 3.2352 | 13.6035 | 45000 | 3.5596 | 0.3716 |
101
+ | 3.2477 | 13.9059 | 46000 | 3.5575 | 0.3720 |
102
+ | 3.1954 | 14.2080 | 47000 | 3.5643 | 0.3716 |
103
+ | 3.2191 | 14.5104 | 48000 | 3.5579 | 0.3722 |
104
+ | 3.2221 | 14.8128 | 49000 | 3.5516 | 0.3726 |
105
+ | 3.1676 | 15.1149 | 50000 | 3.5629 | 0.3722 |
106
+ | 3.1728 | 15.4173 | 51000 | 3.5584 | 0.3727 |
107
+ | 3.183 | 15.7196 | 52000 | 3.5534 | 0.3728 |
108
+ | 3.1125 | 16.0218 | 53000 | 3.5569 | 0.3731 |
109
+ | 3.1497 | 16.3241 | 54000 | 3.5563 | 0.3730 |
110
+ | 3.1772 | 16.6265 | 55000 | 3.5500 | 0.3737 |
111
+ | 3.1714 | 16.9289 | 56000 | 3.5468 | 0.3739 |
112
+ | 3.1313 | 17.2310 | 57000 | 3.5568 | 0.3734 |
113
+ | 3.149 | 17.5334 | 58000 | 3.5513 | 0.3739 |
114
+ | 3.1502 | 17.8358 | 59000 | 3.5477 | 0.3743 |
115
+ | 3.0986 | 18.1379 | 60000 | 3.5537 | 0.3741 |
116
+ | 3.1118 | 18.4403 | 61000 | 3.5528 | 0.3743 |
117
+ | 3.1232 | 18.7426 | 62000 | 3.5491 | 0.3744 |
118
+ | 3.0938 | 19.0448 | 63000 | 3.5510 | 0.3746 |
119
+ | 3.0827 | 19.3471 | 64000 | 3.5510 | 0.3746 |
120
+ | 3.0918 | 19.6495 | 65000 | 3.5489 | 0.3748 |
121
+ | 3.1018 | 19.9519 | 66000 | 3.5475 | 0.3749 |
122
+
123
+
124
+ ### Framework versions
125
+
126
+ - Transformers 4.55.2
127
+ - Pytorch 2.8.0+cu128
128
+ - Datasets 4.0.0
129
+ - Tokenizers 0.21.4
cost_to_hit_frequency_40817/all_results.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 20.0,
3
+ "eval_accuracy": 0.3621478635601953,
4
+ "eval_loss": 3.6226511001586914,
5
+ "eval_runtime": 181.8581,
6
+ "eval_samples": 16644,
7
+ "eval_samples_per_second": 91.522,
8
+ "eval_steps_per_second": 5.724,
9
+ "perplexity": 37.43668478047171,
10
+ "total_flos": 1.38263201316864e+18,
11
+ "train_loss": 3.4285881825485136,
12
+ "train_runtime": 132302.3346,
13
+ "train_samples": 264576,
14
+ "train_samples_per_second": 39.996,
15
+ "train_steps_per_second": 0.5
16
+ }
cost_to_hit_frequency_40817/checkpoint-20000/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "embd_pdrop": 0.1,
9
+ "eos_token_id": 50256,
10
+ "initializer_range": 0.02,
11
+ "layer_norm_epsilon": 1e-05,
12
+ "model_type": "gpt2",
13
+ "n_embd": 768,
14
+ "n_head": 12,
15
+ "n_inner": null,
16
+ "n_layer": 12,
17
+ "n_positions": 1024,
18
+ "reorder_and_upcast_attn": false,
19
+ "resid_pdrop": 0.1,
20
+ "scale_attn_by_inverse_layer_idx": false,
21
+ "scale_attn_weights": true,
22
+ "summary_activation": null,
23
+ "summary_first_dropout": 0.1,
24
+ "summary_proj_to_labels": true,
25
+ "summary_type": "cls_index",
26
+ "summary_use_proj": true,
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.55.2",
29
+ "use_cache": true,
30
+ "vocab_size": 50257
31
+ }
cost_to_hit_frequency_40817/checkpoint-20000/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.55.2"
6
+ }
cost_to_hit_frequency_40817/checkpoint-20000/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-20000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e93b84ac4a3e7c608b049acb5db8ed818af38f77a8ea8aeefdc0fba8743353a2
3
+ size 497774208
cost_to_hit_frequency_40817/checkpoint-20000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2864c9d4d482e1ba5e1adc9e4e77e1f1d333fda095a59e90c775014f6ac3b15b
3
+ size 995644811
cost_to_hit_frequency_40817/checkpoint-20000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:47d030e407e31192206727a4894006fe845234e06fd43b84c7e9c0656f622294
3
+ size 14581
cost_to_hit_frequency_40817/checkpoint-20000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbe0baec0b603be43d431555809b412ae8dc73a7d567f7a7d9f1be83e9398304
3
+ size 1383
cost_to_hit_frequency_40817/checkpoint-20000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9676c46215067dd7920c9040f12c45bab7b9accf31d8b3f20dd245c46329d03
3
+ size 1465
cost_to_hit_frequency_40817/checkpoint-20000/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
cost_to_hit_frequency_40817/checkpoint-20000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-20000/tokenizer_config.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "50256": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ }
12
+ },
13
+ "bos_token": "<|endoftext|>",
14
+ "clean_up_tokenization_spaces": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "extra_special_tokens": {},
17
+ "model_max_length": 1024,
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
cost_to_hit_frequency_40817/checkpoint-20000/trainer_state.json ADDED
@@ -0,0 +1,3023 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 20000,
3
+ "best_metric": 3.6226511001586914,
4
+ "best_model_checkpoint": "/scratch/cl5625/exceptions/models/cost_to_hit_frequency_40817/checkpoint-20000",
5
+ "epoch": 6.045960328979197,
6
+ "eval_steps": 1000,
7
+ "global_step": 20000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.015118529269472665,
14
+ "grad_norm": 0.6569534540176392,
15
+ "learning_rate": 0.000294,
16
+ "loss": 8.48,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.03023705853894533,
21
+ "grad_norm": 0.9759527444839478,
22
+ "learning_rate": 0.0005939999999999999,
23
+ "loss": 6.7387,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.045355587808417995,
28
+ "grad_norm": 0.4822550415992737,
29
+ "learning_rate": 0.0005995549500454132,
30
+ "loss": 6.3406,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.06047411707789066,
35
+ "grad_norm": 0.4631986618041992,
36
+ "learning_rate": 0.000599100817438692,
37
+ "loss": 6.137,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.07559264634736333,
42
+ "grad_norm": 0.4041077494621277,
43
+ "learning_rate": 0.0005986466848319708,
44
+ "loss": 5.9853,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.09071117561683599,
49
+ "grad_norm": 0.3731299042701721,
50
+ "learning_rate": 0.0005981925522252497,
51
+ "loss": 5.8362,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.10582970488630866,
56
+ "grad_norm": 0.44940078258514404,
57
+ "learning_rate": 0.0005977384196185286,
58
+ "loss": 5.7301,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 0.12094823415578132,
63
+ "grad_norm": 0.4535612165927887,
64
+ "learning_rate": 0.0005972842870118073,
65
+ "loss": 5.6198,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 0.13606676342525398,
70
+ "grad_norm": 0.47181379795074463,
71
+ "learning_rate": 0.0005968301544050862,
72
+ "loss": 5.5126,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 0.15118529269472666,
77
+ "grad_norm": 0.47652333974838257,
78
+ "learning_rate": 0.0005963760217983651,
79
+ "loss": 5.4013,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 0.16630382196419932,
84
+ "grad_norm": 0.45077475905418396,
85
+ "learning_rate": 0.0005959218891916439,
86
+ "loss": 5.3135,
87
+ "step": 550
88
+ },
89
+ {
90
+ "epoch": 0.18142235123367198,
91
+ "grad_norm": 0.4728659987449646,
92
+ "learning_rate": 0.0005954677565849228,
93
+ "loss": 5.2565,
94
+ "step": 600
95
+ },
96
+ {
97
+ "epoch": 0.19654088050314467,
98
+ "grad_norm": 0.447640985250473,
99
+ "learning_rate": 0.0005950136239782016,
100
+ "loss": 5.1786,
101
+ "step": 650
102
+ },
103
+ {
104
+ "epoch": 0.21165940977261732,
105
+ "grad_norm": 0.4660065174102783,
106
+ "learning_rate": 0.0005945594913714804,
107
+ "loss": 5.1181,
108
+ "step": 700
109
+ },
110
+ {
111
+ "epoch": 0.22677793904208998,
112
+ "grad_norm": 0.4138477146625519,
113
+ "learning_rate": 0.0005941053587647593,
114
+ "loss": 5.0609,
115
+ "step": 750
116
+ },
117
+ {
118
+ "epoch": 0.24189646831156264,
119
+ "grad_norm": 0.43173298239707947,
120
+ "learning_rate": 0.0005936512261580381,
121
+ "loss": 5.0094,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 0.2570149975810353,
126
+ "grad_norm": 0.41261234879493713,
127
+ "learning_rate": 0.0005931970935513169,
128
+ "loss": 4.9896,
129
+ "step": 850
130
+ },
131
+ {
132
+ "epoch": 0.27213352685050796,
133
+ "grad_norm": 0.4491656720638275,
134
+ "learning_rate": 0.0005927429609445958,
135
+ "loss": 4.9266,
136
+ "step": 900
137
+ },
138
+ {
139
+ "epoch": 0.28725205611998067,
140
+ "grad_norm": 0.4212245047092438,
141
+ "learning_rate": 0.0005922888283378746,
142
+ "loss": 4.8791,
143
+ "step": 950
144
+ },
145
+ {
146
+ "epoch": 0.30237058538945333,
147
+ "grad_norm": 0.4332761764526367,
148
+ "learning_rate": 0.0005918346957311534,
149
+ "loss": 4.8382,
150
+ "step": 1000
151
+ },
152
+ {
153
+ "epoch": 0.30237058538945333,
154
+ "eval_accuracy": 0.2525349543872818,
155
+ "eval_loss": 4.76720666885376,
156
+ "eval_runtime": 182.6488,
157
+ "eval_samples_per_second": 91.126,
158
+ "eval_steps_per_second": 5.699,
159
+ "step": 1000
160
+ },
161
+ {
162
+ "epoch": 0.317489114658926,
163
+ "grad_norm": 0.5782724022865295,
164
+ "learning_rate": 0.0005913805631244322,
165
+ "loss": 4.7861,
166
+ "step": 1050
167
+ },
168
+ {
169
+ "epoch": 0.33260764392839864,
170
+ "grad_norm": 0.5215025544166565,
171
+ "learning_rate": 0.0005909264305177112,
172
+ "loss": 4.7449,
173
+ "step": 1100
174
+ },
175
+ {
176
+ "epoch": 0.3477261731978713,
177
+ "grad_norm": 0.5074264407157898,
178
+ "learning_rate": 0.0005904722979109899,
179
+ "loss": 4.7092,
180
+ "step": 1150
181
+ },
182
+ {
183
+ "epoch": 0.36284470246734396,
184
+ "grad_norm": 0.4659997820854187,
185
+ "learning_rate": 0.0005900181653042687,
186
+ "loss": 4.6644,
187
+ "step": 1200
188
+ },
189
+ {
190
+ "epoch": 0.3779632317368166,
191
+ "grad_norm": 0.46156764030456543,
192
+ "learning_rate": 0.0005895640326975477,
193
+ "loss": 4.6383,
194
+ "step": 1250
195
+ },
196
+ {
197
+ "epoch": 0.39308176100628933,
198
+ "grad_norm": 0.43247443437576294,
199
+ "learning_rate": 0.0005891099000908265,
200
+ "loss": 4.5939,
201
+ "step": 1300
202
+ },
203
+ {
204
+ "epoch": 0.408200290275762,
205
+ "grad_norm": 0.43695470690727234,
206
+ "learning_rate": 0.0005886557674841053,
207
+ "loss": 4.545,
208
+ "step": 1350
209
+ },
210
+ {
211
+ "epoch": 0.42331881954523465,
212
+ "grad_norm": 0.4135619103908539,
213
+ "learning_rate": 0.0005882016348773842,
214
+ "loss": 4.5432,
215
+ "step": 1400
216
+ },
217
+ {
218
+ "epoch": 0.4384373488147073,
219
+ "grad_norm": 0.3893912136554718,
220
+ "learning_rate": 0.000587747502270663,
221
+ "loss": 4.5358,
222
+ "step": 1450
223
+ },
224
+ {
225
+ "epoch": 0.45355587808417996,
226
+ "grad_norm": 0.4318796694278717,
227
+ "learning_rate": 0.0005872933696639418,
228
+ "loss": 4.5063,
229
+ "step": 1500
230
+ },
231
+ {
232
+ "epoch": 0.4686744073536526,
233
+ "grad_norm": 0.404938668012619,
234
+ "learning_rate": 0.0005868392370572207,
235
+ "loss": 4.4833,
236
+ "step": 1550
237
+ },
238
+ {
239
+ "epoch": 0.4837929366231253,
240
+ "grad_norm": 0.39407220482826233,
241
+ "learning_rate": 0.0005863851044504995,
242
+ "loss": 4.4573,
243
+ "step": 1600
244
+ },
245
+ {
246
+ "epoch": 0.498911465892598,
247
+ "grad_norm": 0.4312690496444702,
248
+ "learning_rate": 0.0005859309718437783,
249
+ "loss": 4.4493,
250
+ "step": 1650
251
+ },
252
+ {
253
+ "epoch": 0.5140299951620706,
254
+ "grad_norm": 0.42534348368644714,
255
+ "learning_rate": 0.0005854768392370572,
256
+ "loss": 4.4358,
257
+ "step": 1700
258
+ },
259
+ {
260
+ "epoch": 0.5291485244315433,
261
+ "grad_norm": 0.3989129066467285,
262
+ "learning_rate": 0.000585022706630336,
263
+ "loss": 4.4004,
264
+ "step": 1750
265
+ },
266
+ {
267
+ "epoch": 0.5442670537010159,
268
+ "grad_norm": 0.4206331968307495,
269
+ "learning_rate": 0.0005845685740236149,
270
+ "loss": 4.3903,
271
+ "step": 1800
272
+ },
273
+ {
274
+ "epoch": 0.5593855829704887,
275
+ "grad_norm": 0.39128994941711426,
276
+ "learning_rate": 0.0005841144414168936,
277
+ "loss": 4.3716,
278
+ "step": 1850
279
+ },
280
+ {
281
+ "epoch": 0.5745041122399613,
282
+ "grad_norm": 0.4127030670642853,
283
+ "learning_rate": 0.0005836603088101725,
284
+ "loss": 4.3646,
285
+ "step": 1900
286
+ },
287
+ {
288
+ "epoch": 0.589622641509434,
289
+ "grad_norm": 0.47868236899375916,
290
+ "learning_rate": 0.0005832061762034513,
291
+ "loss": 4.3474,
292
+ "step": 1950
293
+ },
294
+ {
295
+ "epoch": 0.6047411707789067,
296
+ "grad_norm": 0.41013041138648987,
297
+ "learning_rate": 0.0005827520435967303,
298
+ "loss": 4.3297,
299
+ "step": 2000
300
+ },
301
+ {
302
+ "epoch": 0.6047411707789067,
303
+ "eval_accuracy": 0.29954742363508696,
304
+ "eval_loss": 4.283868312835693,
305
+ "eval_runtime": 182.004,
306
+ "eval_samples_per_second": 91.449,
307
+ "eval_steps_per_second": 5.72,
308
+ "step": 2000
309
+ },
310
+ {
311
+ "epoch": 0.6198597000483793,
312
+ "grad_norm": 0.3787741959095001,
313
+ "learning_rate": 0.0005822979109900091,
314
+ "loss": 4.3142,
315
+ "step": 2050
316
+ },
317
+ {
318
+ "epoch": 0.634978229317852,
319
+ "grad_norm": 0.3691158890724182,
320
+ "learning_rate": 0.0005818437783832879,
321
+ "loss": 4.3103,
322
+ "step": 2100
323
+ },
324
+ {
325
+ "epoch": 0.6500967585873246,
326
+ "grad_norm": 0.38457924127578735,
327
+ "learning_rate": 0.0005813896457765666,
328
+ "loss": 4.2851,
329
+ "step": 2150
330
+ },
331
+ {
332
+ "epoch": 0.6652152878567973,
333
+ "grad_norm": 0.40187668800354004,
334
+ "learning_rate": 0.0005809355131698456,
335
+ "loss": 4.2813,
336
+ "step": 2200
337
+ },
338
+ {
339
+ "epoch": 0.68033381712627,
340
+ "grad_norm": 0.38091689348220825,
341
+ "learning_rate": 0.0005804813805631244,
342
+ "loss": 4.2651,
343
+ "step": 2250
344
+ },
345
+ {
346
+ "epoch": 0.6954523463957426,
347
+ "grad_norm": 0.4001523554325104,
348
+ "learning_rate": 0.0005800272479564032,
349
+ "loss": 4.2552,
350
+ "step": 2300
351
+ },
352
+ {
353
+ "epoch": 0.7105708756652153,
354
+ "grad_norm": 0.3821207880973816,
355
+ "learning_rate": 0.0005795731153496821,
356
+ "loss": 4.2479,
357
+ "step": 2350
358
+ },
359
+ {
360
+ "epoch": 0.7256894049346879,
361
+ "grad_norm": 0.3614497482776642,
362
+ "learning_rate": 0.0005791189827429609,
363
+ "loss": 4.2514,
364
+ "step": 2400
365
+ },
366
+ {
367
+ "epoch": 0.7408079342041606,
368
+ "grad_norm": 0.40655964612960815,
369
+ "learning_rate": 0.0005786648501362397,
370
+ "loss": 4.2292,
371
+ "step": 2450
372
+ },
373
+ {
374
+ "epoch": 0.7559264634736332,
375
+ "grad_norm": 0.3501703441143036,
376
+ "learning_rate": 0.0005782107175295186,
377
+ "loss": 4.2267,
378
+ "step": 2500
379
+ },
380
+ {
381
+ "epoch": 0.771044992743106,
382
+ "grad_norm": 0.41427701711654663,
383
+ "learning_rate": 0.0005777565849227974,
384
+ "loss": 4.2247,
385
+ "step": 2550
386
+ },
387
+ {
388
+ "epoch": 0.7861635220125787,
389
+ "grad_norm": 0.44100141525268555,
390
+ "learning_rate": 0.0005773024523160762,
391
+ "loss": 4.2089,
392
+ "step": 2600
393
+ },
394
+ {
395
+ "epoch": 0.8012820512820513,
396
+ "grad_norm": 0.3879771828651428,
397
+ "learning_rate": 0.000576848319709355,
398
+ "loss": 4.2073,
399
+ "step": 2650
400
+ },
401
+ {
402
+ "epoch": 0.816400580551524,
403
+ "grad_norm": 0.3788333833217621,
404
+ "learning_rate": 0.0005763941871026339,
405
+ "loss": 4.2015,
406
+ "step": 2700
407
+ },
408
+ {
409
+ "epoch": 0.8315191098209966,
410
+ "grad_norm": 0.366184800863266,
411
+ "learning_rate": 0.0005759400544959128,
412
+ "loss": 4.1719,
413
+ "step": 2750
414
+ },
415
+ {
416
+ "epoch": 0.8466376390904693,
417
+ "grad_norm": 0.3733968436717987,
418
+ "learning_rate": 0.0005754859218891917,
419
+ "loss": 4.1749,
420
+ "step": 2800
421
+ },
422
+ {
423
+ "epoch": 0.861756168359942,
424
+ "grad_norm": 0.37004509568214417,
425
+ "learning_rate": 0.0005750317892824704,
426
+ "loss": 4.1642,
427
+ "step": 2850
428
+ },
429
+ {
430
+ "epoch": 0.8768746976294146,
431
+ "grad_norm": 0.36916959285736084,
432
+ "learning_rate": 0.0005745776566757492,
433
+ "loss": 4.1666,
434
+ "step": 2900
435
+ },
436
+ {
437
+ "epoch": 0.8919932268988873,
438
+ "grad_norm": 0.39038950204849243,
439
+ "learning_rate": 0.0005741235240690281,
440
+ "loss": 4.1523,
441
+ "step": 2950
442
+ },
443
+ {
444
+ "epoch": 0.9071117561683599,
445
+ "grad_norm": 0.3550504744052887,
446
+ "learning_rate": 0.000573669391462307,
447
+ "loss": 4.1437,
448
+ "step": 3000
449
+ },
450
+ {
451
+ "epoch": 0.9071117561683599,
452
+ "eval_accuracy": 0.3157440890648464,
453
+ "eval_loss": 4.091999530792236,
454
+ "eval_runtime": 182.3379,
455
+ "eval_samples_per_second": 91.281,
456
+ "eval_steps_per_second": 5.709,
457
+ "step": 3000
458
+ },
459
+ {
460
+ "epoch": 0.9222302854378326,
461
+ "grad_norm": 0.3827008903026581,
462
+ "learning_rate": 0.0005732152588555858,
463
+ "loss": 4.1333,
464
+ "step": 3050
465
+ },
466
+ {
467
+ "epoch": 0.9373488147073052,
468
+ "grad_norm": 0.379339337348938,
469
+ "learning_rate": 0.0005727611262488646,
470
+ "loss": 4.1309,
471
+ "step": 3100
472
+ },
473
+ {
474
+ "epoch": 0.9524673439767779,
475
+ "grad_norm": 0.38624098896980286,
476
+ "learning_rate": 0.0005723069936421435,
477
+ "loss": 4.1218,
478
+ "step": 3150
479
+ },
480
+ {
481
+ "epoch": 0.9675858732462506,
482
+ "grad_norm": 0.36264657974243164,
483
+ "learning_rate": 0.0005718528610354223,
484
+ "loss": 4.1127,
485
+ "step": 3200
486
+ },
487
+ {
488
+ "epoch": 0.9827044025157232,
489
+ "grad_norm": 0.3813261389732361,
490
+ "learning_rate": 0.0005713987284287011,
491
+ "loss": 4.1095,
492
+ "step": 3250
493
+ },
494
+ {
495
+ "epoch": 0.997822931785196,
496
+ "grad_norm": 0.46720796823501587,
497
+ "learning_rate": 0.00057094459582198,
498
+ "loss": 4.0923,
499
+ "step": 3300
500
+ },
501
+ {
502
+ "epoch": 1.012699564586357,
503
+ "grad_norm": 0.3777443766593933,
504
+ "learning_rate": 0.0005704904632152588,
505
+ "loss": 4.0438,
506
+ "step": 3350
507
+ },
508
+ {
509
+ "epoch": 1.0278180938558297,
510
+ "grad_norm": 0.35427603125572205,
511
+ "learning_rate": 0.0005700363306085376,
512
+ "loss": 4.0293,
513
+ "step": 3400
514
+ },
515
+ {
516
+ "epoch": 1.0429366231253023,
517
+ "grad_norm": 0.36464598774909973,
518
+ "learning_rate": 0.0005695821980018164,
519
+ "loss": 4.0156,
520
+ "step": 3450
521
+ },
522
+ {
523
+ "epoch": 1.058055152394775,
524
+ "grad_norm": 0.39198657870292664,
525
+ "learning_rate": 0.0005691280653950954,
526
+ "loss": 4.0147,
527
+ "step": 3500
528
+ },
529
+ {
530
+ "epoch": 1.0731736816642476,
531
+ "grad_norm": 0.3432103395462036,
532
+ "learning_rate": 0.0005686739327883741,
533
+ "loss": 4.0026,
534
+ "step": 3550
535
+ },
536
+ {
537
+ "epoch": 1.0882922109337203,
538
+ "grad_norm": 0.370108962059021,
539
+ "learning_rate": 0.0005682198001816529,
540
+ "loss": 4.0078,
541
+ "step": 3600
542
+ },
543
+ {
544
+ "epoch": 1.103410740203193,
545
+ "grad_norm": 0.3692167103290558,
546
+ "learning_rate": 0.0005677656675749318,
547
+ "loss": 4.0147,
548
+ "step": 3650
549
+ },
550
+ {
551
+ "epoch": 1.1185292694726656,
552
+ "grad_norm": 0.354033887386322,
553
+ "learning_rate": 0.0005673115349682107,
554
+ "loss": 3.9974,
555
+ "step": 3700
556
+ },
557
+ {
558
+ "epoch": 1.1336477987421383,
559
+ "grad_norm": 0.3713250160217285,
560
+ "learning_rate": 0.0005668574023614895,
561
+ "loss": 3.9928,
562
+ "step": 3750
563
+ },
564
+ {
565
+ "epoch": 1.148766328011611,
566
+ "grad_norm": 0.364036500453949,
567
+ "learning_rate": 0.0005664032697547684,
568
+ "loss": 4.0023,
569
+ "step": 3800
570
+ },
571
+ {
572
+ "epoch": 1.1638848572810838,
573
+ "grad_norm": 0.3609767258167267,
574
+ "learning_rate": 0.0005659491371480471,
575
+ "loss": 3.9908,
576
+ "step": 3850
577
+ },
578
+ {
579
+ "epoch": 1.1790033865505563,
580
+ "grad_norm": 0.3580855131149292,
581
+ "learning_rate": 0.000565495004541326,
582
+ "loss": 3.9918,
583
+ "step": 3900
584
+ },
585
+ {
586
+ "epoch": 1.1941219158200291,
587
+ "grad_norm": 0.35853010416030884,
588
+ "learning_rate": 0.0005650408719346049,
589
+ "loss": 3.9891,
590
+ "step": 3950
591
+ },
592
+ {
593
+ "epoch": 1.2092404450895016,
594
+ "grad_norm": 0.3546011745929718,
595
+ "learning_rate": 0.0005645867393278837,
596
+ "loss": 3.9999,
597
+ "step": 4000
598
+ },
599
+ {
600
+ "epoch": 1.2092404450895016,
601
+ "eval_accuracy": 0.32564334461599675,
602
+ "eval_loss": 3.9847805500030518,
603
+ "eval_runtime": 182.3722,
604
+ "eval_samples_per_second": 91.264,
605
+ "eval_steps_per_second": 5.708,
606
+ "step": 4000
607
+ },
608
+ {
609
+ "epoch": 1.2243589743589745,
610
+ "grad_norm": 0.3494962453842163,
611
+ "learning_rate": 0.0005641326067211625,
612
+ "loss": 3.9874,
613
+ "step": 4050
614
+ },
615
+ {
616
+ "epoch": 1.239477503628447,
617
+ "grad_norm": 0.37261122465133667,
618
+ "learning_rate": 0.0005636784741144414,
619
+ "loss": 3.9849,
620
+ "step": 4100
621
+ },
622
+ {
623
+ "epoch": 1.2545960328979198,
624
+ "grad_norm": 0.3606380224227905,
625
+ "learning_rate": 0.0005632243415077202,
626
+ "loss": 3.9697,
627
+ "step": 4150
628
+ },
629
+ {
630
+ "epoch": 1.2697145621673924,
631
+ "grad_norm": 0.3712509572505951,
632
+ "learning_rate": 0.000562770208900999,
633
+ "loss": 3.9746,
634
+ "step": 4200
635
+ },
636
+ {
637
+ "epoch": 1.284833091436865,
638
+ "grad_norm": 0.3441880941390991,
639
+ "learning_rate": 0.000562316076294278,
640
+ "loss": 3.9746,
641
+ "step": 4250
642
+ },
643
+ {
644
+ "epoch": 1.2999516207063377,
645
+ "grad_norm": 0.37153229117393494,
646
+ "learning_rate": 0.0005618619436875567,
647
+ "loss": 3.9607,
648
+ "step": 4300
649
+ },
650
+ {
651
+ "epoch": 1.3150701499758104,
652
+ "grad_norm": 0.3481687605381012,
653
+ "learning_rate": 0.0005614078110808355,
654
+ "loss": 3.9699,
655
+ "step": 4350
656
+ },
657
+ {
658
+ "epoch": 1.330188679245283,
659
+ "grad_norm": 0.3434485197067261,
660
+ "learning_rate": 0.0005609536784741143,
661
+ "loss": 3.955,
662
+ "step": 4400
663
+ },
664
+ {
665
+ "epoch": 1.3453072085147557,
666
+ "grad_norm": 0.36444127559661865,
667
+ "learning_rate": 0.0005604995458673933,
668
+ "loss": 3.959,
669
+ "step": 4450
670
+ },
671
+ {
672
+ "epoch": 1.3604257377842284,
673
+ "grad_norm": 0.3475540280342102,
674
+ "learning_rate": 0.0005600454132606721,
675
+ "loss": 3.9587,
676
+ "step": 4500
677
+ },
678
+ {
679
+ "epoch": 1.375544267053701,
680
+ "grad_norm": 0.35136139392852783,
681
+ "learning_rate": 0.0005595912806539509,
682
+ "loss": 3.9493,
683
+ "step": 4550
684
+ },
685
+ {
686
+ "epoch": 1.3906627963231737,
687
+ "grad_norm": 0.3854037821292877,
688
+ "learning_rate": 0.0005591371480472297,
689
+ "loss": 3.9523,
690
+ "step": 4600
691
+ },
692
+ {
693
+ "epoch": 1.4057813255926463,
694
+ "grad_norm": 0.32536762952804565,
695
+ "learning_rate": 0.0005586830154405086,
696
+ "loss": 3.9458,
697
+ "step": 4650
698
+ },
699
+ {
700
+ "epoch": 1.420899854862119,
701
+ "grad_norm": 0.33678901195526123,
702
+ "learning_rate": 0.0005582288828337874,
703
+ "loss": 3.9349,
704
+ "step": 4700
705
+ },
706
+ {
707
+ "epoch": 1.4360183841315917,
708
+ "grad_norm": 0.38056421279907227,
709
+ "learning_rate": 0.0005577747502270663,
710
+ "loss": 3.9357,
711
+ "step": 4750
712
+ },
713
+ {
714
+ "epoch": 1.4511369134010643,
715
+ "grad_norm": 0.3556210994720459,
716
+ "learning_rate": 0.0005573206176203451,
717
+ "loss": 3.9439,
718
+ "step": 4800
719
+ },
720
+ {
721
+ "epoch": 1.466255442670537,
722
+ "grad_norm": 0.3364984691143036,
723
+ "learning_rate": 0.0005568664850136239,
724
+ "loss": 3.9258,
725
+ "step": 4850
726
+ },
727
+ {
728
+ "epoch": 1.4813739719400096,
729
+ "grad_norm": 0.34836122393608093,
730
+ "learning_rate": 0.0005564123524069028,
731
+ "loss": 3.9111,
732
+ "step": 4900
733
+ },
734
+ {
735
+ "epoch": 1.4964925012094823,
736
+ "grad_norm": 0.32042673230171204,
737
+ "learning_rate": 0.0005559582198001816,
738
+ "loss": 3.9368,
739
+ "step": 4950
740
+ },
741
+ {
742
+ "epoch": 1.511611030478955,
743
+ "grad_norm": 0.34929823875427246,
744
+ "learning_rate": 0.0005555040871934604,
745
+ "loss": 3.9251,
746
+ "step": 5000
747
+ },
748
+ {
749
+ "epoch": 1.511611030478955,
750
+ "eval_accuracy": 0.33236250223983677,
751
+ "eval_loss": 3.909000873565674,
752
+ "eval_runtime": 182.2679,
753
+ "eval_samples_per_second": 91.316,
754
+ "eval_steps_per_second": 5.711,
755
+ "step": 5000
756
+ },
757
+ {
758
+ "epoch": 1.5267295597484276,
759
+ "grad_norm": 0.3525310158729553,
760
+ "learning_rate": 0.0005550499545867392,
761
+ "loss": 3.9314,
762
+ "step": 5050
763
+ },
764
+ {
765
+ "epoch": 1.5418480890179005,
766
+ "grad_norm": 0.34068188071250916,
767
+ "learning_rate": 0.0005545958219800181,
768
+ "loss": 3.9343,
769
+ "step": 5100
770
+ },
771
+ {
772
+ "epoch": 1.556966618287373,
773
+ "grad_norm": 0.3572525084018707,
774
+ "learning_rate": 0.0005541416893732969,
775
+ "loss": 3.9272,
776
+ "step": 5150
777
+ },
778
+ {
779
+ "epoch": 1.5720851475568458,
780
+ "grad_norm": 0.3401585519313812,
781
+ "learning_rate": 0.0005536875567665758,
782
+ "loss": 3.9159,
783
+ "step": 5200
784
+ },
785
+ {
786
+ "epoch": 1.5872036768263182,
787
+ "grad_norm": 0.3368934094905853,
788
+ "learning_rate": 0.0005532334241598547,
789
+ "loss": 3.9024,
790
+ "step": 5250
791
+ },
792
+ {
793
+ "epoch": 1.6023222060957911,
794
+ "grad_norm": 0.33803030848503113,
795
+ "learning_rate": 0.0005527792915531334,
796
+ "loss": 3.9133,
797
+ "step": 5300
798
+ },
799
+ {
800
+ "epoch": 1.6174407353652636,
801
+ "grad_norm": 0.3491084575653076,
802
+ "learning_rate": 0.0005523251589464122,
803
+ "loss": 3.9049,
804
+ "step": 5350
805
+ },
806
+ {
807
+ "epoch": 1.6325592646347364,
808
+ "grad_norm": 0.3191988170146942,
809
+ "learning_rate": 0.0005518710263396912,
810
+ "loss": 3.897,
811
+ "step": 5400
812
+ },
813
+ {
814
+ "epoch": 1.6476777939042089,
815
+ "grad_norm": 0.33832481503486633,
816
+ "learning_rate": 0.00055141689373297,
817
+ "loss": 3.8915,
818
+ "step": 5450
819
+ },
820
+ {
821
+ "epoch": 1.6627963231736818,
822
+ "grad_norm": 0.3492324650287628,
823
+ "learning_rate": 0.0005509627611262488,
824
+ "loss": 3.8975,
825
+ "step": 5500
826
+ },
827
+ {
828
+ "epoch": 1.6779148524431542,
829
+ "grad_norm": 0.3200375735759735,
830
+ "learning_rate": 0.0005505086285195277,
831
+ "loss": 3.8861,
832
+ "step": 5550
833
+ },
834
+ {
835
+ "epoch": 1.693033381712627,
836
+ "grad_norm": 0.3539038896560669,
837
+ "learning_rate": 0.0005500544959128065,
838
+ "loss": 3.8972,
839
+ "step": 5600
840
+ },
841
+ {
842
+ "epoch": 1.7081519109820995,
843
+ "grad_norm": 0.34983664751052856,
844
+ "learning_rate": 0.0005496003633060853,
845
+ "loss": 3.876,
846
+ "step": 5650
847
+ },
848
+ {
849
+ "epoch": 1.7232704402515724,
850
+ "grad_norm": 0.33722037076950073,
851
+ "learning_rate": 0.0005491462306993642,
852
+ "loss": 3.8811,
853
+ "step": 5700
854
+ },
855
+ {
856
+ "epoch": 1.738388969521045,
857
+ "grad_norm": 0.34622326493263245,
858
+ "learning_rate": 0.000548692098092643,
859
+ "loss": 3.8913,
860
+ "step": 5750
861
+ },
862
+ {
863
+ "epoch": 1.7535074987905177,
864
+ "grad_norm": 0.3671897351741791,
865
+ "learning_rate": 0.0005482379654859218,
866
+ "loss": 3.8912,
867
+ "step": 5800
868
+ },
869
+ {
870
+ "epoch": 1.7686260280599904,
871
+ "grad_norm": 0.3275216817855835,
872
+ "learning_rate": 0.0005477838328792006,
873
+ "loss": 3.8837,
874
+ "step": 5850
875
+ },
876
+ {
877
+ "epoch": 1.783744557329463,
878
+ "grad_norm": 0.3239729404449463,
879
+ "learning_rate": 0.0005473297002724795,
880
+ "loss": 3.8796,
881
+ "step": 5900
882
+ },
883
+ {
884
+ "epoch": 1.7988630865989357,
885
+ "grad_norm": 0.3572208881378174,
886
+ "learning_rate": 0.0005468755676657584,
887
+ "loss": 3.8735,
888
+ "step": 5950
889
+ },
890
+ {
891
+ "epoch": 1.8139816158684083,
892
+ "grad_norm": 0.3316764831542969,
893
+ "learning_rate": 0.0005464214350590371,
894
+ "loss": 3.8666,
895
+ "step": 6000
896
+ },
897
+ {
898
+ "epoch": 1.8139816158684083,
899
+ "eval_accuracy": 0.33760983430616326,
900
+ "eval_loss": 3.853510856628418,
901
+ "eval_runtime": 182.1185,
902
+ "eval_samples_per_second": 91.391,
903
+ "eval_steps_per_second": 5.716,
904
+ "step": 6000
905
+ },
906
+ {
907
+ "epoch": 1.829100145137881,
908
+ "grad_norm": 0.3352029025554657,
909
+ "learning_rate": 0.000545967302452316,
910
+ "loss": 3.8716,
911
+ "step": 6050
912
+ },
913
+ {
914
+ "epoch": 1.8442186744073537,
915
+ "grad_norm": 0.33704429864883423,
916
+ "learning_rate": 0.0005455131698455948,
917
+ "loss": 3.8742,
918
+ "step": 6100
919
+ },
920
+ {
921
+ "epoch": 1.8593372036768263,
922
+ "grad_norm": 0.32570329308509827,
923
+ "learning_rate": 0.0005450590372388737,
924
+ "loss": 3.8545,
925
+ "step": 6150
926
+ },
927
+ {
928
+ "epoch": 1.874455732946299,
929
+ "grad_norm": 0.3355255424976349,
930
+ "learning_rate": 0.0005446049046321526,
931
+ "loss": 3.8584,
932
+ "step": 6200
933
+ },
934
+ {
935
+ "epoch": 1.8895742622157716,
936
+ "grad_norm": 0.3264237642288208,
937
+ "learning_rate": 0.0005441507720254314,
938
+ "loss": 3.8647,
939
+ "step": 6250
940
+ },
941
+ {
942
+ "epoch": 1.9046927914852443,
943
+ "grad_norm": 0.34331536293029785,
944
+ "learning_rate": 0.0005436966394187102,
945
+ "loss": 3.8587,
946
+ "step": 6300
947
+ },
948
+ {
949
+ "epoch": 1.919811320754717,
950
+ "grad_norm": 0.3500491976737976,
951
+ "learning_rate": 0.0005432425068119891,
952
+ "loss": 3.853,
953
+ "step": 6350
954
+ },
955
+ {
956
+ "epoch": 1.9349298500241896,
957
+ "grad_norm": 0.33102309703826904,
958
+ "learning_rate": 0.0005427883742052679,
959
+ "loss": 3.8585,
960
+ "step": 6400
961
+ },
962
+ {
963
+ "epoch": 1.9500483792936623,
964
+ "grad_norm": 0.3271898329257965,
965
+ "learning_rate": 0.0005423342415985467,
966
+ "loss": 3.8577,
967
+ "step": 6450
968
+ },
969
+ {
970
+ "epoch": 1.965166908563135,
971
+ "grad_norm": 0.328890323638916,
972
+ "learning_rate": 0.0005418801089918256,
973
+ "loss": 3.8379,
974
+ "step": 6500
975
+ },
976
+ {
977
+ "epoch": 1.9802854378326078,
978
+ "grad_norm": 0.31617918610572815,
979
+ "learning_rate": 0.0005414259763851044,
980
+ "loss": 3.8392,
981
+ "step": 6550
982
+ },
983
+ {
984
+ "epoch": 1.9954039671020802,
985
+ "grad_norm": 0.324520081281662,
986
+ "learning_rate": 0.0005409718437783832,
987
+ "loss": 3.8461,
988
+ "step": 6600
989
+ },
990
+ {
991
+ "epoch": 2.0102805999032416,
992
+ "grad_norm": 0.324241042137146,
993
+ "learning_rate": 0.000540517711171662,
994
+ "loss": 3.7744,
995
+ "step": 6650
996
+ },
997
+ {
998
+ "epoch": 2.025399129172714,
999
+ "grad_norm": 0.36148601770401,
1000
+ "learning_rate": 0.000540063578564941,
1001
+ "loss": 3.7479,
1002
+ "step": 6700
1003
+ },
1004
+ {
1005
+ "epoch": 2.040517658442187,
1006
+ "grad_norm": 0.32816359400749207,
1007
+ "learning_rate": 0.0005396094459582197,
1008
+ "loss": 3.74,
1009
+ "step": 6750
1010
+ },
1011
+ {
1012
+ "epoch": 2.0556361877116593,
1013
+ "grad_norm": 0.33710262179374695,
1014
+ "learning_rate": 0.0005391553133514985,
1015
+ "loss": 3.7454,
1016
+ "step": 6800
1017
+ },
1018
+ {
1019
+ "epoch": 2.0707547169811322,
1020
+ "grad_norm": 0.3189140558242798,
1021
+ "learning_rate": 0.0005387011807447775,
1022
+ "loss": 3.7555,
1023
+ "step": 6850
1024
+ },
1025
+ {
1026
+ "epoch": 2.0858732462506047,
1027
+ "grad_norm": 0.34328320622444153,
1028
+ "learning_rate": 0.0005382470481380563,
1029
+ "loss": 3.7461,
1030
+ "step": 6900
1031
+ },
1032
+ {
1033
+ "epoch": 2.1009917755200775,
1034
+ "grad_norm": 0.3447844982147217,
1035
+ "learning_rate": 0.0005377929155313351,
1036
+ "loss": 3.7688,
1037
+ "step": 6950
1038
+ },
1039
+ {
1040
+ "epoch": 2.11611030478955,
1041
+ "grad_norm": 0.3423329293727875,
1042
+ "learning_rate": 0.000537338782924614,
1043
+ "loss": 3.748,
1044
+ "step": 7000
1045
+ },
1046
+ {
1047
+ "epoch": 2.11611030478955,
1048
+ "eval_accuracy": 0.3412945715762478,
1049
+ "eval_loss": 3.8153622150421143,
1050
+ "eval_runtime": 182.3197,
1051
+ "eval_samples_per_second": 91.29,
1052
+ "eval_steps_per_second": 5.71,
1053
+ "step": 7000
1054
+ },
1055
+ {
1056
+ "epoch": 2.131228834059023,
1057
+ "grad_norm": 0.3441483974456787,
1058
+ "learning_rate": 0.0005368846503178928,
1059
+ "loss": 3.7481,
1060
+ "step": 7050
1061
+ },
1062
+ {
1063
+ "epoch": 2.1463473633284953,
1064
+ "grad_norm": 0.3269549608230591,
1065
+ "learning_rate": 0.0005364305177111716,
1066
+ "loss": 3.7588,
1067
+ "step": 7100
1068
+ },
1069
+ {
1070
+ "epoch": 2.161465892597968,
1071
+ "grad_norm": 0.330601304769516,
1072
+ "learning_rate": 0.0005359763851044505,
1073
+ "loss": 3.7418,
1074
+ "step": 7150
1075
+ },
1076
+ {
1077
+ "epoch": 2.1765844218674406,
1078
+ "grad_norm": 0.31895920634269714,
1079
+ "learning_rate": 0.0005355222524977293,
1080
+ "loss": 3.7563,
1081
+ "step": 7200
1082
+ },
1083
+ {
1084
+ "epoch": 2.1917029511369135,
1085
+ "grad_norm": 0.33730292320251465,
1086
+ "learning_rate": 0.0005350681198910081,
1087
+ "loss": 3.7622,
1088
+ "step": 7250
1089
+ },
1090
+ {
1091
+ "epoch": 2.206821480406386,
1092
+ "grad_norm": 0.32292884588241577,
1093
+ "learning_rate": 0.000534613987284287,
1094
+ "loss": 3.7546,
1095
+ "step": 7300
1096
+ },
1097
+ {
1098
+ "epoch": 2.221940009675859,
1099
+ "grad_norm": 0.33400362730026245,
1100
+ "learning_rate": 0.0005341598546775658,
1101
+ "loss": 3.7655,
1102
+ "step": 7350
1103
+ },
1104
+ {
1105
+ "epoch": 2.2370585389453312,
1106
+ "grad_norm": 0.3238161504268646,
1107
+ "learning_rate": 0.0005337057220708446,
1108
+ "loss": 3.7483,
1109
+ "step": 7400
1110
+ },
1111
+ {
1112
+ "epoch": 2.252177068214804,
1113
+ "grad_norm": 0.3304459750652313,
1114
+ "learning_rate": 0.0005332515894641234,
1115
+ "loss": 3.7577,
1116
+ "step": 7450
1117
+ },
1118
+ {
1119
+ "epoch": 2.2672955974842766,
1120
+ "grad_norm": 0.34201282262802124,
1121
+ "learning_rate": 0.0005327974568574023,
1122
+ "loss": 3.7453,
1123
+ "step": 7500
1124
+ },
1125
+ {
1126
+ "epoch": 2.2824141267537494,
1127
+ "grad_norm": 0.3679095208644867,
1128
+ "learning_rate": 0.0005323433242506811,
1129
+ "loss": 3.7635,
1130
+ "step": 7550
1131
+ },
1132
+ {
1133
+ "epoch": 2.297532656023222,
1134
+ "grad_norm": 0.3323363959789276,
1135
+ "learning_rate": 0.00053188919164396,
1136
+ "loss": 3.7442,
1137
+ "step": 7600
1138
+ },
1139
+ {
1140
+ "epoch": 2.3126511852926948,
1141
+ "grad_norm": 0.3315403163433075,
1142
+ "learning_rate": 0.0005314350590372389,
1143
+ "loss": 3.751,
1144
+ "step": 7650
1145
+ },
1146
+ {
1147
+ "epoch": 2.3277697145621676,
1148
+ "grad_norm": 0.35335952043533325,
1149
+ "learning_rate": 0.0005309809264305177,
1150
+ "loss": 3.7569,
1151
+ "step": 7700
1152
+ },
1153
+ {
1154
+ "epoch": 2.34288824383164,
1155
+ "grad_norm": 0.33651435375213623,
1156
+ "learning_rate": 0.0005305267938237964,
1157
+ "loss": 3.7466,
1158
+ "step": 7750
1159
+ },
1160
+ {
1161
+ "epoch": 2.3580067731011125,
1162
+ "grad_norm": 0.31409308314323425,
1163
+ "learning_rate": 0.0005300726612170754,
1164
+ "loss": 3.7599,
1165
+ "step": 7800
1166
+ },
1167
+ {
1168
+ "epoch": 2.3731253023705854,
1169
+ "grad_norm": 0.3148058354854584,
1170
+ "learning_rate": 0.0005296185286103542,
1171
+ "loss": 3.7423,
1172
+ "step": 7850
1173
+ },
1174
+ {
1175
+ "epoch": 2.3882438316400583,
1176
+ "grad_norm": 0.34514766931533813,
1177
+ "learning_rate": 0.000529164396003633,
1178
+ "loss": 3.7253,
1179
+ "step": 7900
1180
+ },
1181
+ {
1182
+ "epoch": 2.4033623609095307,
1183
+ "grad_norm": 0.32518208026885986,
1184
+ "learning_rate": 0.0005287102633969119,
1185
+ "loss": 3.7365,
1186
+ "step": 7950
1187
+ },
1188
+ {
1189
+ "epoch": 2.418480890179003,
1190
+ "grad_norm": 0.3367462158203125,
1191
+ "learning_rate": 0.0005282561307901907,
1192
+ "loss": 3.7526,
1193
+ "step": 8000
1194
+ },
1195
+ {
1196
+ "epoch": 2.418480890179003,
1197
+ "eval_accuracy": 0.3448259887850608,
1198
+ "eval_loss": 3.7824463844299316,
1199
+ "eval_runtime": 182.5201,
1200
+ "eval_samples_per_second": 91.19,
1201
+ "eval_steps_per_second": 5.703,
1202
+ "step": 8000
1203
+ },
1204
+ {
1205
+ "epoch": 2.433599419448476,
1206
+ "grad_norm": 0.3492235541343689,
1207
+ "learning_rate": 0.0005278019981834695,
1208
+ "loss": 3.7511,
1209
+ "step": 8050
1210
+ },
1211
+ {
1212
+ "epoch": 2.448717948717949,
1213
+ "grad_norm": 0.3320801556110382,
1214
+ "learning_rate": 0.0005273478655767484,
1215
+ "loss": 3.7335,
1216
+ "step": 8100
1217
+ },
1218
+ {
1219
+ "epoch": 2.4638364779874213,
1220
+ "grad_norm": 0.32247304916381836,
1221
+ "learning_rate": 0.0005268937329700272,
1222
+ "loss": 3.7359,
1223
+ "step": 8150
1224
+ },
1225
+ {
1226
+ "epoch": 2.478955007256894,
1227
+ "grad_norm": 0.3158959448337555,
1228
+ "learning_rate": 0.000526439600363306,
1229
+ "loss": 3.7384,
1230
+ "step": 8200
1231
+ },
1232
+ {
1233
+ "epoch": 2.4940735365263667,
1234
+ "grad_norm": 0.3267774283885956,
1235
+ "learning_rate": 0.0005259854677565848,
1236
+ "loss": 3.7429,
1237
+ "step": 8250
1238
+ },
1239
+ {
1240
+ "epoch": 2.5091920657958395,
1241
+ "grad_norm": 0.34227943420410156,
1242
+ "learning_rate": 0.0005255313351498637,
1243
+ "loss": 3.7212,
1244
+ "step": 8300
1245
+ },
1246
+ {
1247
+ "epoch": 2.524310595065312,
1248
+ "grad_norm": 0.3359690010547638,
1249
+ "learning_rate": 0.0005250772025431426,
1250
+ "loss": 3.7497,
1251
+ "step": 8350
1252
+ },
1253
+ {
1254
+ "epoch": 2.539429124334785,
1255
+ "grad_norm": 0.3244568407535553,
1256
+ "learning_rate": 0.0005246230699364214,
1257
+ "loss": 3.738,
1258
+ "step": 8400
1259
+ },
1260
+ {
1261
+ "epoch": 2.5545476536042573,
1262
+ "grad_norm": 0.3494398891925812,
1263
+ "learning_rate": 0.0005241689373297002,
1264
+ "loss": 3.7482,
1265
+ "step": 8450
1266
+ },
1267
+ {
1268
+ "epoch": 2.56966618287373,
1269
+ "grad_norm": 0.33279120922088623,
1270
+ "learning_rate": 0.000523714804722979,
1271
+ "loss": 3.7474,
1272
+ "step": 8500
1273
+ },
1274
+ {
1275
+ "epoch": 2.5847847121432026,
1276
+ "grad_norm": 0.3192063868045807,
1277
+ "learning_rate": 0.0005232606721162579,
1278
+ "loss": 3.7344,
1279
+ "step": 8550
1280
+ },
1281
+ {
1282
+ "epoch": 2.5999032414126755,
1283
+ "grad_norm": 0.32627859711647034,
1284
+ "learning_rate": 0.0005228065395095368,
1285
+ "loss": 3.7326,
1286
+ "step": 8600
1287
+ },
1288
+ {
1289
+ "epoch": 2.615021770682148,
1290
+ "grad_norm": 0.32838842272758484,
1291
+ "learning_rate": 0.0005223524069028156,
1292
+ "loss": 3.7382,
1293
+ "step": 8650
1294
+ },
1295
+ {
1296
+ "epoch": 2.630140299951621,
1297
+ "grad_norm": 0.30039238929748535,
1298
+ "learning_rate": 0.0005218982742960944,
1299
+ "loss": 3.7325,
1300
+ "step": 8700
1301
+ },
1302
+ {
1303
+ "epoch": 2.6452588292210932,
1304
+ "grad_norm": 0.35245180130004883,
1305
+ "learning_rate": 0.0005214441416893733,
1306
+ "loss": 3.7375,
1307
+ "step": 8750
1308
+ },
1309
+ {
1310
+ "epoch": 2.660377358490566,
1311
+ "grad_norm": 0.3139996826648712,
1312
+ "learning_rate": 0.0005209900090826521,
1313
+ "loss": 3.7314,
1314
+ "step": 8800
1315
+ },
1316
+ {
1317
+ "epoch": 2.6754958877600385,
1318
+ "grad_norm": 0.32654982805252075,
1319
+ "learning_rate": 0.0005205358764759309,
1320
+ "loss": 3.7347,
1321
+ "step": 8850
1322
+ },
1323
+ {
1324
+ "epoch": 2.6906144170295114,
1325
+ "grad_norm": 0.3103334307670593,
1326
+ "learning_rate": 0.0005200817438692098,
1327
+ "loss": 3.7339,
1328
+ "step": 8900
1329
+ },
1330
+ {
1331
+ "epoch": 2.7057329462989843,
1332
+ "grad_norm": 0.3158484101295471,
1333
+ "learning_rate": 0.0005196276112624886,
1334
+ "loss": 3.7294,
1335
+ "step": 8950
1336
+ },
1337
+ {
1338
+ "epoch": 2.7208514755684567,
1339
+ "grad_norm": 0.3339202404022217,
1340
+ "learning_rate": 0.0005191734786557674,
1341
+ "loss": 3.7241,
1342
+ "step": 9000
1343
+ },
1344
+ {
1345
+ "epoch": 2.7208514755684567,
1346
+ "eval_accuracy": 0.3476167901457528,
1347
+ "eval_loss": 3.7497520446777344,
1348
+ "eval_runtime": 182.7073,
1349
+ "eval_samples_per_second": 91.097,
1350
+ "eval_steps_per_second": 5.698,
1351
+ "step": 9000
1352
+ },
1353
+ {
1354
+ "epoch": 2.735970004837929,
1355
+ "grad_norm": 0.33506032824516296,
1356
+ "learning_rate": 0.0005187193460490462,
1357
+ "loss": 3.7369,
1358
+ "step": 9050
1359
+ },
1360
+ {
1361
+ "epoch": 2.751088534107402,
1362
+ "grad_norm": 0.34870266914367676,
1363
+ "learning_rate": 0.0005182652134423252,
1364
+ "loss": 3.7235,
1365
+ "step": 9100
1366
+ },
1367
+ {
1368
+ "epoch": 2.766207063376875,
1369
+ "grad_norm": 0.33400285243988037,
1370
+ "learning_rate": 0.000517811080835604,
1371
+ "loss": 3.7238,
1372
+ "step": 9150
1373
+ },
1374
+ {
1375
+ "epoch": 2.7813255926463474,
1376
+ "grad_norm": 0.32307007908821106,
1377
+ "learning_rate": 0.0005173569482288827,
1378
+ "loss": 3.7244,
1379
+ "step": 9200
1380
+ },
1381
+ {
1382
+ "epoch": 2.79644412191582,
1383
+ "grad_norm": 0.3319950997829437,
1384
+ "learning_rate": 0.0005169028156221616,
1385
+ "loss": 3.7157,
1386
+ "step": 9250
1387
+ },
1388
+ {
1389
+ "epoch": 2.8115626511852927,
1390
+ "grad_norm": 0.3344840705394745,
1391
+ "learning_rate": 0.0005164486830154405,
1392
+ "loss": 3.7192,
1393
+ "step": 9300
1394
+ },
1395
+ {
1396
+ "epoch": 2.8266811804547656,
1397
+ "grad_norm": 0.30677810311317444,
1398
+ "learning_rate": 0.0005159945504087193,
1399
+ "loss": 3.7245,
1400
+ "step": 9350
1401
+ },
1402
+ {
1403
+ "epoch": 2.841799709724238,
1404
+ "grad_norm": 0.3162977695465088,
1405
+ "learning_rate": 0.0005155404178019982,
1406
+ "loss": 3.7228,
1407
+ "step": 9400
1408
+ },
1409
+ {
1410
+ "epoch": 2.8569182389937104,
1411
+ "grad_norm": 0.3243446946144104,
1412
+ "learning_rate": 0.000515086285195277,
1413
+ "loss": 3.716,
1414
+ "step": 9450
1415
+ },
1416
+ {
1417
+ "epoch": 2.8720367682631833,
1418
+ "grad_norm": 0.3386097252368927,
1419
+ "learning_rate": 0.0005146321525885558,
1420
+ "loss": 3.7195,
1421
+ "step": 9500
1422
+ },
1423
+ {
1424
+ "epoch": 2.887155297532656,
1425
+ "grad_norm": 0.3524998128414154,
1426
+ "learning_rate": 0.0005141780199818347,
1427
+ "loss": 3.7071,
1428
+ "step": 9550
1429
+ },
1430
+ {
1431
+ "epoch": 2.9022738268021286,
1432
+ "grad_norm": 0.3099507689476013,
1433
+ "learning_rate": 0.0005137238873751135,
1434
+ "loss": 3.7196,
1435
+ "step": 9600
1436
+ },
1437
+ {
1438
+ "epoch": 2.917392356071601,
1439
+ "grad_norm": 0.31758880615234375,
1440
+ "learning_rate": 0.0005132697547683923,
1441
+ "loss": 3.7128,
1442
+ "step": 9650
1443
+ },
1444
+ {
1445
+ "epoch": 2.932510885341074,
1446
+ "grad_norm": 0.3277053236961365,
1447
+ "learning_rate": 0.0005128156221616712,
1448
+ "loss": 3.7313,
1449
+ "step": 9700
1450
+ },
1451
+ {
1452
+ "epoch": 2.947629414610547,
1453
+ "grad_norm": 0.30629098415374756,
1454
+ "learning_rate": 0.00051236148955495,
1455
+ "loss": 3.7135,
1456
+ "step": 9750
1457
+ },
1458
+ {
1459
+ "epoch": 2.9627479438800193,
1460
+ "grad_norm": 0.3403231203556061,
1461
+ "learning_rate": 0.0005119073569482288,
1462
+ "loss": 3.7087,
1463
+ "step": 9800
1464
+ },
1465
+ {
1466
+ "epoch": 2.977866473149492,
1467
+ "grad_norm": 0.3153614103794098,
1468
+ "learning_rate": 0.0005114532243415078,
1469
+ "loss": 3.7156,
1470
+ "step": 9850
1471
+ },
1472
+ {
1473
+ "epoch": 2.9929850024189646,
1474
+ "grad_norm": 0.3428856134414673,
1475
+ "learning_rate": 0.0005109990917347865,
1476
+ "loss": 3.7188,
1477
+ "step": 9900
1478
+ },
1479
+ {
1480
+ "epoch": 3.007861635220126,
1481
+ "grad_norm": 0.32290196418762207,
1482
+ "learning_rate": 0.0005105449591280653,
1483
+ "loss": 3.657,
1484
+ "step": 9950
1485
+ },
1486
+ {
1487
+ "epoch": 3.0229801644895984,
1488
+ "grad_norm": 0.3294488787651062,
1489
+ "learning_rate": 0.0005100908265213441,
1490
+ "loss": 3.5917,
1491
+ "step": 10000
1492
+ },
1493
+ {
1494
+ "epoch": 3.0229801644895984,
1495
+ "eval_accuracy": 0.35003781267768785,
1496
+ "eval_loss": 3.730295181274414,
1497
+ "eval_runtime": 182.2568,
1498
+ "eval_samples_per_second": 91.322,
1499
+ "eval_steps_per_second": 5.712,
1500
+ "step": 10000
1501
+ },
1502
+ {
1503
+ "epoch": 3.0380986937590713,
1504
+ "grad_norm": 0.3343505561351776,
1505
+ "learning_rate": 0.0005096366939146231,
1506
+ "loss": 3.5981,
1507
+ "step": 10050
1508
+ },
1509
+ {
1510
+ "epoch": 3.0532172230285437,
1511
+ "grad_norm": 0.32677552103996277,
1512
+ "learning_rate": 0.0005091825613079019,
1513
+ "loss": 3.6168,
1514
+ "step": 10100
1515
+ },
1516
+ {
1517
+ "epoch": 3.0683357522980166,
1518
+ "grad_norm": 0.32819923758506775,
1519
+ "learning_rate": 0.0005087284287011807,
1520
+ "loss": 3.6285,
1521
+ "step": 10150
1522
+ },
1523
+ {
1524
+ "epoch": 3.083454281567489,
1525
+ "grad_norm": 0.34550225734710693,
1526
+ "learning_rate": 0.0005082742960944595,
1527
+ "loss": 3.6147,
1528
+ "step": 10200
1529
+ },
1530
+ {
1531
+ "epoch": 3.098572810836962,
1532
+ "grad_norm": 0.3343297839164734,
1533
+ "learning_rate": 0.0005078201634877384,
1534
+ "loss": 3.6216,
1535
+ "step": 10250
1536
+ },
1537
+ {
1538
+ "epoch": 3.1136913401064343,
1539
+ "grad_norm": 0.3187820315361023,
1540
+ "learning_rate": 0.0005073660308810172,
1541
+ "loss": 3.6035,
1542
+ "step": 10300
1543
+ },
1544
+ {
1545
+ "epoch": 3.128809869375907,
1546
+ "grad_norm": 0.31732386350631714,
1547
+ "learning_rate": 0.0005069118982742961,
1548
+ "loss": 3.6181,
1549
+ "step": 10350
1550
+ },
1551
+ {
1552
+ "epoch": 3.1439283986453797,
1553
+ "grad_norm": 0.33218568563461304,
1554
+ "learning_rate": 0.0005064577656675749,
1555
+ "loss": 3.6207,
1556
+ "step": 10400
1557
+ },
1558
+ {
1559
+ "epoch": 3.1590469279148525,
1560
+ "grad_norm": 0.3314531147480011,
1561
+ "learning_rate": 0.0005060036330608537,
1562
+ "loss": 3.6108,
1563
+ "step": 10450
1564
+ },
1565
+ {
1566
+ "epoch": 3.174165457184325,
1567
+ "grad_norm": 0.3125026226043701,
1568
+ "learning_rate": 0.0005055495004541326,
1569
+ "loss": 3.6393,
1570
+ "step": 10500
1571
+ },
1572
+ {
1573
+ "epoch": 3.189283986453798,
1574
+ "grad_norm": 0.3300221562385559,
1575
+ "learning_rate": 0.0005050953678474114,
1576
+ "loss": 3.6234,
1577
+ "step": 10550
1578
+ },
1579
+ {
1580
+ "epoch": 3.2044025157232703,
1581
+ "grad_norm": 0.3192563056945801,
1582
+ "learning_rate": 0.0005046412352406902,
1583
+ "loss": 3.6309,
1584
+ "step": 10600
1585
+ },
1586
+ {
1587
+ "epoch": 3.219521044992743,
1588
+ "grad_norm": 0.3273380994796753,
1589
+ "learning_rate": 0.000504187102633969,
1590
+ "loss": 3.6277,
1591
+ "step": 10650
1592
+ },
1593
+ {
1594
+ "epoch": 3.2346395742622156,
1595
+ "grad_norm": 0.3166305422782898,
1596
+ "learning_rate": 0.0005037329700272479,
1597
+ "loss": 3.6222,
1598
+ "step": 10700
1599
+ },
1600
+ {
1601
+ "epoch": 3.2497581035316885,
1602
+ "grad_norm": 0.34870728850364685,
1603
+ "learning_rate": 0.0005032788374205267,
1604
+ "loss": 3.6306,
1605
+ "step": 10750
1606
+ },
1607
+ {
1608
+ "epoch": 3.264876632801161,
1609
+ "grad_norm": 0.3375926911830902,
1610
+ "learning_rate": 0.0005028247048138056,
1611
+ "loss": 3.6342,
1612
+ "step": 10800
1613
+ },
1614
+ {
1615
+ "epoch": 3.279995162070634,
1616
+ "grad_norm": 0.3100663721561432,
1617
+ "learning_rate": 0.0005023705722070845,
1618
+ "loss": 3.6279,
1619
+ "step": 10850
1620
+ },
1621
+ {
1622
+ "epoch": 3.2951136913401067,
1623
+ "grad_norm": 0.31706514954566956,
1624
+ "learning_rate": 0.0005019164396003632,
1625
+ "loss": 3.6229,
1626
+ "step": 10900
1627
+ },
1628
+ {
1629
+ "epoch": 3.310232220609579,
1630
+ "grad_norm": 0.32962343096733093,
1631
+ "learning_rate": 0.000501462306993642,
1632
+ "loss": 3.6332,
1633
+ "step": 10950
1634
+ },
1635
+ {
1636
+ "epoch": 3.3253507498790515,
1637
+ "grad_norm": 0.33387279510498047,
1638
+ "learning_rate": 0.000501008174386921,
1639
+ "loss": 3.6298,
1640
+ "step": 11000
1641
+ },
1642
+ {
1643
+ "epoch": 3.3253507498790515,
1644
+ "eval_accuracy": 0.3517179842080337,
1645
+ "eval_loss": 3.7141380310058594,
1646
+ "eval_runtime": 182.4417,
1647
+ "eval_samples_per_second": 91.229,
1648
+ "eval_steps_per_second": 5.706,
1649
+ "step": 11000
1650
+ },
1651
+ {
1652
+ "epoch": 3.3404692791485244,
1653
+ "grad_norm": 0.31837111711502075,
1654
+ "learning_rate": 0.0005005540417801998,
1655
+ "loss": 3.637,
1656
+ "step": 11050
1657
+ },
1658
+ {
1659
+ "epoch": 3.3555878084179973,
1660
+ "grad_norm": 0.33772799372673035,
1661
+ "learning_rate": 0.0005000999091734786,
1662
+ "loss": 3.6324,
1663
+ "step": 11100
1664
+ },
1665
+ {
1666
+ "epoch": 3.3707063376874697,
1667
+ "grad_norm": 0.3386647403240204,
1668
+ "learning_rate": 0.0004996457765667575,
1669
+ "loss": 3.629,
1670
+ "step": 11150
1671
+ },
1672
+ {
1673
+ "epoch": 3.385824866956942,
1674
+ "grad_norm": 0.31823232769966125,
1675
+ "learning_rate": 0.0004991916439600363,
1676
+ "loss": 3.6385,
1677
+ "step": 11200
1678
+ },
1679
+ {
1680
+ "epoch": 3.400943396226415,
1681
+ "grad_norm": 0.32417285442352295,
1682
+ "learning_rate": 0.0004987375113533151,
1683
+ "loss": 3.6336,
1684
+ "step": 11250
1685
+ },
1686
+ {
1687
+ "epoch": 3.416061925495888,
1688
+ "grad_norm": 0.34577304124832153,
1689
+ "learning_rate": 0.000498283378746594,
1690
+ "loss": 3.6371,
1691
+ "step": 11300
1692
+ },
1693
+ {
1694
+ "epoch": 3.4311804547653604,
1695
+ "grad_norm": 0.35057663917541504,
1696
+ "learning_rate": 0.0004978292461398728,
1697
+ "loss": 3.6232,
1698
+ "step": 11350
1699
+ },
1700
+ {
1701
+ "epoch": 3.4462989840348333,
1702
+ "grad_norm": 0.33663854002952576,
1703
+ "learning_rate": 0.0004973751135331516,
1704
+ "loss": 3.6322,
1705
+ "step": 11400
1706
+ },
1707
+ {
1708
+ "epoch": 3.4614175133043057,
1709
+ "grad_norm": 0.32272565364837646,
1710
+ "learning_rate": 0.0004969209809264304,
1711
+ "loss": 3.6351,
1712
+ "step": 11450
1713
+ },
1714
+ {
1715
+ "epoch": 3.4765360425737786,
1716
+ "grad_norm": 0.34051501750946045,
1717
+ "learning_rate": 0.0004964668483197093,
1718
+ "loss": 3.6287,
1719
+ "step": 11500
1720
+ },
1721
+ {
1722
+ "epoch": 3.491654571843251,
1723
+ "grad_norm": 0.3307649493217468,
1724
+ "learning_rate": 0.0004960127157129882,
1725
+ "loss": 3.6389,
1726
+ "step": 11550
1727
+ },
1728
+ {
1729
+ "epoch": 3.506773101112724,
1730
+ "grad_norm": 0.3179891109466553,
1731
+ "learning_rate": 0.0004955585831062669,
1732
+ "loss": 3.6281,
1733
+ "step": 11600
1734
+ },
1735
+ {
1736
+ "epoch": 3.5218916303821963,
1737
+ "grad_norm": 0.32227689027786255,
1738
+ "learning_rate": 0.0004951044504995458,
1739
+ "loss": 3.6423,
1740
+ "step": 11650
1741
+ },
1742
+ {
1743
+ "epoch": 3.537010159651669,
1744
+ "grad_norm": 0.35890889167785645,
1745
+ "learning_rate": 0.0004946503178928246,
1746
+ "loss": 3.6468,
1747
+ "step": 11700
1748
+ },
1749
+ {
1750
+ "epoch": 3.5521286889211416,
1751
+ "grad_norm": 0.36508727073669434,
1752
+ "learning_rate": 0.0004941961852861035,
1753
+ "loss": 3.6366,
1754
+ "step": 11750
1755
+ },
1756
+ {
1757
+ "epoch": 3.5672472181906145,
1758
+ "grad_norm": 0.32733410596847534,
1759
+ "learning_rate": 0.0004937420526793824,
1760
+ "loss": 3.6355,
1761
+ "step": 11800
1762
+ },
1763
+ {
1764
+ "epoch": 3.582365747460087,
1765
+ "grad_norm": 0.3346957564353943,
1766
+ "learning_rate": 0.0004932879200726612,
1767
+ "loss": 3.6388,
1768
+ "step": 11850
1769
+ },
1770
+ {
1771
+ "epoch": 3.59748427672956,
1772
+ "grad_norm": 0.32891422510147095,
1773
+ "learning_rate": 0.00049283378746594,
1774
+ "loss": 3.629,
1775
+ "step": 11900
1776
+ },
1777
+ {
1778
+ "epoch": 3.6126028059990323,
1779
+ "grad_norm": 0.3372475206851959,
1780
+ "learning_rate": 0.0004923796548592189,
1781
+ "loss": 3.6244,
1782
+ "step": 11950
1783
+ },
1784
+ {
1785
+ "epoch": 3.627721335268505,
1786
+ "grad_norm": 0.3311806917190552,
1787
+ "learning_rate": 0.0004919255222524977,
1788
+ "loss": 3.6253,
1789
+ "step": 12000
1790
+ },
1791
+ {
1792
+ "epoch": 3.627721335268505,
1793
+ "eval_accuracy": 0.35409750215283,
1794
+ "eval_loss": 3.6940906047821045,
1795
+ "eval_runtime": 182.4197,
1796
+ "eval_samples_per_second": 91.24,
1797
+ "eval_steps_per_second": 5.707,
1798
+ "step": 12000
1799
+ },
1800
+ {
1801
+ "epoch": 3.6428398645379776,
1802
+ "grad_norm": 0.31895172595977783,
1803
+ "learning_rate": 0.0004914713896457765,
1804
+ "loss": 3.6391,
1805
+ "step": 12050
1806
+ },
1807
+ {
1808
+ "epoch": 3.6579583938074505,
1809
+ "grad_norm": 0.3537571430206299,
1810
+ "learning_rate": 0.0004910172570390554,
1811
+ "loss": 3.6328,
1812
+ "step": 12100
1813
+ },
1814
+ {
1815
+ "epoch": 3.6730769230769234,
1816
+ "grad_norm": 0.336915522813797,
1817
+ "learning_rate": 0.0004905631244323342,
1818
+ "loss": 3.6389,
1819
+ "step": 12150
1820
+ },
1821
+ {
1822
+ "epoch": 3.688195452346396,
1823
+ "grad_norm": 0.33801692724227905,
1824
+ "learning_rate": 0.000490108991825613,
1825
+ "loss": 3.6309,
1826
+ "step": 12200
1827
+ },
1828
+ {
1829
+ "epoch": 3.7033139816158682,
1830
+ "grad_norm": 0.3218366801738739,
1831
+ "learning_rate": 0.0004896548592188918,
1832
+ "loss": 3.625,
1833
+ "step": 12250
1834
+ },
1835
+ {
1836
+ "epoch": 3.718432510885341,
1837
+ "grad_norm": 0.33120447397232056,
1838
+ "learning_rate": 0.0004892007266121708,
1839
+ "loss": 3.6343,
1840
+ "step": 12300
1841
+ },
1842
+ {
1843
+ "epoch": 3.733551040154814,
1844
+ "grad_norm": 0.32918858528137207,
1845
+ "learning_rate": 0.0004887465940054495,
1846
+ "loss": 3.6272,
1847
+ "step": 12350
1848
+ },
1849
+ {
1850
+ "epoch": 3.7486695694242864,
1851
+ "grad_norm": 0.3185270130634308,
1852
+ "learning_rate": 0.0004882924613987283,
1853
+ "loss": 3.6325,
1854
+ "step": 12400
1855
+ },
1856
+ {
1857
+ "epoch": 3.763788098693759,
1858
+ "grad_norm": 0.31657835841178894,
1859
+ "learning_rate": 0.00048783832879200717,
1860
+ "loss": 3.6286,
1861
+ "step": 12450
1862
+ },
1863
+ {
1864
+ "epoch": 3.7789066279632317,
1865
+ "grad_norm": 0.31327733397483826,
1866
+ "learning_rate": 0.00048738419618528606,
1867
+ "loss": 3.6213,
1868
+ "step": 12500
1869
+ },
1870
+ {
1871
+ "epoch": 3.7940251572327046,
1872
+ "grad_norm": 0.32215988636016846,
1873
+ "learning_rate": 0.0004869300635785649,
1874
+ "loss": 3.6203,
1875
+ "step": 12550
1876
+ },
1877
+ {
1878
+ "epoch": 3.809143686502177,
1879
+ "grad_norm": 0.3287181556224823,
1880
+ "learning_rate": 0.0004864759309718437,
1881
+ "loss": 3.6269,
1882
+ "step": 12600
1883
+ },
1884
+ {
1885
+ "epoch": 3.8242622157716495,
1886
+ "grad_norm": 0.3348924219608307,
1887
+ "learning_rate": 0.0004860217983651226,
1888
+ "loss": 3.6146,
1889
+ "step": 12650
1890
+ },
1891
+ {
1892
+ "epoch": 3.8393807450411224,
1893
+ "grad_norm": 0.3488834500312805,
1894
+ "learning_rate": 0.00048556766575840143,
1895
+ "loss": 3.6344,
1896
+ "step": 12700
1897
+ },
1898
+ {
1899
+ "epoch": 3.8544992743105952,
1900
+ "grad_norm": 0.32925039529800415,
1901
+ "learning_rate": 0.00048511353315168026,
1902
+ "loss": 3.6386,
1903
+ "step": 12750
1904
+ },
1905
+ {
1906
+ "epoch": 3.8696178035800677,
1907
+ "grad_norm": 0.33648958802223206,
1908
+ "learning_rate": 0.00048465940054495904,
1909
+ "loss": 3.6341,
1910
+ "step": 12800
1911
+ },
1912
+ {
1913
+ "epoch": 3.8847363328495406,
1914
+ "grad_norm": 0.3223407566547394,
1915
+ "learning_rate": 0.0004842052679382379,
1916
+ "loss": 3.6252,
1917
+ "step": 12850
1918
+ },
1919
+ {
1920
+ "epoch": 3.899854862119013,
1921
+ "grad_norm": 0.3572065830230713,
1922
+ "learning_rate": 0.00048375113533151676,
1923
+ "loss": 3.6246,
1924
+ "step": 12900
1925
+ },
1926
+ {
1927
+ "epoch": 3.914973391388486,
1928
+ "grad_norm": 0.34701117873191833,
1929
+ "learning_rate": 0.0004832970027247956,
1930
+ "loss": 3.6167,
1931
+ "step": 12950
1932
+ },
1933
+ {
1934
+ "epoch": 3.9300919206579583,
1935
+ "grad_norm": 0.33134695887565613,
1936
+ "learning_rate": 0.0004828428701180744,
1937
+ "loss": 3.6323,
1938
+ "step": 13000
1939
+ },
1940
+ {
1941
+ "epoch": 3.9300919206579583,
1942
+ "eval_accuracy": 0.3554904337217598,
1943
+ "eval_loss": 3.676191568374634,
1944
+ "eval_runtime": 182.5829,
1945
+ "eval_samples_per_second": 91.159,
1946
+ "eval_steps_per_second": 5.702,
1947
+ "step": 13000
1948
+ },
1949
+ {
1950
+ "epoch": 3.945210449927431,
1951
+ "grad_norm": 0.3298724293708801,
1952
+ "learning_rate": 0.0004823887375113533,
1953
+ "loss": 3.6291,
1954
+ "step": 13050
1955
+ },
1956
+ {
1957
+ "epoch": 3.9603289791969036,
1958
+ "grad_norm": 0.3415639102458954,
1959
+ "learning_rate": 0.00048193460490463213,
1960
+ "loss": 3.6256,
1961
+ "step": 13100
1962
+ },
1963
+ {
1964
+ "epoch": 3.9754475084663765,
1965
+ "grad_norm": 0.3238424062728882,
1966
+ "learning_rate": 0.0004814804722979109,
1967
+ "loss": 3.6304,
1968
+ "step": 13150
1969
+ },
1970
+ {
1971
+ "epoch": 3.990566037735849,
1972
+ "grad_norm": 0.31136244535446167,
1973
+ "learning_rate": 0.00048102633969118974,
1974
+ "loss": 3.6211,
1975
+ "step": 13200
1976
+ },
1977
+ {
1978
+ "epoch": 4.00544267053701,
1979
+ "grad_norm": 0.3414509892463684,
1980
+ "learning_rate": 0.0004805722070844686,
1981
+ "loss": 3.5841,
1982
+ "step": 13250
1983
+ },
1984
+ {
1985
+ "epoch": 4.020561199806483,
1986
+ "grad_norm": 0.3479083776473999,
1987
+ "learning_rate": 0.00048011807447774746,
1988
+ "loss": 3.5103,
1989
+ "step": 13300
1990
+ },
1991
+ {
1992
+ "epoch": 4.035679729075955,
1993
+ "grad_norm": 0.32956740260124207,
1994
+ "learning_rate": 0.0004796639418710263,
1995
+ "loss": 3.5115,
1996
+ "step": 13350
1997
+ },
1998
+ {
1999
+ "epoch": 4.050798258345428,
2000
+ "grad_norm": 0.32761216163635254,
2001
+ "learning_rate": 0.00047920980926430517,
2002
+ "loss": 3.5252,
2003
+ "step": 13400
2004
+ },
2005
+ {
2006
+ "epoch": 4.065916787614901,
2007
+ "grad_norm": 0.32798823714256287,
2008
+ "learning_rate": 0.000478755676657584,
2009
+ "loss": 3.529,
2010
+ "step": 13450
2011
+ },
2012
+ {
2013
+ "epoch": 4.081035316884374,
2014
+ "grad_norm": 0.3446704149246216,
2015
+ "learning_rate": 0.00047830154405086283,
2016
+ "loss": 3.5224,
2017
+ "step": 13500
2018
+ },
2019
+ {
2020
+ "epoch": 4.096153846153846,
2021
+ "grad_norm": 0.3277835547924042,
2022
+ "learning_rate": 0.0004778474114441416,
2023
+ "loss": 3.5262,
2024
+ "step": 13550
2025
+ },
2026
+ {
2027
+ "epoch": 4.111272375423319,
2028
+ "grad_norm": 0.3266860544681549,
2029
+ "learning_rate": 0.0004773932788374205,
2030
+ "loss": 3.5234,
2031
+ "step": 13600
2032
+ },
2033
+ {
2034
+ "epoch": 4.126390904692792,
2035
+ "grad_norm": 0.33199459314346313,
2036
+ "learning_rate": 0.0004769391462306993,
2037
+ "loss": 3.5329,
2038
+ "step": 13650
2039
+ },
2040
+ {
2041
+ "epoch": 4.1415094339622645,
2042
+ "grad_norm": 0.3167458176612854,
2043
+ "learning_rate": 0.00047648501362397816,
2044
+ "loss": 3.5315,
2045
+ "step": 13700
2046
+ },
2047
+ {
2048
+ "epoch": 4.1566279632317364,
2049
+ "grad_norm": 0.33163997530937195,
2050
+ "learning_rate": 0.000476030881017257,
2051
+ "loss": 3.539,
2052
+ "step": 13750
2053
+ },
2054
+ {
2055
+ "epoch": 4.171746492501209,
2056
+ "grad_norm": 0.3242467939853668,
2057
+ "learning_rate": 0.00047557674841053587,
2058
+ "loss": 3.5322,
2059
+ "step": 13800
2060
+ },
2061
+ {
2062
+ "epoch": 4.186865021770682,
2063
+ "grad_norm": 0.3438900113105774,
2064
+ "learning_rate": 0.0004751226158038147,
2065
+ "loss": 3.5399,
2066
+ "step": 13850
2067
+ },
2068
+ {
2069
+ "epoch": 4.201983551040155,
2070
+ "grad_norm": 0.34759825468063354,
2071
+ "learning_rate": 0.0004746684831970935,
2072
+ "loss": 3.5452,
2073
+ "step": 13900
2074
+ },
2075
+ {
2076
+ "epoch": 4.217102080309627,
2077
+ "grad_norm": 0.3428172171115875,
2078
+ "learning_rate": 0.0004742143505903723,
2079
+ "loss": 3.5468,
2080
+ "step": 13950
2081
+ },
2082
+ {
2083
+ "epoch": 4.2322206095791,
2084
+ "grad_norm": 0.3243090808391571,
2085
+ "learning_rate": 0.0004737602179836512,
2086
+ "loss": 3.5399,
2087
+ "step": 14000
2088
+ },
2089
+ {
2090
+ "epoch": 4.2322206095791,
2091
+ "eval_accuracy": 0.35642869605990957,
2092
+ "eval_loss": 3.6728808879852295,
2093
+ "eval_runtime": 182.5974,
2094
+ "eval_samples_per_second": 91.151,
2095
+ "eval_steps_per_second": 5.701,
2096
+ "step": 14000
2097
+ },
2098
+ {
2099
+ "epoch": 4.247339138848573,
2100
+ "grad_norm": 0.3331063985824585,
2101
+ "learning_rate": 0.00047330608537693,
2102
+ "loss": 3.5417,
2103
+ "step": 14050
2104
+ },
2105
+ {
2106
+ "epoch": 4.262457668118046,
2107
+ "grad_norm": 0.32661089301109314,
2108
+ "learning_rate": 0.00047285195277020886,
2109
+ "loss": 3.5406,
2110
+ "step": 14100
2111
+ },
2112
+ {
2113
+ "epoch": 4.277576197387518,
2114
+ "grad_norm": 0.33398354053497314,
2115
+ "learning_rate": 0.00047239782016348774,
2116
+ "loss": 3.5467,
2117
+ "step": 14150
2118
+ },
2119
+ {
2120
+ "epoch": 4.292694726656991,
2121
+ "grad_norm": 0.3145262897014618,
2122
+ "learning_rate": 0.00047194368755676657,
2123
+ "loss": 3.5434,
2124
+ "step": 14200
2125
+ },
2126
+ {
2127
+ "epoch": 4.3078132559264635,
2128
+ "grad_norm": 0.3353433609008789,
2129
+ "learning_rate": 0.00047148955495004535,
2130
+ "loss": 3.5485,
2131
+ "step": 14250
2132
+ },
2133
+ {
2134
+ "epoch": 4.322931785195936,
2135
+ "grad_norm": 0.33363988995552063,
2136
+ "learning_rate": 0.0004710354223433242,
2137
+ "loss": 3.5426,
2138
+ "step": 14300
2139
+ },
2140
+ {
2141
+ "epoch": 4.338050314465409,
2142
+ "grad_norm": 0.33351442217826843,
2143
+ "learning_rate": 0.00047058128973660306,
2144
+ "loss": 3.5549,
2145
+ "step": 14350
2146
+ },
2147
+ {
2148
+ "epoch": 4.353168843734881,
2149
+ "grad_norm": 0.3099449872970581,
2150
+ "learning_rate": 0.0004701271571298819,
2151
+ "loss": 3.5518,
2152
+ "step": 14400
2153
+ },
2154
+ {
2155
+ "epoch": 4.368287373004354,
2156
+ "grad_norm": 0.3316916823387146,
2157
+ "learning_rate": 0.0004696730245231607,
2158
+ "loss": 3.557,
2159
+ "step": 14450
2160
+ },
2161
+ {
2162
+ "epoch": 4.383405902273827,
2163
+ "grad_norm": 0.32294154167175293,
2164
+ "learning_rate": 0.00046921889191643956,
2165
+ "loss": 3.5462,
2166
+ "step": 14500
2167
+ },
2168
+ {
2169
+ "epoch": 4.3985244315433,
2170
+ "grad_norm": 0.32864052057266235,
2171
+ "learning_rate": 0.00046876475930971844,
2172
+ "loss": 3.5644,
2173
+ "step": 14550
2174
+ },
2175
+ {
2176
+ "epoch": 4.413642960812772,
2177
+ "grad_norm": 0.3232649564743042,
2178
+ "learning_rate": 0.0004683106267029972,
2179
+ "loss": 3.5485,
2180
+ "step": 14600
2181
+ },
2182
+ {
2183
+ "epoch": 4.428761490082245,
2184
+ "grad_norm": 0.33318766951560974,
2185
+ "learning_rate": 0.00046785649409627605,
2186
+ "loss": 3.5564,
2187
+ "step": 14650
2188
+ },
2189
+ {
2190
+ "epoch": 4.443880019351718,
2191
+ "grad_norm": 0.3175857663154602,
2192
+ "learning_rate": 0.0004674023614895549,
2193
+ "loss": 3.5574,
2194
+ "step": 14700
2195
+ },
2196
+ {
2197
+ "epoch": 4.4589985486211905,
2198
+ "grad_norm": 0.3410588502883911,
2199
+ "learning_rate": 0.00046694822888283376,
2200
+ "loss": 3.5549,
2201
+ "step": 14750
2202
+ },
2203
+ {
2204
+ "epoch": 4.4741170778906625,
2205
+ "grad_norm": 0.33438578248023987,
2206
+ "learning_rate": 0.0004664940962761126,
2207
+ "loss": 3.5644,
2208
+ "step": 14800
2209
+ },
2210
+ {
2211
+ "epoch": 4.489235607160135,
2212
+ "grad_norm": 0.3205425441265106,
2213
+ "learning_rate": 0.0004660399636693914,
2214
+ "loss": 3.5554,
2215
+ "step": 14850
2216
+ },
2217
+ {
2218
+ "epoch": 4.504354136429608,
2219
+ "grad_norm": 0.3268115222454071,
2220
+ "learning_rate": 0.0004655858310626703,
2221
+ "loss": 3.568,
2222
+ "step": 14900
2223
+ },
2224
+ {
2225
+ "epoch": 4.519472665699081,
2226
+ "grad_norm": 0.32924684882164,
2227
+ "learning_rate": 0.0004651316984559491,
2228
+ "loss": 3.5596,
2229
+ "step": 14950
2230
+ },
2231
+ {
2232
+ "epoch": 4.534591194968553,
2233
+ "grad_norm": 0.3317072093486786,
2234
+ "learning_rate": 0.0004646775658492279,
2235
+ "loss": 3.5538,
2236
+ "step": 15000
2237
+ },
2238
+ {
2239
+ "epoch": 4.534591194968553,
2240
+ "eval_accuracy": 0.35780751842074693,
2241
+ "eval_loss": 3.6592884063720703,
2242
+ "eval_runtime": 182.5425,
2243
+ "eval_samples_per_second": 91.179,
2244
+ "eval_steps_per_second": 5.703,
2245
+ "step": 15000
2246
+ },
2247
+ {
2248
+ "epoch": 4.549709724238026,
2249
+ "grad_norm": 0.3319750428199768,
2250
+ "learning_rate": 0.00046422343324250675,
2251
+ "loss": 3.5586,
2252
+ "step": 15050
2253
+ },
2254
+ {
2255
+ "epoch": 4.564828253507499,
2256
+ "grad_norm": 0.33184248208999634,
2257
+ "learning_rate": 0.00046376930063578563,
2258
+ "loss": 3.5611,
2259
+ "step": 15100
2260
+ },
2261
+ {
2262
+ "epoch": 4.579946782776972,
2263
+ "grad_norm": 0.3535155653953552,
2264
+ "learning_rate": 0.00046331516802906446,
2265
+ "loss": 3.56,
2266
+ "step": 15150
2267
+ },
2268
+ {
2269
+ "epoch": 4.595065312046444,
2270
+ "grad_norm": 0.322519987821579,
2271
+ "learning_rate": 0.0004628610354223433,
2272
+ "loss": 3.5594,
2273
+ "step": 15200
2274
+ },
2275
+ {
2276
+ "epoch": 4.610183841315917,
2277
+ "grad_norm": 0.31665462255477905,
2278
+ "learning_rate": 0.00046240690281562207,
2279
+ "loss": 3.5673,
2280
+ "step": 15250
2281
+ },
2282
+ {
2283
+ "epoch": 4.6253023705853895,
2284
+ "grad_norm": 0.341170072555542,
2285
+ "learning_rate": 0.00046195277020890096,
2286
+ "loss": 3.5536,
2287
+ "step": 15300
2288
+ },
2289
+ {
2290
+ "epoch": 4.640420899854862,
2291
+ "grad_norm": 0.3396996557712555,
2292
+ "learning_rate": 0.0004614986376021798,
2293
+ "loss": 3.5653,
2294
+ "step": 15350
2295
+ },
2296
+ {
2297
+ "epoch": 4.655539429124335,
2298
+ "grad_norm": 0.33313286304473877,
2299
+ "learning_rate": 0.0004610445049954586,
2300
+ "loss": 3.5667,
2301
+ "step": 15400
2302
+ },
2303
+ {
2304
+ "epoch": 4.670657958393807,
2305
+ "grad_norm": 0.3346698582172394,
2306
+ "learning_rate": 0.00046059037238873745,
2307
+ "loss": 3.5552,
2308
+ "step": 15450
2309
+ },
2310
+ {
2311
+ "epoch": 4.68577648766328,
2312
+ "grad_norm": 0.32503342628479004,
2313
+ "learning_rate": 0.00046013623978201633,
2314
+ "loss": 3.5628,
2315
+ "step": 15500
2316
+ },
2317
+ {
2318
+ "epoch": 4.700895016932753,
2319
+ "grad_norm": 0.3317323327064514,
2320
+ "learning_rate": 0.00045968210717529516,
2321
+ "loss": 3.5598,
2322
+ "step": 15550
2323
+ },
2324
+ {
2325
+ "epoch": 4.716013546202225,
2326
+ "grad_norm": 0.31406593322753906,
2327
+ "learning_rate": 0.00045922797456857394,
2328
+ "loss": 3.5652,
2329
+ "step": 15600
2330
+ },
2331
+ {
2332
+ "epoch": 4.731132075471698,
2333
+ "grad_norm": 0.35221582651138306,
2334
+ "learning_rate": 0.0004587738419618528,
2335
+ "loss": 3.5625,
2336
+ "step": 15650
2337
+ },
2338
+ {
2339
+ "epoch": 4.746250604741171,
2340
+ "grad_norm": 0.3168243169784546,
2341
+ "learning_rate": 0.00045831970935513166,
2342
+ "loss": 3.5446,
2343
+ "step": 15700
2344
+ },
2345
+ {
2346
+ "epoch": 4.761369134010644,
2347
+ "grad_norm": 0.3266247510910034,
2348
+ "learning_rate": 0.0004578655767484105,
2349
+ "loss": 3.5824,
2350
+ "step": 15750
2351
+ },
2352
+ {
2353
+ "epoch": 4.7764876632801165,
2354
+ "grad_norm": 0.3345303535461426,
2355
+ "learning_rate": 0.0004574114441416893,
2356
+ "loss": 3.5577,
2357
+ "step": 15800
2358
+ },
2359
+ {
2360
+ "epoch": 4.7916061925495885,
2361
+ "grad_norm": 0.36775925755500793,
2362
+ "learning_rate": 0.0004569573115349682,
2363
+ "loss": 3.5736,
2364
+ "step": 15850
2365
+ },
2366
+ {
2367
+ "epoch": 4.806724721819061,
2368
+ "grad_norm": 0.31279629468917847,
2369
+ "learning_rate": 0.00045650317892824703,
2370
+ "loss": 3.5719,
2371
+ "step": 15900
2372
+ },
2373
+ {
2374
+ "epoch": 4.821843251088534,
2375
+ "grad_norm": 0.32946261763572693,
2376
+ "learning_rate": 0.0004560490463215258,
2377
+ "loss": 3.5472,
2378
+ "step": 15950
2379
+ },
2380
+ {
2381
+ "epoch": 4.836961780358006,
2382
+ "grad_norm": 0.3353503942489624,
2383
+ "learning_rate": 0.00045559491371480464,
2384
+ "loss": 3.5512,
2385
+ "step": 16000
2386
+ },
2387
+ {
2388
+ "epoch": 4.836961780358006,
2389
+ "eval_accuracy": 0.3589908106727694,
2390
+ "eval_loss": 3.6440577507019043,
2391
+ "eval_runtime": 182.5579,
2392
+ "eval_samples_per_second": 91.171,
2393
+ "eval_steps_per_second": 5.702,
2394
+ "step": 16000
2395
+ },
2396
+ {
2397
+ "epoch": 4.852080309627479,
2398
+ "grad_norm": 0.3155944347381592,
2399
+ "learning_rate": 0.0004551407811080835,
2400
+ "loss": 3.5697,
2401
+ "step": 16050
2402
+ },
2403
+ {
2404
+ "epoch": 4.867198838896952,
2405
+ "grad_norm": 0.34022998809814453,
2406
+ "learning_rate": 0.00045468664850136235,
2407
+ "loss": 3.5635,
2408
+ "step": 16100
2409
+ },
2410
+ {
2411
+ "epoch": 4.882317368166425,
2412
+ "grad_norm": 0.349436491727829,
2413
+ "learning_rate": 0.0004542325158946412,
2414
+ "loss": 3.5606,
2415
+ "step": 16150
2416
+ },
2417
+ {
2418
+ "epoch": 4.897435897435898,
2419
+ "grad_norm": 0.32045167684555054,
2420
+ "learning_rate": 0.00045377838328792,
2421
+ "loss": 3.5483,
2422
+ "step": 16200
2423
+ },
2424
+ {
2425
+ "epoch": 4.91255442670537,
2426
+ "grad_norm": 0.3273170292377472,
2427
+ "learning_rate": 0.0004533242506811989,
2428
+ "loss": 3.561,
2429
+ "step": 16250
2430
+ },
2431
+ {
2432
+ "epoch": 4.927672955974843,
2433
+ "grad_norm": 0.3492163121700287,
2434
+ "learning_rate": 0.00045287011807447773,
2435
+ "loss": 3.5538,
2436
+ "step": 16300
2437
+ },
2438
+ {
2439
+ "epoch": 4.9427914852443156,
2440
+ "grad_norm": 0.3319571018218994,
2441
+ "learning_rate": 0.0004524159854677565,
2442
+ "loss": 3.5678,
2443
+ "step": 16350
2444
+ },
2445
+ {
2446
+ "epoch": 4.957910014513788,
2447
+ "grad_norm": 0.3193274438381195,
2448
+ "learning_rate": 0.0004519618528610354,
2449
+ "loss": 3.5604,
2450
+ "step": 16400
2451
+ },
2452
+ {
2453
+ "epoch": 4.97302854378326,
2454
+ "grad_norm": 0.3205724060535431,
2455
+ "learning_rate": 0.0004515077202543142,
2456
+ "loss": 3.5501,
2457
+ "step": 16450
2458
+ },
2459
+ {
2460
+ "epoch": 4.988147073052733,
2461
+ "grad_norm": 0.32597634196281433,
2462
+ "learning_rate": 0.00045105358764759305,
2463
+ "loss": 3.5648,
2464
+ "step": 16500
2465
+ },
2466
+ {
2467
+ "epoch": 5.003023705853894,
2468
+ "grad_norm": 0.3266983926296234,
2469
+ "learning_rate": 0.0004505994550408719,
2470
+ "loss": 3.5279,
2471
+ "step": 16550
2472
+ },
2473
+ {
2474
+ "epoch": 5.018142235123367,
2475
+ "grad_norm": 0.32244426012039185,
2476
+ "learning_rate": 0.00045014532243415077,
2477
+ "loss": 3.4542,
2478
+ "step": 16600
2479
+ },
2480
+ {
2481
+ "epoch": 5.03326076439284,
2482
+ "grad_norm": 0.32874855399131775,
2483
+ "learning_rate": 0.0004496911898274296,
2484
+ "loss": 3.4671,
2485
+ "step": 16650
2486
+ },
2487
+ {
2488
+ "epoch": 5.048379293662313,
2489
+ "grad_norm": 0.34664833545684814,
2490
+ "learning_rate": 0.0004492370572207084,
2491
+ "loss": 3.4582,
2492
+ "step": 16700
2493
+ },
2494
+ {
2495
+ "epoch": 5.063497822931785,
2496
+ "grad_norm": 0.3226701021194458,
2497
+ "learning_rate": 0.0004487829246139872,
2498
+ "loss": 3.4471,
2499
+ "step": 16750
2500
+ },
2501
+ {
2502
+ "epoch": 5.078616352201258,
2503
+ "grad_norm": 0.3484433889389038,
2504
+ "learning_rate": 0.0004483287920072661,
2505
+ "loss": 3.4685,
2506
+ "step": 16800
2507
+ },
2508
+ {
2509
+ "epoch": 5.093734881470731,
2510
+ "grad_norm": 0.338821142911911,
2511
+ "learning_rate": 0.0004478746594005449,
2512
+ "loss": 3.4738,
2513
+ "step": 16850
2514
+ },
2515
+ {
2516
+ "epoch": 5.1088534107402035,
2517
+ "grad_norm": 0.3581444323062897,
2518
+ "learning_rate": 0.00044742052679382375,
2519
+ "loss": 3.4712,
2520
+ "step": 16900
2521
+ },
2522
+ {
2523
+ "epoch": 5.1239719400096755,
2524
+ "grad_norm": 0.34923458099365234,
2525
+ "learning_rate": 0.00044696639418710264,
2526
+ "loss": 3.4607,
2527
+ "step": 16950
2528
+ },
2529
+ {
2530
+ "epoch": 5.139090469279148,
2531
+ "grad_norm": 0.33646735548973083,
2532
+ "learning_rate": 0.00044651226158038147,
2533
+ "loss": 3.4722,
2534
+ "step": 17000
2535
+ },
2536
+ {
2537
+ "epoch": 5.139090469279148,
2538
+ "eval_accuracy": 0.35967545999545686,
2539
+ "eval_loss": 3.646660327911377,
2540
+ "eval_runtime": 182.5899,
2541
+ "eval_samples_per_second": 91.155,
2542
+ "eval_steps_per_second": 5.701,
2543
+ "step": 17000
2544
+ },
2545
+ {
2546
+ "epoch": 5.154208998548621,
2547
+ "grad_norm": 0.33612239360809326,
2548
+ "learning_rate": 0.00044605812897366025,
2549
+ "loss": 3.4686,
2550
+ "step": 17050
2551
+ },
2552
+ {
2553
+ "epoch": 5.169327527818094,
2554
+ "grad_norm": 0.3267473876476288,
2555
+ "learning_rate": 0.0004456039963669391,
2556
+ "loss": 3.474,
2557
+ "step": 17100
2558
+ },
2559
+ {
2560
+ "epoch": 5.184446057087566,
2561
+ "grad_norm": 0.33728793263435364,
2562
+ "learning_rate": 0.00044514986376021796,
2563
+ "loss": 3.4742,
2564
+ "step": 17150
2565
+ },
2566
+ {
2567
+ "epoch": 5.199564586357039,
2568
+ "grad_norm": 0.33321622014045715,
2569
+ "learning_rate": 0.0004446957311534968,
2570
+ "loss": 3.4825,
2571
+ "step": 17200
2572
+ },
2573
+ {
2574
+ "epoch": 5.214683115626512,
2575
+ "grad_norm": 0.3262038230895996,
2576
+ "learning_rate": 0.0004442415985467756,
2577
+ "loss": 3.4769,
2578
+ "step": 17250
2579
+ },
2580
+ {
2581
+ "epoch": 5.229801644895985,
2582
+ "grad_norm": 0.3380536139011383,
2583
+ "learning_rate": 0.00044378746594005445,
2584
+ "loss": 3.4779,
2585
+ "step": 17300
2586
+ },
2587
+ {
2588
+ "epoch": 5.244920174165458,
2589
+ "grad_norm": 0.32373255491256714,
2590
+ "learning_rate": 0.00044333333333333334,
2591
+ "loss": 3.4721,
2592
+ "step": 17350
2593
+ },
2594
+ {
2595
+ "epoch": 5.26003870343493,
2596
+ "grad_norm": 0.34807828068733215,
2597
+ "learning_rate": 0.0004428792007266121,
2598
+ "loss": 3.4817,
2599
+ "step": 17400
2600
+ },
2601
+ {
2602
+ "epoch": 5.2751572327044025,
2603
+ "grad_norm": 0.34097328782081604,
2604
+ "learning_rate": 0.00044242506811989095,
2605
+ "loss": 3.4845,
2606
+ "step": 17450
2607
+ },
2608
+ {
2609
+ "epoch": 5.290275761973875,
2610
+ "grad_norm": 0.33616551756858826,
2611
+ "learning_rate": 0.0004419709355131698,
2612
+ "loss": 3.5016,
2613
+ "step": 17500
2614
+ },
2615
+ {
2616
+ "epoch": 5.305394291243347,
2617
+ "grad_norm": 0.3330867886543274,
2618
+ "learning_rate": 0.00044151680290644866,
2619
+ "loss": 3.4844,
2620
+ "step": 17550
2621
+ },
2622
+ {
2623
+ "epoch": 5.32051282051282,
2624
+ "grad_norm": 0.3476291298866272,
2625
+ "learning_rate": 0.0004410626702997275,
2626
+ "loss": 3.4877,
2627
+ "step": 17600
2628
+ },
2629
+ {
2630
+ "epoch": 5.335631349782293,
2631
+ "grad_norm": 0.3363373577594757,
2632
+ "learning_rate": 0.0004406085376930063,
2633
+ "loss": 3.4931,
2634
+ "step": 17650
2635
+ },
2636
+ {
2637
+ "epoch": 5.350749879051766,
2638
+ "grad_norm": 0.35095226764678955,
2639
+ "learning_rate": 0.0004401544050862852,
2640
+ "loss": 3.4904,
2641
+ "step": 17700
2642
+ },
2643
+ {
2644
+ "epoch": 5.365868408321239,
2645
+ "grad_norm": 0.35196179151535034,
2646
+ "learning_rate": 0.000439700272479564,
2647
+ "loss": 3.5003,
2648
+ "step": 17750
2649
+ },
2650
+ {
2651
+ "epoch": 5.380986937590711,
2652
+ "grad_norm": 0.3490413427352905,
2653
+ "learning_rate": 0.0004392461398728428,
2654
+ "loss": 3.4914,
2655
+ "step": 17800
2656
+ },
2657
+ {
2658
+ "epoch": 5.396105466860184,
2659
+ "grad_norm": 0.3287891447544098,
2660
+ "learning_rate": 0.00043879200726612165,
2661
+ "loss": 3.482,
2662
+ "step": 17850
2663
+ },
2664
+ {
2665
+ "epoch": 5.411223996129657,
2666
+ "grad_norm": 0.3373957574367523,
2667
+ "learning_rate": 0.00043833787465940053,
2668
+ "loss": 3.4928,
2669
+ "step": 17900
2670
+ },
2671
+ {
2672
+ "epoch": 5.4263425253991295,
2673
+ "grad_norm": 0.3305628299713135,
2674
+ "learning_rate": 0.00043788374205267936,
2675
+ "loss": 3.4918,
2676
+ "step": 17950
2677
+ },
2678
+ {
2679
+ "epoch": 5.4414610546686015,
2680
+ "grad_norm": 0.33179208636283875,
2681
+ "learning_rate": 0.0004374296094459582,
2682
+ "loss": 3.5072,
2683
+ "step": 18000
2684
+ },
2685
+ {
2686
+ "epoch": 5.4414610546686015,
2687
+ "eval_accuracy": 0.36049544013909796,
2688
+ "eval_loss": 3.6349031925201416,
2689
+ "eval_runtime": 182.6046,
2690
+ "eval_samples_per_second": 91.148,
2691
+ "eval_steps_per_second": 5.701,
2692
+ "step": 18000
2693
+ },
2694
+ {
2695
+ "epoch": 5.456579583938074,
2696
+ "grad_norm": 0.3481210172176361,
2697
+ "learning_rate": 0.00043697547683923697,
2698
+ "loss": 3.5004,
2699
+ "step": 18050
2700
+ },
2701
+ {
2702
+ "epoch": 5.471698113207547,
2703
+ "grad_norm": 0.34053078293800354,
2704
+ "learning_rate": 0.00043652134423251585,
2705
+ "loss": 3.4979,
2706
+ "step": 18100
2707
+ },
2708
+ {
2709
+ "epoch": 5.48681664247702,
2710
+ "grad_norm": 0.3321893811225891,
2711
+ "learning_rate": 0.0004360672116257947,
2712
+ "loss": 3.5042,
2713
+ "step": 18150
2714
+ },
2715
+ {
2716
+ "epoch": 5.501935171746492,
2717
+ "grad_norm": 0.3595956861972809,
2718
+ "learning_rate": 0.0004356130790190735,
2719
+ "loss": 3.4927,
2720
+ "step": 18200
2721
+ },
2722
+ {
2723
+ "epoch": 5.517053701015965,
2724
+ "grad_norm": 0.33346107602119446,
2725
+ "learning_rate": 0.00043515894641235235,
2726
+ "loss": 3.4992,
2727
+ "step": 18250
2728
+ },
2729
+ {
2730
+ "epoch": 5.532172230285438,
2731
+ "grad_norm": 0.3272981345653534,
2732
+ "learning_rate": 0.00043470481380563123,
2733
+ "loss": 3.4941,
2734
+ "step": 18300
2735
+ },
2736
+ {
2737
+ "epoch": 5.547290759554911,
2738
+ "grad_norm": 0.3254975378513336,
2739
+ "learning_rate": 0.00043425068119891006,
2740
+ "loss": 3.5031,
2741
+ "step": 18350
2742
+ },
2743
+ {
2744
+ "epoch": 5.562409288824383,
2745
+ "grad_norm": 0.3381115198135376,
2746
+ "learning_rate": 0.00043379654859218884,
2747
+ "loss": 3.5026,
2748
+ "step": 18400
2749
+ },
2750
+ {
2751
+ "epoch": 5.577527818093856,
2752
+ "grad_norm": 0.35261309146881104,
2753
+ "learning_rate": 0.0004333424159854677,
2754
+ "loss": 3.5006,
2755
+ "step": 18450
2756
+ },
2757
+ {
2758
+ "epoch": 5.5926463473633286,
2759
+ "grad_norm": 0.3541949987411499,
2760
+ "learning_rate": 0.00043288828337874655,
2761
+ "loss": 3.5055,
2762
+ "step": 18500
2763
+ },
2764
+ {
2765
+ "epoch": 5.607764876632801,
2766
+ "grad_norm": 0.34113311767578125,
2767
+ "learning_rate": 0.0004324341507720254,
2768
+ "loss": 3.5077,
2769
+ "step": 18550
2770
+ },
2771
+ {
2772
+ "epoch": 5.622883405902273,
2773
+ "grad_norm": 0.32559531927108765,
2774
+ "learning_rate": 0.0004319800181653042,
2775
+ "loss": 3.509,
2776
+ "step": 18600
2777
+ },
2778
+ {
2779
+ "epoch": 5.638001935171746,
2780
+ "grad_norm": 0.3446948528289795,
2781
+ "learning_rate": 0.0004315258855585831,
2782
+ "loss": 3.504,
2783
+ "step": 18650
2784
+ },
2785
+ {
2786
+ "epoch": 5.653120464441219,
2787
+ "grad_norm": 0.33571872115135193,
2788
+ "learning_rate": 0.00043107175295186193,
2789
+ "loss": 3.4972,
2790
+ "step": 18700
2791
+ },
2792
+ {
2793
+ "epoch": 5.668238993710692,
2794
+ "grad_norm": 0.331429660320282,
2795
+ "learning_rate": 0.0004306176203451407,
2796
+ "loss": 3.4938,
2797
+ "step": 18750
2798
+ },
2799
+ {
2800
+ "epoch": 5.683357522980165,
2801
+ "grad_norm": 0.3403501510620117,
2802
+ "learning_rate": 0.00043016348773841954,
2803
+ "loss": 3.5087,
2804
+ "step": 18800
2805
+ },
2806
+ {
2807
+ "epoch": 5.698476052249637,
2808
+ "grad_norm": 0.3406428098678589,
2809
+ "learning_rate": 0.0004297093551316984,
2810
+ "loss": 3.5111,
2811
+ "step": 18850
2812
+ },
2813
+ {
2814
+ "epoch": 5.71359458151911,
2815
+ "grad_norm": 0.33290889859199524,
2816
+ "learning_rate": 0.00042925522252497725,
2817
+ "loss": 3.4946,
2818
+ "step": 18900
2819
+ },
2820
+ {
2821
+ "epoch": 5.728713110788583,
2822
+ "grad_norm": 0.3506568670272827,
2823
+ "learning_rate": 0.0004288010899182561,
2824
+ "loss": 3.4977,
2825
+ "step": 18950
2826
+ },
2827
+ {
2828
+ "epoch": 5.743831640058055,
2829
+ "grad_norm": 0.32340604066848755,
2830
+ "learning_rate": 0.0004283469573115349,
2831
+ "loss": 3.501,
2832
+ "step": 19000
2833
+ },
2834
+ {
2835
+ "epoch": 5.743831640058055,
2836
+ "eval_accuracy": 0.3617825526473342,
2837
+ "eval_loss": 3.6229019165039062,
2838
+ "eval_runtime": 182.4602,
2839
+ "eval_samples_per_second": 91.22,
2840
+ "eval_steps_per_second": 5.705,
2841
+ "step": 19000
2842
+ },
2843
+ {
2844
+ "epoch": 5.758950169327528,
2845
+ "grad_norm": 0.3353550434112549,
2846
+ "learning_rate": 0.0004278928247048138,
2847
+ "loss": 3.4941,
2848
+ "step": 19050
2849
+ },
2850
+ {
2851
+ "epoch": 5.7740686985970004,
2852
+ "grad_norm": 0.33119282126426697,
2853
+ "learning_rate": 0.00042743869209809263,
2854
+ "loss": 3.5137,
2855
+ "step": 19100
2856
+ },
2857
+ {
2858
+ "epoch": 5.789187227866473,
2859
+ "grad_norm": 0.33732518553733826,
2860
+ "learning_rate": 0.0004269845594913714,
2861
+ "loss": 3.5052,
2862
+ "step": 19150
2863
+ },
2864
+ {
2865
+ "epoch": 5.804305757135946,
2866
+ "grad_norm": 0.32491031289100647,
2867
+ "learning_rate": 0.0004265304268846503,
2868
+ "loss": 3.505,
2869
+ "step": 19200
2870
+ },
2871
+ {
2872
+ "epoch": 5.819424286405418,
2873
+ "grad_norm": 0.328688383102417,
2874
+ "learning_rate": 0.0004260762942779291,
2875
+ "loss": 3.5022,
2876
+ "step": 19250
2877
+ },
2878
+ {
2879
+ "epoch": 5.834542815674891,
2880
+ "grad_norm": 0.33090704679489136,
2881
+ "learning_rate": 0.00042562216167120795,
2882
+ "loss": 3.5023,
2883
+ "step": 19300
2884
+ },
2885
+ {
2886
+ "epoch": 5.849661344944364,
2887
+ "grad_norm": 0.3403943181037903,
2888
+ "learning_rate": 0.0004251680290644868,
2889
+ "loss": 3.4987,
2890
+ "step": 19350
2891
+ },
2892
+ {
2893
+ "epoch": 5.864779874213837,
2894
+ "grad_norm": 0.33841657638549805,
2895
+ "learning_rate": 0.00042471389645776567,
2896
+ "loss": 3.5105,
2897
+ "step": 19400
2898
+ },
2899
+ {
2900
+ "epoch": 5.879898403483309,
2901
+ "grad_norm": 0.31852486729621887,
2902
+ "learning_rate": 0.0004242597638510445,
2903
+ "loss": 3.5051,
2904
+ "step": 19450
2905
+ },
2906
+ {
2907
+ "epoch": 5.895016932752782,
2908
+ "grad_norm": 0.3277070224285126,
2909
+ "learning_rate": 0.0004238056312443233,
2910
+ "loss": 3.4959,
2911
+ "step": 19500
2912
+ },
2913
+ {
2914
+ "epoch": 5.910135462022255,
2915
+ "grad_norm": 0.32635408639907837,
2916
+ "learning_rate": 0.0004233514986376021,
2917
+ "loss": 3.5042,
2918
+ "step": 19550
2919
+ },
2920
+ {
2921
+ "epoch": 5.9252539912917275,
2922
+ "grad_norm": 0.32446935772895813,
2923
+ "learning_rate": 0.000422897366030881,
2924
+ "loss": 3.5092,
2925
+ "step": 19600
2926
+ },
2927
+ {
2928
+ "epoch": 5.9403725205611995,
2929
+ "grad_norm": 0.33674439787864685,
2930
+ "learning_rate": 0.0004224432334241598,
2931
+ "loss": 3.4991,
2932
+ "step": 19650
2933
+ },
2934
+ {
2935
+ "epoch": 5.955491049830672,
2936
+ "grad_norm": 0.32446402311325073,
2937
+ "learning_rate": 0.00042198910081743865,
2938
+ "loss": 3.5077,
2939
+ "step": 19700
2940
+ },
2941
+ {
2942
+ "epoch": 5.970609579100145,
2943
+ "grad_norm": 0.3281049132347107,
2944
+ "learning_rate": 0.0004215349682107175,
2945
+ "loss": 3.5125,
2946
+ "step": 19750
2947
+ },
2948
+ {
2949
+ "epoch": 5.985728108369618,
2950
+ "grad_norm": 0.33700883388519287,
2951
+ "learning_rate": 0.00042108083560399637,
2952
+ "loss": 3.5084,
2953
+ "step": 19800
2954
+ },
2955
+ {
2956
+ "epoch": 6.000604741170779,
2957
+ "grad_norm": 0.35103023052215576,
2958
+ "learning_rate": 0.00042062670299727515,
2959
+ "loss": 3.5109,
2960
+ "step": 19850
2961
+ },
2962
+ {
2963
+ "epoch": 6.015723270440252,
2964
+ "grad_norm": 0.33696576952934265,
2965
+ "learning_rate": 0.000420172570390554,
2966
+ "loss": 3.4014,
2967
+ "step": 19900
2968
+ },
2969
+ {
2970
+ "epoch": 6.030841799709724,
2971
+ "grad_norm": 0.3408699631690979,
2972
+ "learning_rate": 0.00041971843778383286,
2973
+ "loss": 3.3988,
2974
+ "step": 19950
2975
+ },
2976
+ {
2977
+ "epoch": 6.045960328979197,
2978
+ "grad_norm": 0.340057909488678,
2979
+ "learning_rate": 0.0004192643051771117,
2980
+ "loss": 3.411,
2981
+ "step": 20000
2982
+ },
2983
+ {
2984
+ "epoch": 6.045960328979197,
2985
+ "eval_accuracy": 0.3621478635601953,
2986
+ "eval_loss": 3.6226511001586914,
2987
+ "eval_runtime": 181.7969,
2988
+ "eval_samples_per_second": 91.553,
2989
+ "eval_steps_per_second": 5.726,
2990
+ "step": 20000
2991
+ }
2992
+ ],
2993
+ "logging_steps": 50,
2994
+ "max_steps": 66160,
2995
+ "num_input_tokens_seen": 0,
2996
+ "num_train_epochs": 20,
2997
+ "save_steps": 10000,
2998
+ "stateful_callbacks": {
2999
+ "EarlyStoppingCallback": {
3000
+ "args": {
3001
+ "early_stopping_patience": 20,
3002
+ "early_stopping_threshold": 0.0
3003
+ },
3004
+ "attributes": {
3005
+ "early_stopping_patience_counter": 0
3006
+ }
3007
+ },
3008
+ "TrainerControl": {
3009
+ "args": {
3010
+ "should_epoch_stop": false,
3011
+ "should_evaluate": false,
3012
+ "should_log": false,
3013
+ "should_save": true,
3014
+ "should_training_stop": false
3015
+ },
3016
+ "attributes": {}
3017
+ }
3018
+ },
3019
+ "total_flos": 4.17966915059712e+17,
3020
+ "train_batch_size": 16,
3021
+ "trial_name": null,
3022
+ "trial_params": null
3023
+ }
cost_to_hit_frequency_40817/checkpoint-20000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8546300ffb9b427a753c00bff33ba1fe16815471c1a5cb2b8bab250adeda22b1
3
+ size 5969
cost_to_hit_frequency_40817/checkpoint-20000/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-40000/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "embd_pdrop": 0.1,
9
+ "eos_token_id": 50256,
10
+ "initializer_range": 0.02,
11
+ "layer_norm_epsilon": 1e-05,
12
+ "model_type": "gpt2",
13
+ "n_embd": 768,
14
+ "n_head": 12,
15
+ "n_inner": null,
16
+ "n_layer": 12,
17
+ "n_positions": 1024,
18
+ "reorder_and_upcast_attn": false,
19
+ "resid_pdrop": 0.1,
20
+ "scale_attn_by_inverse_layer_idx": false,
21
+ "scale_attn_weights": true,
22
+ "summary_activation": null,
23
+ "summary_first_dropout": 0.1,
24
+ "summary_proj_to_labels": true,
25
+ "summary_type": "cls_index",
26
+ "summary_use_proj": true,
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.55.2",
29
+ "use_cache": true,
30
+ "vocab_size": 50257
31
+ }
cost_to_hit_frequency_40817/checkpoint-40000/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.55.2"
6
+ }
cost_to_hit_frequency_40817/checkpoint-40000/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-40000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cfe66f173631bbe317ebc3f3bca2cac1e720714fd07372bd1272558aee49324
3
+ size 497774208
cost_to_hit_frequency_40817/checkpoint-40000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ced5d4bb12099f14858627b9f68d7e724eabeac93f4d322c22ba1067f13d3fe8
3
+ size 995644811
cost_to_hit_frequency_40817/checkpoint-40000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c00c6d979a905e7518e13569aa610783bd4f467482b60c14d918b67dd030e3f1
3
+ size 14581
cost_to_hit_frequency_40817/checkpoint-40000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a9fe57b6e1cf172625c121450482a78cbe2481811cf12789233ec07c4c838a7
3
+ size 1383
cost_to_hit_frequency_40817/checkpoint-40000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62019b5edec75b6bf4a0330b70f3e92cd7e517dea57dc53d6e9e2fb1a74ef6e5
3
+ size 1465
cost_to_hit_frequency_40817/checkpoint-40000/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
cost_to_hit_frequency_40817/checkpoint-40000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-40000/tokenizer_config.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "50256": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ }
12
+ },
13
+ "bos_token": "<|endoftext|>",
14
+ "clean_up_tokenization_spaces": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "extra_special_tokens": {},
17
+ "model_max_length": 1024,
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
cost_to_hit_frequency_40817/checkpoint-40000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-40000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8546300ffb9b427a753c00bff33ba1fe16815471c1a5cb2b8bab250adeda22b1
3
+ size 5969
cost_to_hit_frequency_40817/checkpoint-40000/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-50000/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "embd_pdrop": 0.1,
9
+ "eos_token_id": 50256,
10
+ "initializer_range": 0.02,
11
+ "layer_norm_epsilon": 1e-05,
12
+ "model_type": "gpt2",
13
+ "n_embd": 768,
14
+ "n_head": 12,
15
+ "n_inner": null,
16
+ "n_layer": 12,
17
+ "n_positions": 1024,
18
+ "reorder_and_upcast_attn": false,
19
+ "resid_pdrop": 0.1,
20
+ "scale_attn_by_inverse_layer_idx": false,
21
+ "scale_attn_weights": true,
22
+ "summary_activation": null,
23
+ "summary_first_dropout": 0.1,
24
+ "summary_proj_to_labels": true,
25
+ "summary_type": "cls_index",
26
+ "summary_use_proj": true,
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.55.2",
29
+ "use_cache": true,
30
+ "vocab_size": 50257
31
+ }
cost_to_hit_frequency_40817/checkpoint-50000/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.55.2"
6
+ }
cost_to_hit_frequency_40817/checkpoint-50000/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-50000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4494ce94070fa0f364ca697e38289787f6733313bfef8b82bc0d405d586a85b1
3
+ size 497774208
cost_to_hit_frequency_40817/checkpoint-50000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6f09f61eaf4a15837b48bddcb939900e6a614ba28ef0847698635b6698c5e181
3
+ size 995644811
cost_to_hit_frequency_40817/checkpoint-50000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22b19b1fee3ff594a3337743ee57fdd43ffdbf4703f07bab621ddfceb3d67983
3
+ size 14581
cost_to_hit_frequency_40817/checkpoint-50000/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e1fba39ca35d203b6b453472c51ba43cdee1639cb09cd1a79870eb4a6956167
3
+ size 1383
cost_to_hit_frequency_40817/checkpoint-50000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cab4f6b7a0053a24007cf8c31d7b503dbe60a742c74df50345e22bd4277921d5
3
+ size 1465
cost_to_hit_frequency_40817/checkpoint-50000/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
cost_to_hit_frequency_40817/checkpoint-50000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-50000/tokenizer_config.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "50256": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ }
12
+ },
13
+ "bos_token": "<|endoftext|>",
14
+ "clean_up_tokenization_spaces": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "extra_special_tokens": {},
17
+ "model_max_length": 1024,
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
cost_to_hit_frequency_40817/checkpoint-50000/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-50000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8546300ffb9b427a753c00bff33ba1fe16815471c1a5cb2b8bab250adeda22b1
3
+ size 5969
cost_to_hit_frequency_40817/checkpoint-50000/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-60000/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "embd_pdrop": 0.1,
9
+ "eos_token_id": 50256,
10
+ "initializer_range": 0.02,
11
+ "layer_norm_epsilon": 1e-05,
12
+ "model_type": "gpt2",
13
+ "n_embd": 768,
14
+ "n_head": 12,
15
+ "n_inner": null,
16
+ "n_layer": 12,
17
+ "n_positions": 1024,
18
+ "reorder_and_upcast_attn": false,
19
+ "resid_pdrop": 0.1,
20
+ "scale_attn_by_inverse_layer_idx": false,
21
+ "scale_attn_weights": true,
22
+ "summary_activation": null,
23
+ "summary_first_dropout": 0.1,
24
+ "summary_proj_to_labels": true,
25
+ "summary_type": "cls_index",
26
+ "summary_use_proj": true,
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.55.2",
29
+ "use_cache": true,
30
+ "vocab_size": 50257
31
+ }
cost_to_hit_frequency_40817/checkpoint-60000/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.55.2"
6
+ }
cost_to_hit_frequency_40817/checkpoint-60000/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
cost_to_hit_frequency_40817/checkpoint-60000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:309a79b7aa151ca9463e920bb94cf1baebc984d12b6ea081b83193e9fe2fba3d
3
+ size 497774208
cost_to_hit_frequency_40817/checkpoint-60000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e55882fae70f12a3b439d16a325bb8fb3431717f63bd4048ec6e27a871bbbf7
3
+ size 995644811
cost_to_hit_frequency_40817/checkpoint-60000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77c12a7f4c17947023ca2a8a66613aab64446fdc2655f6779b274cd0eee71672
3
+ size 14581