Hugging Face's logo

Johnny050407
/
QCQC

Safetensors
Model card Files
xet
 Community
Johnny050407 commited on
Commit
80510f0
·
verified ·
1 Parent(s): 8e1dc0b

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. .gitattributes +1 -0
  2. .gitignore +1 -8
  3. coco_faiss_indexes/CLIP-Huge-Flickr-Flat/faiss_IVPQ_PCA.index +3 -0
  4. coco_faiss_indexes/CLIP-Huge-Flickr-Flat/img_ids.npy +3 -0
  5. coco_faiss_indexes/CLIP-Huge-Flickr-Flat/norm1.bin +3 -0
  6. coco_feats/CLIP-Huge-Flickr-Flat_feats.npy +3 -0
  7. outputs/gpt2cocoADS/model_ADS/checkpoint/added_tokens.json +4 -0
  8. outputs/gpt2cocoADS/model_ADS/checkpoint/config.json +40 -0
  9. outputs/gpt2cocoADS/model_ADS/checkpoint/generation_config.json +9 -0
  10. outputs/gpt2cocoADS/model_ADS/checkpoint/merges.txt +0 -0
  11. outputs/gpt2cocoADS/model_ADS/checkpoint/model.safetensors +3 -0
  12. outputs/gpt2cocoADS/model_ADS/checkpoint/optimizer.pt +3 -0
  13. outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_0.pth +3 -0
  14. outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_1.pth +3 -0
  15. outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_2.pth +3 -0
  16. outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_3.pth +3 -0
  17. outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_4.pth +3 -0
  18. outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_5.pth +3 -0
  19. outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_6.pth +3 -0
  20. outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_7.pth +3 -0
  21. outputs/gpt2cocoADS/model_ADS/checkpoint/scaler.pt +3 -0
  22. outputs/gpt2cocoADS/model_ADS/checkpoint/scheduler.pt +3 -0
  23. outputs/gpt2cocoADS/model_ADS/checkpoint/special_tokens_map.json +37 -0
  24. outputs/gpt2cocoADS/model_ADS/checkpoint/tokenizer_config.json +40 -0
  25. outputs/gpt2cocoADS/model_ADS/checkpoint/trainer_state.json +3499 -0
  26. outputs/gpt2cocoADS/model_ADS/checkpoint/training_args.bin +3 -0
  27. outputs/gpt2cocoADS/model_ADS/checkpoint/vocab.json +0 -0
  28. outputs/gpt2cocoADS/tokenized_data_ADS/dataset_dict.json +1 -0
  29. outputs/gpt2cocoADS/tokenized_data_ADS/test/data-00000-of-00001.arrow +3 -0
  30. outputs/gpt2cocoADS/tokenized_data_ADS/test/dataset_info.json +49 -0
  31. outputs/gpt2cocoADS/tokenized_data_ADS/test/state.json +22 -0
  32. outputs/gpt2cocoADS/tokenized_data_ADS/train/data-00000-of-00001.arrow +3 -0
  33. outputs/gpt2cocoADS/tokenized_data_ADS/train/dataset_info.json +49 -0
  34. outputs/gpt2cocoADS/tokenized_data_ADS/train/state.json +22 -0
  35. outputs/gpt2cocoDSA/model_DSA/checkpoint/added_tokens.json +4 -0
  36. outputs/gpt2cocoDSA/model_DSA/checkpoint/config.json +40 -0
  37. outputs/gpt2cocoDSA/model_DSA/checkpoint/generation_config.json +9 -0
  38. outputs/gpt2cocoDSA/model_DSA/checkpoint/merges.txt +0 -0
  39. outputs/gpt2cocoDSA/model_DSA/checkpoint/model.safetensors +3 -0
  40. outputs/gpt2cocoDSA/model_DSA/checkpoint/optimizer.pt +3 -0
  41. outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_0.pth +3 -0
  42. outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_1.pth +3 -0
  43. outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_2.pth +3 -0
  44. outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_3.pth +3 -0
  45. outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_4.pth +3 -0
  46. outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_5.pth +3 -0
  47. outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_6.pth +3 -0
  48. outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_7.pth +3 -0
  49. outputs/gpt2cocoDSA/model_DSA/checkpoint/scaler.pt +3 -0
  50. outputs/gpt2cocoDSA/model_DSA/checkpoint/scheduler.pt +3 -0
.gitattributes CHANGED
@@ -40,3 +40,4 @@ DeQA-Score/fig/singapore_flyer.jpg filter=lfs diff=lfs merge=lfs -text
40
  DeQA-Score/fig/teaser.png filter=lfs diff=lfs merge=lfs -text
41
  imgs/results1.png filter=lfs diff=lfs merge=lfs -text
42
  imgs/results2.png filter=lfs diff=lfs merge=lfs -text
 
 
40
  DeQA-Score/fig/teaser.png filter=lfs diff=lfs merge=lfs -text
41
  imgs/results1.png filter=lfs diff=lfs merge=lfs -text
42
  imgs/results2.png filter=lfs diff=lfs merge=lfs -text
43
+ coco_faiss_indexes/CLIP-Huge-Flickr-Flat/faiss_IVPQ_PCA.index filter=lfs diff=lfs merge=lfs -text
.gitignore CHANGED
@@ -1,9 +1,2 @@
1
- coco_data/
2
  __pycache__/
3
- coco_feats/
4
- coco_faiss_indexes/
5
- faiss_indexes/
6
- processed_data/
7
- outputs/
8
- wandb/
9
- results/
 
1
+
2
  __pycache__/
 
 
 
 
 
 
 
coco_faiss_indexes/CLIP-Huge-Flickr-Flat/faiss_IVPQ_PCA.index ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ae15cd2c1daf3f2e16d6597ac39b26c59210608a8027865115f94267daf1b0e3
3
+ size 484503597
coco_faiss_indexes/CLIP-Huge-Flickr-Flat/img_ids.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:551e049563421284be10773bc30fcf45218965290a32703d435388cecf488c4d
3
+ size 7570496
coco_faiss_indexes/CLIP-Huge-Flickr-Flat/norm1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eaf16b352749296f5b8064af81714b0c06fcbfce682e2924d63469f5e1bbba85
3
+ size 17
coco_feats/CLIP-Huge-Flickr-Flat_feats.npy ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a9045ab207a390cd8d0d7427e0575b3345213ea639f7094b2291b271f0b9804
3
+ size 484503680
outputs/gpt2cocoADS/model_ADS/checkpoint/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<pad>": 50258,
3
+ "<|startoftext|>": 50257
4
+ }
outputs/gpt2cocoADS/model_ADS/checkpoint/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "cls_token_id": 50257,
9
+ "dtype": "float32",
10
+ "embd_pdrop": 0.1,
11
+ "eos_token_id": 50256,
12
+ "initializer_range": 0.02,
13
+ "layer_norm_epsilon": 1e-05,
14
+ "model_type": "gpt2",
15
+ "n_ctx": 1024,
16
+ "n_embd": 768,
17
+ "n_head": 12,
18
+ "n_inner": null,
19
+ "n_layer": 12,
20
+ "n_positions": 1024,
21
+ "pad_token_id": 50258,
22
+ "reorder_and_upcast_attn": false,
23
+ "resid_pdrop": 0.1,
24
+ "scale_attn_by_inverse_layer_idx": false,
25
+ "scale_attn_weights": true,
26
+ "summary_activation": null,
27
+ "summary_first_dropout": 0.1,
28
+ "summary_proj_to_labels": true,
29
+ "summary_type": "cls_index",
30
+ "summary_use_proj": true,
31
+ "task_specific_params": {
32
+ "text-generation": {
33
+ "do_sample": true,
34
+ "max_length": 50
35
+ }
36
+ },
37
+ "transformers_version": "4.57.1",
38
+ "use_cache": true,
39
+ "vocab_size": 50259
40
+ }
outputs/gpt2cocoADS/model_ADS/checkpoint/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": [
5
+ 50256
6
+ ],
7
+ "pad_token_id": 50258,
8
+ "transformers_version": "4.57.1"
9
+ }
outputs/gpt2cocoADS/model_ADS/checkpoint/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
outputs/gpt2cocoADS/model_ADS/checkpoint/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9a2bdf2f1ce3b946aee0ed2a01808ff745f9a3806dbf856e0d3de357d622367
3
+ size 497780352
outputs/gpt2cocoADS/model_ADS/checkpoint/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6fbdf0a95ca9e72831de47fb65fb3534eaac1a5692d6cefcc95940f0087e9c91
3
+ size 995654586
outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aebbbdd9f6b6458006140b9399df973322b39193c3491688f309b28eda7f6516
3
+ size 15984
outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02092b11072872275f41067881e0ff9c3ebb4f71cf3347a14575ee0c3bb4dc37
3
+ size 15984
outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:834873e0297393e1fa57f1b955d4a9198d35ae71d5b1b5b0661327390e456a6a
3
+ size 15984
outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6756722a563897612ae24bdf113ccc0f63bb646ff00b0de94943e7815ca03cf9
3
+ size 15984
outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:811a877d2d10d1034d381ef47c53f9a523828b58a68b612cd1b0858ee81e6c1e
3
+ size 15984
outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65538c95dd42ed5ff89d120cad03ed208d8388db3902f5332493f3f25f8ac51a
3
+ size 15984
outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44952594d0b6ac119c313df4062415014d62be491e303928f3548bb0f2c5c55f
3
+ size 15984
outputs/gpt2cocoADS/model_ADS/checkpoint/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2081cb2f3fdd83c5b77b31a5b51f1f3492a4f66d21e8c736d093ea56b1461995
3
+ size 15984
outputs/gpt2cocoADS/model_ADS/checkpoint/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:110a77af7f5b3083bf8bd7e576fddab43fd1bc1ddb140b9c8edbb7ced2781a44
3
+ size 988
outputs/gpt2cocoADS/model_ADS/checkpoint/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f86d4debfeccbc9cc7f34fa8036cc4f47e09cf0625bf34c4a4cb3f392bd25f62
3
+ size 1064
outputs/gpt2cocoADS/model_ADS/checkpoint/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|endoftext|>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<|startoftext|>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "<|endoftext|>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "pad_token": {
24
+ "content": "<pad>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "<|endoftext|>",
32
+ "lstrip": false,
33
+ "normalized": true,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
outputs/gpt2cocoADS/model_ADS/checkpoint/tokenizer_config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "50256": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "50257": {
14
+ "content": "<|startoftext|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "50258": {
22
+ "content": "<pad>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "bos_token": "<|endoftext|>",
31
+ "clean_up_tokenization_spaces": false,
32
+ "cls_token": "<|startoftext|>",
33
+ "eos_token": "<|endoftext|>",
34
+ "errors": "replace",
35
+ "extra_special_tokens": {},
36
+ "model_max_length": 1024,
37
+ "pad_token": "<pad>",
38
+ "tokenizer_class": "GPT2Tokenizer",
39
+ "unk_token": "<|endoftext|>"
40
+ }
outputs/gpt2cocoADS/model_ADS/checkpoint/trainer_state.json ADDED
@@ -0,0 +1,3499 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 600,
3
+ "best_metric": 2.524237871170044,
4
+ "best_model_checkpoint": "../a1_text/fl/gpt2/pt3lgpt2cocoADS/finetuned_on_pt3lgpt2cocoADS_ICLR_rebuttal_prompt_pt3lgpt2cocoADS_percent_0.01/checkpoint-600",
5
+ "epoch": 99.56896551724138,
6
+ "eval_steps": 100,
7
+ "global_step": 23100,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.43103448275862066,
14
+ "grad_norm": 0.3606378138065338,
15
+ "learning_rate": 0.00188,
16
+ "loss": 3.1881,
17
+ "step": 100
18
+ },
19
+ {
20
+ "epoch": 0.43103448275862066,
21
+ "eval_loss": 2.688020944595337,
22
+ "eval_runtime": 9.2605,
23
+ "eval_samples_per_second": 12780.164,
24
+ "eval_steps_per_second": 6.263,
25
+ "step": 100
26
+ },
27
+ {
28
+ "epoch": 0.8620689655172413,
29
+ "grad_norm": 0.35380175709724426,
30
+ "learning_rate": 0.0019999182861670534,
31
+ "loss": 2.675,
32
+ "step": 200
33
+ },
34
+ {
35
+ "epoch": 0.8620689655172413,
36
+ "eval_loss": 2.594282388687134,
37
+ "eval_runtime": 9.2689,
38
+ "eval_samples_per_second": 12768.561,
39
+ "eval_steps_per_second": 6.257,
40
+ "step": 200
41
+ },
42
+ {
43
+ "epoch": 1.293103448275862,
44
+ "grad_norm": 0.22037602961063385,
45
+ "learning_rate": 0.0019996519640898547,
46
+ "loss": 2.5388,
47
+ "step": 300
48
+ },
49
+ {
50
+ "epoch": 1.293103448275862,
51
+ "eval_loss": 2.560807943344116,
52
+ "eval_runtime": 9.7881,
53
+ "eval_samples_per_second": 12091.346,
54
+ "eval_steps_per_second": 5.926,
55
+ "step": 300
56
+ },
57
+ {
58
+ "epoch": 1.7241379310344827,
59
+ "grad_norm": 0.21512971818447113,
60
+ "learning_rate": 0.0019992007500999214,
61
+ "loss": 2.5051,
62
+ "step": 400
63
+ },
64
+ {
65
+ "epoch": 1.7241379310344827,
66
+ "eval_loss": 2.533008575439453,
67
+ "eval_runtime": 9.2432,
68
+ "eval_samples_per_second": 12804.106,
69
+ "eval_steps_per_second": 6.275,
70
+ "step": 400
71
+ },
72
+ {
73
+ "epoch": 2.1551724137931036,
74
+ "grad_norm": 0.21010388433933258,
75
+ "learning_rate": 0.001998564727652117,
76
+ "loss": 2.447,
77
+ "step": 500
78
+ },
79
+ {
80
+ "epoch": 2.1551724137931036,
81
+ "eval_loss": 2.54014253616333,
82
+ "eval_runtime": 9.2465,
83
+ "eval_samples_per_second": 12799.491,
84
+ "eval_steps_per_second": 6.273,
85
+ "step": 500
86
+ },
87
+ {
88
+ "epoch": 2.586206896551724,
89
+ "grad_norm": 0.20196236670017242,
90
+ "learning_rate": 0.001997744014382789,
91
+ "loss": 2.3909,
92
+ "step": 600
93
+ },
94
+ {
95
+ "epoch": 2.586206896551724,
96
+ "eval_loss": 2.524237871170044,
97
+ "eval_runtime": 9.4244,
98
+ "eval_samples_per_second": 12557.87,
99
+ "eval_steps_per_second": 6.154,
100
+ "step": 600
101
+ },
102
+ {
103
+ "epoch": 3.0172413793103448,
104
+ "grad_norm": 0.2814484238624573,
105
+ "learning_rate": 0.0019967387620880147,
106
+ "loss": 2.401,
107
+ "step": 700
108
+ },
109
+ {
110
+ "epoch": 3.0172413793103448,
111
+ "eval_loss": 2.5517618656158447,
112
+ "eval_runtime": 9.297,
113
+ "eval_samples_per_second": 12730.004,
114
+ "eval_steps_per_second": 6.239,
115
+ "step": 700
116
+ },
117
+ {
118
+ "epoch": 3.4482758620689653,
119
+ "grad_norm": 0.20441880822181702,
120
+ "learning_rate": 0.0019955491566955234,
121
+ "loss": 2.2847,
122
+ "step": 800
123
+ },
124
+ {
125
+ "epoch": 3.4482758620689653,
126
+ "eval_loss": 2.543337345123291,
127
+ "eval_runtime": 9.2762,
128
+ "eval_samples_per_second": 12758.516,
129
+ "eval_steps_per_second": 6.253,
130
+ "step": 800
131
+ },
132
+ {
133
+ "epoch": 3.8793103448275863,
134
+ "grad_norm": 0.17433364689350128,
135
+ "learning_rate": 0.0019941754182303075,
136
+ "loss": 2.3263,
137
+ "step": 900
138
+ },
139
+ {
140
+ "epoch": 3.8793103448275863,
141
+ "eval_loss": 2.524676561355591,
142
+ "eval_runtime": 9.5276,
143
+ "eval_samples_per_second": 12421.935,
144
+ "eval_steps_per_second": 6.088,
145
+ "step": 900
146
+ },
147
+ {
148
+ "epoch": 4.310344827586207,
149
+ "grad_norm": 0.17612139880657196,
150
+ "learning_rate": 0.0019926178007739293,
151
+ "loss": 2.2276,
152
+ "step": 1000
153
+ },
154
+ {
155
+ "epoch": 4.310344827586207,
156
+ "eval_loss": 2.5814201831817627,
157
+ "eval_runtime": 9.2823,
158
+ "eval_samples_per_second": 12750.129,
159
+ "eval_steps_per_second": 6.248,
160
+ "step": 1000
161
+ },
162
+ {
163
+ "epoch": 4.741379310344827,
164
+ "grad_norm": 0.20014843344688416,
165
+ "learning_rate": 0.0019908765924175256,
166
+ "loss": 2.2433,
167
+ "step": 1100
168
+ },
169
+ {
170
+ "epoch": 4.741379310344827,
171
+ "eval_loss": 2.5539886951446533,
172
+ "eval_runtime": 9.2771,
173
+ "eval_samples_per_second": 12757.267,
174
+ "eval_steps_per_second": 6.252,
175
+ "step": 1100
176
+ },
177
+ {
178
+ "epoch": 5.172413793103448,
179
+ "grad_norm": 0.2258746325969696,
180
+ "learning_rate": 0.001988952115208524,
181
+ "loss": 2.195,
182
+ "step": 1200
183
+ },
184
+ {
185
+ "epoch": 5.172413793103448,
186
+ "eval_loss": 2.6236913204193115,
187
+ "eval_runtime": 9.404,
188
+ "eval_samples_per_second": 12585.164,
189
+ "eval_steps_per_second": 6.168,
190
+ "step": 1200
191
+ },
192
+ {
193
+ "epoch": 5.603448275862069,
194
+ "grad_norm": 0.20413880050182343,
195
+ "learning_rate": 0.001986844725091078,
196
+ "loss": 2.1551,
197
+ "step": 1300
198
+ },
199
+ {
200
+ "epoch": 5.603448275862069,
201
+ "eval_loss": 2.5979394912719727,
202
+ "eval_runtime": 9.2548,
203
+ "eval_samples_per_second": 12788.087,
204
+ "eval_steps_per_second": 6.267,
205
+ "step": 1300
206
+ },
207
+ {
208
+ "epoch": 6.0344827586206895,
209
+ "grad_norm": 0.24910715222358704,
210
+ "learning_rate": 0.0019845548118402334,
211
+ "loss": 2.1765,
212
+ "step": 1400
213
+ },
214
+ {
215
+ "epoch": 6.0344827586206895,
216
+ "eval_loss": 2.672145366668701,
217
+ "eval_runtime": 9.3198,
218
+ "eval_samples_per_second": 12698.817,
219
+ "eval_steps_per_second": 6.223,
220
+ "step": 1400
221
+ },
222
+ {
223
+ "epoch": 6.4655172413793105,
224
+ "grad_norm": 0.18814320862293243,
225
+ "learning_rate": 0.0019820827989898353,
226
+ "loss": 2.0557,
227
+ "step": 1500
228
+ },
229
+ {
230
+ "epoch": 6.4655172413793105,
231
+ "eval_loss": 2.6465389728546143,
232
+ "eval_runtime": 9.5906,
233
+ "eval_samples_per_second": 12340.267,
234
+ "eval_steps_per_second": 6.048,
235
+ "step": 1500
236
+ },
237
+ {
238
+ "epoch": 6.896551724137931,
239
+ "grad_norm": 0.1877131313085556,
240
+ "learning_rate": 0.0019794291437541956,
241
+ "loss": 2.118,
242
+ "step": 1600
243
+ },
244
+ {
245
+ "epoch": 6.896551724137931,
246
+ "eval_loss": 2.6222500801086426,
247
+ "eval_runtime": 9.3092,
248
+ "eval_samples_per_second": 12713.276,
249
+ "eval_steps_per_second": 6.23,
250
+ "step": 1600
251
+ },
252
+ {
253
+ "epoch": 7.327586206896552,
254
+ "grad_norm": 0.24426060914993286,
255
+ "learning_rate": 0.001976594336943526,
256
+ "loss": 2.0044,
257
+ "step": 1700
258
+ },
259
+ {
260
+ "epoch": 7.327586206896552,
261
+ "eval_loss": 2.7189927101135254,
262
+ "eval_runtime": 9.2714,
263
+ "eval_samples_per_second": 12765.194,
264
+ "eval_steps_per_second": 6.256,
265
+ "step": 1700
266
+ },
267
+ {
268
+ "epoch": 7.758620689655173,
269
+ "grad_norm": 0.2100994735956192,
270
+ "learning_rate": 0.0019735789028731605,
271
+ "loss": 2.0362,
272
+ "step": 1800
273
+ },
274
+ {
275
+ "epoch": 7.758620689655173,
276
+ "eval_loss": 2.6856698989868164,
277
+ "eval_runtime": 9.4232,
278
+ "eval_samples_per_second": 12559.482,
279
+ "eval_steps_per_second": 6.155,
280
+ "step": 1800
281
+ },
282
+ {
283
+ "epoch": 8.189655172413794,
284
+ "grad_norm": 0.23137305676937103,
285
+ "learning_rate": 0.0019703833992665795,
286
+ "loss": 1.9819,
287
+ "step": 1900
288
+ },
289
+ {
290
+ "epoch": 8.189655172413794,
291
+ "eval_loss": 2.7954814434051514,
292
+ "eval_runtime": 9.2746,
293
+ "eval_samples_per_second": 12760.72,
294
+ "eval_steps_per_second": 6.254,
295
+ "step": 1900
296
+ },
297
+ {
298
+ "epoch": 8.620689655172415,
299
+ "grad_norm": 0.19724325835704803,
300
+ "learning_rate": 0.0019670084171522564,
301
+ "loss": 1.9498,
302
+ "step": 2000
303
+ },
304
+ {
305
+ "epoch": 8.620689655172415,
306
+ "eval_loss": 2.7527308464050293,
307
+ "eval_runtime": 9.427,
308
+ "eval_samples_per_second": 12554.456,
309
+ "eval_steps_per_second": 6.153,
310
+ "step": 2000
311
+ },
312
+ {
313
+ "epoch": 9.051724137931034,
314
+ "grad_norm": 0.21562309563159943,
315
+ "learning_rate": 0.0019634545807543425,
316
+ "loss": 1.9719,
317
+ "step": 2100
318
+ },
319
+ {
320
+ "epoch": 9.051724137931034,
321
+ "eval_loss": 2.885232448577881,
322
+ "eval_runtime": 9.2621,
323
+ "eval_samples_per_second": 12778.035,
324
+ "eval_steps_per_second": 6.262,
325
+ "step": 2100
326
+ },
327
+ {
328
+ "epoch": 9.482758620689655,
329
+ "grad_norm": 0.210566908121109,
330
+ "learning_rate": 0.0019597225473772122,
331
+ "loss": 1.8528,
332
+ "step": 2200
333
+ },
334
+ {
335
+ "epoch": 9.482758620689655,
336
+ "eval_loss": 2.8175652027130127,
337
+ "eval_runtime": 9.4866,
338
+ "eval_samples_per_second": 12475.549,
339
+ "eval_steps_per_second": 6.114,
340
+ "step": 2200
341
+ },
342
+ {
343
+ "epoch": 9.913793103448276,
344
+ "grad_norm": 0.23581582307815552,
345
+ "learning_rate": 0.001955813007283891,
346
+ "loss": 1.9251,
347
+ "step": 2300
348
+ },
349
+ {
350
+ "epoch": 9.913793103448276,
351
+ "eval_loss": 2.768282651901245,
352
+ "eval_runtime": 9.2621,
353
+ "eval_samples_per_second": 12777.955,
354
+ "eval_steps_per_second": 6.262,
355
+ "step": 2300
356
+ },
357
+ {
358
+ "epoch": 10.344827586206897,
359
+ "grad_norm": 0.22041216492652893,
360
+ "learning_rate": 0.001951726683568388,
361
+ "loss": 1.7995,
362
+ "step": 2400
363
+ },
364
+ {
365
+ "epoch": 10.344827586206897,
366
+ "eval_loss": 2.912801504135132,
367
+ "eval_runtime": 9.2692,
368
+ "eval_samples_per_second": 12768.227,
369
+ "eval_steps_per_second": 6.257,
370
+ "step": 2400
371
+ },
372
+ {
373
+ "epoch": 10.775862068965518,
374
+ "grad_norm": 0.2417748123407364,
375
+ "learning_rate": 0.0019474643320219532,
376
+ "loss": 1.8457,
377
+ "step": 2500
378
+ },
379
+ {
380
+ "epoch": 10.775862068965518,
381
+ "eval_loss": 2.867093324661255,
382
+ "eval_runtime": 9.6387,
383
+ "eval_samples_per_second": 12278.739,
384
+ "eval_steps_per_second": 6.017,
385
+ "step": 2500
386
+ },
387
+ {
388
+ "epoch": 11.206896551724139,
389
+ "grad_norm": 0.2587749660015106,
390
+ "learning_rate": 0.00194302674099329,
391
+ "loss": 1.7872,
392
+ "step": 2600
393
+ },
394
+ {
395
+ "epoch": 11.206896551724139,
396
+ "eval_loss": 3.023223400115967,
397
+ "eval_runtime": 9.3365,
398
+ "eval_samples_per_second": 12676.102,
399
+ "eval_steps_per_second": 6.212,
400
+ "step": 2600
401
+ },
402
+ {
403
+ "epoch": 11.637931034482758,
404
+ "grad_norm": 0.2514320909976959,
405
+ "learning_rate": 0.0019384147312427463,
406
+ "loss": 1.7656,
407
+ "step": 2700
408
+ },
409
+ {
410
+ "epoch": 11.637931034482758,
411
+ "eval_loss": 2.9454939365386963,
412
+ "eval_runtime": 9.2584,
413
+ "eval_samples_per_second": 12783.153,
414
+ "eval_steps_per_second": 6.265,
415
+ "step": 2700
416
+ },
417
+ {
418
+ "epoch": 12.068965517241379,
419
+ "grad_norm": 0.25401562452316284,
420
+ "learning_rate": 0.001933629155790508,
421
+ "loss": 1.7835,
422
+ "step": 2800
423
+ },
424
+ {
425
+ "epoch": 12.068965517241379,
426
+ "eval_loss": 3.1060149669647217,
427
+ "eval_runtime": 9.2595,
428
+ "eval_samples_per_second": 12781.591,
429
+ "eval_steps_per_second": 6.264,
430
+ "step": 2800
431
+ },
432
+ {
433
+ "epoch": 12.5,
434
+ "grad_norm": 0.24663059413433075,
435
+ "learning_rate": 0.0019286708997588277,
436
+ "loss": 1.6779,
437
+ "step": 2900
438
+ },
439
+ {
440
+ "epoch": 12.5,
441
+ "eval_loss": 3.0567786693573,
442
+ "eval_runtime": 9.4661,
443
+ "eval_samples_per_second": 12502.63,
444
+ "eval_steps_per_second": 6.127,
445
+ "step": 2900
446
+ },
447
+ {
448
+ "epoch": 12.931034482758621,
449
+ "grad_norm": 0.23438771069049835,
450
+ "learning_rate": 0.0019235408802083177,
451
+ "loss": 1.7516,
452
+ "step": 3000
453
+ },
454
+ {
455
+ "epoch": 12.931034482758621,
456
+ "eval_loss": 2.974315881729126,
457
+ "eval_runtime": 9.2511,
458
+ "eval_samples_per_second": 12793.237,
459
+ "eval_steps_per_second": 6.27,
460
+ "step": 3000
461
+ },
462
+ {
463
+ "epoch": 13.362068965517242,
464
+ "grad_norm": 0.2801017761230469,
465
+ "learning_rate": 0.0019182400459683319,
466
+ "loss": 1.6276,
467
+ "step": 3100
468
+ },
469
+ {
470
+ "epoch": 13.362068965517242,
471
+ "eval_loss": 3.134687662124634,
472
+ "eval_runtime": 9.3858,
473
+ "eval_samples_per_second": 12609.573,
474
+ "eval_steps_per_second": 6.18,
475
+ "step": 3100
476
+ },
477
+ {
478
+ "epoch": 13.793103448275861,
479
+ "grad_norm": 0.2482592910528183,
480
+ "learning_rate": 0.0019127693774614737,
481
+ "loss": 1.6826,
482
+ "step": 3200
483
+ },
484
+ {
485
+ "epoch": 13.793103448275861,
486
+ "eval_loss": 3.0509395599365234,
487
+ "eval_runtime": 9.554,
488
+ "eval_samples_per_second": 12387.633,
489
+ "eval_steps_per_second": 6.071,
490
+ "step": 3200
491
+ },
492
+ {
493
+ "epoch": 14.224137931034482,
494
+ "grad_norm": 0.24827790260314941,
495
+ "learning_rate": 0.0019071298865222619,
496
+ "loss": 1.6232,
497
+ "step": 3300
498
+ },
499
+ {
500
+ "epoch": 14.224137931034482,
501
+ "eval_loss": 3.2457408905029297,
502
+ "eval_runtime": 9.4432,
503
+ "eval_samples_per_second": 12532.974,
504
+ "eval_steps_per_second": 6.142,
505
+ "step": 3300
506
+ },
507
+ {
508
+ "epoch": 14.655172413793103,
509
+ "grad_norm": 0.25374507904052734,
510
+ "learning_rate": 0.0019013226162099856,
511
+ "loss": 1.6105,
512
+ "step": 3400
513
+ },
514
+ {
515
+ "epoch": 14.655172413793103,
516
+ "eval_loss": 3.1603097915649414,
517
+ "eval_runtime": 9.3143,
518
+ "eval_samples_per_second": 12706.354,
519
+ "eval_steps_per_second": 6.227,
520
+ "step": 3400
521
+ },
522
+ {
523
+ "epoch": 15.086206896551724,
524
+ "grad_norm": 0.2645600140094757,
525
+ "learning_rate": 0.001895348640615783,
526
+ "loss": 1.6263,
527
+ "step": 3500
528
+ },
529
+ {
530
+ "epoch": 15.086206896551724,
531
+ "eval_loss": 3.357069730758667,
532
+ "eval_runtime": 9.3861,
533
+ "eval_samples_per_second": 12609.156,
534
+ "eval_steps_per_second": 6.179,
535
+ "step": 3500
536
+ },
537
+ {
538
+ "epoch": 15.517241379310345,
539
+ "grad_norm": 0.2520388960838318,
540
+ "learning_rate": 0.0018892090646639814,
541
+ "loss": 1.5382,
542
+ "step": 3600
543
+ },
544
+ {
545
+ "epoch": 15.517241379310345,
546
+ "eval_loss": 3.2382264137268066,
547
+ "eval_runtime": 9.4086,
548
+ "eval_samples_per_second": 12579.086,
549
+ "eval_steps_per_second": 6.165,
550
+ "step": 3600
551
+ },
552
+ {
553
+ "epoch": 15.948275862068966,
554
+ "grad_norm": 0.24910874664783478,
555
+ "learning_rate": 0.001882905023907735,
556
+ "loss": 1.6089,
557
+ "step": 3700
558
+ },
559
+ {
560
+ "epoch": 15.948275862068966,
561
+ "eval_loss": 3.1936817169189453,
562
+ "eval_runtime": 9.3017,
563
+ "eval_samples_per_second": 12723.611,
564
+ "eval_steps_per_second": 6.235,
565
+ "step": 3700
566
+ },
567
+ {
568
+ "epoch": 16.379310344827587,
569
+ "grad_norm": 0.24739116430282593,
570
+ "learning_rate": 0.0018764376843189978,
571
+ "loss": 1.4912,
572
+ "step": 3800
573
+ },
574
+ {
575
+ "epoch": 16.379310344827587,
576
+ "eval_loss": 3.353105068206787,
577
+ "eval_runtime": 9.6086,
578
+ "eval_samples_per_second": 12317.176,
579
+ "eval_steps_per_second": 6.036,
580
+ "step": 3800
581
+ },
582
+ {
583
+ "epoch": 16.810344827586206,
584
+ "grad_norm": 0.24842214584350586,
585
+ "learning_rate": 0.0018698082420728684,
586
+ "loss": 1.5502,
587
+ "step": 3900
588
+ },
589
+ {
590
+ "epoch": 16.810344827586206,
591
+ "eval_loss": 3.259680986404419,
592
+ "eval_runtime": 9.3091,
593
+ "eval_samples_per_second": 12713.446,
594
+ "eval_steps_per_second": 6.23,
595
+ "step": 3900
596
+ },
597
+ {
598
+ "epoch": 17.24137931034483,
599
+ "grad_norm": 0.27025410532951355,
600
+ "learning_rate": 0.0018630179233263503,
601
+ "loss": 1.4927,
602
+ "step": 4000
603
+ },
604
+ {
605
+ "epoch": 17.24137931034483,
606
+ "eval_loss": 3.4672367572784424,
607
+ "eval_runtime": 9.2704,
608
+ "eval_samples_per_second": 12766.549,
609
+ "eval_steps_per_second": 6.256,
610
+ "step": 4000
611
+ },
612
+ {
613
+ "epoch": 17.67241379310345,
614
+ "grad_norm": 0.25576576590538025,
615
+ "learning_rate": 0.0018560679839915678,
616
+ "loss": 1.4883,
617
+ "step": 4100
618
+ },
619
+ {
620
+ "epoch": 17.67241379310345,
621
+ "eval_loss": 3.34930157661438,
622
+ "eval_runtime": 9.2868,
623
+ "eval_samples_per_second": 12743.996,
624
+ "eval_steps_per_second": 6.245,
625
+ "step": 4100
626
+ },
627
+ {
628
+ "epoch": 18.103448275862068,
629
+ "grad_norm": 0.27449163794517517,
630
+ "learning_rate": 0.0018489597095034747,
631
+ "loss": 1.4985,
632
+ "step": 4200
633
+ },
634
+ {
635
+ "epoch": 18.103448275862068,
636
+ "eval_loss": 3.5502684116363525,
637
+ "eval_runtime": 9.3869,
638
+ "eval_samples_per_second": 12608.154,
639
+ "eval_steps_per_second": 6.179,
640
+ "step": 4200
641
+ },
642
+ {
643
+ "epoch": 18.53448275862069,
644
+ "grad_norm": 0.24876263737678528,
645
+ "learning_rate": 0.0018416944145821076,
646
+ "loss": 1.4296,
647
+ "step": 4300
648
+ },
649
+ {
650
+ "epoch": 18.53448275862069,
651
+ "eval_loss": 3.463294506072998,
652
+ "eval_runtime": 9.5784,
653
+ "eval_samples_per_second": 12356.09,
654
+ "eval_steps_per_second": 6.055,
655
+ "step": 4300
656
+ },
657
+ {
658
+ "epoch": 18.96551724137931,
659
+ "grad_norm": 0.2824811339378357,
660
+ "learning_rate": 0.0018342734429894186,
661
+ "loss": 1.4945,
662
+ "step": 4400
663
+ },
664
+ {
665
+ "epoch": 18.96551724137931,
666
+ "eval_loss": 3.3413078784942627,
667
+ "eval_runtime": 9.2525,
668
+ "eval_samples_per_second": 12791.199,
669
+ "eval_steps_per_second": 6.269,
670
+ "step": 4400
671
+ },
672
+ {
673
+ "epoch": 19.396551724137932,
674
+ "grad_norm": 0.24018803238868713,
675
+ "learning_rate": 0.0018266981672807382,
676
+ "loss": 1.3855,
677
+ "step": 4500
678
+ },
679
+ {
680
+ "epoch": 19.396551724137932,
681
+ "eval_loss": 3.5319149494171143,
682
+ "eval_runtime": 9.4198,
683
+ "eval_samples_per_second": 12564.07,
684
+ "eval_steps_per_second": 6.157,
685
+ "step": 4500
686
+ },
687
+ {
688
+ "epoch": 19.82758620689655,
689
+ "grad_norm": 0.27930110692977905,
690
+ "learning_rate": 0.0018189699885509127,
691
+ "loss": 1.4438,
692
+ "step": 4600
693
+ },
694
+ {
695
+ "epoch": 19.82758620689655,
696
+ "eval_loss": 3.430851936340332,
697
+ "eval_runtime": 9.4099,
698
+ "eval_samples_per_second": 12577.273,
699
+ "eval_steps_per_second": 6.164,
700
+ "step": 4600
701
+ },
702
+ {
703
+ "epoch": 20.25862068965517,
704
+ "grad_norm": 0.2738105058670044,
705
+ "learning_rate": 0.0018110903361751639,
706
+ "loss": 1.3895,
707
+ "step": 4700
708
+ },
709
+ {
710
+ "epoch": 20.25862068965517,
711
+ "eval_loss": 3.641018867492676,
712
+ "eval_runtime": 9.2705,
713
+ "eval_samples_per_second": 12766.367,
714
+ "eval_steps_per_second": 6.256,
715
+ "step": 4700
716
+ },
717
+ {
718
+ "epoch": 20.689655172413794,
719
+ "grad_norm": 0.2917766571044922,
720
+ "learning_rate": 0.0018030606675447152,
721
+ "loss": 1.395,
722
+ "step": 4800
723
+ },
724
+ {
725
+ "epoch": 20.689655172413794,
726
+ "eval_loss": 3.5762739181518555,
727
+ "eval_runtime": 9.5847,
728
+ "eval_samples_per_second": 12347.853,
729
+ "eval_steps_per_second": 6.051,
730
+ "step": 4800
731
+ },
732
+ {
733
+ "epoch": 21.120689655172413,
734
+ "grad_norm": 0.2610531747341156,
735
+ "learning_rate": 0.0017948824677972398,
736
+ "loss": 1.3993,
737
+ "step": 4900
738
+ },
739
+ {
740
+ "epoch": 21.120689655172413,
741
+ "eval_loss": 3.7315163612365723,
742
+ "eval_runtime": 9.2601,
743
+ "eval_samples_per_second": 12780.78,
744
+ "eval_steps_per_second": 6.263,
745
+ "step": 4900
746
+ },
747
+ {
748
+ "epoch": 21.551724137931036,
749
+ "grad_norm": 0.26478374004364014,
750
+ "learning_rate": 0.0017865572495421743,
751
+ "loss": 1.3454,
752
+ "step": 5000
753
+ },
754
+ {
755
+ "epoch": 21.551724137931036,
756
+ "eval_loss": 3.64851975440979,
757
+ "eval_runtime": 9.2862,
758
+ "eval_samples_per_second": 12744.84,
759
+ "eval_steps_per_second": 6.246,
760
+ "step": 5000
761
+ },
762
+ {
763
+ "epoch": 21.982758620689655,
764
+ "grad_norm": 0.2939194440841675,
765
+ "learning_rate": 0.0017780865525809514,
766
+ "loss": 1.4046,
767
+ "step": 5100
768
+ },
769
+ {
770
+ "epoch": 21.982758620689655,
771
+ "eval_loss": 3.543053388595581,
772
+ "eval_runtime": 9.4277,
773
+ "eval_samples_per_second": 12553.504,
774
+ "eval_steps_per_second": 6.152,
775
+ "step": 5100
776
+ },
777
+ {
778
+ "epoch": 22.413793103448278,
779
+ "grad_norm": 0.27193760871887207,
780
+ "learning_rate": 0.001769471943622206,
781
+ "loss": 1.3036,
782
+ "step": 5200
783
+ },
784
+ {
785
+ "epoch": 22.413793103448278,
786
+ "eval_loss": 3.7068560123443604,
787
+ "eval_runtime": 9.2617,
788
+ "eval_samples_per_second": 12778.557,
789
+ "eval_steps_per_second": 6.262,
790
+ "step": 5200
791
+ },
792
+ {
793
+ "epoch": 22.844827586206897,
794
+ "grad_norm": 0.2961808145046234,
795
+ "learning_rate": 0.001760715015992,
796
+ "loss": 1.3629,
797
+ "step": 5300
798
+ },
799
+ {
800
+ "epoch": 22.844827586206897,
801
+ "eval_loss": 3.6374130249023438,
802
+ "eval_runtime": 9.6151,
803
+ "eval_samples_per_second": 12308.828,
804
+ "eval_steps_per_second": 6.032,
805
+ "step": 5300
806
+ },
807
+ {
808
+ "epoch": 23.275862068965516,
809
+ "grad_norm": 0.24808737635612488,
810
+ "learning_rate": 0.0017518173893391294,
811
+ "loss": 1.3103,
812
+ "step": 5400
813
+ },
814
+ {
815
+ "epoch": 23.275862068965516,
816
+ "eval_loss": 3.787332057952881,
817
+ "eval_runtime": 9.2998,
818
+ "eval_samples_per_second": 12726.238,
819
+ "eval_steps_per_second": 6.237,
820
+ "step": 5400
821
+ },
822
+ {
823
+ "epoch": 23.70689655172414,
824
+ "grad_norm": 0.2666555643081665,
825
+ "learning_rate": 0.0017427807093355573,
826
+ "loss": 1.3231,
827
+ "step": 5500
828
+ },
829
+ {
830
+ "epoch": 23.70689655172414,
831
+ "eval_loss": 3.7004289627075195,
832
+ "eval_runtime": 9.2874,
833
+ "eval_samples_per_second": 12743.207,
834
+ "eval_steps_per_second": 6.245,
835
+ "step": 5500
836
+ },
837
+ {
838
+ "epoch": 24.137931034482758,
839
+ "grad_norm": 0.2620752453804016,
840
+ "learning_rate": 0.0017336066473720379,
841
+ "loss": 1.3198,
842
+ "step": 5600
843
+ },
844
+ {
845
+ "epoch": 24.137931034482758,
846
+ "eval_loss": 3.9016237258911133,
847
+ "eval_runtime": 9.3821,
848
+ "eval_samples_per_second": 12614.569,
849
+ "eval_steps_per_second": 6.182,
850
+ "step": 5600
851
+ },
852
+ {
853
+ "epoch": 24.56896551724138,
854
+ "grad_norm": 0.27390095591545105,
855
+ "learning_rate": 0.0017242969002489832,
856
+ "loss": 1.2811,
857
+ "step": 5700
858
+ },
859
+ {
860
+ "epoch": 24.56896551724138,
861
+ "eval_loss": 3.827505111694336,
862
+ "eval_runtime": 9.2529,
863
+ "eval_samples_per_second": 12790.715,
864
+ "eval_steps_per_second": 6.268,
865
+ "step": 5700
866
+ },
867
+ {
868
+ "epoch": 25.0,
869
+ "grad_norm": 0.27188900113105774,
870
+ "learning_rate": 0.0017148531898626277,
871
+ "loss": 1.3311,
872
+ "step": 5800
873
+ },
874
+ {
875
+ "epoch": 25.0,
876
+ "eval_loss": 3.7333984375,
877
+ "eval_runtime": 9.476,
878
+ "eval_samples_per_second": 12489.578,
879
+ "eval_steps_per_second": 6.121,
880
+ "step": 5800
881
+ },
882
+ {
883
+ "epoch": 25.43103448275862,
884
+ "grad_norm": 0.24834686517715454,
885
+ "learning_rate": 0.0017052772628865526,
886
+ "loss": 1.2388,
887
+ "step": 5900
888
+ },
889
+ {
890
+ "epoch": 25.43103448275862,
891
+ "eval_loss": 3.8666064739227295,
892
+ "eval_runtime": 9.4345,
893
+ "eval_samples_per_second": 12544.426,
894
+ "eval_steps_per_second": 6.148,
895
+ "step": 5900
896
+ },
897
+ {
898
+ "epoch": 25.862068965517242,
899
+ "grad_norm": 0.2640489339828491,
900
+ "learning_rate": 0.0016955708904486295,
901
+ "loss": 1.2972,
902
+ "step": 6000
903
+ },
904
+ {
905
+ "epoch": 25.862068965517242,
906
+ "eval_loss": 3.787432909011841,
907
+ "eval_runtime": 9.2777,
908
+ "eval_samples_per_second": 12756.54,
909
+ "eval_steps_per_second": 6.252,
910
+ "step": 6000
911
+ },
912
+ {
913
+ "epoch": 26.29310344827586,
914
+ "grad_norm": 0.2595365643501282,
915
+ "learning_rate": 0.0016857358678034368,
916
+ "loss": 1.2452,
917
+ "step": 6100
918
+ },
919
+ {
920
+ "epoch": 26.29310344827586,
921
+ "eval_loss": 3.9664251804351807,
922
+ "eval_runtime": 9.5344,
923
+ "eval_samples_per_second": 12413.04,
924
+ "eval_steps_per_second": 6.083,
925
+ "step": 6100
926
+ },
927
+ {
928
+ "epoch": 26.724137931034484,
929
+ "grad_norm": 0.268236368894577,
930
+ "learning_rate": 0.0016757740140002178,
931
+ "loss": 1.2629,
932
+ "step": 6200
933
+ },
934
+ {
935
+ "epoch": 26.724137931034484,
936
+ "eval_loss": 3.8767783641815186,
937
+ "eval_runtime": 9.2567,
938
+ "eval_samples_per_second": 12785.418,
939
+ "eval_steps_per_second": 6.266,
940
+ "step": 6200
941
+ },
942
+ {
943
+ "epoch": 27.155172413793103,
944
+ "grad_norm": 0.2526521682739258,
945
+ "learning_rate": 0.0016656871715464352,
946
+ "loss": 1.2554,
947
+ "step": 6300
948
+ },
949
+ {
950
+ "epoch": 27.155172413793103,
951
+ "eval_loss": 4.029198169708252,
952
+ "eval_runtime": 9.4821,
953
+ "eval_samples_per_second": 12481.487,
954
+ "eval_steps_per_second": 6.117,
955
+ "step": 6300
956
+ },
957
+ {
958
+ "epoch": 27.586206896551722,
959
+ "grad_norm": 0.27689260244369507,
960
+ "learning_rate": 0.001655477206066988,
961
+ "loss": 1.2282,
962
+ "step": 6400
963
+ },
964
+ {
965
+ "epoch": 27.586206896551722,
966
+ "eval_loss": 3.9616963863372803,
967
+ "eval_runtime": 9.2226,
968
+ "eval_samples_per_second": 12832.698,
969
+ "eval_steps_per_second": 6.289,
970
+ "step": 6400
971
+ },
972
+ {
973
+ "epoch": 28.017241379310345,
974
+ "grad_norm": 0.26988276839256287,
975
+ "learning_rate": 0.0016451460059591532,
976
+ "loss": 1.2676,
977
+ "step": 6500
978
+ },
979
+ {
980
+ "epoch": 28.017241379310345,
981
+ "eval_loss": 4.142755031585693,
982
+ "eval_runtime": 9.3655,
983
+ "eval_samples_per_second": 12636.926,
984
+ "eval_steps_per_second": 6.193,
985
+ "step": 6500
986
+ },
987
+ {
988
+ "epoch": 28.448275862068964,
989
+ "grad_norm": 0.25296926498413086,
990
+ "learning_rate": 0.001634695482043313,
991
+ "loss": 1.1919,
992
+ "step": 6600
993
+ },
994
+ {
995
+ "epoch": 28.448275862068964,
996
+ "eval_loss": 3.995730400085449,
997
+ "eval_runtime": 9.3822,
998
+ "eval_samples_per_second": 12614.418,
999
+ "eval_steps_per_second": 6.182,
1000
+ "step": 6600
1001
+ },
1002
+ {
1003
+ "epoch": 28.879310344827587,
1004
+ "grad_norm": 0.2919161021709442,
1005
+ "learning_rate": 0.0016241275672095395,
1006
+ "loss": 1.2444,
1007
+ "step": 6700
1008
+ },
1009
+ {
1010
+ "epoch": 28.879310344827587,
1011
+ "eval_loss": 3.9436511993408203,
1012
+ "eval_runtime": 9.2939,
1013
+ "eval_samples_per_second": 12734.257,
1014
+ "eval_steps_per_second": 6.241,
1015
+ "step": 6700
1016
+ },
1017
+ {
1018
+ "epoch": 29.310344827586206,
1019
+ "grad_norm": 0.2617655396461487,
1020
+ "learning_rate": 0.0016134442160600903,
1021
+ "loss": 1.1936,
1022
+ "step": 6800
1023
+ },
1024
+ {
1025
+ "epoch": 29.310344827586206,
1026
+ "eval_loss": 4.093754291534424,
1027
+ "eval_runtime": 9.4237,
1028
+ "eval_samples_per_second": 12558.924,
1029
+ "eval_steps_per_second": 6.155,
1030
+ "step": 6800
1031
+ },
1032
+ {
1033
+ "epoch": 29.74137931034483,
1034
+ "grad_norm": 0.2699144184589386,
1035
+ "learning_rate": 0.0016026474045478978,
1036
+ "loss": 1.2141,
1037
+ "step": 6900
1038
+ },
1039
+ {
1040
+ "epoch": 29.74137931034483,
1041
+ "eval_loss": 4.007887363433838,
1042
+ "eval_runtime": 9.2733,
1043
+ "eval_samples_per_second": 12762.564,
1044
+ "eval_steps_per_second": 6.255,
1045
+ "step": 6900
1046
+ },
1047
+ {
1048
+ "epoch": 30.17241379310345,
1049
+ "grad_norm": 0.259067177772522,
1050
+ "learning_rate": 0.0015917391296110992,
1051
+ "loss": 1.2032,
1052
+ "step": 7000
1053
+ },
1054
+ {
1055
+ "epoch": 30.17241379310345,
1056
+ "eval_loss": 4.160769462585449,
1057
+ "eval_runtime": 9.2534,
1058
+ "eval_samples_per_second": 12790.067,
1059
+ "eval_steps_per_second": 6.268,
1060
+ "step": 7000
1061
+ },
1062
+ {
1063
+ "epoch": 30.603448275862068,
1064
+ "grad_norm": 0.2688118517398834,
1065
+ "learning_rate": 0.0015807214088036938,
1066
+ "loss": 1.1848,
1067
+ "step": 7100
1068
+ },
1069
+ {
1070
+ "epoch": 30.603448275862068,
1071
+ "eval_loss": 4.0972747802734375,
1072
+ "eval_runtime": 9.4548,
1073
+ "eval_samples_per_second": 12517.588,
1074
+ "eval_steps_per_second": 6.134,
1075
+ "step": 7100
1076
+ },
1077
+ {
1078
+ "epoch": 31.03448275862069,
1079
+ "grad_norm": 0.2539116144180298,
1080
+ "learning_rate": 0.0015695962799223819,
1081
+ "loss": 1.2147,
1082
+ "step": 7200
1083
+ },
1084
+ {
1085
+ "epoch": 31.03448275862069,
1086
+ "eval_loss": 4.176488399505615,
1087
+ "eval_runtime": 9.3756,
1088
+ "eval_samples_per_second": 12623.336,
1089
+ "eval_steps_per_second": 6.186,
1090
+ "step": 7200
1091
+ },
1092
+ {
1093
+ "epoch": 31.46551724137931,
1094
+ "grad_norm": 0.2756901681423187,
1095
+ "learning_rate": 0.0015583658006296623,
1096
+ "loss": 1.1523,
1097
+ "step": 7300
1098
+ },
1099
+ {
1100
+ "epoch": 31.46551724137931,
1101
+ "eval_loss": 4.189959526062012,
1102
+ "eval_runtime": 9.399,
1103
+ "eval_samples_per_second": 12591.915,
1104
+ "eval_steps_per_second": 6.171,
1105
+ "step": 7300
1106
+ },
1107
+ {
1108
+ "epoch": 31.896551724137932,
1109
+ "grad_norm": 0.26899629831314087,
1110
+ "learning_rate": 0.0015470320480732548,
1111
+ "loss": 1.2002,
1112
+ "step": 7400
1113
+ },
1114
+ {
1115
+ "epoch": 31.896551724137932,
1116
+ "eval_loss": 4.085175037384033,
1117
+ "eval_runtime": 9.2558,
1118
+ "eval_samples_per_second": 12786.624,
1119
+ "eval_steps_per_second": 6.266,
1120
+ "step": 7400
1121
+ },
1122
+ {
1123
+ "epoch": 32.327586206896555,
1124
+ "grad_norm": 0.2548110783100128,
1125
+ "learning_rate": 0.0015355971185019174,
1126
+ "loss": 1.149,
1127
+ "step": 7500
1128
+ },
1129
+ {
1130
+ "epoch": 32.327586206896555,
1131
+ "eval_loss": 4.262933254241943,
1132
+ "eval_runtime": 9.463,
1133
+ "eval_samples_per_second": 12506.72,
1134
+ "eval_steps_per_second": 6.129,
1135
+ "step": 7500
1136
+ },
1137
+ {
1138
+ "epoch": 32.758620689655174,
1139
+ "grad_norm": 0.25835704803466797,
1140
+ "learning_rate": 0.0015240631268777327,
1141
+ "loss": 1.1741,
1142
+ "step": 7600
1143
+ },
1144
+ {
1145
+ "epoch": 32.758620689655174,
1146
+ "eval_loss": 4.147505283355713,
1147
+ "eval_runtime": 9.37,
1148
+ "eval_samples_per_second": 12630.822,
1149
+ "eval_steps_per_second": 6.19,
1150
+ "step": 7600
1151
+ },
1152
+ {
1153
+ "epoch": 33.189655172413794,
1154
+ "grad_norm": 0.25628194212913513,
1155
+ "learning_rate": 0.0015124322064849342,
1156
+ "loss": 1.1596,
1157
+ "step": 7700
1158
+ },
1159
+ {
1160
+ "epoch": 33.189655172413794,
1161
+ "eval_loss": 4.304795742034912,
1162
+ "eval_runtime": 9.2212,
1163
+ "eval_samples_per_second": 12834.665,
1164
+ "eval_steps_per_second": 6.29,
1165
+ "step": 7700
1166
+ },
1167
+ {
1168
+ "epoch": 33.62068965517241,
1169
+ "grad_norm": 0.28352490067481995,
1170
+ "learning_rate": 0.0015007065085353387,
1171
+ "loss": 1.1461,
1172
+ "step": 7800
1173
+ },
1174
+ {
1175
+ "epoch": 33.62068965517241,
1176
+ "eval_loss": 4.201080322265625,
1177
+ "eval_runtime": 9.5708,
1178
+ "eval_samples_per_second": 12365.811,
1179
+ "eval_steps_per_second": 6.06,
1180
+ "step": 7800
1181
+ },
1182
+ {
1183
+ "epoch": 34.05172413793103,
1184
+ "grad_norm": 0.2512941360473633,
1185
+ "learning_rate": 0.001488888201770468,
1186
+ "loss": 1.1706,
1187
+ "step": 7900
1188
+ },
1189
+ {
1190
+ "epoch": 34.05172413793103,
1191
+ "eval_loss": 4.385834693908691,
1192
+ "eval_runtime": 9.2927,
1193
+ "eval_samples_per_second": 12735.903,
1194
+ "eval_steps_per_second": 6.241,
1195
+ "step": 7900
1196
+ },
1197
+ {
1198
+ "epoch": 34.48275862068966,
1199
+ "grad_norm": 0.26079025864601135,
1200
+ "learning_rate": 0.0014769794720604263,
1201
+ "loss": 1.1193,
1202
+ "step": 8000
1203
+ },
1204
+ {
1205
+ "epoch": 34.48275862068966,
1206
+ "eval_loss": 4.2830634117126465,
1207
+ "eval_runtime": 9.284,
1208
+ "eval_samples_per_second": 12747.87,
1209
+ "eval_steps_per_second": 6.247,
1210
+ "step": 8000
1211
+ },
1212
+ {
1213
+ "epoch": 34.91379310344828,
1214
+ "grad_norm": 0.26432371139526367,
1215
+ "learning_rate": 0.0014649825219996107,
1216
+ "loss": 1.1619,
1217
+ "step": 8100
1218
+ },
1219
+ {
1220
+ "epoch": 34.91379310344828,
1221
+ "eval_loss": 4.2099223136901855,
1222
+ "eval_runtime": 9.4403,
1223
+ "eval_samples_per_second": 12536.732,
1224
+ "eval_steps_per_second": 6.144,
1225
+ "step": 8100
1226
+ },
1227
+ {
1228
+ "epoch": 35.3448275862069,
1229
+ "grad_norm": 0.24524888396263123,
1230
+ "learning_rate": 0.0014528995704993248,
1231
+ "loss": 1.1126,
1232
+ "step": 8200
1233
+ },
1234
+ {
1235
+ "epoch": 35.3448275862069,
1236
+ "eval_loss": 4.350750923156738,
1237
+ "eval_runtime": 9.2959,
1238
+ "eval_samples_per_second": 12731.57,
1239
+ "eval_steps_per_second": 6.239,
1240
+ "step": 8200
1241
+ },
1242
+ {
1243
+ "epoch": 35.775862068965516,
1244
+ "grad_norm": 0.3057849705219269,
1245
+ "learning_rate": 0.0014407328523773804,
1246
+ "loss": 1.1379,
1247
+ "step": 8300
1248
+ },
1249
+ {
1250
+ "epoch": 35.775862068965516,
1251
+ "eval_loss": 4.252375602722168,
1252
+ "eval_runtime": 9.5853,
1253
+ "eval_samples_per_second": 12347.118,
1254
+ "eval_steps_per_second": 6.051,
1255
+ "step": 8300
1256
+ },
1257
+ {
1258
+ "epoch": 36.206896551724135,
1259
+ "grad_norm": 0.2398909628391266,
1260
+ "learning_rate": 0.0014284846179447514,
1261
+ "loss": 1.1216,
1262
+ "step": 8400
1263
+ },
1264
+ {
1265
+ "epoch": 36.206896551724135,
1266
+ "eval_loss": 4.42500638961792,
1267
+ "eval_runtime": 9.2921,
1268
+ "eval_samples_per_second": 12736.794,
1269
+ "eval_steps_per_second": 6.242,
1270
+ "step": 8400
1271
+ },
1272
+ {
1273
+ "epoch": 36.63793103448276,
1274
+ "grad_norm": 0.24997535347938538,
1275
+ "learning_rate": 0.0014161571325893665,
1276
+ "loss": 1.1152,
1277
+ "step": 8500
1278
+ },
1279
+ {
1280
+ "epoch": 36.63793103448276,
1281
+ "eval_loss": 4.350626468658447,
1282
+ "eval_runtime": 9.272,
1283
+ "eval_samples_per_second": 12764.285,
1284
+ "eval_steps_per_second": 6.255,
1285
+ "step": 8500
1286
+ },
1287
+ {
1288
+ "epoch": 37.06896551724138,
1289
+ "grad_norm": 0.23204432427883148,
1290
+ "learning_rate": 0.0014037526763571095,
1291
+ "loss": 1.1316,
1292
+ "step": 8600
1293
+ },
1294
+ {
1295
+ "epoch": 37.06896551724138,
1296
+ "eval_loss": 4.5020670890808105,
1297
+ "eval_runtime": 9.4681,
1298
+ "eval_samples_per_second": 12500.031,
1299
+ "eval_steps_per_second": 6.126,
1300
+ "step": 8600
1301
+ },
1302
+ {
1303
+ "epoch": 37.5,
1304
+ "grad_norm": 0.24823765456676483,
1305
+ "learning_rate": 0.0013912735435301126,
1306
+ "loss": 1.09,
1307
+ "step": 8700
1308
+ },
1309
+ {
1310
+ "epoch": 37.5,
1311
+ "eval_loss": 4.4144792556762695,
1312
+ "eval_runtime": 9.2529,
1313
+ "eval_samples_per_second": 12790.65,
1314
+ "eval_steps_per_second": 6.268,
1315
+ "step": 8700
1316
+ },
1317
+ {
1318
+ "epoch": 37.93103448275862,
1319
+ "grad_norm": 0.2573583722114563,
1320
+ "learning_rate": 0.0013787220422024134,
1321
+ "loss": 1.1286,
1322
+ "step": 8800
1323
+ },
1324
+ {
1325
+ "epoch": 37.93103448275862,
1326
+ "eval_loss": 4.3031487464904785,
1327
+ "eval_runtime": 9.5833,
1328
+ "eval_samples_per_second": 12349.751,
1329
+ "eval_steps_per_second": 6.052,
1330
+ "step": 8800
1331
+ },
1332
+ {
1333
+ "epoch": 38.36206896551724,
1334
+ "grad_norm": 0.2552712559700012,
1335
+ "learning_rate": 0.0013661004938530573,
1336
+ "loss": 1.0802,
1337
+ "step": 8900
1338
+ },
1339
+ {
1340
+ "epoch": 38.36206896551724,
1341
+ "eval_loss": 4.499365329742432,
1342
+ "eval_runtime": 9.2698,
1343
+ "eval_samples_per_second": 12767.347,
1344
+ "eval_steps_per_second": 6.257,
1345
+ "step": 8900
1346
+ },
1347
+ {
1348
+ "epoch": 38.793103448275865,
1349
+ "grad_norm": 0.25154897570610046,
1350
+ "learning_rate": 0.0013534112329167277,
1351
+ "loss": 1.1087,
1352
+ "step": 9000
1353
+ },
1354
+ {
1355
+ "epoch": 38.793103448275865,
1356
+ "eval_loss": 4.370322227478027,
1357
+ "eval_runtime": 9.271,
1358
+ "eval_samples_per_second": 12765.709,
1359
+ "eval_steps_per_second": 6.256,
1360
+ "step": 9000
1361
+ },
1362
+ {
1363
+ "epoch": 39.224137931034484,
1364
+ "grad_norm": 0.2541041970252991,
1365
+ "learning_rate": 0.0013406566063519764,
1366
+ "loss": 1.0898,
1367
+ "step": 9100
1368
+ },
1369
+ {
1370
+ "epoch": 39.224137931034484,
1371
+ "eval_loss": 4.526272296905518,
1372
+ "eval_runtime": 9.3993,
1373
+ "eval_samples_per_second": 12591.497,
1374
+ "eval_steps_per_second": 6.171,
1375
+ "step": 9100
1376
+ },
1377
+ {
1378
+ "epoch": 39.6551724137931,
1379
+ "grad_norm": 0.2621018886566162,
1380
+ "learning_rate": 0.0013278389732071388,
1381
+ "loss": 1.0871,
1382
+ "step": 9200
1383
+ },
1384
+ {
1385
+ "epoch": 39.6551724137931,
1386
+ "eval_loss": 4.441021919250488,
1387
+ "eval_runtime": 9.2797,
1388
+ "eval_samples_per_second": 12753.725,
1389
+ "eval_steps_per_second": 6.25,
1390
+ "step": 9200
1391
+ },
1392
+ {
1393
+ "epoch": 40.08620689655172,
1394
+ "grad_norm": 0.21874089539051056,
1395
+ "learning_rate": 0.0013149607041840125,
1396
+ "loss": 1.0995,
1397
+ "step": 9300
1398
+ },
1399
+ {
1400
+ "epoch": 40.08620689655172,
1401
+ "eval_loss": 4.613419055938721,
1402
+ "eval_runtime": 9.3717,
1403
+ "eval_samples_per_second": 12628.601,
1404
+ "eval_steps_per_second": 6.189,
1405
+ "step": 9300
1406
+ },
1407
+ {
1408
+ "epoch": 40.51724137931034,
1409
+ "grad_norm": 0.23630797863006592,
1410
+ "learning_rate": 0.0013020241811993831,
1411
+ "loss": 1.0644,
1412
+ "step": 9400
1413
+ },
1414
+ {
1415
+ "epoch": 40.51724137931034,
1416
+ "eval_loss": 4.484513759613037,
1417
+ "eval_runtime": 9.2248,
1418
+ "eval_samples_per_second": 12829.63,
1419
+ "eval_steps_per_second": 6.287,
1420
+ "step": 9400
1421
+ },
1422
+ {
1423
+ "epoch": 40.94827586206897,
1424
+ "grad_norm": 0.2795443534851074,
1425
+ "learning_rate": 0.0012890317969444715,
1426
+ "loss": 1.1001,
1427
+ "step": 9500
1428
+ },
1429
+ {
1430
+ "epoch": 40.94827586206897,
1431
+ "eval_loss": 4.466510772705078,
1432
+ "eval_runtime": 9.3955,
1433
+ "eval_samples_per_second": 12596.615,
1434
+ "eval_steps_per_second": 6.173,
1435
+ "step": 9500
1436
+ },
1437
+ {
1438
+ "epoch": 41.37931034482759,
1439
+ "grad_norm": 0.25362640619277954,
1440
+ "learning_rate": 0.0012759859544423924,
1441
+ "loss": 1.052,
1442
+ "step": 9600
1443
+ },
1444
+ {
1445
+ "epoch": 41.37931034482759,
1446
+ "eval_loss": 4.589365482330322,
1447
+ "eval_runtime": 9.4467,
1448
+ "eval_samples_per_second": 12528.312,
1449
+ "eval_steps_per_second": 6.14,
1450
+ "step": 9600
1451
+ },
1452
+ {
1453
+ "epoch": 41.810344827586206,
1454
+ "grad_norm": 0.24765975773334503,
1455
+ "learning_rate": 0.001262889066603698,
1456
+ "loss": 1.0813,
1457
+ "step": 9700
1458
+ },
1459
+ {
1460
+ "epoch": 41.810344827586206,
1461
+ "eval_loss": 4.512991905212402,
1462
+ "eval_runtime": 9.2365,
1463
+ "eval_samples_per_second": 12813.371,
1464
+ "eval_steps_per_second": 6.279,
1465
+ "step": 9700
1466
+ },
1467
+ {
1468
+ "epoch": 42.241379310344826,
1469
+ "grad_norm": 0.23028719425201416,
1470
+ "learning_rate": 0.0012497435557800973,
1471
+ "loss": 1.0599,
1472
+ "step": 9800
1473
+ },
1474
+ {
1475
+ "epoch": 42.241379310344826,
1476
+ "eval_loss": 4.664239406585693,
1477
+ "eval_runtime": 9.4013,
1478
+ "eval_samples_per_second": 12588.838,
1479
+ "eval_steps_per_second": 6.169,
1480
+ "step": 9800
1481
+ },
1482
+ {
1483
+ "epoch": 42.672413793103445,
1484
+ "grad_norm": 0.21719814836978912,
1485
+ "learning_rate": 0.0012365518533164264,
1486
+ "loss": 1.0614,
1487
+ "step": 9900
1488
+ },
1489
+ {
1490
+ "epoch": 42.672413793103445,
1491
+ "eval_loss": 4.562778472900391,
1492
+ "eval_runtime": 9.2788,
1493
+ "eval_samples_per_second": 12754.966,
1494
+ "eval_steps_per_second": 6.251,
1495
+ "step": 9900
1496
+ },
1497
+ {
1498
+ "epoch": 43.10344827586207,
1499
+ "grad_norm": 0.21051952242851257,
1500
+ "learning_rate": 0.0012233163991009536,
1501
+ "loss": 1.0701,
1502
+ "step": 10000
1503
+ },
1504
+ {
1505
+ "epoch": 43.10344827586207,
1506
+ "eval_loss": 4.727095127105713,
1507
+ "eval_runtime": 9.2509,
1508
+ "eval_samples_per_second": 12793.434,
1509
+ "eval_steps_per_second": 6.27,
1510
+ "step": 10000
1511
+ },
1512
+ {
1513
+ "epoch": 43.53448275862069,
1514
+ "grad_norm": 0.21511335670948029,
1515
+ "learning_rate": 0.0012100396411141104,
1516
+ "loss": 1.0409,
1517
+ "step": 10100
1518
+ },
1519
+ {
1520
+ "epoch": 43.53448275862069,
1521
+ "eval_loss": 4.6477861404418945,
1522
+ "eval_runtime": 9.5173,
1523
+ "eval_samples_per_second": 12435.318,
1524
+ "eval_steps_per_second": 6.094,
1525
+ "step": 10100
1526
+ },
1527
+ {
1528
+ "epoch": 43.96551724137931,
1529
+ "grad_norm": 0.24457387626171112,
1530
+ "learning_rate": 0.0011967240349757202,
1531
+ "loss": 1.0744,
1532
+ "step": 10200
1533
+ },
1534
+ {
1535
+ "epoch": 43.96551724137931,
1536
+ "eval_loss": 4.5467848777771,
1537
+ "eval_runtime": 9.2312,
1538
+ "eval_samples_per_second": 12820.738,
1539
+ "eval_steps_per_second": 6.283,
1540
+ "step": 10200
1541
+ },
1542
+ {
1543
+ "epoch": 44.39655172413793,
1544
+ "grad_norm": 0.21022631227970123,
1545
+ "learning_rate": 0.001183372043490815,
1546
+ "loss": 1.0275,
1547
+ "step": 10300
1548
+ },
1549
+ {
1550
+ "epoch": 44.39655172413793,
1551
+ "eval_loss": 4.740345001220703,
1552
+ "eval_runtime": 9.431,
1553
+ "eval_samples_per_second": 12549.089,
1554
+ "eval_steps_per_second": 6.15,
1555
+ "step": 10300
1556
+ },
1557
+ {
1558
+ "epoch": 44.827586206896555,
1559
+ "grad_norm": 0.23863334953784943,
1560
+ "learning_rate": 0.0011699861361941259,
1561
+ "loss": 1.057,
1562
+ "step": 10400
1563
+ },
1564
+ {
1565
+ "epoch": 44.827586206896555,
1566
+ "eval_loss": 4.651246547698975,
1567
+ "eval_runtime": 9.2754,
1568
+ "eval_samples_per_second": 12759.723,
1569
+ "eval_steps_per_second": 6.253,
1570
+ "step": 10400
1571
+ },
1572
+ {
1573
+ "epoch": 45.258620689655174,
1574
+ "grad_norm": 0.2264426201581955,
1575
+ "learning_rate": 0.0011565687888933276,
1576
+ "loss": 1.0348,
1577
+ "step": 10500
1578
+ },
1579
+ {
1580
+ "epoch": 45.258620689655174,
1581
+ "eval_loss": 4.769204139709473,
1582
+ "eval_runtime": 9.4254,
1583
+ "eval_samples_per_second": 12556.649,
1584
+ "eval_steps_per_second": 6.154,
1585
+ "step": 10500
1586
+ },
1587
+ {
1588
+ "epoch": 45.689655172413794,
1589
+ "grad_norm": 0.2344275563955307,
1590
+ "learning_rate": 0.0011431224832111197,
1591
+ "loss": 1.0406,
1592
+ "step": 10600
1593
+ },
1594
+ {
1595
+ "epoch": 45.689655172413794,
1596
+ "eval_loss": 4.6706976890563965,
1597
+ "eval_runtime": 9.2785,
1598
+ "eval_samples_per_second": 12755.412,
1599
+ "eval_steps_per_second": 6.251,
1600
+ "step": 10600
1601
+ },
1602
+ {
1603
+ "epoch": 46.12068965517241,
1604
+ "grad_norm": 0.20913919806480408,
1605
+ "learning_rate": 0.0011296497061262364,
1606
+ "loss": 1.0419,
1607
+ "step": 10700
1608
+ },
1609
+ {
1610
+ "epoch": 46.12068965517241,
1611
+ "eval_loss": 4.827648162841797,
1612
+ "eval_runtime": 9.2894,
1613
+ "eval_samples_per_second": 12740.445,
1614
+ "eval_steps_per_second": 6.244,
1615
+ "step": 10700
1616
+ },
1617
+ {
1618
+ "epoch": 46.55172413793103,
1619
+ "grad_norm": 0.22326400876045227,
1620
+ "learning_rate": 0.0011161529495134655,
1621
+ "loss": 1.0222,
1622
+ "step": 10800
1623
+ },
1624
+ {
1625
+ "epoch": 46.55172413793103,
1626
+ "eval_loss": 4.75519323348999,
1627
+ "eval_runtime": 9.5215,
1628
+ "eval_samples_per_second": 12429.892,
1629
+ "eval_steps_per_second": 6.091,
1630
+ "step": 10800
1631
+ },
1632
+ {
1633
+ "epoch": 46.98275862068966,
1634
+ "grad_norm": 0.23432683944702148,
1635
+ "learning_rate": 0.0011026347096827578,
1636
+ "loss": 1.05,
1637
+ "step": 10900
1638
+ },
1639
+ {
1640
+ "epoch": 46.98275862068966,
1641
+ "eval_loss": 4.685484886169434,
1642
+ "eval_runtime": 9.3692,
1643
+ "eval_samples_per_second": 12631.957,
1644
+ "eval_steps_per_second": 6.191,
1645
+ "step": 10900
1646
+ },
1647
+ {
1648
+ "epoch": 47.41379310344828,
1649
+ "grad_norm": 0.22489097714424133,
1650
+ "learning_rate": 0.0010890974869175213,
1651
+ "loss": 1.0049,
1652
+ "step": 11000
1653
+ },
1654
+ {
1655
+ "epoch": 47.41379310344828,
1656
+ "eval_loss": 4.797884464263916,
1657
+ "eval_runtime": 9.2664,
1658
+ "eval_samples_per_second": 12772.008,
1659
+ "eval_steps_per_second": 6.259,
1660
+ "step": 11000
1661
+ },
1662
+ {
1663
+ "epoch": 47.8448275862069,
1664
+ "grad_norm": 0.2433350831270218,
1665
+ "learning_rate": 0.001075543785012176,
1666
+ "loss": 1.0345,
1667
+ "step": 11100
1668
+ },
1669
+ {
1670
+ "epoch": 47.8448275862069,
1671
+ "eval_loss": 4.75103759765625,
1672
+ "eval_runtime": 9.2767,
1673
+ "eval_samples_per_second": 12757.82,
1674
+ "eval_steps_per_second": 6.252,
1675
+ "step": 11100
1676
+ },
1677
+ {
1678
+ "epoch": 48.275862068965516,
1679
+ "grad_norm": 0.20576754212379456,
1680
+ "learning_rate": 0.0010619761108090632,
1681
+ "loss": 1.0123,
1682
+ "step": 11200
1683
+ },
1684
+ {
1685
+ "epoch": 48.275862068965516,
1686
+ "eval_loss": 4.8856587409973145,
1687
+ "eval_runtime": 9.2938,
1688
+ "eval_samples_per_second": 12734.377,
1689
+ "eval_steps_per_second": 6.241,
1690
+ "step": 11200
1691
+ },
1692
+ {
1693
+ "epoch": 48.706896551724135,
1694
+ "grad_norm": 0.20991961658000946,
1695
+ "learning_rate": 0.0010483969737347882,
1696
+ "loss": 1.0197,
1697
+ "step": 11300
1698
+ },
1699
+ {
1700
+ "epoch": 48.706896551724135,
1701
+ "eval_loss": 4.79866886138916,
1702
+ "eval_runtime": 9.2399,
1703
+ "eval_samples_per_second": 12808.67,
1704
+ "eval_steps_per_second": 6.277,
1705
+ "step": 11300
1706
+ },
1707
+ {
1708
+ "epoch": 49.13793103448276,
1709
+ "grad_norm": 0.22874431312084198,
1710
+ "learning_rate": 0.0010348088853360865,
1711
+ "loss": 1.0184,
1712
+ "step": 11400
1713
+ },
1714
+ {
1715
+ "epoch": 49.13793103448276,
1716
+ "eval_loss": 4.938732147216797,
1717
+ "eval_runtime": 9.5524,
1718
+ "eval_samples_per_second": 12389.725,
1719
+ "eval_steps_per_second": 6.072,
1720
+ "step": 11400
1721
+ },
1722
+ {
1723
+ "epoch": 49.56896551724138,
1724
+ "grad_norm": 0.2126045823097229,
1725
+ "learning_rate": 0.0010212143588152972,
1726
+ "loss": 1.0027,
1727
+ "step": 11500
1728
+ },
1729
+ {
1730
+ "epoch": 49.56896551724138,
1731
+ "eval_loss": 4.877725124359131,
1732
+ "eval_runtime": 9.4117,
1733
+ "eval_samples_per_second": 12574.862,
1734
+ "eval_steps_per_second": 6.163,
1735
+ "step": 11500
1736
+ },
1737
+ {
1738
+ "epoch": 50.0,
1739
+ "grad_norm": 0.22109200060367584,
1740
+ "learning_rate": 0.0010076159085655307,
1741
+ "loss": 1.0285,
1742
+ "step": 11600
1743
+ },
1744
+ {
1745
+ "epoch": 50.0,
1746
+ "eval_loss": 4.814455986022949,
1747
+ "eval_runtime": 9.4248,
1748
+ "eval_samples_per_second": 12557.425,
1749
+ "eval_steps_per_second": 6.154,
1750
+ "step": 11600
1751
+ },
1752
+ {
1753
+ "epoch": 50.43103448275862,
1754
+ "grad_norm": 0.19808194041252136,
1755
+ "learning_rate": 0.0009940160497056154,
1756
+ "loss": 0.9852,
1757
+ "step": 11700
1758
+ },
1759
+ {
1760
+ "epoch": 50.43103448275862,
1761
+ "eval_loss": 4.916428089141846,
1762
+ "eval_runtime": 9.2601,
1763
+ "eval_samples_per_second": 12780.803,
1764
+ "eval_steps_per_second": 6.263,
1765
+ "step": 11700
1766
+ },
1767
+ {
1768
+ "epoch": 50.86206896551724,
1769
+ "grad_norm": 0.19970214366912842,
1770
+ "learning_rate": 0.0009804172976149107,
1771
+ "loss": 1.0142,
1772
+ "step": 11800
1773
+ },
1774
+ {
1775
+ "epoch": 50.86206896551724,
1776
+ "eval_loss": 4.871610641479492,
1777
+ "eval_runtime": 9.2823,
1778
+ "eval_samples_per_second": 12750.17,
1779
+ "eval_steps_per_second": 6.248,
1780
+ "step": 11800
1781
+ },
1782
+ {
1783
+ "epoch": 51.293103448275865,
1784
+ "grad_norm": 0.21319642663002014,
1785
+ "learning_rate": 0.0009668221674680736,
1786
+ "loss": 0.9917,
1787
+ "step": 11900
1788
+ },
1789
+ {
1790
+ "epoch": 51.293103448275865,
1791
+ "eval_loss": 4.9852070808410645,
1792
+ "eval_runtime": 9.4078,
1793
+ "eval_samples_per_second": 12580.067,
1794
+ "eval_steps_per_second": 6.165,
1795
+ "step": 11900
1796
+ },
1797
+ {
1798
+ "epoch": 51.724137931034484,
1799
+ "grad_norm": 0.20660990476608276,
1800
+ "learning_rate": 0.0009532331737698578,
1801
+ "loss": 1.0004,
1802
+ "step": 12000
1803
+ },
1804
+ {
1805
+ "epoch": 51.724137931034484,
1806
+ "eval_loss": 4.884436130523682,
1807
+ "eval_runtime": 9.3887,
1808
+ "eval_samples_per_second": 12605.72,
1809
+ "eval_steps_per_second": 6.178,
1810
+ "step": 12000
1811
+ },
1812
+ {
1813
+ "epoch": 52.1551724137931,
1814
+ "grad_norm": 0.21437087655067444,
1815
+ "learning_rate": 0.0009396528298900436,
1816
+ "loss": 0.9971,
1817
+ "step": 12100
1818
+ },
1819
+ {
1820
+ "epoch": 52.1551724137931,
1821
+ "eval_loss": 5.011310577392578,
1822
+ "eval_runtime": 9.3814,
1823
+ "eval_samples_per_second": 12615.449,
1824
+ "eval_steps_per_second": 6.182,
1825
+ "step": 12100
1826
+ },
1827
+ {
1828
+ "epoch": 52.58620689655172,
1829
+ "grad_norm": 0.19599591195583344,
1830
+ "learning_rate": 0.0009260836475985731,
1831
+ "loss": 0.986,
1832
+ "step": 12200
1833
+ },
1834
+ {
1835
+ "epoch": 52.58620689655172,
1836
+ "eval_loss": 4.954803943634033,
1837
+ "eval_runtime": 9.217,
1838
+ "eval_samples_per_second": 12840.507,
1839
+ "eval_steps_per_second": 6.293,
1840
+ "step": 12200
1841
+ },
1842
+ {
1843
+ "epoch": 53.01724137931034,
1844
+ "grad_norm": 0.2128545045852661,
1845
+ "learning_rate": 0.0009125281366009814,
1846
+ "loss": 1.0058,
1847
+ "step": 12300
1848
+ },
1849
+ {
1850
+ "epoch": 53.01724137931034,
1851
+ "eval_loss": 5.177644729614258,
1852
+ "eval_runtime": 9.2288,
1853
+ "eval_samples_per_second": 12824.123,
1854
+ "eval_steps_per_second": 6.285,
1855
+ "step": 12300
1856
+ },
1857
+ {
1858
+ "epoch": 53.44827586206897,
1859
+ "grad_norm": 0.2027304619550705,
1860
+ "learning_rate": 0.0008989888040742127,
1861
+ "loss": 0.9701,
1862
+ "step": 12400
1863
+ },
1864
+ {
1865
+ "epoch": 53.44827586206897,
1866
+ "eval_loss": 5.006808280944824,
1867
+ "eval_runtime": 9.3934,
1868
+ "eval_samples_per_second": 12599.44,
1869
+ "eval_steps_per_second": 6.175,
1870
+ "step": 12400
1871
+ },
1872
+ {
1873
+ "epoch": 53.87931034482759,
1874
+ "grad_norm": 0.20546956360340118,
1875
+ "learning_rate": 0.0008854681542029017,
1876
+ "loss": 0.9968,
1877
+ "step": 12500
1878
+ },
1879
+ {
1880
+ "epoch": 53.87931034482759,
1881
+ "eval_loss": 4.943872928619385,
1882
+ "eval_runtime": 9.4255,
1883
+ "eval_samples_per_second": 12556.411,
1884
+ "eval_steps_per_second": 6.153,
1885
+ "step": 12500
1886
+ },
1887
+ {
1888
+ "epoch": 54.310344827586206,
1889
+ "grad_norm": 0.19119194149971008,
1890
+ "learning_rate": 0.0008719686877162076,
1891
+ "loss": 0.971,
1892
+ "step": 12600
1893
+ },
1894
+ {
1895
+ "epoch": 54.310344827586206,
1896
+ "eval_loss": 5.05370569229126,
1897
+ "eval_runtime": 9.4461,
1898
+ "eval_samples_per_second": 12529.13,
1899
+ "eval_steps_per_second": 6.14,
1900
+ "step": 12600
1901
+ },
1902
+ {
1903
+ "epoch": 54.741379310344826,
1904
+ "grad_norm": 0.20589964091777802,
1905
+ "learning_rate": 0.0008584929014252898,
1906
+ "loss": 0.9848,
1907
+ "step": 12700
1908
+ },
1909
+ {
1910
+ "epoch": 54.741379310344826,
1911
+ "eval_loss": 4.991157531738281,
1912
+ "eval_runtime": 9.2436,
1913
+ "eval_samples_per_second": 12803.575,
1914
+ "eval_steps_per_second": 6.275,
1915
+ "step": 12700
1916
+ },
1917
+ {
1918
+ "epoch": 55.172413793103445,
1919
+ "grad_norm": 0.20905207097530365,
1920
+ "learning_rate": 0.0008450432877615072,
1921
+ "loss": 0.9775,
1922
+ "step": 12800
1923
+ },
1924
+ {
1925
+ "epoch": 55.172413793103445,
1926
+ "eval_loss": 5.12783670425415,
1927
+ "eval_runtime": 9.227,
1928
+ "eval_samples_per_second": 12826.543,
1929
+ "eval_steps_per_second": 6.286,
1930
+ "step": 12800
1931
+ },
1932
+ {
1933
+ "epoch": 55.60344827586207,
1934
+ "grad_norm": 0.2055678814649582,
1935
+ "learning_rate": 0.0008316223343154266,
1936
+ "loss": 0.9707,
1937
+ "step": 12900
1938
+ },
1939
+ {
1940
+ "epoch": 55.60344827586207,
1941
+ "eval_loss": 5.068799018859863,
1942
+ "eval_runtime": 9.4304,
1943
+ "eval_samples_per_second": 12549.927,
1944
+ "eval_steps_per_second": 6.15,
1945
+ "step": 12900
1946
+ },
1947
+ {
1948
+ "epoch": 56.03448275862069,
1949
+ "grad_norm": 0.2250879853963852,
1950
+ "learning_rate": 0.0008182325233767267,
1951
+ "loss": 0.9858,
1952
+ "step": 13000
1953
+ },
1954
+ {
1955
+ "epoch": 56.03448275862069,
1956
+ "eval_loss": 5.237853527069092,
1957
+ "eval_runtime": 9.3164,
1958
+ "eval_samples_per_second": 12703.444,
1959
+ "eval_steps_per_second": 6.226,
1960
+ "step": 13000
1961
+ },
1962
+ {
1963
+ "epoch": 56.46551724137931,
1964
+ "grad_norm": 0.19878874719142914,
1965
+ "learning_rate": 0.0008048763314750851,
1966
+ "loss": 0.9562,
1967
+ "step": 13100
1968
+ },
1969
+ {
1970
+ "epoch": 56.46551724137931,
1971
+ "eval_loss": 5.091078281402588,
1972
+ "eval_runtime": 9.4103,
1973
+ "eval_samples_per_second": 12576.787,
1974
+ "eval_steps_per_second": 6.163,
1975
+ "step": 13100
1976
+ },
1977
+ {
1978
+ "epoch": 56.89655172413793,
1979
+ "grad_norm": 0.19659073650836945,
1980
+ "learning_rate": 0.0007915562289221262,
1981
+ "loss": 0.9789,
1982
+ "step": 13200
1983
+ },
1984
+ {
1985
+ "epoch": 56.89655172413793,
1986
+ "eval_loss": 5.069735527038574,
1987
+ "eval_runtime": 9.2932,
1988
+ "eval_samples_per_second": 12735.178,
1989
+ "eval_steps_per_second": 6.241,
1990
+ "step": 13200
1991
+ },
1992
+ {
1993
+ "epoch": 57.327586206896555,
1994
+ "grad_norm": 0.1879882961511612,
1995
+ "learning_rate": 0.0007782746793545225,
1996
+ "loss": 0.9552,
1997
+ "step": 13300
1998
+ },
1999
+ {
2000
+ "epoch": 57.327586206896555,
2001
+ "eval_loss": 5.13462495803833,
2002
+ "eval_runtime": 9.3255,
2003
+ "eval_samples_per_second": 12691.174,
2004
+ "eval_steps_per_second": 6.22,
2005
+ "step": 13300
2006
+ },
2007
+ {
2008
+ "epoch": 57.758620689655174,
2009
+ "grad_norm": 0.1820300966501236,
2010
+ "learning_rate": 0.0007650341392783306,
2011
+ "loss": 0.9687,
2012
+ "step": 13400
2013
+ },
2014
+ {
2015
+ "epoch": 57.758620689655174,
2016
+ "eval_loss": 5.058874130249023,
2017
+ "eval_runtime": 9.5918,
2018
+ "eval_samples_per_second": 12338.797,
2019
+ "eval_steps_per_second": 6.047,
2020
+ "step": 13400
2021
+ },
2022
+ {
2023
+ "epoch": 58.189655172413794,
2024
+ "grad_norm": 0.19287799298763275,
2025
+ "learning_rate": 0.0007518370576146431,
2026
+ "loss": 0.9605,
2027
+ "step": 13500
2028
+ },
2029
+ {
2030
+ "epoch": 58.189655172413794,
2031
+ "eval_loss": 5.197060585021973,
2032
+ "eval_runtime": 9.2272,
2033
+ "eval_samples_per_second": 12826.269,
2034
+ "eval_steps_per_second": 6.286,
2035
+ "step": 13500
2036
+ },
2037
+ {
2038
+ "epoch": 58.62068965517241,
2039
+ "grad_norm": 0.18384908139705658,
2040
+ "learning_rate": 0.0007386858752466458,
2041
+ "loss": 0.9561,
2042
+ "step": 13600
2043
+ },
2044
+ {
2045
+ "epoch": 58.62068965517241,
2046
+ "eval_loss": 5.117424488067627,
2047
+ "eval_runtime": 9.3805,
2048
+ "eval_samples_per_second": 12616.772,
2049
+ "eval_steps_per_second": 6.183,
2050
+ "step": 13600
2051
+ },
2052
+ {
2053
+ "epoch": 59.05172413793103,
2054
+ "grad_norm": 0.18120047450065613,
2055
+ "learning_rate": 0.0007255830245681625,
2056
+ "loss": 0.9681,
2057
+ "step": 13700
2058
+ },
2059
+ {
2060
+ "epoch": 59.05172413793103,
2061
+ "eval_loss": 5.253732204437256,
2062
+ "eval_runtime": 9.3436,
2063
+ "eval_samples_per_second": 12666.473,
2064
+ "eval_steps_per_second": 6.207,
2065
+ "step": 13700
2066
+ },
2067
+ {
2068
+ "epoch": 59.48275862068966,
2069
+ "grad_norm": 0.16635186970233917,
2070
+ "learning_rate": 0.0007125309290337666,
2071
+ "loss": 0.9428,
2072
+ "step": 13800
2073
+ },
2074
+ {
2075
+ "epoch": 59.48275862068966,
2076
+ "eval_loss": 5.163188934326172,
2077
+ "eval_runtime": 9.2127,
2078
+ "eval_samples_per_second": 12846.546,
2079
+ "eval_steps_per_second": 6.296,
2080
+ "step": 13800
2081
+ },
2082
+ {
2083
+ "epoch": 59.91379310344828,
2084
+ "grad_norm": 0.1886281669139862,
2085
+ "learning_rate": 0.000699532002710548,
2086
+ "loss": 0.9645,
2087
+ "step": 13900
2088
+ },
2089
+ {
2090
+ "epoch": 59.91379310344828,
2091
+ "eval_loss": 5.143974781036377,
2092
+ "eval_runtime": 9.3873,
2093
+ "eval_samples_per_second": 12607.627,
2094
+ "eval_steps_per_second": 6.179,
2095
+ "step": 13900
2096
+ },
2097
+ {
2098
+ "epoch": 60.3448275862069,
2099
+ "grad_norm": 0.19770479202270508,
2100
+ "learning_rate": 0.0006865886498316185,
2101
+ "loss": 0.9397,
2102
+ "step": 14000
2103
+ },
2104
+ {
2105
+ "epoch": 60.3448275862069,
2106
+ "eval_loss": 5.253783702850342,
2107
+ "eval_runtime": 9.4005,
2108
+ "eval_samples_per_second": 12589.827,
2109
+ "eval_steps_per_second": 6.17,
2110
+ "step": 14000
2111
+ },
2112
+ {
2113
+ "epoch": 60.775862068965516,
2114
+ "grad_norm": 0.1776113659143448,
2115
+ "learning_rate": 0.0006737032643514306,
2116
+ "loss": 0.9543,
2117
+ "step": 14100
2118
+ },
2119
+ {
2120
+ "epoch": 60.775862068965516,
2121
+ "eval_loss": 5.162989616394043,
2122
+ "eval_runtime": 9.4016,
2123
+ "eval_samples_per_second": 12588.403,
2124
+ "eval_steps_per_second": 6.169,
2125
+ "step": 14100
2126
+ },
2127
+ {
2128
+ "epoch": 61.206896551724135,
2129
+ "grad_norm": 0.17530174553394318,
2130
+ "learning_rate": 0.000660878229503003,
2131
+ "loss": 0.9453,
2132
+ "step": 14200
2133
+ },
2134
+ {
2135
+ "epoch": 61.206896551724135,
2136
+ "eval_loss": 5.27924919128418,
2137
+ "eval_runtime": 9.2266,
2138
+ "eval_samples_per_second": 12827.132,
2139
+ "eval_steps_per_second": 6.286,
2140
+ "step": 14200
2141
+ },
2142
+ {
2143
+ "epoch": 61.63793103448276,
2144
+ "grad_norm": 0.176809623837471,
2145
+ "learning_rate": 0.0006481159173571249,
2146
+ "loss": 0.943,
2147
+ "step": 14300
2148
+ },
2149
+ {
2150
+ "epoch": 61.63793103448276,
2151
+ "eval_loss": 5.216311931610107,
2152
+ "eval_runtime": 9.2805,
2153
+ "eval_samples_per_second": 12752.68,
2154
+ "eval_steps_per_second": 6.25,
2155
+ "step": 14300
2156
+ },
2157
+ {
2158
+ "epoch": 62.06896551724138,
2159
+ "grad_norm": 0.17062318325042725,
2160
+ "learning_rate": 0.0006354186883836291,
2161
+ "loss": 0.9508,
2162
+ "step": 14400
2163
+ },
2164
+ {
2165
+ "epoch": 62.06896551724138,
2166
+ "eval_loss": 5.327795505523682,
2167
+ "eval_runtime": 9.4204,
2168
+ "eval_samples_per_second": 12563.306,
2169
+ "eval_steps_per_second": 6.157,
2170
+ "step": 14400
2171
+ },
2172
+ {
2173
+ "epoch": 62.5,
2174
+ "grad_norm": 0.17021775245666504,
2175
+ "learning_rate": 0.0006227888910148052,
2176
+ "loss": 0.9313,
2177
+ "step": 14500
2178
+ },
2179
+ {
2180
+ "epoch": 62.5,
2181
+ "eval_loss": 5.2860636711120605,
2182
+ "eval_runtime": 9.2268,
2183
+ "eval_samples_per_second": 12826.933,
2184
+ "eval_steps_per_second": 6.286,
2185
+ "step": 14500
2186
+ },
2187
+ {
2188
+ "epoch": 62.93103448275862,
2189
+ "grad_norm": 0.17448383569717407,
2190
+ "learning_rate": 0.000610228861211044,
2191
+ "loss": 0.9511,
2192
+ "step": 14600
2193
+ },
2194
+ {
2195
+ "epoch": 62.93103448275862,
2196
+ "eval_loss": 5.205625057220459,
2197
+ "eval_runtime": 9.3599,
2198
+ "eval_samples_per_second": 12644.424,
2199
+ "eval_steps_per_second": 6.197,
2200
+ "step": 14600
2201
+ },
2202
+ {
2203
+ "epoch": 63.36206896551724,
2204
+ "grad_norm": 0.1694514900445938,
2205
+ "learning_rate": 0.0005977409220287874,
2206
+ "loss": 0.9273,
2207
+ "step": 14700
2208
+ },
2209
+ {
2210
+ "epoch": 63.36206896551724,
2211
+ "eval_loss": 5.312868595123291,
2212
+ "eval_runtime": 9.2195,
2213
+ "eval_samples_per_second": 12836.973,
2214
+ "eval_steps_per_second": 6.291,
2215
+ "step": 14700
2216
+ },
2217
+ {
2218
+ "epoch": 63.793103448275865,
2219
+ "grad_norm": 0.16801249980926514,
2220
+ "learning_rate": 0.000585451142747127,
2221
+ "loss": 0.941,
2222
+ "step": 14800
2223
+ },
2224
+ {
2225
+ "epoch": 63.793103448275865,
2226
+ "eval_loss": 5.279133319854736,
2227
+ "eval_runtime": 9.2736,
2228
+ "eval_samples_per_second": 12762.13,
2229
+ "eval_steps_per_second": 6.254,
2230
+ "step": 14800
2231
+ },
2232
+ {
2233
+ "epoch": 64.22413793103448,
2234
+ "grad_norm": 0.1708194762468338,
2235
+ "learning_rate": 0.000573113521933704,
2236
+ "loss": 0.932,
2237
+ "step": 14900
2238
+ },
2239
+ {
2240
+ "epoch": 64.22413793103448,
2241
+ "eval_loss": 5.35509729385376,
2242
+ "eval_runtime": 9.4022,
2243
+ "eval_samples_per_second": 12587.588,
2244
+ "eval_steps_per_second": 6.169,
2245
+ "step": 14900
2246
+ },
2247
+ {
2248
+ "epoch": 64.65517241379311,
2249
+ "grad_norm": 0.164137065410614,
2250
+ "learning_rate": 0.0005608548564570237,
2251
+ "loss": 0.9312,
2252
+ "step": 15000
2253
+ },
2254
+ {
2255
+ "epoch": 64.65517241379311,
2256
+ "eval_loss": 5.286611557006836,
2257
+ "eval_runtime": 9.3646,
2258
+ "eval_samples_per_second": 12638.17,
2259
+ "eval_steps_per_second": 6.194,
2260
+ "step": 15000
2261
+ },
2262
+ {
2263
+ "epoch": 65.08620689655173,
2264
+ "grad_norm": 0.16543178260326385,
2265
+ "learning_rate": 0.0005486774136343013,
2266
+ "loss": 0.9363,
2267
+ "step": 15100
2268
+ },
2269
+ {
2270
+ "epoch": 65.08620689655173,
2271
+ "eval_loss": 5.419331073760986,
2272
+ "eval_runtime": 9.3584,
2273
+ "eval_samples_per_second": 12646.509,
2274
+ "eval_steps_per_second": 6.198,
2275
+ "step": 15100
2276
+ },
2277
+ {
2278
+ "epoch": 65.51724137931035,
2279
+ "grad_norm": 0.16081272065639496,
2280
+ "learning_rate": 0.0005365834457601122,
2281
+ "loss": 0.9208,
2282
+ "step": 15200
2283
+ },
2284
+ {
2285
+ "epoch": 65.51724137931035,
2286
+ "eval_loss": 5.305998802185059,
2287
+ "eval_runtime": 9.2006,
2288
+ "eval_samples_per_second": 12863.448,
2289
+ "eval_steps_per_second": 6.304,
2290
+ "step": 15200
2291
+ },
2292
+ {
2293
+ "epoch": 65.94827586206897,
2294
+ "grad_norm": 0.16733773052692413,
2295
+ "learning_rate": 0.000524575189689815,
2296
+ "loss": 0.9373,
2297
+ "step": 15300
2298
+ },
2299
+ {
2300
+ "epoch": 65.94827586206897,
2301
+ "eval_loss": 5.309116363525391,
2302
+ "eval_runtime": 9.2273,
2303
+ "eval_samples_per_second": 12826.184,
2304
+ "eval_steps_per_second": 6.286,
2305
+ "step": 15300
2306
+ },
2307
+ {
2308
+ "epoch": 66.37931034482759,
2309
+ "grad_norm": 0.16292522847652435,
2310
+ "learning_rate": 0.0005126548664258318,
2311
+ "loss": 0.9144,
2312
+ "step": 15400
2313
+ },
2314
+ {
2315
+ "epoch": 66.37931034482759,
2316
+ "eval_loss": 5.348775386810303,
2317
+ "eval_runtime": 9.4291,
2318
+ "eval_samples_per_second": 12551.736,
2319
+ "eval_steps_per_second": 6.151,
2320
+ "step": 15400
2321
+ },
2322
+ {
2323
+ "epoch": 66.8103448275862,
2324
+ "grad_norm": 0.15373744070529938,
2325
+ "learning_rate": 0.0005008246807068595,
2326
+ "loss": 0.9293,
2327
+ "step": 15500
2328
+ },
2329
+ {
2330
+ "epoch": 66.8103448275862,
2331
+ "eval_loss": 5.3392744064331055,
2332
+ "eval_runtime": 9.2056,
2333
+ "eval_samples_per_second": 12856.353,
2334
+ "eval_steps_per_second": 6.3,
2335
+ "step": 15500
2336
+ },
2337
+ {
2338
+ "epoch": 67.24137931034483,
2339
+ "grad_norm": 0.15608009696006775,
2340
+ "learning_rate": 0.0004890868206000897,
2341
+ "loss": 0.919,
2342
+ "step": 15600
2343
+ },
2344
+ {
2345
+ "epoch": 67.24137931034483,
2346
+ "eval_loss": 5.408057689666748,
2347
+ "eval_runtime": 9.3819,
2348
+ "eval_samples_per_second": 12614.855,
2349
+ "eval_steps_per_second": 6.182,
2350
+ "step": 15600
2351
+ },
2352
+ {
2353
+ "epoch": 67.67241379310344,
2354
+ "grad_norm": 0.149125337600708,
2355
+ "learning_rate": 0.00047744345709651215,
2356
+ "loss": 0.9199,
2357
+ "step": 15700
2358
+ },
2359
+ {
2360
+ "epoch": 67.67241379310344,
2361
+ "eval_loss": 5.338372230529785,
2362
+ "eval_runtime": 9.2621,
2363
+ "eval_samples_per_second": 12777.995,
2364
+ "eval_steps_per_second": 6.262,
2365
+ "step": 15700
2366
+ },
2367
+ {
2368
+ "epoch": 68.10344827586206,
2369
+ "grad_norm": 0.16265520453453064,
2370
+ "learning_rate": 0.0004658967437093745,
2371
+ "loss": 0.9232,
2372
+ "step": 15800
2373
+ },
2374
+ {
2375
+ "epoch": 68.10344827586206,
2376
+ "eval_loss": 5.49737024307251,
2377
+ "eval_runtime": 9.4007,
2378
+ "eval_samples_per_second": 12589.636,
2379
+ "eval_steps_per_second": 6.17,
2380
+ "step": 15800
2381
+ },
2382
+ {
2383
+ "epoch": 68.53448275862068,
2384
+ "grad_norm": 0.1510947048664093,
2385
+ "learning_rate": 0.0004544488160758802,
2386
+ "loss": 0.9108,
2387
+ "step": 15900
2388
+ },
2389
+ {
2390
+ "epoch": 68.53448275862068,
2391
+ "eval_loss": 5.427255630493164,
2392
+ "eval_runtime": 9.4251,
2393
+ "eval_samples_per_second": 12557.063,
2394
+ "eval_steps_per_second": 6.154,
2395
+ "step": 15900
2396
+ },
2397
+ {
2398
+ "epoch": 68.96551724137932,
2399
+ "grad_norm": 0.15238577127456665,
2400
+ "learning_rate": 0.00044310179156218564,
2401
+ "loss": 0.9258,
2402
+ "step": 16000
2403
+ },
2404
+ {
2405
+ "epoch": 68.96551724137932,
2406
+ "eval_loss": 5.356872081756592,
2407
+ "eval_runtime": 9.2412,
2408
+ "eval_samples_per_second": 12806.847,
2409
+ "eval_steps_per_second": 6.276,
2410
+ "step": 16000
2411
+ },
2412
+ {
2413
+ "epoch": 69.39655172413794,
2414
+ "grad_norm": 0.1629515439271927,
2415
+ "learning_rate": 0.0004318577688717803,
2416
+ "loss": 0.9037,
2417
+ "step": 16100
2418
+ },
2419
+ {
2420
+ "epoch": 69.39655172413794,
2421
+ "eval_loss": 5.416144371032715,
2422
+ "eval_runtime": 9.5794,
2423
+ "eval_samples_per_second": 12354.676,
2424
+ "eval_steps_per_second": 6.055,
2425
+ "step": 16100
2426
+ },
2427
+ {
2428
+ "epoch": 69.82758620689656,
2429
+ "grad_norm": 0.15684714913368225,
2430
+ "learning_rate": 0.0004207188276573214,
2431
+ "loss": 0.9186,
2432
+ "step": 16200
2433
+ },
2434
+ {
2435
+ "epoch": 69.82758620689656,
2436
+ "eval_loss": 5.367498397827148,
2437
+ "eval_runtime": 9.2206,
2438
+ "eval_samples_per_second": 12835.473,
2439
+ "eval_steps_per_second": 6.29,
2440
+ "step": 16200
2441
+ },
2442
+ {
2443
+ "epoch": 70.25862068965517,
2444
+ "grad_norm": 0.14983102679252625,
2445
+ "learning_rate": 0.00040968702813598437,
2446
+ "loss": 0.9068,
2447
+ "step": 16300
2448
+ },
2449
+ {
2450
+ "epoch": 70.25862068965517,
2451
+ "eval_loss": 5.481048107147217,
2452
+ "eval_runtime": 9.2193,
2453
+ "eval_samples_per_second": 12837.283,
2454
+ "eval_steps_per_second": 6.291,
2455
+ "step": 16300
2456
+ },
2457
+ {
2458
+ "epoch": 70.6896551724138,
2459
+ "grad_norm": 0.14381574094295502,
2460
+ "learning_rate": 0.0003987644107084164,
2461
+ "loss": 0.91,
2462
+ "step": 16400
2463
+ },
2464
+ {
2465
+ "epoch": 70.6896551724138,
2466
+ "eval_loss": 5.443052291870117,
2467
+ "eval_runtime": 9.4115,
2468
+ "eval_samples_per_second": 12575.096,
2469
+ "eval_steps_per_second": 6.163,
2470
+ "step": 16400
2471
+ },
2472
+ {
2473
+ "epoch": 71.12068965517241,
2474
+ "grad_norm": 0.15462049841880798,
2475
+ "learning_rate": 0.0003879529955813511,
2476
+ "loss": 0.9124,
2477
+ "step": 16500
2478
+ },
2479
+ {
2480
+ "epoch": 71.12068965517241,
2481
+ "eval_loss": 5.550149440765381,
2482
+ "eval_runtime": 9.2298,
2483
+ "eval_samples_per_second": 12822.764,
2484
+ "eval_steps_per_second": 6.284,
2485
+ "step": 16500
2486
+ },
2487
+ {
2488
+ "epoch": 71.55172413793103,
2489
+ "grad_norm": 0.155101016163826,
2490
+ "learning_rate": 0.00037725478239395604,
2491
+ "loss": 0.9031,
2492
+ "step": 16600
2493
+ },
2494
+ {
2495
+ "epoch": 71.55172413793103,
2496
+ "eval_loss": 5.477322578430176,
2497
+ "eval_runtime": 9.5729,
2498
+ "eval_samples_per_second": 12363.111,
2499
+ "eval_steps_per_second": 6.059,
2500
+ "step": 16600
2501
+ },
2502
+ {
2503
+ "epoch": 71.98275862068965,
2504
+ "grad_norm": 0.15142673254013062,
2505
+ "learning_rate": 0.0003666717498479895,
2506
+ "loss": 0.9144,
2507
+ "step": 16700
2508
+ },
2509
+ {
2510
+ "epoch": 71.98275862068965,
2511
+ "eval_loss": 5.421406269073486,
2512
+ "eval_runtime": 9.234,
2513
+ "eval_samples_per_second": 12816.868,
2514
+ "eval_steps_per_second": 6.281,
2515
+ "step": 16700
2516
+ },
2517
+ {
2518
+ "epoch": 72.41379310344827,
2519
+ "grad_norm": 0.1433909386396408,
2520
+ "learning_rate": 0.0003562058553418274,
2521
+ "loss": 0.8949,
2522
+ "step": 16800
2523
+ },
2524
+ {
2525
+ "epoch": 72.41379310344827,
2526
+ "eval_loss": 5.500460624694824,
2527
+ "eval_runtime": 9.2277,
2528
+ "eval_samples_per_second": 12825.623,
2529
+ "eval_steps_per_second": 6.285,
2530
+ "step": 16800
2531
+ },
2532
+ {
2533
+ "epoch": 72.84482758620689,
2534
+ "grad_norm": 0.15289029479026794,
2535
+ "learning_rate": 0.0003458590346084269,
2536
+ "loss": 0.9083,
2537
+ "step": 16900
2538
+ },
2539
+ {
2540
+ "epoch": 72.84482758620689,
2541
+ "eval_loss": 5.461334228515625,
2542
+ "eval_runtime": 9.4103,
2543
+ "eval_samples_per_second": 12576.69,
2544
+ "eval_steps_per_second": 6.163,
2545
+ "step": 16900
2546
+ },
2547
+ {
2548
+ "epoch": 73.27586206896552,
2549
+ "grad_norm": 0.1488148272037506,
2550
+ "learning_rate": 0.0003356332013573039,
2551
+ "loss": 0.8971,
2552
+ "step": 17000
2553
+ },
2554
+ {
2555
+ "epoch": 73.27586206896552,
2556
+ "eval_loss": 5.540443420410156,
2557
+ "eval_runtime": 9.2113,
2558
+ "eval_samples_per_second": 12848.519,
2559
+ "eval_steps_per_second": 6.297,
2560
+ "step": 17000
2561
+ },
2562
+ {
2563
+ "epoch": 73.70689655172414,
2564
+ "grad_norm": 0.14399707317352295,
2565
+ "learning_rate": 0.0003255302469205801,
2566
+ "loss": 0.9015,
2567
+ "step": 17100
2568
+ },
2569
+ {
2570
+ "epoch": 73.70689655172414,
2571
+ "eval_loss": 5.492214679718018,
2572
+ "eval_runtime": 9.3487,
2573
+ "eval_samples_per_second": 12659.559,
2574
+ "eval_steps_per_second": 6.204,
2575
+ "step": 17100
2576
+ },
2577
+ {
2578
+ "epoch": 74.13793103448276,
2579
+ "grad_norm": 0.14008432626724243,
2580
+ "learning_rate": 0.0003155520399031659,
2581
+ "loss": 0.9007,
2582
+ "step": 17200
2583
+ },
2584
+ {
2585
+ "epoch": 74.13793103448276,
2586
+ "eval_loss": 5.615210056304932,
2587
+ "eval_runtime": 9.2197,
2588
+ "eval_samples_per_second": 12836.75,
2589
+ "eval_steps_per_second": 6.291,
2590
+ "step": 17200
2591
+ },
2592
+ {
2593
+ "epoch": 74.56896551724138,
2594
+ "grad_norm": 0.14556729793548584,
2595
+ "learning_rate": 0.00030570042583715406,
2596
+ "loss": 0.8943,
2597
+ "step": 17300
2598
+ },
2599
+ {
2600
+ "epoch": 74.56896551724138,
2601
+ "eval_loss": 5.521312236785889,
2602
+ "eval_runtime": 9.2696,
2603
+ "eval_samples_per_second": 12767.677,
2604
+ "eval_steps_per_second": 6.257,
2605
+ "step": 17300
2606
+ },
2607
+ {
2608
+ "epoch": 75.0,
2609
+ "grad_norm": 0.14520053565502167,
2610
+ "learning_rate": 0.00029597722684047626,
2611
+ "loss": 0.9053,
2612
+ "step": 17400
2613
+ },
2614
+ {
2615
+ "epoch": 75.0,
2616
+ "eval_loss": 5.499870300292969,
2617
+ "eval_runtime": 9.556,
2618
+ "eval_samples_per_second": 12385.056,
2619
+ "eval_steps_per_second": 6.07,
2620
+ "step": 17400
2621
+ },
2622
+ {
2623
+ "epoch": 75.43103448275862,
2624
+ "grad_norm": 0.1396595537662506,
2625
+ "learning_rate": 0.00028638424127988793,
2626
+ "loss": 0.8865,
2627
+ "step": 17500
2628
+ },
2629
+ {
2630
+ "epoch": 75.43103448275862,
2631
+ "eval_loss": 5.5359368324279785,
2632
+ "eval_runtime": 9.2502,
2633
+ "eval_samples_per_second": 12794.465,
2634
+ "eval_steps_per_second": 6.27,
2635
+ "step": 17500
2636
+ },
2637
+ {
2638
+ "epoch": 75.86206896551724,
2639
+ "grad_norm": 0.14295458793640137,
2640
+ "learning_rate": 0.00027692324343835285,
2641
+ "loss": 0.8994,
2642
+ "step": 17600
2643
+ },
2644
+ {
2645
+ "epoch": 75.86206896551724,
2646
+ "eval_loss": 5.517573356628418,
2647
+ "eval_runtime": 9.3995,
2648
+ "eval_samples_per_second": 12591.242,
2649
+ "eval_steps_per_second": 6.171,
2650
+ "step": 17600
2651
+ },
2652
+ {
2653
+ "epoch": 76.29310344827586,
2654
+ "grad_norm": 0.13252246379852295,
2655
+ "learning_rate": 0.0002675959831868763,
2656
+ "loss": 0.8885,
2657
+ "step": 17700
2658
+ },
2659
+ {
2660
+ "epoch": 76.29310344827586,
2661
+ "eval_loss": 5.573907375335693,
2662
+ "eval_runtime": 9.2866,
2663
+ "eval_samples_per_second": 12744.237,
2664
+ "eval_steps_per_second": 6.246,
2665
+ "step": 17700
2666
+ },
2667
+ {
2668
+ "epoch": 76.72413793103448,
2669
+ "grad_norm": 0.13853983581066132,
2670
+ "learning_rate": 0.00025840418566085567,
2671
+ "loss": 0.8931,
2672
+ "step": 17800
2673
+ },
2674
+ {
2675
+ "epoch": 76.72413793103448,
2676
+ "eval_loss": 5.549348831176758,
2677
+ "eval_runtime": 9.2026,
2678
+ "eval_samples_per_second": 12860.571,
2679
+ "eval_steps_per_second": 6.303,
2680
+ "step": 17800
2681
+ },
2682
+ {
2683
+ "epoch": 77.15517241379311,
2684
+ "grad_norm": 0.13731436431407928,
2685
+ "learning_rate": 0.0002493495509410063,
2686
+ "loss": 0.8918,
2687
+ "step": 17900
2688
+ },
2689
+ {
2690
+ "epoch": 77.15517241379311,
2691
+ "eval_loss": 5.61083984375,
2692
+ "eval_runtime": 9.3473,
2693
+ "eval_samples_per_second": 12661.489,
2694
+ "eval_steps_per_second": 6.205,
2695
+ "step": 17900
2696
+ },
2697
+ {
2698
+ "epoch": 77.58620689655173,
2699
+ "grad_norm": 0.14138437807559967,
2700
+ "learning_rate": 0.00024043375373892028,
2701
+ "loss": 0.8866,
2702
+ "step": 18000
2703
+ },
2704
+ {
2705
+ "epoch": 77.58620689655173,
2706
+ "eval_loss": 5.592303276062012,
2707
+ "eval_runtime": 9.1912,
2708
+ "eval_samples_per_second": 12876.581,
2709
+ "eval_steps_per_second": 6.31,
2710
+ "step": 18000
2711
+ },
2712
+ {
2713
+ "epoch": 78.01724137931035,
2714
+ "grad_norm": 0.13110779225826263,
2715
+ "learning_rate": 0.00023165844308731864,
2716
+ "loss": 0.8951,
2717
+ "step": 18100
2718
+ },
2719
+ {
2720
+ "epoch": 78.01724137931035,
2721
+ "eval_loss": 5.612074375152588,
2722
+ "eval_runtime": 9.2148,
2723
+ "eval_samples_per_second": 12843.529,
2724
+ "eval_steps_per_second": 6.294,
2725
+ "step": 18100
2726
+ },
2727
+ {
2728
+ "epoch": 78.44827586206897,
2729
+ "grad_norm": 0.1357899308204651,
2730
+ "learning_rate": 0.00022302524203505182,
2731
+ "loss": 0.8804,
2732
+ "step": 18200
2733
+ },
2734
+ {
2735
+ "epoch": 78.44827586206897,
2736
+ "eval_loss": 5.599719047546387,
2737
+ "eval_runtime": 9.2495,
2738
+ "eval_samples_per_second": 12795.407,
2739
+ "eval_steps_per_second": 6.271,
2740
+ "step": 18200
2741
+ },
2742
+ {
2743
+ "epoch": 78.87931034482759,
2744
+ "grad_norm": 0.12847194075584412,
2745
+ "learning_rate": 0.00021453574734690674,
2746
+ "loss": 0.891,
2747
+ "step": 18300
2748
+ },
2749
+ {
2750
+ "epoch": 78.87931034482759,
2751
+ "eval_loss": 5.547791004180908,
2752
+ "eval_runtime": 9.2953,
2753
+ "eval_samples_per_second": 12732.34,
2754
+ "eval_steps_per_second": 6.24,
2755
+ "step": 18300
2756
+ },
2757
+ {
2758
+ "epoch": 79.3103448275862,
2759
+ "grad_norm": 0.13394750654697418,
2760
+ "learning_rate": 0.00020619152920827555,
2761
+ "loss": 0.8815,
2762
+ "step": 18400
2763
+ },
2764
+ {
2765
+ "epoch": 79.3103448275862,
2766
+ "eval_loss": 5.631216049194336,
2767
+ "eval_runtime": 9.5355,
2768
+ "eval_samples_per_second": 12411.571,
2769
+ "eval_steps_per_second": 6.083,
2770
+ "step": 18400
2771
+ },
2772
+ {
2773
+ "epoch": 79.74137931034483,
2774
+ "grad_norm": 0.13404041528701782,
2775
+ "learning_rate": 0.00019799413093473962,
2776
+ "loss": 0.8863,
2777
+ "step": 18500
2778
+ },
2779
+ {
2780
+ "epoch": 79.74137931034483,
2781
+ "eval_loss": 5.588956356048584,
2782
+ "eval_runtime": 9.2275,
2783
+ "eval_samples_per_second": 12825.837,
2784
+ "eval_steps_per_second": 6.286,
2785
+ "step": 18500
2786
+ },
2787
+ {
2788
+ "epoch": 80.17241379310344,
2789
+ "grad_norm": 0.12581737339496613,
2790
+ "learning_rate": 0.00018994506868662254,
2791
+ "loss": 0.8839,
2792
+ "step": 18600
2793
+ },
2794
+ {
2795
+ "epoch": 80.17241379310344,
2796
+ "eval_loss": 5.659803867340088,
2797
+ "eval_runtime": 9.2909,
2798
+ "eval_samples_per_second": 12738.337,
2799
+ "eval_steps_per_second": 6.243,
2800
+ "step": 18600
2801
+ },
2802
+ {
2803
+ "epoch": 80.60344827586206,
2804
+ "grad_norm": 0.13299065828323364,
2805
+ "learning_rate": 0.00018204583118856953,
2806
+ "loss": 0.8797,
2807
+ "step": 18700
2808
+ },
2809
+ {
2810
+ "epoch": 80.60344827586206,
2811
+ "eval_loss": 5.609480381011963,
2812
+ "eval_runtime": 9.2047,
2813
+ "eval_samples_per_second": 12857.738,
2814
+ "eval_steps_per_second": 6.301,
2815
+ "step": 18700
2816
+ },
2817
+ {
2818
+ "epoch": 81.03448275862068,
2819
+ "grad_norm": 0.13382546603679657,
2820
+ "learning_rate": 0.00017429787945419606,
2821
+ "loss": 0.8877,
2822
+ "step": 18800
2823
+ },
2824
+ {
2825
+ "epoch": 81.03448275862068,
2826
+ "eval_loss": 5.6830644607543945,
2827
+ "eval_runtime": 9.2004,
2828
+ "eval_samples_per_second": 12863.686,
2829
+ "eval_steps_per_second": 6.304,
2830
+ "step": 18800
2831
+ },
2832
+ {
2833
+ "epoch": 81.46551724137932,
2834
+ "grad_norm": 0.12398296594619751,
2835
+ "learning_rate": 0.00016670264651586397,
2836
+ "loss": 0.8757,
2837
+ "step": 18900
2838
+ },
2839
+ {
2840
+ "epoch": 81.46551724137932,
2841
+ "eval_loss": 5.642313480377197,
2842
+ "eval_runtime": 9.3934,
2843
+ "eval_samples_per_second": 12599.406,
2844
+ "eval_steps_per_second": 6.175,
2845
+ "step": 18900
2846
+ },
2847
+ {
2848
+ "epoch": 81.89655172413794,
2849
+ "grad_norm": 0.13016852736473083,
2850
+ "learning_rate": 0.00015926153715963565,
2851
+ "loss": 0.8831,
2852
+ "step": 19000
2853
+ },
2854
+ {
2855
+ "epoch": 81.89655172413794,
2856
+ "eval_loss": 5.612121105194092,
2857
+ "eval_runtime": 9.2714,
2858
+ "eval_samples_per_second": 12765.106,
2859
+ "eval_steps_per_second": 6.256,
2860
+ "step": 19000
2861
+ },
2862
+ {
2863
+ "epoch": 82.32758620689656,
2864
+ "grad_norm": 0.1265469789505005,
2865
+ "learning_rate": 0.00015197592766544755,
2866
+ "loss": 0.8754,
2867
+ "step": 19100
2868
+ },
2869
+ {
2870
+ "epoch": 82.32758620689656,
2871
+ "eval_loss": 5.6395673751831055,
2872
+ "eval_runtime": 9.3621,
2873
+ "eval_samples_per_second": 12641.446,
2874
+ "eval_steps_per_second": 6.195,
2875
+ "step": 19100
2876
+ },
2877
+ {
2878
+ "epoch": 82.75862068965517,
2879
+ "grad_norm": 0.12860514223575592,
2880
+ "learning_rate": 0.00014484716555255916,
2881
+ "loss": 0.879,
2882
+ "step": 19200
2883
+ },
2884
+ {
2885
+ "epoch": 82.75862068965517,
2886
+ "eval_loss": 5.6297993659973145,
2887
+ "eval_runtime": 9.2093,
2888
+ "eval_samples_per_second": 12851.218,
2889
+ "eval_steps_per_second": 6.298,
2890
+ "step": 19200
2891
+ },
2892
+ {
2893
+ "epoch": 83.1896551724138,
2894
+ "grad_norm": 0.12957799434661865,
2895
+ "learning_rate": 0.000137876569330322,
2896
+ "loss": 0.8766,
2897
+ "step": 19300
2898
+ },
2899
+ {
2900
+ "epoch": 83.1896551724138,
2901
+ "eval_loss": 5.6777119636535645,
2902
+ "eval_runtime": 9.2378,
2903
+ "eval_samples_per_second": 12811.648,
2904
+ "eval_steps_per_second": 6.279,
2905
+ "step": 19300
2906
+ },
2907
+ {
2908
+ "epoch": 83.62068965517241,
2909
+ "grad_norm": 0.13022403419017792,
2910
+ "learning_rate": 0.0001310654282543118,
2911
+ "loss": 0.8737,
2912
+ "step": 19400
2913
+ },
2914
+ {
2915
+ "epoch": 83.62068965517241,
2916
+ "eval_loss": 5.648399829864502,
2917
+ "eval_runtime": 9.3858,
2918
+ "eval_samples_per_second": 12609.571,
2919
+ "eval_steps_per_second": 6.18,
2920
+ "step": 19400
2921
+ },
2922
+ {
2923
+ "epoch": 84.05172413793103,
2924
+ "grad_norm": 0.13024047017097473,
2925
+ "learning_rate": 0.00012441500208787225,
2926
+ "loss": 0.8801,
2927
+ "step": 19500
2928
+ },
2929
+ {
2930
+ "epoch": 84.05172413793103,
2931
+ "eval_loss": 5.718311786651611,
2932
+ "eval_runtime": 9.2041,
2933
+ "eval_samples_per_second": 12858.468,
2934
+ "eval_steps_per_second": 6.302,
2935
+ "step": 19500
2936
+ },
2937
+ {
2938
+ "epoch": 84.48275862068965,
2939
+ "grad_norm": 0.12808039784431458,
2940
+ "learning_rate": 0.00011799060008838792,
2941
+ "loss": 0.8706,
2942
+ "step": 19600
2943
+ },
2944
+ {
2945
+ "epoch": 84.48275862068965,
2946
+ "eval_loss": 5.671109676361084,
2947
+ "eval_runtime": 9.3429,
2948
+ "eval_samples_per_second": 12667.454,
2949
+ "eval_steps_per_second": 6.208,
2950
+ "step": 19600
2951
+ },
2952
+ {
2953
+ "epoch": 84.91379310344827,
2954
+ "grad_norm": 0.1306673139333725,
2955
+ "learning_rate": 0.00011166362661083462,
2956
+ "loss": 0.8775,
2957
+ "step": 19700
2958
+ },
2959
+ {
2960
+ "epoch": 84.91379310344827,
2961
+ "eval_loss": 5.6501569747924805,
2962
+ "eval_runtime": 9.2122,
2963
+ "eval_samples_per_second": 12847.207,
2964
+ "eval_steps_per_second": 6.296,
2965
+ "step": 19700
2966
+ },
2967
+ {
2968
+ "epoch": 85.34482758620689,
2969
+ "grad_norm": 0.12629491090774536,
2970
+ "learning_rate": 0.00010550095652799053,
2971
+ "loss": 0.87,
2972
+ "step": 19800
2973
+ },
2974
+ {
2975
+ "epoch": 85.34482758620689,
2976
+ "eval_loss": 5.678717136383057,
2977
+ "eval_runtime": 9.2402,
2978
+ "eval_samples_per_second": 12808.275,
2979
+ "eval_steps_per_second": 6.277,
2980
+ "step": 19800
2981
+ },
2982
+ {
2983
+ "epoch": 85.77586206896552,
2984
+ "grad_norm": 0.12458854168653488,
2985
+ "learning_rate": 9.950372966441468e-05,
2986
+ "loss": 0.8734,
2987
+ "step": 19900
2988
+ },
2989
+ {
2990
+ "epoch": 85.77586206896552,
2991
+ "eval_loss": 5.6672797203063965,
2992
+ "eval_runtime": 9.4106,
2993
+ "eval_samples_per_second": 12576.309,
2994
+ "eval_steps_per_second": 6.163,
2995
+ "step": 19900
2996
+ },
2997
+ {
2998
+ "epoch": 86.20689655172414,
2999
+ "grad_norm": 0.13032549619674683,
3000
+ "learning_rate": 9.367305524490289e-05,
3001
+ "loss": 0.8715,
3002
+ "step": 20000
3003
+ },
3004
+ {
3005
+ "epoch": 86.20689655172414,
3006
+ "eval_loss": 5.714916229248047,
3007
+ "eval_runtime": 9.2405,
3008
+ "eval_samples_per_second": 12807.829,
3009
+ "eval_steps_per_second": 6.277,
3010
+ "step": 20000
3011
+ },
3012
+ {
3013
+ "epoch": 86.63793103448276,
3014
+ "grad_norm": 0.12013848125934601,
3015
+ "learning_rate": 8.801001168933009e-05,
3016
+ "loss": 0.8702,
3017
+ "step": 20100
3018
+ },
3019
+ {
3020
+ "epoch": 86.63793103448276,
3021
+ "eval_loss": 5.669819355010986,
3022
+ "eval_runtime": 9.4322,
3023
+ "eval_samples_per_second": 12547.573,
3024
+ "eval_steps_per_second": 6.149,
3025
+ "step": 20100
3026
+ },
3027
+ {
3028
+ "epoch": 87.06896551724138,
3029
+ "grad_norm": 0.12465270608663559,
3030
+ "learning_rate": 8.251564641318931e-05,
3031
+ "loss": 0.8729,
3032
+ "step": 20200
3033
+ },
3034
+ {
3035
+ "epoch": 87.06896551724138,
3036
+ "eval_loss": 5.735557556152344,
3037
+ "eval_runtime": 9.2721,
3038
+ "eval_samples_per_second": 12764.157,
3039
+ "eval_steps_per_second": 6.255,
3040
+ "step": 20200
3041
+ },
3042
+ {
3043
+ "epoch": 87.5,
3044
+ "grad_norm": 0.1301339566707611,
3045
+ "learning_rate": 7.71909756338659e-05,
3046
+ "loss": 0.8668,
3047
+ "step": 20300
3048
+ },
3049
+ {
3050
+ "epoch": 87.5,
3051
+ "eval_loss": 5.69363260269165,
3052
+ "eval_runtime": 9.1947,
3053
+ "eval_samples_per_second": 12871.683,
3054
+ "eval_steps_per_second": 6.308,
3055
+ "step": 20300
3056
+ },
3057
+ {
3058
+ "epoch": 87.93103448275862,
3059
+ "grad_norm": 0.12156441807746887,
3060
+ "learning_rate": 7.203698418268168e-05,
3061
+ "loss": 0.8717,
3062
+ "step": 20400
3063
+ },
3064
+ {
3065
+ "epoch": 87.93103448275862,
3066
+ "eval_loss": 5.678969860076904,
3067
+ "eval_runtime": 9.339,
3068
+ "eval_samples_per_second": 12672.825,
3069
+ "eval_steps_per_second": 6.211,
3070
+ "step": 20400
3071
+ },
3072
+ {
3073
+ "epoch": 88.36206896551724,
3074
+ "grad_norm": 0.12576840817928314,
3075
+ "learning_rate": 6.705462532274386e-05,
3076
+ "loss": 0.8657,
3077
+ "step": 20500
3078
+ },
3079
+ {
3080
+ "epoch": 88.36206896551724,
3081
+ "eval_loss": 5.71215295791626,
3082
+ "eval_runtime": 9.2042,
3083
+ "eval_samples_per_second": 12858.316,
3084
+ "eval_steps_per_second": 6.301,
3085
+ "step": 20500
3086
+ },
3087
+ {
3088
+ "epoch": 88.79310344827586,
3089
+ "grad_norm": 0.12462034821510315,
3090
+ "learning_rate": 6.22448205726337e-05,
3091
+ "loss": 0.8686,
3092
+ "step": 20600
3093
+ },
3094
+ {
3095
+ "epoch": 88.79310344827586,
3096
+ "eval_loss": 5.695542335510254,
3097
+ "eval_runtime": 9.3624,
3098
+ "eval_samples_per_second": 12641.117,
3099
+ "eval_steps_per_second": 6.195,
3100
+ "step": 20600
3101
+ },
3102
+ {
3103
+ "epoch": 89.22413793103448,
3104
+ "grad_norm": 0.12270428240299225,
3105
+ "learning_rate": 5.760845953596527e-05,
3106
+ "loss": 0.8667,
3107
+ "step": 20700
3108
+ },
3109
+ {
3110
+ "epoch": 89.22413793103448,
3111
+ "eval_loss": 5.725538730621338,
3112
+ "eval_runtime": 9.258,
3113
+ "eval_samples_per_second": 12783.647,
3114
+ "eval_steps_per_second": 6.265,
3115
+ "step": 20700
3116
+ },
3117
+ {
3118
+ "epoch": 89.65517241379311,
3119
+ "grad_norm": 0.13276109099388123,
3120
+ "learning_rate": 5.314639973684787e-05,
3121
+ "loss": 0.8664,
3122
+ "step": 20800
3123
+ },
3124
+ {
3125
+ "epoch": 89.65517241379311,
3126
+ "eval_loss": 5.707296371459961,
3127
+ "eval_runtime": 9.2285,
3128
+ "eval_samples_per_second": 12824.489,
3129
+ "eval_steps_per_second": 6.285,
3130
+ "step": 20800
3131
+ },
3132
+ {
3133
+ "epoch": 90.08620689655173,
3134
+ "grad_norm": 0.12904435396194458,
3135
+ "learning_rate": 4.885946646128203e-05,
3136
+ "loss": 0.8672,
3137
+ "step": 20900
3138
+ },
3139
+ {
3140
+ "epoch": 90.08620689655173,
3141
+ "eval_loss": 5.738578796386719,
3142
+ "eval_runtime": 9.3955,
3143
+ "eval_samples_per_second": 12596.524,
3144
+ "eval_steps_per_second": 6.173,
3145
+ "step": 20900
3146
+ },
3147
+ {
3148
+ "epoch": 90.51724137931035,
3149
+ "grad_norm": 0.12599405646324158,
3150
+ "learning_rate": 4.474845260451621e-05,
3151
+ "loss": 0.8636,
3152
+ "step": 21000
3153
+ },
3154
+ {
3155
+ "epoch": 90.51724137931035,
3156
+ "eval_loss": 5.716848373413086,
3157
+ "eval_runtime": 9.278,
3158
+ "eval_samples_per_second": 12756.107,
3159
+ "eval_steps_per_second": 6.251,
3160
+ "step": 21000
3161
+ },
3162
+ {
3163
+ "epoch": 90.94827586206897,
3164
+ "grad_norm": 0.11657002568244934,
3165
+ "learning_rate": 4.081411852439809e-05,
3166
+ "loss": 0.8667,
3167
+ "step": 21100
3168
+ },
3169
+ {
3170
+ "epoch": 90.94827586206897,
3171
+ "eval_loss": 5.702765941619873,
3172
+ "eval_runtime": 9.2293,
3173
+ "eval_samples_per_second": 12823.399,
3174
+ "eval_steps_per_second": 6.284,
3175
+ "step": 21100
3176
+ },
3177
+ {
3178
+ "epoch": 91.37931034482759,
3179
+ "grad_norm": 0.12212695181369781,
3180
+ "learning_rate": 3.705719190073964e-05,
3181
+ "loss": 0.8633,
3182
+ "step": 21200
3183
+ },
3184
+ {
3185
+ "epoch": 91.37931034482759,
3186
+ "eval_loss": 5.724610805511475,
3187
+ "eval_runtime": 9.4423,
3188
+ "eval_samples_per_second": 12534.095,
3189
+ "eval_steps_per_second": 6.143,
3190
+ "step": 21200
3191
+ },
3192
+ {
3193
+ "epoch": 91.8103448275862,
3194
+ "grad_norm": 0.1268962025642395,
3195
+ "learning_rate": 3.3478367600728754e-05,
3196
+ "loss": 0.8638,
3197
+ "step": 21300
3198
+ },
3199
+ {
3200
+ "epoch": 91.8103448275862,
3201
+ "eval_loss": 5.716207981109619,
3202
+ "eval_runtime": 9.2366,
3203
+ "eval_samples_per_second": 12813.288,
3204
+ "eval_steps_per_second": 6.279,
3205
+ "step": 21300
3206
+ },
3207
+ {
3208
+ "epoch": 92.24137931034483,
3209
+ "grad_norm": 0.1250430792570114,
3210
+ "learning_rate": 3.0078307550409967e-05,
3211
+ "loss": 0.8641,
3212
+ "step": 21400
3213
+ },
3214
+ {
3215
+ "epoch": 92.24137931034483,
3216
+ "eval_loss": 5.734896183013916,
3217
+ "eval_runtime": 9.3959,
3218
+ "eval_samples_per_second": 12596.041,
3219
+ "eval_steps_per_second": 6.173,
3220
+ "step": 21400
3221
+ },
3222
+ {
3223
+ "epoch": 92.67241379310344,
3224
+ "grad_norm": 0.1324104219675064,
3225
+ "learning_rate": 2.685764061225615e-05,
3226
+ "loss": 0.8629,
3227
+ "step": 21500
3228
+ },
3229
+ {
3230
+ "epoch": 92.67241379310344,
3231
+ "eval_loss": 5.724818706512451,
3232
+ "eval_runtime": 9.2529,
3233
+ "eval_samples_per_second": 12790.698,
3234
+ "eval_steps_per_second": 6.268,
3235
+ "step": 21500
3236
+ },
3237
+ {
3238
+ "epoch": 93.10344827586206,
3239
+ "grad_norm": 0.11621604114770889,
3240
+ "learning_rate": 2.381696246885634e-05,
3241
+ "loss": 0.8635,
3242
+ "step": 21600
3243
+ },
3244
+ {
3245
+ "epoch": 93.10344827586206,
3246
+ "eval_loss": 5.741239070892334,
3247
+ "eval_runtime": 9.2022,
3248
+ "eval_samples_per_second": 12861.136,
3249
+ "eval_steps_per_second": 6.303,
3250
+ "step": 21600
3251
+ },
3252
+ {
3253
+ "epoch": 93.53448275862068,
3254
+ "grad_norm": 0.12569952011108398,
3255
+ "learning_rate": 2.0956835512741523e-05,
3256
+ "loss": 0.8616,
3257
+ "step": 21700
3258
+ },
3259
+ {
3260
+ "epoch": 93.53448275862068,
3261
+ "eval_loss": 5.736214637756348,
3262
+ "eval_runtime": 9.3493,
3263
+ "eval_samples_per_second": 12658.796,
3264
+ "eval_steps_per_second": 6.204,
3265
+ "step": 21700
3266
+ },
3267
+ {
3268
+ "epoch": 93.96551724137932,
3269
+ "grad_norm": 0.123323455452919,
3270
+ "learning_rate": 1.8277788742365964e-05,
3271
+ "loss": 0.8634,
3272
+ "step": 21800
3273
+ },
3274
+ {
3275
+ "epoch": 93.96551724137932,
3276
+ "eval_loss": 5.7303266525268555,
3277
+ "eval_runtime": 9.2176,
3278
+ "eval_samples_per_second": 12839.714,
3279
+ "eval_steps_per_second": 6.292,
3280
+ "step": 21800
3281
+ },
3282
+ {
3283
+ "epoch": 94.39655172413794,
3284
+ "grad_norm": 0.12398874014616013,
3285
+ "learning_rate": 1.5780317664264798e-05,
3286
+ "loss": 0.8613,
3287
+ "step": 21900
3288
+ },
3289
+ {
3290
+ "epoch": 94.39655172413794,
3291
+ "eval_loss": 5.74062967300415,
3292
+ "eval_runtime": 9.4411,
3293
+ "eval_samples_per_second": 12535.686,
3294
+ "eval_steps_per_second": 6.143,
3295
+ "step": 21900
3296
+ },
3297
+ {
3298
+ "epoch": 94.82758620689656,
3299
+ "grad_norm": 0.12468830496072769,
3300
+ "learning_rate": 1.3464884201409122e-05,
3301
+ "loss": 0.8618,
3302
+ "step": 22000
3303
+ },
3304
+ {
3305
+ "epoch": 94.82758620689656,
3306
+ "eval_loss": 5.7382001876831055,
3307
+ "eval_runtime": 9.2391,
3308
+ "eval_samples_per_second": 12809.79,
3309
+ "eval_steps_per_second": 6.278,
3310
+ "step": 22000
3311
+ },
3312
+ {
3313
+ "epoch": 95.25862068965517,
3314
+ "grad_norm": 0.12150618433952332,
3315
+ "learning_rate": 1.1331916607768001e-05,
3316
+ "loss": 0.8616,
3317
+ "step": 22100
3318
+ },
3319
+ {
3320
+ "epoch": 95.25862068965517,
3321
+ "eval_loss": 5.741961479187012,
3322
+ "eval_runtime": 9.2711,
3323
+ "eval_samples_per_second": 12765.649,
3324
+ "eval_steps_per_second": 6.256,
3325
+ "step": 22100
3326
+ },
3327
+ {
3328
+ "epoch": 95.6896551724138,
3329
+ "grad_norm": 0.11873551458120346,
3330
+ "learning_rate": 9.381809389101825e-06,
3331
+ "loss": 0.8605,
3332
+ "step": 22200
3333
+ },
3334
+ {
3335
+ "epoch": 95.6896551724138,
3336
+ "eval_loss": 5.7379045486450195,
3337
+ "eval_runtime": 9.4373,
3338
+ "eval_samples_per_second": 12540.807,
3339
+ "eval_steps_per_second": 6.146,
3340
+ "step": 22200
3341
+ },
3342
+ {
3343
+ "epoch": 96.12068965517241,
3344
+ "grad_norm": 0.11970551311969757,
3345
+ "learning_rate": 7.614923229995796e-06,
3346
+ "loss": 0.8613,
3347
+ "step": 22300
3348
+ },
3349
+ {
3350
+ "epoch": 96.12068965517241,
3351
+ "eval_loss": 5.7379302978515625,
3352
+ "eval_runtime": 9.2566,
3353
+ "eval_samples_per_second": 12785.598,
3354
+ "eval_steps_per_second": 6.266,
3355
+ "step": 22300
3356
+ },
3357
+ {
3358
+ "epoch": 96.55172413793103,
3359
+ "grad_norm": 0.12629768252372742,
3360
+ "learning_rate": 6.03158492714806e-06,
3361
+ "loss": 0.8606,
3362
+ "step": 22400
3363
+ },
3364
+ {
3365
+ "epoch": 96.55172413793103,
3366
+ "eval_loss": 5.744095802307129,
3367
+ "eval_runtime": 9.4095,
3368
+ "eval_samples_per_second": 12577.802,
3369
+ "eval_steps_per_second": 6.164,
3370
+ "step": 22400
3371
+ },
3372
+ {
3373
+ "epoch": 96.98275862068965,
3374
+ "grad_norm": 0.12744753062725067,
3375
+ "learning_rate": 4.632087328927947e-06,
3376
+ "loss": 0.8613,
3377
+ "step": 22500
3378
+ },
3379
+ {
3380
+ "epoch": 96.98275862068965,
3381
+ "eval_loss": 5.739793300628662,
3382
+ "eval_runtime": 9.2487,
3383
+ "eval_samples_per_second": 12796.503,
3384
+ "eval_steps_per_second": 6.271,
3385
+ "step": 22500
3386
+ },
3387
+ {
3388
+ "epoch": 97.41379310344827,
3389
+ "grad_norm": 0.1232086643576622,
3390
+ "learning_rate": 3.4166892812107496e-06,
3391
+ "loss": 0.8599,
3392
+ "step": 22600
3393
+ },
3394
+ {
3395
+ "epoch": 97.41379310344827,
3396
+ "eval_loss": 5.744312763214111,
3397
+ "eval_runtime": 9.2146,
3398
+ "eval_samples_per_second": 12843.856,
3399
+ "eval_steps_per_second": 6.294,
3400
+ "step": 22600
3401
+ },
3402
+ {
3403
+ "epoch": 97.84482758620689,
3404
+ "grad_norm": 0.12403382360935211,
3405
+ "learning_rate": 2.3856155795032307e-06,
3406
+ "loss": 0.8606,
3407
+ "step": 22700
3408
+ },
3409
+ {
3410
+ "epoch": 97.84482758620689,
3411
+ "eval_loss": 5.743608474731445,
3412
+ "eval_runtime": 9.3812,
3413
+ "eval_samples_per_second": 12615.751,
3414
+ "eval_steps_per_second": 6.183,
3415
+ "step": 22700
3416
+ },
3417
+ {
3418
+ "epoch": 98.27586206896552,
3419
+ "grad_norm": 0.1235072985291481,
3420
+ "learning_rate": 1.5390569273660005e-06,
3421
+ "loss": 0.8595,
3422
+ "step": 22800
3423
+ },
3424
+ {
3425
+ "epoch": 98.27586206896552,
3426
+ "eval_loss": 5.745834827423096,
3427
+ "eval_runtime": 9.2722,
3428
+ "eval_samples_per_second": 12764.058,
3429
+ "eval_steps_per_second": 6.255,
3430
+ "step": 22800
3431
+ },
3432
+ {
3433
+ "epoch": 98.70689655172414,
3434
+ "grad_norm": 0.12580013275146484,
3435
+ "learning_rate": 8.771699011416167e-07,
3436
+ "loss": 0.8598,
3437
+ "step": 22900
3438
+ },
3439
+ {
3440
+ "epoch": 98.70689655172414,
3441
+ "eval_loss": 5.746177673339844,
3442
+ "eval_runtime": 9.3655,
3443
+ "eval_samples_per_second": 12636.937,
3444
+ "eval_steps_per_second": 6.193,
3445
+ "step": 22900
3446
+ },
3447
+ {
3448
+ "epoch": 99.13793103448276,
3449
+ "grad_norm": 0.12041032314300537,
3450
+ "learning_rate": 4.0007692099508495e-07,
3451
+ "loss": 0.8599,
3452
+ "step": 23000
3453
+ },
3454
+ {
3455
+ "epoch": 99.13793103448276,
3456
+ "eval_loss": 5.745883941650391,
3457
+ "eval_runtime": 9.3798,
3458
+ "eval_samples_per_second": 12617.71,
3459
+ "eval_steps_per_second": 6.184,
3460
+ "step": 23000
3461
+ },
3462
+ {
3463
+ "epoch": 99.56896551724138,
3464
+ "grad_norm": 0.11941543966531754,
3465
+ "learning_rate": 1.0786622827108072e-07,
3466
+ "loss": 0.8597,
3467
+ "step": 23100
3468
+ },
3469
+ {
3470
+ "epoch": 99.56896551724138,
3471
+ "eval_loss": 5.745979309082031,
3472
+ "eval_runtime": 9.2683,
3473
+ "eval_samples_per_second": 12769.475,
3474
+ "eval_steps_per_second": 6.258,
3475
+ "step": 23100
3476
+ }
3477
+ ],
3478
+ "logging_steps": 100,
3479
+ "max_steps": 23200,
3480
+ "num_input_tokens_seen": 0,
3481
+ "num_train_epochs": 100,
3482
+ "save_steps": 100,
3483
+ "stateful_callbacks": {
3484
+ "TrainerControl": {
3485
+ "args": {
3486
+ "should_epoch_stop": false,
3487
+ "should_evaluate": false,
3488
+ "should_log": false,
3489
+ "should_save": true,
3490
+ "should_training_stop": false
3491
+ },
3492
+ "attributes": {}
3493
+ }
3494
+ },
3495
+ "total_flos": 1.5693199771340636e+18,
3496
+ "train_batch_size": 256,
3497
+ "trial_name": null,
3498
+ "trial_params": null
3499
+ }
outputs/gpt2cocoADS/model_ADS/checkpoint/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aff2abf122a8a2d67f687e591850a09ea423874c3fa242a497ea218ec3e3210b
3
+ size 5432
outputs/gpt2cocoADS/model_ADS/checkpoint/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
outputs/gpt2cocoADS/tokenized_data_ADS/dataset_dict.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"splits": ["train", "test"]}
outputs/gpt2cocoADS/tokenized_data_ADS/test/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:57ce35f69caedf9dbd49cfd8f4bedcc3dd291ce324f6e50c5ee186b0725b90c7
3
+ size 120516976
outputs/gpt2cocoADS/tokenized_data_ADS/test/dataset_info.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "similarity": {
6
+ "dtype": "float32",
7
+ "_type": "Value"
8
+ },
9
+ "aesthetics_score": {
10
+ "dtype": "float32",
11
+ "_type": "Value"
12
+ },
13
+ "IQAs": {
14
+ "dtype": "float32",
15
+ "_type": "Value"
16
+ },
17
+ "prompt": {
18
+ "dtype": "string",
19
+ "_type": "Value"
20
+ },
21
+ "query": {
22
+ "dtype": "string",
23
+ "_type": "Value"
24
+ },
25
+ "input_ids": {
26
+ "feature": {
27
+ "dtype": "int32",
28
+ "_type": "Value"
29
+ },
30
+ "_type": "List"
31
+ },
32
+ "attention_mask": {
33
+ "feature": {
34
+ "dtype": "int8",
35
+ "_type": "Value"
36
+ },
37
+ "_type": "List"
38
+ },
39
+ "labels": {
40
+ "feature": {
41
+ "dtype": "int64",
42
+ "_type": "Value"
43
+ },
44
+ "_type": "List"
45
+ }
46
+ },
47
+ "homepage": "",
48
+ "license": ""
49
+ }
outputs/gpt2cocoADS/tokenized_data_ADS/test/state.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "e63fb29121138be5",
8
+ "_format_columns": [
9
+ "IQAs",
10
+ "aesthetics_score",
11
+ "attention_mask",
12
+ "input_ids",
13
+ "labels",
14
+ "prompt",
15
+ "query",
16
+ "similarity"
17
+ ],
18
+ "_format_kwargs": {},
19
+ "_format_type": "torch",
20
+ "_output_all_columns": false,
21
+ "_split": null
22
+ }
outputs/gpt2cocoADS/tokenized_data_ADS/train/data-00000-of-00001.arrow ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b3e296159771485b96b750cd4816169d716b49e9f96db0fd071b71a5670aae0
3
+ size 482053608
outputs/gpt2cocoADS/tokenized_data_ADS/train/dataset_info.json ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "citation": "",
3
+ "description": "",
4
+ "features": {
5
+ "similarity": {
6
+ "dtype": "float32",
7
+ "_type": "Value"
8
+ },
9
+ "aesthetics_score": {
10
+ "dtype": "float32",
11
+ "_type": "Value"
12
+ },
13
+ "IQAs": {
14
+ "dtype": "float32",
15
+ "_type": "Value"
16
+ },
17
+ "prompt": {
18
+ "dtype": "string",
19
+ "_type": "Value"
20
+ },
21
+ "query": {
22
+ "dtype": "string",
23
+ "_type": "Value"
24
+ },
25
+ "input_ids": {
26
+ "feature": {
27
+ "dtype": "int32",
28
+ "_type": "Value"
29
+ },
30
+ "_type": "List"
31
+ },
32
+ "attention_mask": {
33
+ "feature": {
34
+ "dtype": "int8",
35
+ "_type": "Value"
36
+ },
37
+ "_type": "List"
38
+ },
39
+ "labels": {
40
+ "feature": {
41
+ "dtype": "int64",
42
+ "_type": "Value"
43
+ },
44
+ "_type": "List"
45
+ }
46
+ },
47
+ "homepage": "",
48
+ "license": ""
49
+ }
outputs/gpt2cocoADS/tokenized_data_ADS/train/state.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_data_files": [
3
+ {
4
+ "filename": "data-00000-of-00001.arrow"
5
+ }
6
+ ],
7
+ "_fingerprint": "b20a08d5b42f0265",
8
+ "_format_columns": [
9
+ "IQAs",
10
+ "aesthetics_score",
11
+ "attention_mask",
12
+ "input_ids",
13
+ "labels",
14
+ "prompt",
15
+ "query",
16
+ "similarity"
17
+ ],
18
+ "_format_kwargs": {},
19
+ "_format_type": "torch",
20
+ "_output_all_columns": false,
21
+ "_split": null
22
+ }
outputs/gpt2cocoDSA/model_DSA/checkpoint/added_tokens.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "<pad>": 50258,
3
+ "<|startoftext|>": 50257
4
+ }
outputs/gpt2cocoDSA/model_DSA/checkpoint/config.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "cls_token_id": 50257,
9
+ "dtype": "float32",
10
+ "embd_pdrop": 0.1,
11
+ "eos_token_id": 50256,
12
+ "initializer_range": 0.02,
13
+ "layer_norm_epsilon": 1e-05,
14
+ "model_type": "gpt2",
15
+ "n_ctx": 1024,
16
+ "n_embd": 768,
17
+ "n_head": 12,
18
+ "n_inner": null,
19
+ "n_layer": 12,
20
+ "n_positions": 1024,
21
+ "pad_token_id": 50258,
22
+ "reorder_and_upcast_attn": false,
23
+ "resid_pdrop": 0.1,
24
+ "scale_attn_by_inverse_layer_idx": false,
25
+ "scale_attn_weights": true,
26
+ "summary_activation": null,
27
+ "summary_first_dropout": 0.1,
28
+ "summary_proj_to_labels": true,
29
+ "summary_type": "cls_index",
30
+ "summary_use_proj": true,
31
+ "task_specific_params": {
32
+ "text-generation": {
33
+ "do_sample": true,
34
+ "max_length": 50
35
+ }
36
+ },
37
+ "transformers_version": "4.57.1",
38
+ "use_cache": true,
39
+ "vocab_size": 50259
40
+ }
outputs/gpt2cocoDSA/model_DSA/checkpoint/generation_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": [
5
+ 50256
6
+ ],
7
+ "pad_token_id": 50258,
8
+ "transformers_version": "4.57.1"
9
+ }
outputs/gpt2cocoDSA/model_DSA/checkpoint/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
outputs/gpt2cocoDSA/model_DSA/checkpoint/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f80025898ed2d75495a9306b3e7bcbb4d2378202db5bfae250f436158791c4f5
3
+ size 497780352
outputs/gpt2cocoDSA/model_DSA/checkpoint/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53c4a9fff4a4b72fc17394e09017bb6955efe829cf7894648b45d34203fd8c6c
3
+ size 995654586
outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aebbbdd9f6b6458006140b9399df973322b39193c3491688f309b28eda7f6516
3
+ size 15984
outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02092b11072872275f41067881e0ff9c3ebb4f71cf3347a14575ee0c3bb4dc37
3
+ size 15984
outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:834873e0297393e1fa57f1b955d4a9198d35ae71d5b1b5b0661327390e456a6a
3
+ size 15984
outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6756722a563897612ae24bdf113ccc0f63bb646ff00b0de94943e7815ca03cf9
3
+ size 15984
outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_4.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:811a877d2d10d1034d381ef47c53f9a523828b58a68b612cd1b0858ee81e6c1e
3
+ size 15984
outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_5.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65538c95dd42ed5ff89d120cad03ed208d8388db3902f5332493f3f25f8ac51a
3
+ size 15984
outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_6.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44952594d0b6ac119c313df4062415014d62be491e303928f3548bb0f2c5c55f
3
+ size 15984
outputs/gpt2cocoDSA/model_DSA/checkpoint/rng_state_7.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2081cb2f3fdd83c5b77b31a5b51f1f3492a4f66d21e8c736d093ea56b1461995
3
+ size 15984
outputs/gpt2cocoDSA/model_DSA/checkpoint/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b78e1481fc501f93036b39f92a529b1905f6709db01ab54ddda1515ca83b584
3
+ size 988
outputs/gpt2cocoDSA/model_DSA/checkpoint/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f86d4debfeccbc9cc7f34fa8036cc4f47e09cf0625bf34c4a4cb3f392bd25f62
3
+ size 1064