YWZBrandon commited on
Commit
3102e25
·
verified ·
1 Parent(s): 6f521fd

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. added_tokens.json +3 -0
  2. checkpoint-1804/added_tokens.json +3 -0
  3. checkpoint-1804/config.json +46 -0
  4. checkpoint-1804/model.safetensors +3 -0
  5. checkpoint-1804/optimizer.pt +3 -0
  6. checkpoint-1804/rng_state.pth +3 -0
  7. checkpoint-1804/scheduler.pt +3 -0
  8. checkpoint-1804/special_tokens_map.json +15 -0
  9. checkpoint-1804/spm.model +3 -0
  10. checkpoint-1804/tokenizer.json +0 -0
  11. checkpoint-1804/tokenizer_config.json +59 -0
  12. checkpoint-1804/trainer_state.json +1318 -0
  13. checkpoint-1804/training_args.bin +3 -0
  14. checkpoint-2706/added_tokens.json +3 -0
  15. checkpoint-2706/config.json +46 -0
  16. checkpoint-2706/model.safetensors +3 -0
  17. checkpoint-2706/optimizer.pt +3 -0
  18. checkpoint-2706/rng_state.pth +3 -0
  19. checkpoint-2706/scheduler.pt +3 -0
  20. checkpoint-2706/special_tokens_map.json +15 -0
  21. checkpoint-2706/spm.model +3 -0
  22. checkpoint-2706/tokenizer.json +0 -0
  23. checkpoint-2706/tokenizer_config.json +59 -0
  24. checkpoint-2706/trainer_state.json +1960 -0
  25. checkpoint-2706/training_args.bin +3 -0
  26. checkpoint-3608/added_tokens.json +3 -0
  27. checkpoint-3608/config.json +46 -0
  28. checkpoint-3608/model.safetensors +3 -0
  29. checkpoint-3608/optimizer.pt +3 -0
  30. checkpoint-3608/rng_state.pth +3 -0
  31. checkpoint-3608/scheduler.pt +3 -0
  32. checkpoint-3608/special_tokens_map.json +15 -0
  33. checkpoint-3608/spm.model +3 -0
  34. checkpoint-3608/tokenizer.json +0 -0
  35. checkpoint-3608/tokenizer_config.json +59 -0
  36. checkpoint-3608/trainer_state.json +2602 -0
  37. checkpoint-3608/training_args.bin +3 -0
  38. checkpoint-4510/added_tokens.json +3 -0
  39. checkpoint-4510/config.json +46 -0
  40. checkpoint-4510/model.safetensors +3 -0
  41. checkpoint-4510/optimizer.pt +3 -0
  42. checkpoint-4510/rng_state.pth +3 -0
  43. checkpoint-4510/scheduler.pt +3 -0
  44. checkpoint-4510/special_tokens_map.json +15 -0
  45. checkpoint-4510/spm.model +3 -0
  46. checkpoint-4510/tokenizer.json +0 -0
  47. checkpoint-4510/tokenizer_config.json +59 -0
  48. checkpoint-4510/trainer_state.json +3251 -0
  49. checkpoint-4510/training_args.bin +3 -0
  50. checkpoint-5412/added_tokens.json +3 -0
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
checkpoint-1804/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
checkpoint-1804/config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaV2ForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "hidden_act": "gelu",
7
+ "hidden_dropout_prob": 0.1,
8
+ "hidden_size": 1024,
9
+ "id2label": {
10
+ "0": 0,
11
+ "1": 1,
12
+ "2": 2
13
+ },
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 4096,
16
+ "label2id": {
17
+ "0": 0,
18
+ "1": 1,
19
+ "2": 2
20
+ },
21
+ "layer_norm_eps": 1e-07,
22
+ "legacy": true,
23
+ "max_position_embeddings": 512,
24
+ "max_relative_positions": -1,
25
+ "model_type": "deberta-v2",
26
+ "norm_rel_ebd": "layer_norm",
27
+ "num_attention_heads": 16,
28
+ "num_hidden_layers": 24,
29
+ "pad_token_id": 0,
30
+ "pooler_dropout": 0,
31
+ "pooler_hidden_act": "gelu",
32
+ "pooler_hidden_size": 1024,
33
+ "pos_att_type": [
34
+ "p2c",
35
+ "c2p"
36
+ ],
37
+ "position_biased_input": false,
38
+ "position_buckets": 256,
39
+ "problem_type": "multi_label_classification",
40
+ "relative_attention": true,
41
+ "share_att_key": true,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.51.3",
44
+ "type_vocab_size": 0,
45
+ "vocab_size": 128100
46
+ }
checkpoint-1804/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b7f46e7d3b3645227c56871d5367a24aceef9b52b112c057a0310d97337410c
3
+ size 1740308548
checkpoint-1804/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75387d6e3357c2858fdb7df1c35568941b80d3c3d62d1d87cd57f53ed1d961bd
3
+ size 3480848432
checkpoint-1804/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e7a6177d40c4fa527124f6d28f6869b9b7f403d4d18ca2fa952605a14627b87
3
+ size 14244
checkpoint-1804/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb55cb4c7ef32286047fa8f11c947a115042b842f08e32edf5fc9635534f692d
3
+ size 1064
checkpoint-1804/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[SEP]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": {
9
+ "content": "[UNK]",
10
+ "lstrip": false,
11
+ "normalized": true,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ }
15
+ }
checkpoint-1804/spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
3
+ size 2464616
checkpoint-1804/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1804/tokenizer_config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[CLS]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[SEP]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[UNK]",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "128000": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "[CLS]",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "[CLS]",
47
+ "do_lower_case": false,
48
+ "eos_token": "[SEP]",
49
+ "extra_special_tokens": {},
50
+ "mask_token": "[MASK]",
51
+ "model_max_length": 1000000000000000019884624838656,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "sp_model_kwargs": {},
55
+ "split_by_punct": false,
56
+ "tokenizer_class": "DebertaV2Tokenizer",
57
+ "unk_token": "[UNK]",
58
+ "vocab_type": "spm"
59
+ }
checkpoint-1804/trainer_state.json ADDED
@@ -0,0 +1,1318 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 2.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1804,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.011092623405435386,
14
+ "grad_norm": 2.009918689727783,
15
+ "learning_rate": 1.998002219755827e-05,
16
+ "loss": 0.5022,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.022185246810870772,
21
+ "grad_norm": 0.49915584921836853,
22
+ "learning_rate": 1.995782463928968e-05,
23
+ "loss": 0.1943,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.033277870216306155,
28
+ "grad_norm": 0.47061625123023987,
29
+ "learning_rate": 1.993562708102109e-05,
30
+ "loss": 0.1436,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.044370493621741544,
35
+ "grad_norm": 0.2260124832391739,
36
+ "learning_rate": 1.99134295227525e-05,
37
+ "loss": 0.1106,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.05546311702717693,
42
+ "grad_norm": 0.31766048073768616,
43
+ "learning_rate": 1.989123196448391e-05,
44
+ "loss": 0.0994,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.06655574043261231,
49
+ "grad_norm": 0.2788391709327698,
50
+ "learning_rate": 1.9869034406215316e-05,
51
+ "loss": 0.111,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0776483638380477,
56
+ "grad_norm": 0.24376386404037476,
57
+ "learning_rate": 1.9846836847946726e-05,
58
+ "loss": 0.0849,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.08874098724348309,
63
+ "grad_norm": 0.1789788454771042,
64
+ "learning_rate": 1.982463928967814e-05,
65
+ "loss": 0.0971,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.09983361064891846,
70
+ "grad_norm": 0.29698437452316284,
71
+ "learning_rate": 1.9802441731409546e-05,
72
+ "loss": 0.0856,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.11092623405435385,
77
+ "grad_norm": 0.2553412616252899,
78
+ "learning_rate": 1.9780244173140956e-05,
79
+ "loss": 0.0976,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.12201885745978924,
84
+ "grad_norm": 0.20608791708946228,
85
+ "learning_rate": 1.9758046614872365e-05,
86
+ "loss": 0.0947,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.13311148086522462,
91
+ "grad_norm": 0.23551669716835022,
92
+ "learning_rate": 1.9735849056603775e-05,
93
+ "loss": 0.0965,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.14420410427066002,
98
+ "grad_norm": 0.3207148611545563,
99
+ "learning_rate": 1.9713651498335185e-05,
100
+ "loss": 0.0931,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.1552967276760954,
105
+ "grad_norm": 0.34813177585601807,
106
+ "learning_rate": 1.9691453940066595e-05,
107
+ "loss": 0.1,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.16638935108153077,
112
+ "grad_norm": 0.41501104831695557,
113
+ "learning_rate": 1.9669256381798e-05,
114
+ "loss": 0.0922,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.17748197448696618,
119
+ "grad_norm": 0.24550506472587585,
120
+ "learning_rate": 1.964705882352941e-05,
121
+ "loss": 0.0857,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.18857459789240155,
126
+ "grad_norm": 0.4205056130886078,
127
+ "learning_rate": 1.9624861265260825e-05,
128
+ "loss": 0.0831,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.19966722129783693,
133
+ "grad_norm": 0.6002993583679199,
134
+ "learning_rate": 1.9602663706992235e-05,
135
+ "loss": 0.0792,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.21075984470327233,
140
+ "grad_norm": 0.27535638213157654,
141
+ "learning_rate": 1.958046614872364e-05,
142
+ "loss": 0.0786,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.2218524681087077,
147
+ "grad_norm": 0.41602805256843567,
148
+ "learning_rate": 1.955826859045505e-05,
149
+ "loss": 0.0789,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.23294509151414308,
154
+ "grad_norm": 0.41827332973480225,
155
+ "learning_rate": 1.953607103218646e-05,
156
+ "loss": 0.0663,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.24403771491957849,
161
+ "grad_norm": 0.47976112365722656,
162
+ "learning_rate": 1.951387347391787e-05,
163
+ "loss": 0.0727,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.25513033832501386,
168
+ "grad_norm": 0.7771281599998474,
169
+ "learning_rate": 1.949167591564928e-05,
170
+ "loss": 0.0776,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.26622296173044924,
175
+ "grad_norm": 1.0573914051055908,
176
+ "learning_rate": 1.946947835738069e-05,
177
+ "loss": 0.0828,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.2773155851358846,
182
+ "grad_norm": 0.5186040997505188,
183
+ "learning_rate": 1.9447280799112097e-05,
184
+ "loss": 0.0767,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.28840820854132004,
189
+ "grad_norm": 0.5594862699508667,
190
+ "learning_rate": 1.942508324084351e-05,
191
+ "loss": 0.0699,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.2995008319467554,
196
+ "grad_norm": 0.5147818922996521,
197
+ "learning_rate": 1.940288568257492e-05,
198
+ "loss": 0.0591,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.3105934553521908,
203
+ "grad_norm": 0.5510437488555908,
204
+ "learning_rate": 1.938068812430633e-05,
205
+ "loss": 0.0761,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.32168607875762617,
210
+ "grad_norm": 0.4817625880241394,
211
+ "learning_rate": 1.9358490566037736e-05,
212
+ "loss": 0.0688,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.33277870216306155,
217
+ "grad_norm": 0.504362940788269,
218
+ "learning_rate": 1.9336293007769146e-05,
219
+ "loss": 0.0691,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.343871325568497,
224
+ "grad_norm": 0.5732384920120239,
225
+ "learning_rate": 1.9314095449500556e-05,
226
+ "loss": 0.057,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.35496394897393235,
231
+ "grad_norm": 0.5892298221588135,
232
+ "learning_rate": 1.9291897891231966e-05,
233
+ "loss": 0.059,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.36605657237936773,
238
+ "grad_norm": 0.3680512309074402,
239
+ "learning_rate": 1.9269700332963376e-05,
240
+ "loss": 0.0706,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.3771491957848031,
245
+ "grad_norm": 0.41971662640571594,
246
+ "learning_rate": 1.9247502774694786e-05,
247
+ "loss": 0.053,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.3882418191902385,
252
+ "grad_norm": 0.37261390686035156,
253
+ "learning_rate": 1.9225305216426195e-05,
254
+ "loss": 0.0632,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.39933444259567386,
259
+ "grad_norm": 0.48256734013557434,
260
+ "learning_rate": 1.9203107658157605e-05,
261
+ "loss": 0.0651,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.4104270660011093,
266
+ "grad_norm": 0.7914339303970337,
267
+ "learning_rate": 1.9180910099889015e-05,
268
+ "loss": 0.0659,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.42151968940654466,
273
+ "grad_norm": 0.6772429347038269,
274
+ "learning_rate": 1.915871254162042e-05,
275
+ "loss": 0.0655,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.43261231281198004,
280
+ "grad_norm": 0.5621687173843384,
281
+ "learning_rate": 1.913651498335183e-05,
282
+ "loss": 0.0707,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.4437049362174154,
287
+ "grad_norm": 0.30515748262405396,
288
+ "learning_rate": 1.911431742508324e-05,
289
+ "loss": 0.0548,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.4547975596228508,
294
+ "grad_norm": 0.5506859421730042,
295
+ "learning_rate": 1.909211986681465e-05,
296
+ "loss": 0.0554,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.46589018302828616,
301
+ "grad_norm": 0.6387749314308167,
302
+ "learning_rate": 1.906992230854606e-05,
303
+ "loss": 0.0554,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.4769828064337216,
308
+ "grad_norm": 0.18500734865665436,
309
+ "learning_rate": 1.904772475027747e-05,
310
+ "loss": 0.0558,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.48807542983915697,
315
+ "grad_norm": 0.6815407276153564,
316
+ "learning_rate": 1.902552719200888e-05,
317
+ "loss": 0.0537,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.49916805324459235,
322
+ "grad_norm": 0.4826994836330414,
323
+ "learning_rate": 1.900332963374029e-05,
324
+ "loss": 0.0567,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.5102606766500277,
329
+ "grad_norm": 0.24618124961853027,
330
+ "learning_rate": 1.89811320754717e-05,
331
+ "loss": 0.0614,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.5213533000554631,
336
+ "grad_norm": 1.0037415027618408,
337
+ "learning_rate": 1.895893451720311e-05,
338
+ "loss": 0.0604,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.5324459234608985,
343
+ "grad_norm": 0.6000948548316956,
344
+ "learning_rate": 1.8936736958934517e-05,
345
+ "loss": 0.0683,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.5435385468663338,
350
+ "grad_norm": 0.3302474319934845,
351
+ "learning_rate": 1.8914539400665927e-05,
352
+ "loss": 0.0543,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.5546311702717692,
357
+ "grad_norm": 0.5560783743858337,
358
+ "learning_rate": 1.8892341842397337e-05,
359
+ "loss": 0.0569,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.5657237936772047,
364
+ "grad_norm": 0.8041097521781921,
365
+ "learning_rate": 1.8870144284128747e-05,
366
+ "loss": 0.0544,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.5768164170826401,
371
+ "grad_norm": 0.6846103668212891,
372
+ "learning_rate": 1.8847946725860156e-05,
373
+ "loss": 0.058,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.5879090404880755,
378
+ "grad_norm": 0.50434809923172,
379
+ "learning_rate": 1.8825749167591566e-05,
380
+ "loss": 0.0654,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.5990016638935108,
385
+ "grad_norm": 0.54362553358078,
386
+ "learning_rate": 1.8803551609322976e-05,
387
+ "loss": 0.0582,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.6100942872989462,
392
+ "grad_norm": 0.6166839599609375,
393
+ "learning_rate": 1.8781354051054386e-05,
394
+ "loss": 0.0672,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.6211869107043816,
399
+ "grad_norm": 0.4353054165840149,
400
+ "learning_rate": 1.8759156492785796e-05,
401
+ "loss": 0.0532,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.632279534109817,
406
+ "grad_norm": 0.6025580167770386,
407
+ "learning_rate": 1.8736958934517206e-05,
408
+ "loss": 0.0746,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.6433721575152523,
413
+ "grad_norm": 0.7259892225265503,
414
+ "learning_rate": 1.8714761376248612e-05,
415
+ "loss": 0.0642,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.6544647809206877,
420
+ "grad_norm": 0.4940318465232849,
421
+ "learning_rate": 1.8692563817980022e-05,
422
+ "loss": 0.0547,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.6655574043261231,
427
+ "grad_norm": 0.7005699872970581,
428
+ "learning_rate": 1.8670366259711435e-05,
429
+ "loss": 0.0522,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.6766500277315585,
434
+ "grad_norm": 0.4530707895755768,
435
+ "learning_rate": 1.8648168701442845e-05,
436
+ "loss": 0.054,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.687742651136994,
441
+ "grad_norm": 0.9097110629081726,
442
+ "learning_rate": 1.8625971143174252e-05,
443
+ "loss": 0.0622,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.6988352745424293,
448
+ "grad_norm": 0.5374599695205688,
449
+ "learning_rate": 1.860377358490566e-05,
450
+ "loss": 0.0522,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.7099278979478647,
455
+ "grad_norm": 0.3849945664405823,
456
+ "learning_rate": 1.858157602663707e-05,
457
+ "loss": 0.0571,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.7210205213533001,
462
+ "grad_norm": 0.5918008685112,
463
+ "learning_rate": 1.855937846836848e-05,
464
+ "loss": 0.0576,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.7321131447587355,
469
+ "grad_norm": 0.3229956030845642,
470
+ "learning_rate": 1.853718091009989e-05,
471
+ "loss": 0.0537,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.7432057681641708,
476
+ "grad_norm": 0.5264039039611816,
477
+ "learning_rate": 1.85149833518313e-05,
478
+ "loss": 0.06,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.7542983915696062,
483
+ "grad_norm": 0.36795660853385925,
484
+ "learning_rate": 1.8492785793562708e-05,
485
+ "loss": 0.0511,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.7653910149750416,
490
+ "grad_norm": 0.5905130505561829,
491
+ "learning_rate": 1.847058823529412e-05,
492
+ "loss": 0.0636,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.776483638380477,
497
+ "grad_norm": 0.36266571283340454,
498
+ "learning_rate": 1.844839067702553e-05,
499
+ "loss": 0.0598,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.7875762617859123,
504
+ "grad_norm": 0.4978592097759247,
505
+ "learning_rate": 1.8426193118756937e-05,
506
+ "loss": 0.0607,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.7986688851913477,
511
+ "grad_norm": 0.4635021686553955,
512
+ "learning_rate": 1.8403995560488347e-05,
513
+ "loss": 0.0543,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.8097615085967831,
518
+ "grad_norm": 0.44571858644485474,
519
+ "learning_rate": 1.8381798002219757e-05,
520
+ "loss": 0.0678,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.8208541320022186,
525
+ "grad_norm": 0.8265877366065979,
526
+ "learning_rate": 1.8359600443951167e-05,
527
+ "loss": 0.0552,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.831946755407654,
532
+ "grad_norm": 0.5776472091674805,
533
+ "learning_rate": 1.8337402885682577e-05,
534
+ "loss": 0.0505,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.8430393788130893,
539
+ "grad_norm": 0.299274742603302,
540
+ "learning_rate": 1.8315205327413986e-05,
541
+ "loss": 0.0619,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.8541320022185247,
546
+ "grad_norm": 0.587645947933197,
547
+ "learning_rate": 1.8293007769145393e-05,
548
+ "loss": 0.0575,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.8652246256239601,
553
+ "grad_norm": 0.39164137840270996,
554
+ "learning_rate": 1.8270810210876806e-05,
555
+ "loss": 0.0458,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.8763172490293955,
560
+ "grad_norm": 0.4663292169570923,
561
+ "learning_rate": 1.8248612652608216e-05,
562
+ "loss": 0.048,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.8874098724348308,
567
+ "grad_norm": 0.5804581642150879,
568
+ "learning_rate": 1.8226415094339626e-05,
569
+ "loss": 0.0557,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.8985024958402662,
574
+ "grad_norm": 0.4279440939426422,
575
+ "learning_rate": 1.8204217536071032e-05,
576
+ "loss": 0.051,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.9095951192457016,
581
+ "grad_norm": 0.5384302735328674,
582
+ "learning_rate": 1.8182019977802442e-05,
583
+ "loss": 0.0558,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.920687742651137,
588
+ "grad_norm": 0.5049973726272583,
589
+ "learning_rate": 1.8159822419533852e-05,
590
+ "loss": 0.047,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.9317803660565723,
595
+ "grad_norm": 0.8061177134513855,
596
+ "learning_rate": 1.8137624861265262e-05,
597
+ "loss": 0.0572,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.9428729894620078,
602
+ "grad_norm": 0.6056540012359619,
603
+ "learning_rate": 1.8115427302996672e-05,
604
+ "loss": 0.0539,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.9539656128674432,
609
+ "grad_norm": 0.3001384139060974,
610
+ "learning_rate": 1.8093229744728082e-05,
611
+ "loss": 0.053,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.9650582362728786,
616
+ "grad_norm": 0.6709749102592468,
617
+ "learning_rate": 1.807103218645949e-05,
618
+ "loss": 0.0526,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.9761508596783139,
623
+ "grad_norm": 0.8233507871627808,
624
+ "learning_rate": 1.80488346281909e-05,
625
+ "loss": 0.0621,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.9872434830837493,
630
+ "grad_norm": 0.5757150650024414,
631
+ "learning_rate": 1.802663706992231e-05,
632
+ "loss": 0.0546,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.9983361064891847,
637
+ "grad_norm": 0.6834889054298401,
638
+ "learning_rate": 1.800443951165372e-05,
639
+ "loss": 0.0462,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 1.0,
644
+ "eval_accuracy": 0.8720027017899359,
645
+ "eval_f1": 0.5678449258836944,
646
+ "eval_loss": 0.34414270520210266,
647
+ "eval_precision": 0.7929936305732485,
648
+ "eval_recall": 0.4422735346358792,
649
+ "eval_runtime": 2.727,
650
+ "eval_samples_per_second": 361.937,
651
+ "eval_steps_per_second": 11.368,
652
+ "step": 902
653
+ },
654
+ {
655
+ "epoch": 1.0088740987243483,
656
+ "grad_norm": 0.4411003887653351,
657
+ "learning_rate": 1.7982241953385128e-05,
658
+ "loss": 0.0312,
659
+ "step": 910
660
+ },
661
+ {
662
+ "epoch": 1.0199667221297837,
663
+ "grad_norm": 0.40500083565711975,
664
+ "learning_rate": 1.7960044395116538e-05,
665
+ "loss": 0.0345,
666
+ "step": 920
667
+ },
668
+ {
669
+ "epoch": 1.031059345535219,
670
+ "grad_norm": 0.9287449717521667,
671
+ "learning_rate": 1.7937846836847947e-05,
672
+ "loss": 0.0419,
673
+ "step": 930
674
+ },
675
+ {
676
+ "epoch": 1.0421519689406544,
677
+ "grad_norm": 0.7969145178794861,
678
+ "learning_rate": 1.7915649278579357e-05,
679
+ "loss": 0.042,
680
+ "step": 940
681
+ },
682
+ {
683
+ "epoch": 1.0532445923460898,
684
+ "grad_norm": 0.49280259013175964,
685
+ "learning_rate": 1.7893451720310767e-05,
686
+ "loss": 0.0427,
687
+ "step": 950
688
+ },
689
+ {
690
+ "epoch": 1.0643372157515252,
691
+ "grad_norm": 1.1647624969482422,
692
+ "learning_rate": 1.7871254162042177e-05,
693
+ "loss": 0.0341,
694
+ "step": 960
695
+ },
696
+ {
697
+ "epoch": 1.0754298391569606,
698
+ "grad_norm": 0.2908968925476074,
699
+ "learning_rate": 1.7849056603773587e-05,
700
+ "loss": 0.0389,
701
+ "step": 970
702
+ },
703
+ {
704
+ "epoch": 1.086522462562396,
705
+ "grad_norm": 0.521218478679657,
706
+ "learning_rate": 1.7826859045504997e-05,
707
+ "loss": 0.0396,
708
+ "step": 980
709
+ },
710
+ {
711
+ "epoch": 1.0976150859678313,
712
+ "grad_norm": 0.4156598150730133,
713
+ "learning_rate": 1.7804661487236407e-05,
714
+ "loss": 0.0359,
715
+ "step": 990
716
+ },
717
+ {
718
+ "epoch": 1.1087077093732667,
719
+ "grad_norm": 0.7260242104530334,
720
+ "learning_rate": 1.7782463928967813e-05,
721
+ "loss": 0.0384,
722
+ "step": 1000
723
+ },
724
+ {
725
+ "epoch": 1.119800332778702,
726
+ "grad_norm": 0.43658459186553955,
727
+ "learning_rate": 1.7760266370699223e-05,
728
+ "loss": 0.0383,
729
+ "step": 1010
730
+ },
731
+ {
732
+ "epoch": 1.1308929561841374,
733
+ "grad_norm": 0.3585101068019867,
734
+ "learning_rate": 1.7738068812430633e-05,
735
+ "loss": 0.0401,
736
+ "step": 1020
737
+ },
738
+ {
739
+ "epoch": 1.141985579589573,
740
+ "grad_norm": 0.5010389089584351,
741
+ "learning_rate": 1.7715871254162043e-05,
742
+ "loss": 0.0307,
743
+ "step": 1030
744
+ },
745
+ {
746
+ "epoch": 1.1530782029950084,
747
+ "grad_norm": 0.6415812969207764,
748
+ "learning_rate": 1.7693673695893453e-05,
749
+ "loss": 0.0423,
750
+ "step": 1040
751
+ },
752
+ {
753
+ "epoch": 1.1641708264004438,
754
+ "grad_norm": 0.6393259763717651,
755
+ "learning_rate": 1.7671476137624862e-05,
756
+ "loss": 0.0392,
757
+ "step": 1050
758
+ },
759
+ {
760
+ "epoch": 1.1752634498058792,
761
+ "grad_norm": 0.6339041590690613,
762
+ "learning_rate": 1.7649278579356272e-05,
763
+ "loss": 0.0506,
764
+ "step": 1060
765
+ },
766
+ {
767
+ "epoch": 1.1863560732113145,
768
+ "grad_norm": 0.6086763143539429,
769
+ "learning_rate": 1.7627081021087682e-05,
770
+ "loss": 0.0416,
771
+ "step": 1070
772
+ },
773
+ {
774
+ "epoch": 1.19744869661675,
775
+ "grad_norm": 0.6648682355880737,
776
+ "learning_rate": 1.7604883462819092e-05,
777
+ "loss": 0.0379,
778
+ "step": 1080
779
+ },
780
+ {
781
+ "epoch": 1.2085413200221853,
782
+ "grad_norm": 0.6018221378326416,
783
+ "learning_rate": 1.7582685904550502e-05,
784
+ "loss": 0.0354,
785
+ "step": 1090
786
+ },
787
+ {
788
+ "epoch": 1.2196339434276207,
789
+ "grad_norm": 1.2366654872894287,
790
+ "learning_rate": 1.756048834628191e-05,
791
+ "loss": 0.037,
792
+ "step": 1100
793
+ },
794
+ {
795
+ "epoch": 1.230726566833056,
796
+ "grad_norm": 0.5976310968399048,
797
+ "learning_rate": 1.7538290788013318e-05,
798
+ "loss": 0.0361,
799
+ "step": 1110
800
+ },
801
+ {
802
+ "epoch": 1.2418191902384914,
803
+ "grad_norm": 0.8290308117866516,
804
+ "learning_rate": 1.751609322974473e-05,
805
+ "loss": 0.0383,
806
+ "step": 1120
807
+ },
808
+ {
809
+ "epoch": 1.2529118136439268,
810
+ "grad_norm": 0.382548987865448,
811
+ "learning_rate": 1.749389567147614e-05,
812
+ "loss": 0.0449,
813
+ "step": 1130
814
+ },
815
+ {
816
+ "epoch": 1.2640044370493622,
817
+ "grad_norm": 0.5307976007461548,
818
+ "learning_rate": 1.7471698113207548e-05,
819
+ "loss": 0.0364,
820
+ "step": 1140
821
+ },
822
+ {
823
+ "epoch": 1.2750970604547975,
824
+ "grad_norm": 0.5508521795272827,
825
+ "learning_rate": 1.7449500554938958e-05,
826
+ "loss": 0.0312,
827
+ "step": 1150
828
+ },
829
+ {
830
+ "epoch": 1.286189683860233,
831
+ "grad_norm": 0.37057268619537354,
832
+ "learning_rate": 1.7427302996670368e-05,
833
+ "loss": 0.0397,
834
+ "step": 1160
835
+ },
836
+ {
837
+ "epoch": 1.2972823072656683,
838
+ "grad_norm": 0.4806898236274719,
839
+ "learning_rate": 1.7405105438401777e-05,
840
+ "loss": 0.0391,
841
+ "step": 1170
842
+ },
843
+ {
844
+ "epoch": 1.3083749306711037,
845
+ "grad_norm": 0.759772002696991,
846
+ "learning_rate": 1.7382907880133187e-05,
847
+ "loss": 0.0322,
848
+ "step": 1180
849
+ },
850
+ {
851
+ "epoch": 1.319467554076539,
852
+ "grad_norm": 0.39819085597991943,
853
+ "learning_rate": 1.7360710321864597e-05,
854
+ "loss": 0.0345,
855
+ "step": 1190
856
+ },
857
+ {
858
+ "epoch": 1.3305601774819744,
859
+ "grad_norm": 0.7733897566795349,
860
+ "learning_rate": 1.7338512763596004e-05,
861
+ "loss": 0.0424,
862
+ "step": 1200
863
+ },
864
+ {
865
+ "epoch": 1.3416528008874098,
866
+ "grad_norm": 0.7979075908660889,
867
+ "learning_rate": 1.7316315205327417e-05,
868
+ "loss": 0.0403,
869
+ "step": 1210
870
+ },
871
+ {
872
+ "epoch": 1.3527454242928454,
873
+ "grad_norm": 0.5658752918243408,
874
+ "learning_rate": 1.7294117647058827e-05,
875
+ "loss": 0.0276,
876
+ "step": 1220
877
+ },
878
+ {
879
+ "epoch": 1.3638380476982808,
880
+ "grad_norm": 0.25207844376564026,
881
+ "learning_rate": 1.7271920088790237e-05,
882
+ "loss": 0.0347,
883
+ "step": 1230
884
+ },
885
+ {
886
+ "epoch": 1.3749306711037161,
887
+ "grad_norm": 0.6338945031166077,
888
+ "learning_rate": 1.7249722530521643e-05,
889
+ "loss": 0.0416,
890
+ "step": 1240
891
+ },
892
+ {
893
+ "epoch": 1.3860232945091515,
894
+ "grad_norm": 0.8125913143157959,
895
+ "learning_rate": 1.7227524972253053e-05,
896
+ "loss": 0.0442,
897
+ "step": 1250
898
+ },
899
+ {
900
+ "epoch": 1.397115917914587,
901
+ "grad_norm": 0.5067325234413147,
902
+ "learning_rate": 1.7205327413984463e-05,
903
+ "loss": 0.0358,
904
+ "step": 1260
905
+ },
906
+ {
907
+ "epoch": 1.4082085413200223,
908
+ "grad_norm": 0.40132638812065125,
909
+ "learning_rate": 1.7183129855715873e-05,
910
+ "loss": 0.0303,
911
+ "step": 1270
912
+ },
913
+ {
914
+ "epoch": 1.4193011647254576,
915
+ "grad_norm": 0.6337258815765381,
916
+ "learning_rate": 1.7160932297447283e-05,
917
+ "loss": 0.0471,
918
+ "step": 1280
919
+ },
920
+ {
921
+ "epoch": 1.430393788130893,
922
+ "grad_norm": 0.43456393480300903,
923
+ "learning_rate": 1.713873473917869e-05,
924
+ "loss": 0.0354,
925
+ "step": 1290
926
+ },
927
+ {
928
+ "epoch": 1.4414864115363284,
929
+ "grad_norm": 0.3291069269180298,
930
+ "learning_rate": 1.7116537180910102e-05,
931
+ "loss": 0.0312,
932
+ "step": 1300
933
+ },
934
+ {
935
+ "epoch": 1.4525790349417638,
936
+ "grad_norm": 0.2960388660430908,
937
+ "learning_rate": 1.7094339622641512e-05,
938
+ "loss": 0.0362,
939
+ "step": 1310
940
+ },
941
+ {
942
+ "epoch": 1.4636716583471991,
943
+ "grad_norm": 0.820751428604126,
944
+ "learning_rate": 1.7072142064372922e-05,
945
+ "loss": 0.0396,
946
+ "step": 1320
947
+ },
948
+ {
949
+ "epoch": 1.4747642817526345,
950
+ "grad_norm": 0.7335907816886902,
951
+ "learning_rate": 1.704994450610433e-05,
952
+ "loss": 0.0386,
953
+ "step": 1330
954
+ },
955
+ {
956
+ "epoch": 1.48585690515807,
957
+ "grad_norm": 0.40592002868652344,
958
+ "learning_rate": 1.702774694783574e-05,
959
+ "loss": 0.0358,
960
+ "step": 1340
961
+ },
962
+ {
963
+ "epoch": 1.4969495285635053,
964
+ "grad_norm": 0.6194770336151123,
965
+ "learning_rate": 1.7005549389567148e-05,
966
+ "loss": 0.0361,
967
+ "step": 1350
968
+ },
969
+ {
970
+ "epoch": 1.5080421519689406,
971
+ "grad_norm": 0.44590774178504944,
972
+ "learning_rate": 1.6983351831298558e-05,
973
+ "loss": 0.033,
974
+ "step": 1360
975
+ },
976
+ {
977
+ "epoch": 1.519134775374376,
978
+ "grad_norm": 0.5857370495796204,
979
+ "learning_rate": 1.6961154273029968e-05,
980
+ "loss": 0.0423,
981
+ "step": 1370
982
+ },
983
+ {
984
+ "epoch": 1.5302273987798114,
985
+ "grad_norm": 0.6809953451156616,
986
+ "learning_rate": 1.6938956714761378e-05,
987
+ "loss": 0.0345,
988
+ "step": 1380
989
+ },
990
+ {
991
+ "epoch": 1.5413200221852468,
992
+ "grad_norm": 0.7102778553962708,
993
+ "learning_rate": 1.6916759156492788e-05,
994
+ "loss": 0.0372,
995
+ "step": 1390
996
+ },
997
+ {
998
+ "epoch": 1.5524126455906821,
999
+ "grad_norm": 0.31781554222106934,
1000
+ "learning_rate": 1.6894561598224198e-05,
1001
+ "loss": 0.032,
1002
+ "step": 1400
1003
+ },
1004
+ {
1005
+ "epoch": 1.5635052689961175,
1006
+ "grad_norm": 0.4304943382740021,
1007
+ "learning_rate": 1.6872364039955607e-05,
1008
+ "loss": 0.0375,
1009
+ "step": 1410
1010
+ },
1011
+ {
1012
+ "epoch": 1.574597892401553,
1013
+ "grad_norm": 0.918550968170166,
1014
+ "learning_rate": 1.6850166481687017e-05,
1015
+ "loss": 0.0416,
1016
+ "step": 1420
1017
+ },
1018
+ {
1019
+ "epoch": 1.5856905158069883,
1020
+ "grad_norm": 0.722892701625824,
1021
+ "learning_rate": 1.6827968923418424e-05,
1022
+ "loss": 0.0359,
1023
+ "step": 1430
1024
+ },
1025
+ {
1026
+ "epoch": 1.5967831392124237,
1027
+ "grad_norm": 0.508703351020813,
1028
+ "learning_rate": 1.6805771365149834e-05,
1029
+ "loss": 0.0402,
1030
+ "step": 1440
1031
+ },
1032
+ {
1033
+ "epoch": 1.607875762617859,
1034
+ "grad_norm": 0.4725389778614044,
1035
+ "learning_rate": 1.6783573806881244e-05,
1036
+ "loss": 0.0335,
1037
+ "step": 1450
1038
+ },
1039
+ {
1040
+ "epoch": 1.6189683860232944,
1041
+ "grad_norm": 0.4699971079826355,
1042
+ "learning_rate": 1.6761376248612653e-05,
1043
+ "loss": 0.0277,
1044
+ "step": 1460
1045
+ },
1046
+ {
1047
+ "epoch": 1.6300610094287298,
1048
+ "grad_norm": 0.775764524936676,
1049
+ "learning_rate": 1.6739178690344063e-05,
1050
+ "loss": 0.0333,
1051
+ "step": 1470
1052
+ },
1053
+ {
1054
+ "epoch": 1.6411536328341652,
1055
+ "grad_norm": 0.6115106344223022,
1056
+ "learning_rate": 1.6716981132075473e-05,
1057
+ "loss": 0.0301,
1058
+ "step": 1480
1059
+ },
1060
+ {
1061
+ "epoch": 1.6522462562396005,
1062
+ "grad_norm": 0.743077278137207,
1063
+ "learning_rate": 1.6694783573806883e-05,
1064
+ "loss": 0.035,
1065
+ "step": 1490
1066
+ },
1067
+ {
1068
+ "epoch": 1.663338879645036,
1069
+ "grad_norm": 0.5189201235771179,
1070
+ "learning_rate": 1.6672586015538293e-05,
1071
+ "loss": 0.0475,
1072
+ "step": 1500
1073
+ },
1074
+ {
1075
+ "epoch": 1.6744315030504713,
1076
+ "grad_norm": 0.5089607238769531,
1077
+ "learning_rate": 1.6650388457269703e-05,
1078
+ "loss": 0.0377,
1079
+ "step": 1510
1080
+ },
1081
+ {
1082
+ "epoch": 1.6855241264559067,
1083
+ "grad_norm": 0.46067437529563904,
1084
+ "learning_rate": 1.6628190899001113e-05,
1085
+ "loss": 0.0297,
1086
+ "step": 1520
1087
+ },
1088
+ {
1089
+ "epoch": 1.6966167498613423,
1090
+ "grad_norm": 0.5661717653274536,
1091
+ "learning_rate": 1.660599334073252e-05,
1092
+ "loss": 0.0372,
1093
+ "step": 1530
1094
+ },
1095
+ {
1096
+ "epoch": 1.7077093732667776,
1097
+ "grad_norm": 0.45938414335250854,
1098
+ "learning_rate": 1.658379578246393e-05,
1099
+ "loss": 0.0371,
1100
+ "step": 1540
1101
+ },
1102
+ {
1103
+ "epoch": 1.718801996672213,
1104
+ "grad_norm": 0.7390128970146179,
1105
+ "learning_rate": 1.656159822419534e-05,
1106
+ "loss": 0.0332,
1107
+ "step": 1550
1108
+ },
1109
+ {
1110
+ "epoch": 1.7298946200776484,
1111
+ "grad_norm": 0.6475571990013123,
1112
+ "learning_rate": 1.653940066592675e-05,
1113
+ "loss": 0.0352,
1114
+ "step": 1560
1115
+ },
1116
+ {
1117
+ "epoch": 1.7409872434830838,
1118
+ "grad_norm": 0.8529049754142761,
1119
+ "learning_rate": 1.651720310765816e-05,
1120
+ "loss": 0.0386,
1121
+ "step": 1570
1122
+ },
1123
+ {
1124
+ "epoch": 1.7520798668885191,
1125
+ "grad_norm": 0.6980950236320496,
1126
+ "learning_rate": 1.649500554938957e-05,
1127
+ "loss": 0.0358,
1128
+ "step": 1580
1129
+ },
1130
+ {
1131
+ "epoch": 1.7631724902939545,
1132
+ "grad_norm": 0.5858293175697327,
1133
+ "learning_rate": 1.6472807991120978e-05,
1134
+ "loss": 0.0418,
1135
+ "step": 1590
1136
+ },
1137
+ {
1138
+ "epoch": 1.7742651136993899,
1139
+ "grad_norm": 0.49686577916145325,
1140
+ "learning_rate": 1.6450610432852388e-05,
1141
+ "loss": 0.0361,
1142
+ "step": 1600
1143
+ },
1144
+ {
1145
+ "epoch": 1.7853577371048253,
1146
+ "grad_norm": 0.3631349205970764,
1147
+ "learning_rate": 1.6428412874583798e-05,
1148
+ "loss": 0.0352,
1149
+ "step": 1610
1150
+ },
1151
+ {
1152
+ "epoch": 1.7964503605102606,
1153
+ "grad_norm": 0.5329940915107727,
1154
+ "learning_rate": 1.6406215316315204e-05,
1155
+ "loss": 0.04,
1156
+ "step": 1620
1157
+ },
1158
+ {
1159
+ "epoch": 1.807542983915696,
1160
+ "grad_norm": 0.5995050668716431,
1161
+ "learning_rate": 1.6384017758046614e-05,
1162
+ "loss": 0.0368,
1163
+ "step": 1630
1164
+ },
1165
+ {
1166
+ "epoch": 1.8186356073211316,
1167
+ "grad_norm": 1.3379504680633545,
1168
+ "learning_rate": 1.6361820199778028e-05,
1169
+ "loss": 0.0397,
1170
+ "step": 1640
1171
+ },
1172
+ {
1173
+ "epoch": 1.829728230726567,
1174
+ "grad_norm": 0.8026002645492554,
1175
+ "learning_rate": 1.6339622641509437e-05,
1176
+ "loss": 0.0318,
1177
+ "step": 1650
1178
+ },
1179
+ {
1180
+ "epoch": 1.8408208541320024,
1181
+ "grad_norm": 0.917950451374054,
1182
+ "learning_rate": 1.6317425083240844e-05,
1183
+ "loss": 0.0352,
1184
+ "step": 1660
1185
+ },
1186
+ {
1187
+ "epoch": 1.8519134775374377,
1188
+ "grad_norm": 0.6763226389884949,
1189
+ "learning_rate": 1.6295227524972254e-05,
1190
+ "loss": 0.0413,
1191
+ "step": 1670
1192
+ },
1193
+ {
1194
+ "epoch": 1.8630061009428731,
1195
+ "grad_norm": 0.5728912949562073,
1196
+ "learning_rate": 1.6273029966703664e-05,
1197
+ "loss": 0.0291,
1198
+ "step": 1680
1199
+ },
1200
+ {
1201
+ "epoch": 1.8740987243483085,
1202
+ "grad_norm": 0.6339443325996399,
1203
+ "learning_rate": 1.6250832408435074e-05,
1204
+ "loss": 0.0409,
1205
+ "step": 1690
1206
+ },
1207
+ {
1208
+ "epoch": 1.8851913477537439,
1209
+ "grad_norm": 0.6930853128433228,
1210
+ "learning_rate": 1.6228634850166483e-05,
1211
+ "loss": 0.0371,
1212
+ "step": 1700
1213
+ },
1214
+ {
1215
+ "epoch": 1.8962839711591792,
1216
+ "grad_norm": 0.37715452909469604,
1217
+ "learning_rate": 1.6206437291897893e-05,
1218
+ "loss": 0.031,
1219
+ "step": 1710
1220
+ },
1221
+ {
1222
+ "epoch": 1.9073765945646146,
1223
+ "grad_norm": 0.5656572580337524,
1224
+ "learning_rate": 1.61842397336293e-05,
1225
+ "loss": 0.0414,
1226
+ "step": 1720
1227
+ },
1228
+ {
1229
+ "epoch": 1.91846921797005,
1230
+ "grad_norm": 0.3563915491104126,
1231
+ "learning_rate": 1.6162042175360713e-05,
1232
+ "loss": 0.0276,
1233
+ "step": 1730
1234
+ },
1235
+ {
1236
+ "epoch": 1.9295618413754854,
1237
+ "grad_norm": 0.8260300159454346,
1238
+ "learning_rate": 1.6139844617092123e-05,
1239
+ "loss": 0.0337,
1240
+ "step": 1740
1241
+ },
1242
+ {
1243
+ "epoch": 1.9406544647809207,
1244
+ "grad_norm": 0.6907551884651184,
1245
+ "learning_rate": 1.6117647058823533e-05,
1246
+ "loss": 0.031,
1247
+ "step": 1750
1248
+ },
1249
+ {
1250
+ "epoch": 1.9517470881863561,
1251
+ "grad_norm": 0.5782826542854309,
1252
+ "learning_rate": 1.609544950055494e-05,
1253
+ "loss": 0.0419,
1254
+ "step": 1760
1255
+ },
1256
+ {
1257
+ "epoch": 1.9628397115917915,
1258
+ "grad_norm": 0.36368125677108765,
1259
+ "learning_rate": 1.607325194228635e-05,
1260
+ "loss": 0.0356,
1261
+ "step": 1770
1262
+ },
1263
+ {
1264
+ "epoch": 1.9739323349972269,
1265
+ "grad_norm": 0.369911789894104,
1266
+ "learning_rate": 1.605105438401776e-05,
1267
+ "loss": 0.0264,
1268
+ "step": 1780
1269
+ },
1270
+ {
1271
+ "epoch": 1.9850249584026622,
1272
+ "grad_norm": 0.6072190999984741,
1273
+ "learning_rate": 1.602885682574917e-05,
1274
+ "loss": 0.0366,
1275
+ "step": 1790
1276
+ },
1277
+ {
1278
+ "epoch": 1.9961175818080976,
1279
+ "grad_norm": 0.9733797311782837,
1280
+ "learning_rate": 1.600665926748058e-05,
1281
+ "loss": 0.0371,
1282
+ "step": 1800
1283
+ },
1284
+ {
1285
+ "epoch": 2.0,
1286
+ "eval_accuracy": 0.8973319824383654,
1287
+ "eval_f1": 0.6872427983539094,
1288
+ "eval_loss": 0.31070804595947266,
1289
+ "eval_precision": 0.8166259168704156,
1290
+ "eval_recall": 0.5932504440497336,
1291
+ "eval_runtime": 2.7114,
1292
+ "eval_samples_per_second": 364.014,
1293
+ "eval_steps_per_second": 11.433,
1294
+ "step": 1804
1295
+ }
1296
+ ],
1297
+ "logging_steps": 10,
1298
+ "max_steps": 9010,
1299
+ "num_input_tokens_seen": 0,
1300
+ "num_train_epochs": 10,
1301
+ "save_steps": 500,
1302
+ "stateful_callbacks": {
1303
+ "TrainerControl": {
1304
+ "args": {
1305
+ "should_epoch_stop": false,
1306
+ "should_evaluate": false,
1307
+ "should_log": false,
1308
+ "should_save": true,
1309
+ "should_training_stop": false
1310
+ },
1311
+ "attributes": {}
1312
+ }
1313
+ },
1314
+ "total_flos": 1.9234313503792004e+16,
1315
+ "train_batch_size": 32,
1316
+ "trial_name": null,
1317
+ "trial_params": null
1318
+ }
checkpoint-1804/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e64fe8f815fa80a7291289cde9af3ebb27f62206c40a93e8adf97661cd289de3
3
+ size 5368
checkpoint-2706/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
checkpoint-2706/config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaV2ForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "hidden_act": "gelu",
7
+ "hidden_dropout_prob": 0.1,
8
+ "hidden_size": 1024,
9
+ "id2label": {
10
+ "0": 0,
11
+ "1": 1,
12
+ "2": 2
13
+ },
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 4096,
16
+ "label2id": {
17
+ "0": 0,
18
+ "1": 1,
19
+ "2": 2
20
+ },
21
+ "layer_norm_eps": 1e-07,
22
+ "legacy": true,
23
+ "max_position_embeddings": 512,
24
+ "max_relative_positions": -1,
25
+ "model_type": "deberta-v2",
26
+ "norm_rel_ebd": "layer_norm",
27
+ "num_attention_heads": 16,
28
+ "num_hidden_layers": 24,
29
+ "pad_token_id": 0,
30
+ "pooler_dropout": 0,
31
+ "pooler_hidden_act": "gelu",
32
+ "pooler_hidden_size": 1024,
33
+ "pos_att_type": [
34
+ "p2c",
35
+ "c2p"
36
+ ],
37
+ "position_biased_input": false,
38
+ "position_buckets": 256,
39
+ "problem_type": "multi_label_classification",
40
+ "relative_attention": true,
41
+ "share_att_key": true,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.51.3",
44
+ "type_vocab_size": 0,
45
+ "vocab_size": 128100
46
+ }
checkpoint-2706/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3069d4fd091ef8af49a6aef3ce201d078bcb1f9943f1db9c557bfffcd5a4585
3
+ size 1740308548
checkpoint-2706/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af34088966ad696d849af1c510cd5173ed2a7fc75b64054de60420b05962c0ae
3
+ size 3480848432
checkpoint-2706/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7aec63e3c11595a16c72ea2a879b683d39bd91f9e653eaa8d09bf3d5798e951
3
+ size 14244
checkpoint-2706/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2fb444a65f40f002b72e2558a1b56a3a9a379eb0825bd3d6d1de4278b67cac8c
3
+ size 1064
checkpoint-2706/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[SEP]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": {
9
+ "content": "[UNK]",
10
+ "lstrip": false,
11
+ "normalized": true,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ }
15
+ }
checkpoint-2706/spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
3
+ size 2464616
checkpoint-2706/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-2706/tokenizer_config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[CLS]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[SEP]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[UNK]",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "128000": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "[CLS]",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "[CLS]",
47
+ "do_lower_case": false,
48
+ "eos_token": "[SEP]",
49
+ "extra_special_tokens": {},
50
+ "mask_token": "[MASK]",
51
+ "model_max_length": 1000000000000000019884624838656,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "sp_model_kwargs": {},
55
+ "split_by_punct": false,
56
+ "tokenizer_class": "DebertaV2Tokenizer",
57
+ "unk_token": "[UNK]",
58
+ "vocab_type": "spm"
59
+ }
checkpoint-2706/trainer_state.json ADDED
@@ -0,0 +1,1960 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 3.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2706,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.011092623405435386,
14
+ "grad_norm": 2.009918689727783,
15
+ "learning_rate": 1.998002219755827e-05,
16
+ "loss": 0.5022,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.022185246810870772,
21
+ "grad_norm": 0.49915584921836853,
22
+ "learning_rate": 1.995782463928968e-05,
23
+ "loss": 0.1943,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.033277870216306155,
28
+ "grad_norm": 0.47061625123023987,
29
+ "learning_rate": 1.993562708102109e-05,
30
+ "loss": 0.1436,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.044370493621741544,
35
+ "grad_norm": 0.2260124832391739,
36
+ "learning_rate": 1.99134295227525e-05,
37
+ "loss": 0.1106,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.05546311702717693,
42
+ "grad_norm": 0.31766048073768616,
43
+ "learning_rate": 1.989123196448391e-05,
44
+ "loss": 0.0994,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.06655574043261231,
49
+ "grad_norm": 0.2788391709327698,
50
+ "learning_rate": 1.9869034406215316e-05,
51
+ "loss": 0.111,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0776483638380477,
56
+ "grad_norm": 0.24376386404037476,
57
+ "learning_rate": 1.9846836847946726e-05,
58
+ "loss": 0.0849,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.08874098724348309,
63
+ "grad_norm": 0.1789788454771042,
64
+ "learning_rate": 1.982463928967814e-05,
65
+ "loss": 0.0971,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.09983361064891846,
70
+ "grad_norm": 0.29698437452316284,
71
+ "learning_rate": 1.9802441731409546e-05,
72
+ "loss": 0.0856,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.11092623405435385,
77
+ "grad_norm": 0.2553412616252899,
78
+ "learning_rate": 1.9780244173140956e-05,
79
+ "loss": 0.0976,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.12201885745978924,
84
+ "grad_norm": 0.20608791708946228,
85
+ "learning_rate": 1.9758046614872365e-05,
86
+ "loss": 0.0947,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.13311148086522462,
91
+ "grad_norm": 0.23551669716835022,
92
+ "learning_rate": 1.9735849056603775e-05,
93
+ "loss": 0.0965,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.14420410427066002,
98
+ "grad_norm": 0.3207148611545563,
99
+ "learning_rate": 1.9713651498335185e-05,
100
+ "loss": 0.0931,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.1552967276760954,
105
+ "grad_norm": 0.34813177585601807,
106
+ "learning_rate": 1.9691453940066595e-05,
107
+ "loss": 0.1,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.16638935108153077,
112
+ "grad_norm": 0.41501104831695557,
113
+ "learning_rate": 1.9669256381798e-05,
114
+ "loss": 0.0922,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.17748197448696618,
119
+ "grad_norm": 0.24550506472587585,
120
+ "learning_rate": 1.964705882352941e-05,
121
+ "loss": 0.0857,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.18857459789240155,
126
+ "grad_norm": 0.4205056130886078,
127
+ "learning_rate": 1.9624861265260825e-05,
128
+ "loss": 0.0831,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.19966722129783693,
133
+ "grad_norm": 0.6002993583679199,
134
+ "learning_rate": 1.9602663706992235e-05,
135
+ "loss": 0.0792,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.21075984470327233,
140
+ "grad_norm": 0.27535638213157654,
141
+ "learning_rate": 1.958046614872364e-05,
142
+ "loss": 0.0786,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.2218524681087077,
147
+ "grad_norm": 0.41602805256843567,
148
+ "learning_rate": 1.955826859045505e-05,
149
+ "loss": 0.0789,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.23294509151414308,
154
+ "grad_norm": 0.41827332973480225,
155
+ "learning_rate": 1.953607103218646e-05,
156
+ "loss": 0.0663,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.24403771491957849,
161
+ "grad_norm": 0.47976112365722656,
162
+ "learning_rate": 1.951387347391787e-05,
163
+ "loss": 0.0727,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.25513033832501386,
168
+ "grad_norm": 0.7771281599998474,
169
+ "learning_rate": 1.949167591564928e-05,
170
+ "loss": 0.0776,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.26622296173044924,
175
+ "grad_norm": 1.0573914051055908,
176
+ "learning_rate": 1.946947835738069e-05,
177
+ "loss": 0.0828,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.2773155851358846,
182
+ "grad_norm": 0.5186040997505188,
183
+ "learning_rate": 1.9447280799112097e-05,
184
+ "loss": 0.0767,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.28840820854132004,
189
+ "grad_norm": 0.5594862699508667,
190
+ "learning_rate": 1.942508324084351e-05,
191
+ "loss": 0.0699,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.2995008319467554,
196
+ "grad_norm": 0.5147818922996521,
197
+ "learning_rate": 1.940288568257492e-05,
198
+ "loss": 0.0591,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.3105934553521908,
203
+ "grad_norm": 0.5510437488555908,
204
+ "learning_rate": 1.938068812430633e-05,
205
+ "loss": 0.0761,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.32168607875762617,
210
+ "grad_norm": 0.4817625880241394,
211
+ "learning_rate": 1.9358490566037736e-05,
212
+ "loss": 0.0688,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.33277870216306155,
217
+ "grad_norm": 0.504362940788269,
218
+ "learning_rate": 1.9336293007769146e-05,
219
+ "loss": 0.0691,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.343871325568497,
224
+ "grad_norm": 0.5732384920120239,
225
+ "learning_rate": 1.9314095449500556e-05,
226
+ "loss": 0.057,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.35496394897393235,
231
+ "grad_norm": 0.5892298221588135,
232
+ "learning_rate": 1.9291897891231966e-05,
233
+ "loss": 0.059,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.36605657237936773,
238
+ "grad_norm": 0.3680512309074402,
239
+ "learning_rate": 1.9269700332963376e-05,
240
+ "loss": 0.0706,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.3771491957848031,
245
+ "grad_norm": 0.41971662640571594,
246
+ "learning_rate": 1.9247502774694786e-05,
247
+ "loss": 0.053,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.3882418191902385,
252
+ "grad_norm": 0.37261390686035156,
253
+ "learning_rate": 1.9225305216426195e-05,
254
+ "loss": 0.0632,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.39933444259567386,
259
+ "grad_norm": 0.48256734013557434,
260
+ "learning_rate": 1.9203107658157605e-05,
261
+ "loss": 0.0651,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.4104270660011093,
266
+ "grad_norm": 0.7914339303970337,
267
+ "learning_rate": 1.9180910099889015e-05,
268
+ "loss": 0.0659,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.42151968940654466,
273
+ "grad_norm": 0.6772429347038269,
274
+ "learning_rate": 1.915871254162042e-05,
275
+ "loss": 0.0655,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.43261231281198004,
280
+ "grad_norm": 0.5621687173843384,
281
+ "learning_rate": 1.913651498335183e-05,
282
+ "loss": 0.0707,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.4437049362174154,
287
+ "grad_norm": 0.30515748262405396,
288
+ "learning_rate": 1.911431742508324e-05,
289
+ "loss": 0.0548,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.4547975596228508,
294
+ "grad_norm": 0.5506859421730042,
295
+ "learning_rate": 1.909211986681465e-05,
296
+ "loss": 0.0554,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.46589018302828616,
301
+ "grad_norm": 0.6387749314308167,
302
+ "learning_rate": 1.906992230854606e-05,
303
+ "loss": 0.0554,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.4769828064337216,
308
+ "grad_norm": 0.18500734865665436,
309
+ "learning_rate": 1.904772475027747e-05,
310
+ "loss": 0.0558,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.48807542983915697,
315
+ "grad_norm": 0.6815407276153564,
316
+ "learning_rate": 1.902552719200888e-05,
317
+ "loss": 0.0537,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.49916805324459235,
322
+ "grad_norm": 0.4826994836330414,
323
+ "learning_rate": 1.900332963374029e-05,
324
+ "loss": 0.0567,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.5102606766500277,
329
+ "grad_norm": 0.24618124961853027,
330
+ "learning_rate": 1.89811320754717e-05,
331
+ "loss": 0.0614,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.5213533000554631,
336
+ "grad_norm": 1.0037415027618408,
337
+ "learning_rate": 1.895893451720311e-05,
338
+ "loss": 0.0604,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.5324459234608985,
343
+ "grad_norm": 0.6000948548316956,
344
+ "learning_rate": 1.8936736958934517e-05,
345
+ "loss": 0.0683,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.5435385468663338,
350
+ "grad_norm": 0.3302474319934845,
351
+ "learning_rate": 1.8914539400665927e-05,
352
+ "loss": 0.0543,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.5546311702717692,
357
+ "grad_norm": 0.5560783743858337,
358
+ "learning_rate": 1.8892341842397337e-05,
359
+ "loss": 0.0569,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.5657237936772047,
364
+ "grad_norm": 0.8041097521781921,
365
+ "learning_rate": 1.8870144284128747e-05,
366
+ "loss": 0.0544,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.5768164170826401,
371
+ "grad_norm": 0.6846103668212891,
372
+ "learning_rate": 1.8847946725860156e-05,
373
+ "loss": 0.058,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.5879090404880755,
378
+ "grad_norm": 0.50434809923172,
379
+ "learning_rate": 1.8825749167591566e-05,
380
+ "loss": 0.0654,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.5990016638935108,
385
+ "grad_norm": 0.54362553358078,
386
+ "learning_rate": 1.8803551609322976e-05,
387
+ "loss": 0.0582,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.6100942872989462,
392
+ "grad_norm": 0.6166839599609375,
393
+ "learning_rate": 1.8781354051054386e-05,
394
+ "loss": 0.0672,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.6211869107043816,
399
+ "grad_norm": 0.4353054165840149,
400
+ "learning_rate": 1.8759156492785796e-05,
401
+ "loss": 0.0532,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.632279534109817,
406
+ "grad_norm": 0.6025580167770386,
407
+ "learning_rate": 1.8736958934517206e-05,
408
+ "loss": 0.0746,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.6433721575152523,
413
+ "grad_norm": 0.7259892225265503,
414
+ "learning_rate": 1.8714761376248612e-05,
415
+ "loss": 0.0642,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.6544647809206877,
420
+ "grad_norm": 0.4940318465232849,
421
+ "learning_rate": 1.8692563817980022e-05,
422
+ "loss": 0.0547,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.6655574043261231,
427
+ "grad_norm": 0.7005699872970581,
428
+ "learning_rate": 1.8670366259711435e-05,
429
+ "loss": 0.0522,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.6766500277315585,
434
+ "grad_norm": 0.4530707895755768,
435
+ "learning_rate": 1.8648168701442845e-05,
436
+ "loss": 0.054,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.687742651136994,
441
+ "grad_norm": 0.9097110629081726,
442
+ "learning_rate": 1.8625971143174252e-05,
443
+ "loss": 0.0622,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.6988352745424293,
448
+ "grad_norm": 0.5374599695205688,
449
+ "learning_rate": 1.860377358490566e-05,
450
+ "loss": 0.0522,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.7099278979478647,
455
+ "grad_norm": 0.3849945664405823,
456
+ "learning_rate": 1.858157602663707e-05,
457
+ "loss": 0.0571,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.7210205213533001,
462
+ "grad_norm": 0.5918008685112,
463
+ "learning_rate": 1.855937846836848e-05,
464
+ "loss": 0.0576,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.7321131447587355,
469
+ "grad_norm": 0.3229956030845642,
470
+ "learning_rate": 1.853718091009989e-05,
471
+ "loss": 0.0537,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.7432057681641708,
476
+ "grad_norm": 0.5264039039611816,
477
+ "learning_rate": 1.85149833518313e-05,
478
+ "loss": 0.06,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.7542983915696062,
483
+ "grad_norm": 0.36795660853385925,
484
+ "learning_rate": 1.8492785793562708e-05,
485
+ "loss": 0.0511,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.7653910149750416,
490
+ "grad_norm": 0.5905130505561829,
491
+ "learning_rate": 1.847058823529412e-05,
492
+ "loss": 0.0636,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.776483638380477,
497
+ "grad_norm": 0.36266571283340454,
498
+ "learning_rate": 1.844839067702553e-05,
499
+ "loss": 0.0598,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.7875762617859123,
504
+ "grad_norm": 0.4978592097759247,
505
+ "learning_rate": 1.8426193118756937e-05,
506
+ "loss": 0.0607,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.7986688851913477,
511
+ "grad_norm": 0.4635021686553955,
512
+ "learning_rate": 1.8403995560488347e-05,
513
+ "loss": 0.0543,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.8097615085967831,
518
+ "grad_norm": 0.44571858644485474,
519
+ "learning_rate": 1.8381798002219757e-05,
520
+ "loss": 0.0678,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.8208541320022186,
525
+ "grad_norm": 0.8265877366065979,
526
+ "learning_rate": 1.8359600443951167e-05,
527
+ "loss": 0.0552,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.831946755407654,
532
+ "grad_norm": 0.5776472091674805,
533
+ "learning_rate": 1.8337402885682577e-05,
534
+ "loss": 0.0505,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.8430393788130893,
539
+ "grad_norm": 0.299274742603302,
540
+ "learning_rate": 1.8315205327413986e-05,
541
+ "loss": 0.0619,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.8541320022185247,
546
+ "grad_norm": 0.587645947933197,
547
+ "learning_rate": 1.8293007769145393e-05,
548
+ "loss": 0.0575,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.8652246256239601,
553
+ "grad_norm": 0.39164137840270996,
554
+ "learning_rate": 1.8270810210876806e-05,
555
+ "loss": 0.0458,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.8763172490293955,
560
+ "grad_norm": 0.4663292169570923,
561
+ "learning_rate": 1.8248612652608216e-05,
562
+ "loss": 0.048,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.8874098724348308,
567
+ "grad_norm": 0.5804581642150879,
568
+ "learning_rate": 1.8226415094339626e-05,
569
+ "loss": 0.0557,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.8985024958402662,
574
+ "grad_norm": 0.4279440939426422,
575
+ "learning_rate": 1.8204217536071032e-05,
576
+ "loss": 0.051,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.9095951192457016,
581
+ "grad_norm": 0.5384302735328674,
582
+ "learning_rate": 1.8182019977802442e-05,
583
+ "loss": 0.0558,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.920687742651137,
588
+ "grad_norm": 0.5049973726272583,
589
+ "learning_rate": 1.8159822419533852e-05,
590
+ "loss": 0.047,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.9317803660565723,
595
+ "grad_norm": 0.8061177134513855,
596
+ "learning_rate": 1.8137624861265262e-05,
597
+ "loss": 0.0572,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.9428729894620078,
602
+ "grad_norm": 0.6056540012359619,
603
+ "learning_rate": 1.8115427302996672e-05,
604
+ "loss": 0.0539,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.9539656128674432,
609
+ "grad_norm": 0.3001384139060974,
610
+ "learning_rate": 1.8093229744728082e-05,
611
+ "loss": 0.053,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.9650582362728786,
616
+ "grad_norm": 0.6709749102592468,
617
+ "learning_rate": 1.807103218645949e-05,
618
+ "loss": 0.0526,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.9761508596783139,
623
+ "grad_norm": 0.8233507871627808,
624
+ "learning_rate": 1.80488346281909e-05,
625
+ "loss": 0.0621,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.9872434830837493,
630
+ "grad_norm": 0.5757150650024414,
631
+ "learning_rate": 1.802663706992231e-05,
632
+ "loss": 0.0546,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.9983361064891847,
637
+ "grad_norm": 0.6834889054298401,
638
+ "learning_rate": 1.800443951165372e-05,
639
+ "loss": 0.0462,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 1.0,
644
+ "eval_accuracy": 0.8720027017899359,
645
+ "eval_f1": 0.5678449258836944,
646
+ "eval_loss": 0.34414270520210266,
647
+ "eval_precision": 0.7929936305732485,
648
+ "eval_recall": 0.4422735346358792,
649
+ "eval_runtime": 2.727,
650
+ "eval_samples_per_second": 361.937,
651
+ "eval_steps_per_second": 11.368,
652
+ "step": 902
653
+ },
654
+ {
655
+ "epoch": 1.0088740987243483,
656
+ "grad_norm": 0.4411003887653351,
657
+ "learning_rate": 1.7982241953385128e-05,
658
+ "loss": 0.0312,
659
+ "step": 910
660
+ },
661
+ {
662
+ "epoch": 1.0199667221297837,
663
+ "grad_norm": 0.40500083565711975,
664
+ "learning_rate": 1.7960044395116538e-05,
665
+ "loss": 0.0345,
666
+ "step": 920
667
+ },
668
+ {
669
+ "epoch": 1.031059345535219,
670
+ "grad_norm": 0.9287449717521667,
671
+ "learning_rate": 1.7937846836847947e-05,
672
+ "loss": 0.0419,
673
+ "step": 930
674
+ },
675
+ {
676
+ "epoch": 1.0421519689406544,
677
+ "grad_norm": 0.7969145178794861,
678
+ "learning_rate": 1.7915649278579357e-05,
679
+ "loss": 0.042,
680
+ "step": 940
681
+ },
682
+ {
683
+ "epoch": 1.0532445923460898,
684
+ "grad_norm": 0.49280259013175964,
685
+ "learning_rate": 1.7893451720310767e-05,
686
+ "loss": 0.0427,
687
+ "step": 950
688
+ },
689
+ {
690
+ "epoch": 1.0643372157515252,
691
+ "grad_norm": 1.1647624969482422,
692
+ "learning_rate": 1.7871254162042177e-05,
693
+ "loss": 0.0341,
694
+ "step": 960
695
+ },
696
+ {
697
+ "epoch": 1.0754298391569606,
698
+ "grad_norm": 0.2908968925476074,
699
+ "learning_rate": 1.7849056603773587e-05,
700
+ "loss": 0.0389,
701
+ "step": 970
702
+ },
703
+ {
704
+ "epoch": 1.086522462562396,
705
+ "grad_norm": 0.521218478679657,
706
+ "learning_rate": 1.7826859045504997e-05,
707
+ "loss": 0.0396,
708
+ "step": 980
709
+ },
710
+ {
711
+ "epoch": 1.0976150859678313,
712
+ "grad_norm": 0.4156598150730133,
713
+ "learning_rate": 1.7804661487236407e-05,
714
+ "loss": 0.0359,
715
+ "step": 990
716
+ },
717
+ {
718
+ "epoch": 1.1087077093732667,
719
+ "grad_norm": 0.7260242104530334,
720
+ "learning_rate": 1.7782463928967813e-05,
721
+ "loss": 0.0384,
722
+ "step": 1000
723
+ },
724
+ {
725
+ "epoch": 1.119800332778702,
726
+ "grad_norm": 0.43658459186553955,
727
+ "learning_rate": 1.7760266370699223e-05,
728
+ "loss": 0.0383,
729
+ "step": 1010
730
+ },
731
+ {
732
+ "epoch": 1.1308929561841374,
733
+ "grad_norm": 0.3585101068019867,
734
+ "learning_rate": 1.7738068812430633e-05,
735
+ "loss": 0.0401,
736
+ "step": 1020
737
+ },
738
+ {
739
+ "epoch": 1.141985579589573,
740
+ "grad_norm": 0.5010389089584351,
741
+ "learning_rate": 1.7715871254162043e-05,
742
+ "loss": 0.0307,
743
+ "step": 1030
744
+ },
745
+ {
746
+ "epoch": 1.1530782029950084,
747
+ "grad_norm": 0.6415812969207764,
748
+ "learning_rate": 1.7693673695893453e-05,
749
+ "loss": 0.0423,
750
+ "step": 1040
751
+ },
752
+ {
753
+ "epoch": 1.1641708264004438,
754
+ "grad_norm": 0.6393259763717651,
755
+ "learning_rate": 1.7671476137624862e-05,
756
+ "loss": 0.0392,
757
+ "step": 1050
758
+ },
759
+ {
760
+ "epoch": 1.1752634498058792,
761
+ "grad_norm": 0.6339041590690613,
762
+ "learning_rate": 1.7649278579356272e-05,
763
+ "loss": 0.0506,
764
+ "step": 1060
765
+ },
766
+ {
767
+ "epoch": 1.1863560732113145,
768
+ "grad_norm": 0.6086763143539429,
769
+ "learning_rate": 1.7627081021087682e-05,
770
+ "loss": 0.0416,
771
+ "step": 1070
772
+ },
773
+ {
774
+ "epoch": 1.19744869661675,
775
+ "grad_norm": 0.6648682355880737,
776
+ "learning_rate": 1.7604883462819092e-05,
777
+ "loss": 0.0379,
778
+ "step": 1080
779
+ },
780
+ {
781
+ "epoch": 1.2085413200221853,
782
+ "grad_norm": 0.6018221378326416,
783
+ "learning_rate": 1.7582685904550502e-05,
784
+ "loss": 0.0354,
785
+ "step": 1090
786
+ },
787
+ {
788
+ "epoch": 1.2196339434276207,
789
+ "grad_norm": 1.2366654872894287,
790
+ "learning_rate": 1.756048834628191e-05,
791
+ "loss": 0.037,
792
+ "step": 1100
793
+ },
794
+ {
795
+ "epoch": 1.230726566833056,
796
+ "grad_norm": 0.5976310968399048,
797
+ "learning_rate": 1.7538290788013318e-05,
798
+ "loss": 0.0361,
799
+ "step": 1110
800
+ },
801
+ {
802
+ "epoch": 1.2418191902384914,
803
+ "grad_norm": 0.8290308117866516,
804
+ "learning_rate": 1.751609322974473e-05,
805
+ "loss": 0.0383,
806
+ "step": 1120
807
+ },
808
+ {
809
+ "epoch": 1.2529118136439268,
810
+ "grad_norm": 0.382548987865448,
811
+ "learning_rate": 1.749389567147614e-05,
812
+ "loss": 0.0449,
813
+ "step": 1130
814
+ },
815
+ {
816
+ "epoch": 1.2640044370493622,
817
+ "grad_norm": 0.5307976007461548,
818
+ "learning_rate": 1.7471698113207548e-05,
819
+ "loss": 0.0364,
820
+ "step": 1140
821
+ },
822
+ {
823
+ "epoch": 1.2750970604547975,
824
+ "grad_norm": 0.5508521795272827,
825
+ "learning_rate": 1.7449500554938958e-05,
826
+ "loss": 0.0312,
827
+ "step": 1150
828
+ },
829
+ {
830
+ "epoch": 1.286189683860233,
831
+ "grad_norm": 0.37057268619537354,
832
+ "learning_rate": 1.7427302996670368e-05,
833
+ "loss": 0.0397,
834
+ "step": 1160
835
+ },
836
+ {
837
+ "epoch": 1.2972823072656683,
838
+ "grad_norm": 0.4806898236274719,
839
+ "learning_rate": 1.7405105438401777e-05,
840
+ "loss": 0.0391,
841
+ "step": 1170
842
+ },
843
+ {
844
+ "epoch": 1.3083749306711037,
845
+ "grad_norm": 0.759772002696991,
846
+ "learning_rate": 1.7382907880133187e-05,
847
+ "loss": 0.0322,
848
+ "step": 1180
849
+ },
850
+ {
851
+ "epoch": 1.319467554076539,
852
+ "grad_norm": 0.39819085597991943,
853
+ "learning_rate": 1.7360710321864597e-05,
854
+ "loss": 0.0345,
855
+ "step": 1190
856
+ },
857
+ {
858
+ "epoch": 1.3305601774819744,
859
+ "grad_norm": 0.7733897566795349,
860
+ "learning_rate": 1.7338512763596004e-05,
861
+ "loss": 0.0424,
862
+ "step": 1200
863
+ },
864
+ {
865
+ "epoch": 1.3416528008874098,
866
+ "grad_norm": 0.7979075908660889,
867
+ "learning_rate": 1.7316315205327417e-05,
868
+ "loss": 0.0403,
869
+ "step": 1210
870
+ },
871
+ {
872
+ "epoch": 1.3527454242928454,
873
+ "grad_norm": 0.5658752918243408,
874
+ "learning_rate": 1.7294117647058827e-05,
875
+ "loss": 0.0276,
876
+ "step": 1220
877
+ },
878
+ {
879
+ "epoch": 1.3638380476982808,
880
+ "grad_norm": 0.25207844376564026,
881
+ "learning_rate": 1.7271920088790237e-05,
882
+ "loss": 0.0347,
883
+ "step": 1230
884
+ },
885
+ {
886
+ "epoch": 1.3749306711037161,
887
+ "grad_norm": 0.6338945031166077,
888
+ "learning_rate": 1.7249722530521643e-05,
889
+ "loss": 0.0416,
890
+ "step": 1240
891
+ },
892
+ {
893
+ "epoch": 1.3860232945091515,
894
+ "grad_norm": 0.8125913143157959,
895
+ "learning_rate": 1.7227524972253053e-05,
896
+ "loss": 0.0442,
897
+ "step": 1250
898
+ },
899
+ {
900
+ "epoch": 1.397115917914587,
901
+ "grad_norm": 0.5067325234413147,
902
+ "learning_rate": 1.7205327413984463e-05,
903
+ "loss": 0.0358,
904
+ "step": 1260
905
+ },
906
+ {
907
+ "epoch": 1.4082085413200223,
908
+ "grad_norm": 0.40132638812065125,
909
+ "learning_rate": 1.7183129855715873e-05,
910
+ "loss": 0.0303,
911
+ "step": 1270
912
+ },
913
+ {
914
+ "epoch": 1.4193011647254576,
915
+ "grad_norm": 0.6337258815765381,
916
+ "learning_rate": 1.7160932297447283e-05,
917
+ "loss": 0.0471,
918
+ "step": 1280
919
+ },
920
+ {
921
+ "epoch": 1.430393788130893,
922
+ "grad_norm": 0.43456393480300903,
923
+ "learning_rate": 1.713873473917869e-05,
924
+ "loss": 0.0354,
925
+ "step": 1290
926
+ },
927
+ {
928
+ "epoch": 1.4414864115363284,
929
+ "grad_norm": 0.3291069269180298,
930
+ "learning_rate": 1.7116537180910102e-05,
931
+ "loss": 0.0312,
932
+ "step": 1300
933
+ },
934
+ {
935
+ "epoch": 1.4525790349417638,
936
+ "grad_norm": 0.2960388660430908,
937
+ "learning_rate": 1.7094339622641512e-05,
938
+ "loss": 0.0362,
939
+ "step": 1310
940
+ },
941
+ {
942
+ "epoch": 1.4636716583471991,
943
+ "grad_norm": 0.820751428604126,
944
+ "learning_rate": 1.7072142064372922e-05,
945
+ "loss": 0.0396,
946
+ "step": 1320
947
+ },
948
+ {
949
+ "epoch": 1.4747642817526345,
950
+ "grad_norm": 0.7335907816886902,
951
+ "learning_rate": 1.704994450610433e-05,
952
+ "loss": 0.0386,
953
+ "step": 1330
954
+ },
955
+ {
956
+ "epoch": 1.48585690515807,
957
+ "grad_norm": 0.40592002868652344,
958
+ "learning_rate": 1.702774694783574e-05,
959
+ "loss": 0.0358,
960
+ "step": 1340
961
+ },
962
+ {
963
+ "epoch": 1.4969495285635053,
964
+ "grad_norm": 0.6194770336151123,
965
+ "learning_rate": 1.7005549389567148e-05,
966
+ "loss": 0.0361,
967
+ "step": 1350
968
+ },
969
+ {
970
+ "epoch": 1.5080421519689406,
971
+ "grad_norm": 0.44590774178504944,
972
+ "learning_rate": 1.6983351831298558e-05,
973
+ "loss": 0.033,
974
+ "step": 1360
975
+ },
976
+ {
977
+ "epoch": 1.519134775374376,
978
+ "grad_norm": 0.5857370495796204,
979
+ "learning_rate": 1.6961154273029968e-05,
980
+ "loss": 0.0423,
981
+ "step": 1370
982
+ },
983
+ {
984
+ "epoch": 1.5302273987798114,
985
+ "grad_norm": 0.6809953451156616,
986
+ "learning_rate": 1.6938956714761378e-05,
987
+ "loss": 0.0345,
988
+ "step": 1380
989
+ },
990
+ {
991
+ "epoch": 1.5413200221852468,
992
+ "grad_norm": 0.7102778553962708,
993
+ "learning_rate": 1.6916759156492788e-05,
994
+ "loss": 0.0372,
995
+ "step": 1390
996
+ },
997
+ {
998
+ "epoch": 1.5524126455906821,
999
+ "grad_norm": 0.31781554222106934,
1000
+ "learning_rate": 1.6894561598224198e-05,
1001
+ "loss": 0.032,
1002
+ "step": 1400
1003
+ },
1004
+ {
1005
+ "epoch": 1.5635052689961175,
1006
+ "grad_norm": 0.4304943382740021,
1007
+ "learning_rate": 1.6872364039955607e-05,
1008
+ "loss": 0.0375,
1009
+ "step": 1410
1010
+ },
1011
+ {
1012
+ "epoch": 1.574597892401553,
1013
+ "grad_norm": 0.918550968170166,
1014
+ "learning_rate": 1.6850166481687017e-05,
1015
+ "loss": 0.0416,
1016
+ "step": 1420
1017
+ },
1018
+ {
1019
+ "epoch": 1.5856905158069883,
1020
+ "grad_norm": 0.722892701625824,
1021
+ "learning_rate": 1.6827968923418424e-05,
1022
+ "loss": 0.0359,
1023
+ "step": 1430
1024
+ },
1025
+ {
1026
+ "epoch": 1.5967831392124237,
1027
+ "grad_norm": 0.508703351020813,
1028
+ "learning_rate": 1.6805771365149834e-05,
1029
+ "loss": 0.0402,
1030
+ "step": 1440
1031
+ },
1032
+ {
1033
+ "epoch": 1.607875762617859,
1034
+ "grad_norm": 0.4725389778614044,
1035
+ "learning_rate": 1.6783573806881244e-05,
1036
+ "loss": 0.0335,
1037
+ "step": 1450
1038
+ },
1039
+ {
1040
+ "epoch": 1.6189683860232944,
1041
+ "grad_norm": 0.4699971079826355,
1042
+ "learning_rate": 1.6761376248612653e-05,
1043
+ "loss": 0.0277,
1044
+ "step": 1460
1045
+ },
1046
+ {
1047
+ "epoch": 1.6300610094287298,
1048
+ "grad_norm": 0.775764524936676,
1049
+ "learning_rate": 1.6739178690344063e-05,
1050
+ "loss": 0.0333,
1051
+ "step": 1470
1052
+ },
1053
+ {
1054
+ "epoch": 1.6411536328341652,
1055
+ "grad_norm": 0.6115106344223022,
1056
+ "learning_rate": 1.6716981132075473e-05,
1057
+ "loss": 0.0301,
1058
+ "step": 1480
1059
+ },
1060
+ {
1061
+ "epoch": 1.6522462562396005,
1062
+ "grad_norm": 0.743077278137207,
1063
+ "learning_rate": 1.6694783573806883e-05,
1064
+ "loss": 0.035,
1065
+ "step": 1490
1066
+ },
1067
+ {
1068
+ "epoch": 1.663338879645036,
1069
+ "grad_norm": 0.5189201235771179,
1070
+ "learning_rate": 1.6672586015538293e-05,
1071
+ "loss": 0.0475,
1072
+ "step": 1500
1073
+ },
1074
+ {
1075
+ "epoch": 1.6744315030504713,
1076
+ "grad_norm": 0.5089607238769531,
1077
+ "learning_rate": 1.6650388457269703e-05,
1078
+ "loss": 0.0377,
1079
+ "step": 1510
1080
+ },
1081
+ {
1082
+ "epoch": 1.6855241264559067,
1083
+ "grad_norm": 0.46067437529563904,
1084
+ "learning_rate": 1.6628190899001113e-05,
1085
+ "loss": 0.0297,
1086
+ "step": 1520
1087
+ },
1088
+ {
1089
+ "epoch": 1.6966167498613423,
1090
+ "grad_norm": 0.5661717653274536,
1091
+ "learning_rate": 1.660599334073252e-05,
1092
+ "loss": 0.0372,
1093
+ "step": 1530
1094
+ },
1095
+ {
1096
+ "epoch": 1.7077093732667776,
1097
+ "grad_norm": 0.45938414335250854,
1098
+ "learning_rate": 1.658379578246393e-05,
1099
+ "loss": 0.0371,
1100
+ "step": 1540
1101
+ },
1102
+ {
1103
+ "epoch": 1.718801996672213,
1104
+ "grad_norm": 0.7390128970146179,
1105
+ "learning_rate": 1.656159822419534e-05,
1106
+ "loss": 0.0332,
1107
+ "step": 1550
1108
+ },
1109
+ {
1110
+ "epoch": 1.7298946200776484,
1111
+ "grad_norm": 0.6475571990013123,
1112
+ "learning_rate": 1.653940066592675e-05,
1113
+ "loss": 0.0352,
1114
+ "step": 1560
1115
+ },
1116
+ {
1117
+ "epoch": 1.7409872434830838,
1118
+ "grad_norm": 0.8529049754142761,
1119
+ "learning_rate": 1.651720310765816e-05,
1120
+ "loss": 0.0386,
1121
+ "step": 1570
1122
+ },
1123
+ {
1124
+ "epoch": 1.7520798668885191,
1125
+ "grad_norm": 0.6980950236320496,
1126
+ "learning_rate": 1.649500554938957e-05,
1127
+ "loss": 0.0358,
1128
+ "step": 1580
1129
+ },
1130
+ {
1131
+ "epoch": 1.7631724902939545,
1132
+ "grad_norm": 0.5858293175697327,
1133
+ "learning_rate": 1.6472807991120978e-05,
1134
+ "loss": 0.0418,
1135
+ "step": 1590
1136
+ },
1137
+ {
1138
+ "epoch": 1.7742651136993899,
1139
+ "grad_norm": 0.49686577916145325,
1140
+ "learning_rate": 1.6450610432852388e-05,
1141
+ "loss": 0.0361,
1142
+ "step": 1600
1143
+ },
1144
+ {
1145
+ "epoch": 1.7853577371048253,
1146
+ "grad_norm": 0.3631349205970764,
1147
+ "learning_rate": 1.6428412874583798e-05,
1148
+ "loss": 0.0352,
1149
+ "step": 1610
1150
+ },
1151
+ {
1152
+ "epoch": 1.7964503605102606,
1153
+ "grad_norm": 0.5329940915107727,
1154
+ "learning_rate": 1.6406215316315204e-05,
1155
+ "loss": 0.04,
1156
+ "step": 1620
1157
+ },
1158
+ {
1159
+ "epoch": 1.807542983915696,
1160
+ "grad_norm": 0.5995050668716431,
1161
+ "learning_rate": 1.6384017758046614e-05,
1162
+ "loss": 0.0368,
1163
+ "step": 1630
1164
+ },
1165
+ {
1166
+ "epoch": 1.8186356073211316,
1167
+ "grad_norm": 1.3379504680633545,
1168
+ "learning_rate": 1.6361820199778028e-05,
1169
+ "loss": 0.0397,
1170
+ "step": 1640
1171
+ },
1172
+ {
1173
+ "epoch": 1.829728230726567,
1174
+ "grad_norm": 0.8026002645492554,
1175
+ "learning_rate": 1.6339622641509437e-05,
1176
+ "loss": 0.0318,
1177
+ "step": 1650
1178
+ },
1179
+ {
1180
+ "epoch": 1.8408208541320024,
1181
+ "grad_norm": 0.917950451374054,
1182
+ "learning_rate": 1.6317425083240844e-05,
1183
+ "loss": 0.0352,
1184
+ "step": 1660
1185
+ },
1186
+ {
1187
+ "epoch": 1.8519134775374377,
1188
+ "grad_norm": 0.6763226389884949,
1189
+ "learning_rate": 1.6295227524972254e-05,
1190
+ "loss": 0.0413,
1191
+ "step": 1670
1192
+ },
1193
+ {
1194
+ "epoch": 1.8630061009428731,
1195
+ "grad_norm": 0.5728912949562073,
1196
+ "learning_rate": 1.6273029966703664e-05,
1197
+ "loss": 0.0291,
1198
+ "step": 1680
1199
+ },
1200
+ {
1201
+ "epoch": 1.8740987243483085,
1202
+ "grad_norm": 0.6339443325996399,
1203
+ "learning_rate": 1.6250832408435074e-05,
1204
+ "loss": 0.0409,
1205
+ "step": 1690
1206
+ },
1207
+ {
1208
+ "epoch": 1.8851913477537439,
1209
+ "grad_norm": 0.6930853128433228,
1210
+ "learning_rate": 1.6228634850166483e-05,
1211
+ "loss": 0.0371,
1212
+ "step": 1700
1213
+ },
1214
+ {
1215
+ "epoch": 1.8962839711591792,
1216
+ "grad_norm": 0.37715452909469604,
1217
+ "learning_rate": 1.6206437291897893e-05,
1218
+ "loss": 0.031,
1219
+ "step": 1710
1220
+ },
1221
+ {
1222
+ "epoch": 1.9073765945646146,
1223
+ "grad_norm": 0.5656572580337524,
1224
+ "learning_rate": 1.61842397336293e-05,
1225
+ "loss": 0.0414,
1226
+ "step": 1720
1227
+ },
1228
+ {
1229
+ "epoch": 1.91846921797005,
1230
+ "grad_norm": 0.3563915491104126,
1231
+ "learning_rate": 1.6162042175360713e-05,
1232
+ "loss": 0.0276,
1233
+ "step": 1730
1234
+ },
1235
+ {
1236
+ "epoch": 1.9295618413754854,
1237
+ "grad_norm": 0.8260300159454346,
1238
+ "learning_rate": 1.6139844617092123e-05,
1239
+ "loss": 0.0337,
1240
+ "step": 1740
1241
+ },
1242
+ {
1243
+ "epoch": 1.9406544647809207,
1244
+ "grad_norm": 0.6907551884651184,
1245
+ "learning_rate": 1.6117647058823533e-05,
1246
+ "loss": 0.031,
1247
+ "step": 1750
1248
+ },
1249
+ {
1250
+ "epoch": 1.9517470881863561,
1251
+ "grad_norm": 0.5782826542854309,
1252
+ "learning_rate": 1.609544950055494e-05,
1253
+ "loss": 0.0419,
1254
+ "step": 1760
1255
+ },
1256
+ {
1257
+ "epoch": 1.9628397115917915,
1258
+ "grad_norm": 0.36368125677108765,
1259
+ "learning_rate": 1.607325194228635e-05,
1260
+ "loss": 0.0356,
1261
+ "step": 1770
1262
+ },
1263
+ {
1264
+ "epoch": 1.9739323349972269,
1265
+ "grad_norm": 0.369911789894104,
1266
+ "learning_rate": 1.605105438401776e-05,
1267
+ "loss": 0.0264,
1268
+ "step": 1780
1269
+ },
1270
+ {
1271
+ "epoch": 1.9850249584026622,
1272
+ "grad_norm": 0.6072190999984741,
1273
+ "learning_rate": 1.602885682574917e-05,
1274
+ "loss": 0.0366,
1275
+ "step": 1790
1276
+ },
1277
+ {
1278
+ "epoch": 1.9961175818080976,
1279
+ "grad_norm": 0.9733797311782837,
1280
+ "learning_rate": 1.600665926748058e-05,
1281
+ "loss": 0.0371,
1282
+ "step": 1800
1283
+ },
1284
+ {
1285
+ "epoch": 2.0,
1286
+ "eval_accuracy": 0.8973319824383654,
1287
+ "eval_f1": 0.6872427983539094,
1288
+ "eval_loss": 0.31070804595947266,
1289
+ "eval_precision": 0.8166259168704156,
1290
+ "eval_recall": 0.5932504440497336,
1291
+ "eval_runtime": 2.7114,
1292
+ "eval_samples_per_second": 364.014,
1293
+ "eval_steps_per_second": 11.433,
1294
+ "step": 1804
1295
+ },
1296
+ {
1297
+ "epoch": 2.0066555740432612,
1298
+ "grad_norm": 0.2766992449760437,
1299
+ "learning_rate": 1.598446170921199e-05,
1300
+ "loss": 0.0218,
1301
+ "step": 1810
1302
+ },
1303
+ {
1304
+ "epoch": 2.0177481974486966,
1305
+ "grad_norm": 0.6556938290596008,
1306
+ "learning_rate": 1.59622641509434e-05,
1307
+ "loss": 0.0187,
1308
+ "step": 1820
1309
+ },
1310
+ {
1311
+ "epoch": 2.028840820854132,
1312
+ "grad_norm": 0.24634911119937897,
1313
+ "learning_rate": 1.5940066592674808e-05,
1314
+ "loss": 0.0132,
1315
+ "step": 1830
1316
+ },
1317
+ {
1318
+ "epoch": 2.0399334442595674,
1319
+ "grad_norm": 0.683587372303009,
1320
+ "learning_rate": 1.5917869034406218e-05,
1321
+ "loss": 0.0158,
1322
+ "step": 1840
1323
+ },
1324
+ {
1325
+ "epoch": 2.0510260676650027,
1326
+ "grad_norm": 0.47117751836776733,
1327
+ "learning_rate": 1.5895671476137625e-05,
1328
+ "loss": 0.018,
1329
+ "step": 1850
1330
+ },
1331
+ {
1332
+ "epoch": 2.062118691070438,
1333
+ "grad_norm": 0.4102018475532532,
1334
+ "learning_rate": 1.5873473917869034e-05,
1335
+ "loss": 0.0153,
1336
+ "step": 1860
1337
+ },
1338
+ {
1339
+ "epoch": 2.0732113144758735,
1340
+ "grad_norm": 0.4639064371585846,
1341
+ "learning_rate": 1.5851276359600444e-05,
1342
+ "loss": 0.0171,
1343
+ "step": 1870
1344
+ },
1345
+ {
1346
+ "epoch": 2.084303937881309,
1347
+ "grad_norm": 0.49497199058532715,
1348
+ "learning_rate": 1.5829078801331854e-05,
1349
+ "loss": 0.0168,
1350
+ "step": 1880
1351
+ },
1352
+ {
1353
+ "epoch": 2.0953965612867442,
1354
+ "grad_norm": 0.6302080154418945,
1355
+ "learning_rate": 1.5806881243063264e-05,
1356
+ "loss": 0.0191,
1357
+ "step": 1890
1358
+ },
1359
+ {
1360
+ "epoch": 2.1064891846921796,
1361
+ "grad_norm": 0.3464473783969879,
1362
+ "learning_rate": 1.5784683684794674e-05,
1363
+ "loss": 0.0184,
1364
+ "step": 1900
1365
+ },
1366
+ {
1367
+ "epoch": 2.117581808097615,
1368
+ "grad_norm": 0.6507964730262756,
1369
+ "learning_rate": 1.5762486126526084e-05,
1370
+ "loss": 0.0135,
1371
+ "step": 1910
1372
+ },
1373
+ {
1374
+ "epoch": 2.1286744315030504,
1375
+ "grad_norm": 0.1759006232023239,
1376
+ "learning_rate": 1.5740288568257494e-05,
1377
+ "loss": 0.0118,
1378
+ "step": 1920
1379
+ },
1380
+ {
1381
+ "epoch": 2.1397670549084857,
1382
+ "grad_norm": 0.5952832698822021,
1383
+ "learning_rate": 1.5718091009988904e-05,
1384
+ "loss": 0.0157,
1385
+ "step": 1930
1386
+ },
1387
+ {
1388
+ "epoch": 2.150859678313921,
1389
+ "grad_norm": 0.8094580769538879,
1390
+ "learning_rate": 1.5695893451720313e-05,
1391
+ "loss": 0.0125,
1392
+ "step": 1940
1393
+ },
1394
+ {
1395
+ "epoch": 2.1619523017193565,
1396
+ "grad_norm": 0.6284286379814148,
1397
+ "learning_rate": 1.567369589345172e-05,
1398
+ "loss": 0.0178,
1399
+ "step": 1950
1400
+ },
1401
+ {
1402
+ "epoch": 2.173044925124792,
1403
+ "grad_norm": 0.6017957925796509,
1404
+ "learning_rate": 1.565149833518313e-05,
1405
+ "loss": 0.0136,
1406
+ "step": 1960
1407
+ },
1408
+ {
1409
+ "epoch": 2.1841375485302272,
1410
+ "grad_norm": 1.2836827039718628,
1411
+ "learning_rate": 1.562930077691454e-05,
1412
+ "loss": 0.014,
1413
+ "step": 1970
1414
+ },
1415
+ {
1416
+ "epoch": 2.1952301719356626,
1417
+ "grad_norm": 0.6213756203651428,
1418
+ "learning_rate": 1.560710321864595e-05,
1419
+ "loss": 0.0136,
1420
+ "step": 1980
1421
+ },
1422
+ {
1423
+ "epoch": 2.206322795341098,
1424
+ "grad_norm": 0.48122018575668335,
1425
+ "learning_rate": 1.558490566037736e-05,
1426
+ "loss": 0.0147,
1427
+ "step": 1990
1428
+ },
1429
+ {
1430
+ "epoch": 2.2174154187465334,
1431
+ "grad_norm": 0.7161231637001038,
1432
+ "learning_rate": 1.556270810210877e-05,
1433
+ "loss": 0.0105,
1434
+ "step": 2000
1435
+ },
1436
+ {
1437
+ "epoch": 2.2285080421519687,
1438
+ "grad_norm": 0.6598195433616638,
1439
+ "learning_rate": 1.554051054384018e-05,
1440
+ "loss": 0.0172,
1441
+ "step": 2010
1442
+ },
1443
+ {
1444
+ "epoch": 2.239600665557404,
1445
+ "grad_norm": 0.6077088117599487,
1446
+ "learning_rate": 1.551831298557159e-05,
1447
+ "loss": 0.0144,
1448
+ "step": 2020
1449
+ },
1450
+ {
1451
+ "epoch": 2.2506932889628395,
1452
+ "grad_norm": 0.5622262954711914,
1453
+ "learning_rate": 1.5496115427303e-05,
1454
+ "loss": 0.0175,
1455
+ "step": 2030
1456
+ },
1457
+ {
1458
+ "epoch": 2.261785912368275,
1459
+ "grad_norm": 0.1281885802745819,
1460
+ "learning_rate": 1.547391786903441e-05,
1461
+ "loss": 0.018,
1462
+ "step": 2040
1463
+ },
1464
+ {
1465
+ "epoch": 2.2728785357737102,
1466
+ "grad_norm": 0.408607542514801,
1467
+ "learning_rate": 1.5451720310765815e-05,
1468
+ "loss": 0.0095,
1469
+ "step": 2050
1470
+ },
1471
+ {
1472
+ "epoch": 2.283971159179146,
1473
+ "grad_norm": 0.7786557674407959,
1474
+ "learning_rate": 1.5429522752497225e-05,
1475
+ "loss": 0.0163,
1476
+ "step": 2060
1477
+ },
1478
+ {
1479
+ "epoch": 2.2950637825845814,
1480
+ "grad_norm": 0.22975876927375793,
1481
+ "learning_rate": 1.5407325194228635e-05,
1482
+ "loss": 0.0193,
1483
+ "step": 2070
1484
+ },
1485
+ {
1486
+ "epoch": 2.306156405990017,
1487
+ "grad_norm": 0.4313502311706543,
1488
+ "learning_rate": 1.5385127635960048e-05,
1489
+ "loss": 0.0186,
1490
+ "step": 2080
1491
+ },
1492
+ {
1493
+ "epoch": 2.317249029395452,
1494
+ "grad_norm": 0.794448733329773,
1495
+ "learning_rate": 1.5362930077691455e-05,
1496
+ "loss": 0.0159,
1497
+ "step": 2090
1498
+ },
1499
+ {
1500
+ "epoch": 2.3283416528008876,
1501
+ "grad_norm": 0.4191996157169342,
1502
+ "learning_rate": 1.5340732519422865e-05,
1503
+ "loss": 0.0208,
1504
+ "step": 2100
1505
+ },
1506
+ {
1507
+ "epoch": 2.339434276206323,
1508
+ "grad_norm": 0.5219516754150391,
1509
+ "learning_rate": 1.5318534961154274e-05,
1510
+ "loss": 0.0136,
1511
+ "step": 2110
1512
+ },
1513
+ {
1514
+ "epoch": 2.3505268996117583,
1515
+ "grad_norm": 1.1869897842407227,
1516
+ "learning_rate": 1.5296337402885684e-05,
1517
+ "loss": 0.0177,
1518
+ "step": 2120
1519
+ },
1520
+ {
1521
+ "epoch": 2.3616195230171937,
1522
+ "grad_norm": 0.12048923969268799,
1523
+ "learning_rate": 1.5274139844617094e-05,
1524
+ "loss": 0.0129,
1525
+ "step": 2130
1526
+ },
1527
+ {
1528
+ "epoch": 2.372712146422629,
1529
+ "grad_norm": 0.5213373899459839,
1530
+ "learning_rate": 1.5251942286348502e-05,
1531
+ "loss": 0.0127,
1532
+ "step": 2140
1533
+ },
1534
+ {
1535
+ "epoch": 2.3838047698280644,
1536
+ "grad_norm": 0.43704670667648315,
1537
+ "learning_rate": 1.5229744728079912e-05,
1538
+ "loss": 0.0154,
1539
+ "step": 2150
1540
+ },
1541
+ {
1542
+ "epoch": 2.3948973932335,
1543
+ "grad_norm": 0.567150890827179,
1544
+ "learning_rate": 1.5207547169811324e-05,
1545
+ "loss": 0.0117,
1546
+ "step": 2160
1547
+ },
1548
+ {
1549
+ "epoch": 2.405990016638935,
1550
+ "grad_norm": 0.571408212184906,
1551
+ "learning_rate": 1.5185349611542732e-05,
1552
+ "loss": 0.0152,
1553
+ "step": 2170
1554
+ },
1555
+ {
1556
+ "epoch": 2.4170826400443706,
1557
+ "grad_norm": 0.524834394454956,
1558
+ "learning_rate": 1.5163152053274142e-05,
1559
+ "loss": 0.017,
1560
+ "step": 2180
1561
+ },
1562
+ {
1563
+ "epoch": 2.428175263449806,
1564
+ "grad_norm": 0.5026165246963501,
1565
+ "learning_rate": 1.5140954495005552e-05,
1566
+ "loss": 0.0193,
1567
+ "step": 2190
1568
+ },
1569
+ {
1570
+ "epoch": 2.4392678868552413,
1571
+ "grad_norm": 0.3944782614707947,
1572
+ "learning_rate": 1.511875693673696e-05,
1573
+ "loss": 0.0159,
1574
+ "step": 2200
1575
+ },
1576
+ {
1577
+ "epoch": 2.4503605102606767,
1578
+ "grad_norm": 0.6460635662078857,
1579
+ "learning_rate": 1.509655937846837e-05,
1580
+ "loss": 0.0139,
1581
+ "step": 2210
1582
+ },
1583
+ {
1584
+ "epoch": 2.461453133666112,
1585
+ "grad_norm": 0.5573295950889587,
1586
+ "learning_rate": 1.507436182019978e-05,
1587
+ "loss": 0.0183,
1588
+ "step": 2220
1589
+ },
1590
+ {
1591
+ "epoch": 2.4725457570715474,
1592
+ "grad_norm": 0.5001458525657654,
1593
+ "learning_rate": 1.5052164261931188e-05,
1594
+ "loss": 0.0167,
1595
+ "step": 2230
1596
+ },
1597
+ {
1598
+ "epoch": 2.483638380476983,
1599
+ "grad_norm": 1.0214160680770874,
1600
+ "learning_rate": 1.5029966703662598e-05,
1601
+ "loss": 0.0182,
1602
+ "step": 2240
1603
+ },
1604
+ {
1605
+ "epoch": 2.494731003882418,
1606
+ "grad_norm": 0.5756349563598633,
1607
+ "learning_rate": 1.5007769145394009e-05,
1608
+ "loss": 0.0164,
1609
+ "step": 2250
1610
+ },
1611
+ {
1612
+ "epoch": 2.5058236272878536,
1613
+ "grad_norm": 0.771986722946167,
1614
+ "learning_rate": 1.4985571587125419e-05,
1615
+ "loss": 0.0183,
1616
+ "step": 2260
1617
+ },
1618
+ {
1619
+ "epoch": 2.516916250693289,
1620
+ "grad_norm": 0.5560303926467896,
1621
+ "learning_rate": 1.4963374028856827e-05,
1622
+ "loss": 0.0178,
1623
+ "step": 2270
1624
+ },
1625
+ {
1626
+ "epoch": 2.5280088740987243,
1627
+ "grad_norm": 0.8550804853439331,
1628
+ "learning_rate": 1.4941176470588237e-05,
1629
+ "loss": 0.0118,
1630
+ "step": 2280
1631
+ },
1632
+ {
1633
+ "epoch": 2.5391014975041597,
1634
+ "grad_norm": 0.8129355907440186,
1635
+ "learning_rate": 1.4918978912319645e-05,
1636
+ "loss": 0.0126,
1637
+ "step": 2290
1638
+ },
1639
+ {
1640
+ "epoch": 2.550194120909595,
1641
+ "grad_norm": 1.176377773284912,
1642
+ "learning_rate": 1.4896781354051055e-05,
1643
+ "loss": 0.0126,
1644
+ "step": 2300
1645
+ },
1646
+ {
1647
+ "epoch": 2.5612867443150305,
1648
+ "grad_norm": 0.1419341266155243,
1649
+ "learning_rate": 1.4874583795782465e-05,
1650
+ "loss": 0.0196,
1651
+ "step": 2310
1652
+ },
1653
+ {
1654
+ "epoch": 2.572379367720466,
1655
+ "grad_norm": 1.4311761856079102,
1656
+ "learning_rate": 1.4852386237513873e-05,
1657
+ "loss": 0.0143,
1658
+ "step": 2320
1659
+ },
1660
+ {
1661
+ "epoch": 2.583471991125901,
1662
+ "grad_norm": 0.9924690127372742,
1663
+ "learning_rate": 1.4830188679245283e-05,
1664
+ "loss": 0.0198,
1665
+ "step": 2330
1666
+ },
1667
+ {
1668
+ "epoch": 2.5945646145313366,
1669
+ "grad_norm": 0.6496040225028992,
1670
+ "learning_rate": 1.4807991120976695e-05,
1671
+ "loss": 0.014,
1672
+ "step": 2340
1673
+ },
1674
+ {
1675
+ "epoch": 2.605657237936772,
1676
+ "grad_norm": 0.2335842400789261,
1677
+ "learning_rate": 1.4785793562708104e-05,
1678
+ "loss": 0.0128,
1679
+ "step": 2350
1680
+ },
1681
+ {
1682
+ "epoch": 2.6167498613422073,
1683
+ "grad_norm": 0.6549142003059387,
1684
+ "learning_rate": 1.4763596004439513e-05,
1685
+ "loss": 0.019,
1686
+ "step": 2360
1687
+ },
1688
+ {
1689
+ "epoch": 2.6278424847476427,
1690
+ "grad_norm": 0.3264187276363373,
1691
+ "learning_rate": 1.4741398446170922e-05,
1692
+ "loss": 0.0168,
1693
+ "step": 2370
1694
+ },
1695
+ {
1696
+ "epoch": 2.638935108153078,
1697
+ "grad_norm": 0.4175679683685303,
1698
+ "learning_rate": 1.4719200887902332e-05,
1699
+ "loss": 0.0122,
1700
+ "step": 2380
1701
+ },
1702
+ {
1703
+ "epoch": 2.6500277315585135,
1704
+ "grad_norm": 0.2968366742134094,
1705
+ "learning_rate": 1.469700332963374e-05,
1706
+ "loss": 0.0189,
1707
+ "step": 2390
1708
+ },
1709
+ {
1710
+ "epoch": 2.661120354963949,
1711
+ "grad_norm": 1.5489327907562256,
1712
+ "learning_rate": 1.467480577136515e-05,
1713
+ "loss": 0.0158,
1714
+ "step": 2400
1715
+ },
1716
+ {
1717
+ "epoch": 2.672212978369384,
1718
+ "grad_norm": 0.8017039895057678,
1719
+ "learning_rate": 1.465260821309656e-05,
1720
+ "loss": 0.0178,
1721
+ "step": 2410
1722
+ },
1723
+ {
1724
+ "epoch": 2.6833056017748196,
1725
+ "grad_norm": 0.6136099100112915,
1726
+ "learning_rate": 1.4630410654827972e-05,
1727
+ "loss": 0.0164,
1728
+ "step": 2420
1729
+ },
1730
+ {
1731
+ "epoch": 2.6943982251802554,
1732
+ "grad_norm": 0.5461011528968811,
1733
+ "learning_rate": 1.460821309655938e-05,
1734
+ "loss": 0.0168,
1735
+ "step": 2430
1736
+ },
1737
+ {
1738
+ "epoch": 2.7054908485856908,
1739
+ "grad_norm": 0.2680525779724121,
1740
+ "learning_rate": 1.458601553829079e-05,
1741
+ "loss": 0.0138,
1742
+ "step": 2440
1743
+ },
1744
+ {
1745
+ "epoch": 2.716583471991126,
1746
+ "grad_norm": 0.6884289383888245,
1747
+ "learning_rate": 1.45638179800222e-05,
1748
+ "loss": 0.0145,
1749
+ "step": 2450
1750
+ },
1751
+ {
1752
+ "epoch": 2.7276760953965615,
1753
+ "grad_norm": 1.226928472518921,
1754
+ "learning_rate": 1.4541620421753608e-05,
1755
+ "loss": 0.0145,
1756
+ "step": 2460
1757
+ },
1758
+ {
1759
+ "epoch": 2.738768718801997,
1760
+ "grad_norm": 0.447221040725708,
1761
+ "learning_rate": 1.4519422863485018e-05,
1762
+ "loss": 0.0168,
1763
+ "step": 2470
1764
+ },
1765
+ {
1766
+ "epoch": 2.7498613422074323,
1767
+ "grad_norm": 0.47167718410491943,
1768
+ "learning_rate": 1.4497225305216428e-05,
1769
+ "loss": 0.0124,
1770
+ "step": 2480
1771
+ },
1772
+ {
1773
+ "epoch": 2.7609539656128677,
1774
+ "grad_norm": 1.326282262802124,
1775
+ "learning_rate": 1.4475027746947836e-05,
1776
+ "loss": 0.0183,
1777
+ "step": 2490
1778
+ },
1779
+ {
1780
+ "epoch": 2.772046589018303,
1781
+ "grad_norm": 0.9036744236946106,
1782
+ "learning_rate": 1.4452830188679246e-05,
1783
+ "loss": 0.023,
1784
+ "step": 2500
1785
+ },
1786
+ {
1787
+ "epoch": 2.7831392124237384,
1788
+ "grad_norm": 0.30263999104499817,
1789
+ "learning_rate": 1.4430632630410657e-05,
1790
+ "loss": 0.0163,
1791
+ "step": 2510
1792
+ },
1793
+ {
1794
+ "epoch": 2.794231835829174,
1795
+ "grad_norm": 0.7316049933433533,
1796
+ "learning_rate": 1.4408435072142067e-05,
1797
+ "loss": 0.0156,
1798
+ "step": 2520
1799
+ },
1800
+ {
1801
+ "epoch": 2.805324459234609,
1802
+ "grad_norm": 0.8268038630485535,
1803
+ "learning_rate": 1.4386237513873475e-05,
1804
+ "loss": 0.0169,
1805
+ "step": 2530
1806
+ },
1807
+ {
1808
+ "epoch": 2.8164170826400445,
1809
+ "grad_norm": 0.633816659450531,
1810
+ "learning_rate": 1.4364039955604885e-05,
1811
+ "loss": 0.0147,
1812
+ "step": 2540
1813
+ },
1814
+ {
1815
+ "epoch": 2.82750970604548,
1816
+ "grad_norm": 0.43913397192955017,
1817
+ "learning_rate": 1.4341842397336295e-05,
1818
+ "loss": 0.0122,
1819
+ "step": 2550
1820
+ },
1821
+ {
1822
+ "epoch": 2.8386023294509153,
1823
+ "grad_norm": 0.37594836950302124,
1824
+ "learning_rate": 1.4319644839067703e-05,
1825
+ "loss": 0.0149,
1826
+ "step": 2560
1827
+ },
1828
+ {
1829
+ "epoch": 2.8496949528563507,
1830
+ "grad_norm": 0.5090307593345642,
1831
+ "learning_rate": 1.4297447280799113e-05,
1832
+ "loss": 0.015,
1833
+ "step": 2570
1834
+ },
1835
+ {
1836
+ "epoch": 2.860787576261786,
1837
+ "grad_norm": 0.47723984718322754,
1838
+ "learning_rate": 1.4275249722530521e-05,
1839
+ "loss": 0.0166,
1840
+ "step": 2580
1841
+ },
1842
+ {
1843
+ "epoch": 2.8718801996672214,
1844
+ "grad_norm": 1.2512151002883911,
1845
+ "learning_rate": 1.4253052164261931e-05,
1846
+ "loss": 0.0136,
1847
+ "step": 2590
1848
+ },
1849
+ {
1850
+ "epoch": 2.882972823072657,
1851
+ "grad_norm": 0.676337718963623,
1852
+ "learning_rate": 1.4230854605993343e-05,
1853
+ "loss": 0.0152,
1854
+ "step": 2600
1855
+ },
1856
+ {
1857
+ "epoch": 2.894065446478092,
1858
+ "grad_norm": 0.7379328012466431,
1859
+ "learning_rate": 1.4208657047724752e-05,
1860
+ "loss": 0.0182,
1861
+ "step": 2610
1862
+ },
1863
+ {
1864
+ "epoch": 2.9051580698835275,
1865
+ "grad_norm": 0.8266171813011169,
1866
+ "learning_rate": 1.418645948945616e-05,
1867
+ "loss": 0.0207,
1868
+ "step": 2620
1869
+ },
1870
+ {
1871
+ "epoch": 2.916250693288963,
1872
+ "grad_norm": 0.9925495982170105,
1873
+ "learning_rate": 1.416426193118757e-05,
1874
+ "loss": 0.0157,
1875
+ "step": 2630
1876
+ },
1877
+ {
1878
+ "epoch": 2.9273433166943983,
1879
+ "grad_norm": 1.1120530366897583,
1880
+ "learning_rate": 1.414206437291898e-05,
1881
+ "loss": 0.0164,
1882
+ "step": 2640
1883
+ },
1884
+ {
1885
+ "epoch": 2.9384359400998337,
1886
+ "grad_norm": 0.9687446355819702,
1887
+ "learning_rate": 1.4119866814650389e-05,
1888
+ "loss": 0.0211,
1889
+ "step": 2650
1890
+ },
1891
+ {
1892
+ "epoch": 2.949528563505269,
1893
+ "grad_norm": 0.19811566174030304,
1894
+ "learning_rate": 1.4097669256381798e-05,
1895
+ "loss": 0.0181,
1896
+ "step": 2660
1897
+ },
1898
+ {
1899
+ "epoch": 2.9606211869107044,
1900
+ "grad_norm": 0.9253482818603516,
1901
+ "learning_rate": 1.4075471698113208e-05,
1902
+ "loss": 0.0158,
1903
+ "step": 2670
1904
+ },
1905
+ {
1906
+ "epoch": 2.97171381031614,
1907
+ "grad_norm": 0.677994430065155,
1908
+ "learning_rate": 1.405327413984462e-05,
1909
+ "loss": 0.024,
1910
+ "step": 2680
1911
+ },
1912
+ {
1913
+ "epoch": 2.982806433721575,
1914
+ "grad_norm": 0.3959580361843109,
1915
+ "learning_rate": 1.4031076581576028e-05,
1916
+ "loss": 0.009,
1917
+ "step": 2690
1918
+ },
1919
+ {
1920
+ "epoch": 2.9938990571270105,
1921
+ "grad_norm": 1.0569937229156494,
1922
+ "learning_rate": 1.4008879023307438e-05,
1923
+ "loss": 0.0187,
1924
+ "step": 2700
1925
+ },
1926
+ {
1927
+ "epoch": 3.0,
1928
+ "eval_accuracy": 0.8888888888888888,
1929
+ "eval_f1": 0.7171109200343938,
1930
+ "eval_loss": 0.3638182580471039,
1931
+ "eval_precision": 0.695,
1932
+ "eval_recall": 0.7406749555950266,
1933
+ "eval_runtime": 2.7252,
1934
+ "eval_samples_per_second": 362.177,
1935
+ "eval_steps_per_second": 11.375,
1936
+ "step": 2706
1937
+ }
1938
+ ],
1939
+ "logging_steps": 10,
1940
+ "max_steps": 9010,
1941
+ "num_input_tokens_seen": 0,
1942
+ "num_train_epochs": 10,
1943
+ "save_steps": 500,
1944
+ "stateful_callbacks": {
1945
+ "TrainerControl": {
1946
+ "args": {
1947
+ "should_epoch_stop": false,
1948
+ "should_evaluate": false,
1949
+ "should_log": false,
1950
+ "should_save": true,
1951
+ "should_training_stop": false
1952
+ },
1953
+ "attributes": {}
1954
+ }
1955
+ },
1956
+ "total_flos": 2.884237018137645e+16,
1957
+ "train_batch_size": 32,
1958
+ "trial_name": null,
1959
+ "trial_params": null
1960
+ }
checkpoint-2706/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e64fe8f815fa80a7291289cde9af3ebb27f62206c40a93e8adf97661cd289de3
3
+ size 5368
checkpoint-3608/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
checkpoint-3608/config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaV2ForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "hidden_act": "gelu",
7
+ "hidden_dropout_prob": 0.1,
8
+ "hidden_size": 1024,
9
+ "id2label": {
10
+ "0": 0,
11
+ "1": 1,
12
+ "2": 2
13
+ },
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 4096,
16
+ "label2id": {
17
+ "0": 0,
18
+ "1": 1,
19
+ "2": 2
20
+ },
21
+ "layer_norm_eps": 1e-07,
22
+ "legacy": true,
23
+ "max_position_embeddings": 512,
24
+ "max_relative_positions": -1,
25
+ "model_type": "deberta-v2",
26
+ "norm_rel_ebd": "layer_norm",
27
+ "num_attention_heads": 16,
28
+ "num_hidden_layers": 24,
29
+ "pad_token_id": 0,
30
+ "pooler_dropout": 0,
31
+ "pooler_hidden_act": "gelu",
32
+ "pooler_hidden_size": 1024,
33
+ "pos_att_type": [
34
+ "p2c",
35
+ "c2p"
36
+ ],
37
+ "position_biased_input": false,
38
+ "position_buckets": 256,
39
+ "problem_type": "multi_label_classification",
40
+ "relative_attention": true,
41
+ "share_att_key": true,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.51.3",
44
+ "type_vocab_size": 0,
45
+ "vocab_size": 128100
46
+ }
checkpoint-3608/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0891e2cede52fd93e5b1a8218f5df55279ad9923be6849789b1ab2e14c923292
3
+ size 1740308548
checkpoint-3608/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec792cb9e986f379e3975a71e9d8a047750d20b5162cc6e9bfa2aefca17b0410
3
+ size 3480848432
checkpoint-3608/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca27814685a3b867c95e0a3f7bcf8ad712492068d0de996d1862d3beadc0d5b6
3
+ size 14244
checkpoint-3608/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:471cf39ce9e42c624e5374d96c5e15174a9452cbaefa1d887095f2bc277f5261
3
+ size 1064
checkpoint-3608/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[SEP]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": {
9
+ "content": "[UNK]",
10
+ "lstrip": false,
11
+ "normalized": true,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ }
15
+ }
checkpoint-3608/spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
3
+ size 2464616
checkpoint-3608/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-3608/tokenizer_config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[CLS]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[SEP]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[UNK]",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "128000": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "[CLS]",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "[CLS]",
47
+ "do_lower_case": false,
48
+ "eos_token": "[SEP]",
49
+ "extra_special_tokens": {},
50
+ "mask_token": "[MASK]",
51
+ "model_max_length": 1000000000000000019884624838656,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "sp_model_kwargs": {},
55
+ "split_by_punct": false,
56
+ "tokenizer_class": "DebertaV2Tokenizer",
57
+ "unk_token": "[UNK]",
58
+ "vocab_type": "spm"
59
+ }
checkpoint-3608/trainer_state.json ADDED
@@ -0,0 +1,2602 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 4.0,
6
+ "eval_steps": 500,
7
+ "global_step": 3608,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.011092623405435386,
14
+ "grad_norm": 2.009918689727783,
15
+ "learning_rate": 1.998002219755827e-05,
16
+ "loss": 0.5022,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.022185246810870772,
21
+ "grad_norm": 0.49915584921836853,
22
+ "learning_rate": 1.995782463928968e-05,
23
+ "loss": 0.1943,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.033277870216306155,
28
+ "grad_norm": 0.47061625123023987,
29
+ "learning_rate": 1.993562708102109e-05,
30
+ "loss": 0.1436,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.044370493621741544,
35
+ "grad_norm": 0.2260124832391739,
36
+ "learning_rate": 1.99134295227525e-05,
37
+ "loss": 0.1106,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.05546311702717693,
42
+ "grad_norm": 0.31766048073768616,
43
+ "learning_rate": 1.989123196448391e-05,
44
+ "loss": 0.0994,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.06655574043261231,
49
+ "grad_norm": 0.2788391709327698,
50
+ "learning_rate": 1.9869034406215316e-05,
51
+ "loss": 0.111,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0776483638380477,
56
+ "grad_norm": 0.24376386404037476,
57
+ "learning_rate": 1.9846836847946726e-05,
58
+ "loss": 0.0849,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.08874098724348309,
63
+ "grad_norm": 0.1789788454771042,
64
+ "learning_rate": 1.982463928967814e-05,
65
+ "loss": 0.0971,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.09983361064891846,
70
+ "grad_norm": 0.29698437452316284,
71
+ "learning_rate": 1.9802441731409546e-05,
72
+ "loss": 0.0856,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.11092623405435385,
77
+ "grad_norm": 0.2553412616252899,
78
+ "learning_rate": 1.9780244173140956e-05,
79
+ "loss": 0.0976,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.12201885745978924,
84
+ "grad_norm": 0.20608791708946228,
85
+ "learning_rate": 1.9758046614872365e-05,
86
+ "loss": 0.0947,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.13311148086522462,
91
+ "grad_norm": 0.23551669716835022,
92
+ "learning_rate": 1.9735849056603775e-05,
93
+ "loss": 0.0965,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.14420410427066002,
98
+ "grad_norm": 0.3207148611545563,
99
+ "learning_rate": 1.9713651498335185e-05,
100
+ "loss": 0.0931,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.1552967276760954,
105
+ "grad_norm": 0.34813177585601807,
106
+ "learning_rate": 1.9691453940066595e-05,
107
+ "loss": 0.1,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.16638935108153077,
112
+ "grad_norm": 0.41501104831695557,
113
+ "learning_rate": 1.9669256381798e-05,
114
+ "loss": 0.0922,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.17748197448696618,
119
+ "grad_norm": 0.24550506472587585,
120
+ "learning_rate": 1.964705882352941e-05,
121
+ "loss": 0.0857,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.18857459789240155,
126
+ "grad_norm": 0.4205056130886078,
127
+ "learning_rate": 1.9624861265260825e-05,
128
+ "loss": 0.0831,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.19966722129783693,
133
+ "grad_norm": 0.6002993583679199,
134
+ "learning_rate": 1.9602663706992235e-05,
135
+ "loss": 0.0792,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.21075984470327233,
140
+ "grad_norm": 0.27535638213157654,
141
+ "learning_rate": 1.958046614872364e-05,
142
+ "loss": 0.0786,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.2218524681087077,
147
+ "grad_norm": 0.41602805256843567,
148
+ "learning_rate": 1.955826859045505e-05,
149
+ "loss": 0.0789,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.23294509151414308,
154
+ "grad_norm": 0.41827332973480225,
155
+ "learning_rate": 1.953607103218646e-05,
156
+ "loss": 0.0663,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.24403771491957849,
161
+ "grad_norm": 0.47976112365722656,
162
+ "learning_rate": 1.951387347391787e-05,
163
+ "loss": 0.0727,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.25513033832501386,
168
+ "grad_norm": 0.7771281599998474,
169
+ "learning_rate": 1.949167591564928e-05,
170
+ "loss": 0.0776,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.26622296173044924,
175
+ "grad_norm": 1.0573914051055908,
176
+ "learning_rate": 1.946947835738069e-05,
177
+ "loss": 0.0828,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.2773155851358846,
182
+ "grad_norm": 0.5186040997505188,
183
+ "learning_rate": 1.9447280799112097e-05,
184
+ "loss": 0.0767,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.28840820854132004,
189
+ "grad_norm": 0.5594862699508667,
190
+ "learning_rate": 1.942508324084351e-05,
191
+ "loss": 0.0699,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.2995008319467554,
196
+ "grad_norm": 0.5147818922996521,
197
+ "learning_rate": 1.940288568257492e-05,
198
+ "loss": 0.0591,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.3105934553521908,
203
+ "grad_norm": 0.5510437488555908,
204
+ "learning_rate": 1.938068812430633e-05,
205
+ "loss": 0.0761,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.32168607875762617,
210
+ "grad_norm": 0.4817625880241394,
211
+ "learning_rate": 1.9358490566037736e-05,
212
+ "loss": 0.0688,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.33277870216306155,
217
+ "grad_norm": 0.504362940788269,
218
+ "learning_rate": 1.9336293007769146e-05,
219
+ "loss": 0.0691,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.343871325568497,
224
+ "grad_norm": 0.5732384920120239,
225
+ "learning_rate": 1.9314095449500556e-05,
226
+ "loss": 0.057,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.35496394897393235,
231
+ "grad_norm": 0.5892298221588135,
232
+ "learning_rate": 1.9291897891231966e-05,
233
+ "loss": 0.059,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.36605657237936773,
238
+ "grad_norm": 0.3680512309074402,
239
+ "learning_rate": 1.9269700332963376e-05,
240
+ "loss": 0.0706,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.3771491957848031,
245
+ "grad_norm": 0.41971662640571594,
246
+ "learning_rate": 1.9247502774694786e-05,
247
+ "loss": 0.053,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.3882418191902385,
252
+ "grad_norm": 0.37261390686035156,
253
+ "learning_rate": 1.9225305216426195e-05,
254
+ "loss": 0.0632,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.39933444259567386,
259
+ "grad_norm": 0.48256734013557434,
260
+ "learning_rate": 1.9203107658157605e-05,
261
+ "loss": 0.0651,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.4104270660011093,
266
+ "grad_norm": 0.7914339303970337,
267
+ "learning_rate": 1.9180910099889015e-05,
268
+ "loss": 0.0659,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.42151968940654466,
273
+ "grad_norm": 0.6772429347038269,
274
+ "learning_rate": 1.915871254162042e-05,
275
+ "loss": 0.0655,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.43261231281198004,
280
+ "grad_norm": 0.5621687173843384,
281
+ "learning_rate": 1.913651498335183e-05,
282
+ "loss": 0.0707,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.4437049362174154,
287
+ "grad_norm": 0.30515748262405396,
288
+ "learning_rate": 1.911431742508324e-05,
289
+ "loss": 0.0548,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.4547975596228508,
294
+ "grad_norm": 0.5506859421730042,
295
+ "learning_rate": 1.909211986681465e-05,
296
+ "loss": 0.0554,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.46589018302828616,
301
+ "grad_norm": 0.6387749314308167,
302
+ "learning_rate": 1.906992230854606e-05,
303
+ "loss": 0.0554,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.4769828064337216,
308
+ "grad_norm": 0.18500734865665436,
309
+ "learning_rate": 1.904772475027747e-05,
310
+ "loss": 0.0558,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.48807542983915697,
315
+ "grad_norm": 0.6815407276153564,
316
+ "learning_rate": 1.902552719200888e-05,
317
+ "loss": 0.0537,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.49916805324459235,
322
+ "grad_norm": 0.4826994836330414,
323
+ "learning_rate": 1.900332963374029e-05,
324
+ "loss": 0.0567,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.5102606766500277,
329
+ "grad_norm": 0.24618124961853027,
330
+ "learning_rate": 1.89811320754717e-05,
331
+ "loss": 0.0614,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.5213533000554631,
336
+ "grad_norm": 1.0037415027618408,
337
+ "learning_rate": 1.895893451720311e-05,
338
+ "loss": 0.0604,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.5324459234608985,
343
+ "grad_norm": 0.6000948548316956,
344
+ "learning_rate": 1.8936736958934517e-05,
345
+ "loss": 0.0683,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.5435385468663338,
350
+ "grad_norm": 0.3302474319934845,
351
+ "learning_rate": 1.8914539400665927e-05,
352
+ "loss": 0.0543,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.5546311702717692,
357
+ "grad_norm": 0.5560783743858337,
358
+ "learning_rate": 1.8892341842397337e-05,
359
+ "loss": 0.0569,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.5657237936772047,
364
+ "grad_norm": 0.8041097521781921,
365
+ "learning_rate": 1.8870144284128747e-05,
366
+ "loss": 0.0544,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.5768164170826401,
371
+ "grad_norm": 0.6846103668212891,
372
+ "learning_rate": 1.8847946725860156e-05,
373
+ "loss": 0.058,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.5879090404880755,
378
+ "grad_norm": 0.50434809923172,
379
+ "learning_rate": 1.8825749167591566e-05,
380
+ "loss": 0.0654,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.5990016638935108,
385
+ "grad_norm": 0.54362553358078,
386
+ "learning_rate": 1.8803551609322976e-05,
387
+ "loss": 0.0582,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.6100942872989462,
392
+ "grad_norm": 0.6166839599609375,
393
+ "learning_rate": 1.8781354051054386e-05,
394
+ "loss": 0.0672,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.6211869107043816,
399
+ "grad_norm": 0.4353054165840149,
400
+ "learning_rate": 1.8759156492785796e-05,
401
+ "loss": 0.0532,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.632279534109817,
406
+ "grad_norm": 0.6025580167770386,
407
+ "learning_rate": 1.8736958934517206e-05,
408
+ "loss": 0.0746,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.6433721575152523,
413
+ "grad_norm": 0.7259892225265503,
414
+ "learning_rate": 1.8714761376248612e-05,
415
+ "loss": 0.0642,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.6544647809206877,
420
+ "grad_norm": 0.4940318465232849,
421
+ "learning_rate": 1.8692563817980022e-05,
422
+ "loss": 0.0547,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.6655574043261231,
427
+ "grad_norm": 0.7005699872970581,
428
+ "learning_rate": 1.8670366259711435e-05,
429
+ "loss": 0.0522,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.6766500277315585,
434
+ "grad_norm": 0.4530707895755768,
435
+ "learning_rate": 1.8648168701442845e-05,
436
+ "loss": 0.054,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.687742651136994,
441
+ "grad_norm": 0.9097110629081726,
442
+ "learning_rate": 1.8625971143174252e-05,
443
+ "loss": 0.0622,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.6988352745424293,
448
+ "grad_norm": 0.5374599695205688,
449
+ "learning_rate": 1.860377358490566e-05,
450
+ "loss": 0.0522,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.7099278979478647,
455
+ "grad_norm": 0.3849945664405823,
456
+ "learning_rate": 1.858157602663707e-05,
457
+ "loss": 0.0571,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.7210205213533001,
462
+ "grad_norm": 0.5918008685112,
463
+ "learning_rate": 1.855937846836848e-05,
464
+ "loss": 0.0576,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.7321131447587355,
469
+ "grad_norm": 0.3229956030845642,
470
+ "learning_rate": 1.853718091009989e-05,
471
+ "loss": 0.0537,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.7432057681641708,
476
+ "grad_norm": 0.5264039039611816,
477
+ "learning_rate": 1.85149833518313e-05,
478
+ "loss": 0.06,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.7542983915696062,
483
+ "grad_norm": 0.36795660853385925,
484
+ "learning_rate": 1.8492785793562708e-05,
485
+ "loss": 0.0511,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.7653910149750416,
490
+ "grad_norm": 0.5905130505561829,
491
+ "learning_rate": 1.847058823529412e-05,
492
+ "loss": 0.0636,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.776483638380477,
497
+ "grad_norm": 0.36266571283340454,
498
+ "learning_rate": 1.844839067702553e-05,
499
+ "loss": 0.0598,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.7875762617859123,
504
+ "grad_norm": 0.4978592097759247,
505
+ "learning_rate": 1.8426193118756937e-05,
506
+ "loss": 0.0607,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.7986688851913477,
511
+ "grad_norm": 0.4635021686553955,
512
+ "learning_rate": 1.8403995560488347e-05,
513
+ "loss": 0.0543,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.8097615085967831,
518
+ "grad_norm": 0.44571858644485474,
519
+ "learning_rate": 1.8381798002219757e-05,
520
+ "loss": 0.0678,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.8208541320022186,
525
+ "grad_norm": 0.8265877366065979,
526
+ "learning_rate": 1.8359600443951167e-05,
527
+ "loss": 0.0552,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.831946755407654,
532
+ "grad_norm": 0.5776472091674805,
533
+ "learning_rate": 1.8337402885682577e-05,
534
+ "loss": 0.0505,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.8430393788130893,
539
+ "grad_norm": 0.299274742603302,
540
+ "learning_rate": 1.8315205327413986e-05,
541
+ "loss": 0.0619,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.8541320022185247,
546
+ "grad_norm": 0.587645947933197,
547
+ "learning_rate": 1.8293007769145393e-05,
548
+ "loss": 0.0575,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.8652246256239601,
553
+ "grad_norm": 0.39164137840270996,
554
+ "learning_rate": 1.8270810210876806e-05,
555
+ "loss": 0.0458,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.8763172490293955,
560
+ "grad_norm": 0.4663292169570923,
561
+ "learning_rate": 1.8248612652608216e-05,
562
+ "loss": 0.048,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.8874098724348308,
567
+ "grad_norm": 0.5804581642150879,
568
+ "learning_rate": 1.8226415094339626e-05,
569
+ "loss": 0.0557,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.8985024958402662,
574
+ "grad_norm": 0.4279440939426422,
575
+ "learning_rate": 1.8204217536071032e-05,
576
+ "loss": 0.051,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.9095951192457016,
581
+ "grad_norm": 0.5384302735328674,
582
+ "learning_rate": 1.8182019977802442e-05,
583
+ "loss": 0.0558,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.920687742651137,
588
+ "grad_norm": 0.5049973726272583,
589
+ "learning_rate": 1.8159822419533852e-05,
590
+ "loss": 0.047,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.9317803660565723,
595
+ "grad_norm": 0.8061177134513855,
596
+ "learning_rate": 1.8137624861265262e-05,
597
+ "loss": 0.0572,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.9428729894620078,
602
+ "grad_norm": 0.6056540012359619,
603
+ "learning_rate": 1.8115427302996672e-05,
604
+ "loss": 0.0539,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.9539656128674432,
609
+ "grad_norm": 0.3001384139060974,
610
+ "learning_rate": 1.8093229744728082e-05,
611
+ "loss": 0.053,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.9650582362728786,
616
+ "grad_norm": 0.6709749102592468,
617
+ "learning_rate": 1.807103218645949e-05,
618
+ "loss": 0.0526,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.9761508596783139,
623
+ "grad_norm": 0.8233507871627808,
624
+ "learning_rate": 1.80488346281909e-05,
625
+ "loss": 0.0621,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.9872434830837493,
630
+ "grad_norm": 0.5757150650024414,
631
+ "learning_rate": 1.802663706992231e-05,
632
+ "loss": 0.0546,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.9983361064891847,
637
+ "grad_norm": 0.6834889054298401,
638
+ "learning_rate": 1.800443951165372e-05,
639
+ "loss": 0.0462,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 1.0,
644
+ "eval_accuracy": 0.8720027017899359,
645
+ "eval_f1": 0.5678449258836944,
646
+ "eval_loss": 0.34414270520210266,
647
+ "eval_precision": 0.7929936305732485,
648
+ "eval_recall": 0.4422735346358792,
649
+ "eval_runtime": 2.727,
650
+ "eval_samples_per_second": 361.937,
651
+ "eval_steps_per_second": 11.368,
652
+ "step": 902
653
+ },
654
+ {
655
+ "epoch": 1.0088740987243483,
656
+ "grad_norm": 0.4411003887653351,
657
+ "learning_rate": 1.7982241953385128e-05,
658
+ "loss": 0.0312,
659
+ "step": 910
660
+ },
661
+ {
662
+ "epoch": 1.0199667221297837,
663
+ "grad_norm": 0.40500083565711975,
664
+ "learning_rate": 1.7960044395116538e-05,
665
+ "loss": 0.0345,
666
+ "step": 920
667
+ },
668
+ {
669
+ "epoch": 1.031059345535219,
670
+ "grad_norm": 0.9287449717521667,
671
+ "learning_rate": 1.7937846836847947e-05,
672
+ "loss": 0.0419,
673
+ "step": 930
674
+ },
675
+ {
676
+ "epoch": 1.0421519689406544,
677
+ "grad_norm": 0.7969145178794861,
678
+ "learning_rate": 1.7915649278579357e-05,
679
+ "loss": 0.042,
680
+ "step": 940
681
+ },
682
+ {
683
+ "epoch": 1.0532445923460898,
684
+ "grad_norm": 0.49280259013175964,
685
+ "learning_rate": 1.7893451720310767e-05,
686
+ "loss": 0.0427,
687
+ "step": 950
688
+ },
689
+ {
690
+ "epoch": 1.0643372157515252,
691
+ "grad_norm": 1.1647624969482422,
692
+ "learning_rate": 1.7871254162042177e-05,
693
+ "loss": 0.0341,
694
+ "step": 960
695
+ },
696
+ {
697
+ "epoch": 1.0754298391569606,
698
+ "grad_norm": 0.2908968925476074,
699
+ "learning_rate": 1.7849056603773587e-05,
700
+ "loss": 0.0389,
701
+ "step": 970
702
+ },
703
+ {
704
+ "epoch": 1.086522462562396,
705
+ "grad_norm": 0.521218478679657,
706
+ "learning_rate": 1.7826859045504997e-05,
707
+ "loss": 0.0396,
708
+ "step": 980
709
+ },
710
+ {
711
+ "epoch": 1.0976150859678313,
712
+ "grad_norm": 0.4156598150730133,
713
+ "learning_rate": 1.7804661487236407e-05,
714
+ "loss": 0.0359,
715
+ "step": 990
716
+ },
717
+ {
718
+ "epoch": 1.1087077093732667,
719
+ "grad_norm": 0.7260242104530334,
720
+ "learning_rate": 1.7782463928967813e-05,
721
+ "loss": 0.0384,
722
+ "step": 1000
723
+ },
724
+ {
725
+ "epoch": 1.119800332778702,
726
+ "grad_norm": 0.43658459186553955,
727
+ "learning_rate": 1.7760266370699223e-05,
728
+ "loss": 0.0383,
729
+ "step": 1010
730
+ },
731
+ {
732
+ "epoch": 1.1308929561841374,
733
+ "grad_norm": 0.3585101068019867,
734
+ "learning_rate": 1.7738068812430633e-05,
735
+ "loss": 0.0401,
736
+ "step": 1020
737
+ },
738
+ {
739
+ "epoch": 1.141985579589573,
740
+ "grad_norm": 0.5010389089584351,
741
+ "learning_rate": 1.7715871254162043e-05,
742
+ "loss": 0.0307,
743
+ "step": 1030
744
+ },
745
+ {
746
+ "epoch": 1.1530782029950084,
747
+ "grad_norm": 0.6415812969207764,
748
+ "learning_rate": 1.7693673695893453e-05,
749
+ "loss": 0.0423,
750
+ "step": 1040
751
+ },
752
+ {
753
+ "epoch": 1.1641708264004438,
754
+ "grad_norm": 0.6393259763717651,
755
+ "learning_rate": 1.7671476137624862e-05,
756
+ "loss": 0.0392,
757
+ "step": 1050
758
+ },
759
+ {
760
+ "epoch": 1.1752634498058792,
761
+ "grad_norm": 0.6339041590690613,
762
+ "learning_rate": 1.7649278579356272e-05,
763
+ "loss": 0.0506,
764
+ "step": 1060
765
+ },
766
+ {
767
+ "epoch": 1.1863560732113145,
768
+ "grad_norm": 0.6086763143539429,
769
+ "learning_rate": 1.7627081021087682e-05,
770
+ "loss": 0.0416,
771
+ "step": 1070
772
+ },
773
+ {
774
+ "epoch": 1.19744869661675,
775
+ "grad_norm": 0.6648682355880737,
776
+ "learning_rate": 1.7604883462819092e-05,
777
+ "loss": 0.0379,
778
+ "step": 1080
779
+ },
780
+ {
781
+ "epoch": 1.2085413200221853,
782
+ "grad_norm": 0.6018221378326416,
783
+ "learning_rate": 1.7582685904550502e-05,
784
+ "loss": 0.0354,
785
+ "step": 1090
786
+ },
787
+ {
788
+ "epoch": 1.2196339434276207,
789
+ "grad_norm": 1.2366654872894287,
790
+ "learning_rate": 1.756048834628191e-05,
791
+ "loss": 0.037,
792
+ "step": 1100
793
+ },
794
+ {
795
+ "epoch": 1.230726566833056,
796
+ "grad_norm": 0.5976310968399048,
797
+ "learning_rate": 1.7538290788013318e-05,
798
+ "loss": 0.0361,
799
+ "step": 1110
800
+ },
801
+ {
802
+ "epoch": 1.2418191902384914,
803
+ "grad_norm": 0.8290308117866516,
804
+ "learning_rate": 1.751609322974473e-05,
805
+ "loss": 0.0383,
806
+ "step": 1120
807
+ },
808
+ {
809
+ "epoch": 1.2529118136439268,
810
+ "grad_norm": 0.382548987865448,
811
+ "learning_rate": 1.749389567147614e-05,
812
+ "loss": 0.0449,
813
+ "step": 1130
814
+ },
815
+ {
816
+ "epoch": 1.2640044370493622,
817
+ "grad_norm": 0.5307976007461548,
818
+ "learning_rate": 1.7471698113207548e-05,
819
+ "loss": 0.0364,
820
+ "step": 1140
821
+ },
822
+ {
823
+ "epoch": 1.2750970604547975,
824
+ "grad_norm": 0.5508521795272827,
825
+ "learning_rate": 1.7449500554938958e-05,
826
+ "loss": 0.0312,
827
+ "step": 1150
828
+ },
829
+ {
830
+ "epoch": 1.286189683860233,
831
+ "grad_norm": 0.37057268619537354,
832
+ "learning_rate": 1.7427302996670368e-05,
833
+ "loss": 0.0397,
834
+ "step": 1160
835
+ },
836
+ {
837
+ "epoch": 1.2972823072656683,
838
+ "grad_norm": 0.4806898236274719,
839
+ "learning_rate": 1.7405105438401777e-05,
840
+ "loss": 0.0391,
841
+ "step": 1170
842
+ },
843
+ {
844
+ "epoch": 1.3083749306711037,
845
+ "grad_norm": 0.759772002696991,
846
+ "learning_rate": 1.7382907880133187e-05,
847
+ "loss": 0.0322,
848
+ "step": 1180
849
+ },
850
+ {
851
+ "epoch": 1.319467554076539,
852
+ "grad_norm": 0.39819085597991943,
853
+ "learning_rate": 1.7360710321864597e-05,
854
+ "loss": 0.0345,
855
+ "step": 1190
856
+ },
857
+ {
858
+ "epoch": 1.3305601774819744,
859
+ "grad_norm": 0.7733897566795349,
860
+ "learning_rate": 1.7338512763596004e-05,
861
+ "loss": 0.0424,
862
+ "step": 1200
863
+ },
864
+ {
865
+ "epoch": 1.3416528008874098,
866
+ "grad_norm": 0.7979075908660889,
867
+ "learning_rate": 1.7316315205327417e-05,
868
+ "loss": 0.0403,
869
+ "step": 1210
870
+ },
871
+ {
872
+ "epoch": 1.3527454242928454,
873
+ "grad_norm": 0.5658752918243408,
874
+ "learning_rate": 1.7294117647058827e-05,
875
+ "loss": 0.0276,
876
+ "step": 1220
877
+ },
878
+ {
879
+ "epoch": 1.3638380476982808,
880
+ "grad_norm": 0.25207844376564026,
881
+ "learning_rate": 1.7271920088790237e-05,
882
+ "loss": 0.0347,
883
+ "step": 1230
884
+ },
885
+ {
886
+ "epoch": 1.3749306711037161,
887
+ "grad_norm": 0.6338945031166077,
888
+ "learning_rate": 1.7249722530521643e-05,
889
+ "loss": 0.0416,
890
+ "step": 1240
891
+ },
892
+ {
893
+ "epoch": 1.3860232945091515,
894
+ "grad_norm": 0.8125913143157959,
895
+ "learning_rate": 1.7227524972253053e-05,
896
+ "loss": 0.0442,
897
+ "step": 1250
898
+ },
899
+ {
900
+ "epoch": 1.397115917914587,
901
+ "grad_norm": 0.5067325234413147,
902
+ "learning_rate": 1.7205327413984463e-05,
903
+ "loss": 0.0358,
904
+ "step": 1260
905
+ },
906
+ {
907
+ "epoch": 1.4082085413200223,
908
+ "grad_norm": 0.40132638812065125,
909
+ "learning_rate": 1.7183129855715873e-05,
910
+ "loss": 0.0303,
911
+ "step": 1270
912
+ },
913
+ {
914
+ "epoch": 1.4193011647254576,
915
+ "grad_norm": 0.6337258815765381,
916
+ "learning_rate": 1.7160932297447283e-05,
917
+ "loss": 0.0471,
918
+ "step": 1280
919
+ },
920
+ {
921
+ "epoch": 1.430393788130893,
922
+ "grad_norm": 0.43456393480300903,
923
+ "learning_rate": 1.713873473917869e-05,
924
+ "loss": 0.0354,
925
+ "step": 1290
926
+ },
927
+ {
928
+ "epoch": 1.4414864115363284,
929
+ "grad_norm": 0.3291069269180298,
930
+ "learning_rate": 1.7116537180910102e-05,
931
+ "loss": 0.0312,
932
+ "step": 1300
933
+ },
934
+ {
935
+ "epoch": 1.4525790349417638,
936
+ "grad_norm": 0.2960388660430908,
937
+ "learning_rate": 1.7094339622641512e-05,
938
+ "loss": 0.0362,
939
+ "step": 1310
940
+ },
941
+ {
942
+ "epoch": 1.4636716583471991,
943
+ "grad_norm": 0.820751428604126,
944
+ "learning_rate": 1.7072142064372922e-05,
945
+ "loss": 0.0396,
946
+ "step": 1320
947
+ },
948
+ {
949
+ "epoch": 1.4747642817526345,
950
+ "grad_norm": 0.7335907816886902,
951
+ "learning_rate": 1.704994450610433e-05,
952
+ "loss": 0.0386,
953
+ "step": 1330
954
+ },
955
+ {
956
+ "epoch": 1.48585690515807,
957
+ "grad_norm": 0.40592002868652344,
958
+ "learning_rate": 1.702774694783574e-05,
959
+ "loss": 0.0358,
960
+ "step": 1340
961
+ },
962
+ {
963
+ "epoch": 1.4969495285635053,
964
+ "grad_norm": 0.6194770336151123,
965
+ "learning_rate": 1.7005549389567148e-05,
966
+ "loss": 0.0361,
967
+ "step": 1350
968
+ },
969
+ {
970
+ "epoch": 1.5080421519689406,
971
+ "grad_norm": 0.44590774178504944,
972
+ "learning_rate": 1.6983351831298558e-05,
973
+ "loss": 0.033,
974
+ "step": 1360
975
+ },
976
+ {
977
+ "epoch": 1.519134775374376,
978
+ "grad_norm": 0.5857370495796204,
979
+ "learning_rate": 1.6961154273029968e-05,
980
+ "loss": 0.0423,
981
+ "step": 1370
982
+ },
983
+ {
984
+ "epoch": 1.5302273987798114,
985
+ "grad_norm": 0.6809953451156616,
986
+ "learning_rate": 1.6938956714761378e-05,
987
+ "loss": 0.0345,
988
+ "step": 1380
989
+ },
990
+ {
991
+ "epoch": 1.5413200221852468,
992
+ "grad_norm": 0.7102778553962708,
993
+ "learning_rate": 1.6916759156492788e-05,
994
+ "loss": 0.0372,
995
+ "step": 1390
996
+ },
997
+ {
998
+ "epoch": 1.5524126455906821,
999
+ "grad_norm": 0.31781554222106934,
1000
+ "learning_rate": 1.6894561598224198e-05,
1001
+ "loss": 0.032,
1002
+ "step": 1400
1003
+ },
1004
+ {
1005
+ "epoch": 1.5635052689961175,
1006
+ "grad_norm": 0.4304943382740021,
1007
+ "learning_rate": 1.6872364039955607e-05,
1008
+ "loss": 0.0375,
1009
+ "step": 1410
1010
+ },
1011
+ {
1012
+ "epoch": 1.574597892401553,
1013
+ "grad_norm": 0.918550968170166,
1014
+ "learning_rate": 1.6850166481687017e-05,
1015
+ "loss": 0.0416,
1016
+ "step": 1420
1017
+ },
1018
+ {
1019
+ "epoch": 1.5856905158069883,
1020
+ "grad_norm": 0.722892701625824,
1021
+ "learning_rate": 1.6827968923418424e-05,
1022
+ "loss": 0.0359,
1023
+ "step": 1430
1024
+ },
1025
+ {
1026
+ "epoch": 1.5967831392124237,
1027
+ "grad_norm": 0.508703351020813,
1028
+ "learning_rate": 1.6805771365149834e-05,
1029
+ "loss": 0.0402,
1030
+ "step": 1440
1031
+ },
1032
+ {
1033
+ "epoch": 1.607875762617859,
1034
+ "grad_norm": 0.4725389778614044,
1035
+ "learning_rate": 1.6783573806881244e-05,
1036
+ "loss": 0.0335,
1037
+ "step": 1450
1038
+ },
1039
+ {
1040
+ "epoch": 1.6189683860232944,
1041
+ "grad_norm": 0.4699971079826355,
1042
+ "learning_rate": 1.6761376248612653e-05,
1043
+ "loss": 0.0277,
1044
+ "step": 1460
1045
+ },
1046
+ {
1047
+ "epoch": 1.6300610094287298,
1048
+ "grad_norm": 0.775764524936676,
1049
+ "learning_rate": 1.6739178690344063e-05,
1050
+ "loss": 0.0333,
1051
+ "step": 1470
1052
+ },
1053
+ {
1054
+ "epoch": 1.6411536328341652,
1055
+ "grad_norm": 0.6115106344223022,
1056
+ "learning_rate": 1.6716981132075473e-05,
1057
+ "loss": 0.0301,
1058
+ "step": 1480
1059
+ },
1060
+ {
1061
+ "epoch": 1.6522462562396005,
1062
+ "grad_norm": 0.743077278137207,
1063
+ "learning_rate": 1.6694783573806883e-05,
1064
+ "loss": 0.035,
1065
+ "step": 1490
1066
+ },
1067
+ {
1068
+ "epoch": 1.663338879645036,
1069
+ "grad_norm": 0.5189201235771179,
1070
+ "learning_rate": 1.6672586015538293e-05,
1071
+ "loss": 0.0475,
1072
+ "step": 1500
1073
+ },
1074
+ {
1075
+ "epoch": 1.6744315030504713,
1076
+ "grad_norm": 0.5089607238769531,
1077
+ "learning_rate": 1.6650388457269703e-05,
1078
+ "loss": 0.0377,
1079
+ "step": 1510
1080
+ },
1081
+ {
1082
+ "epoch": 1.6855241264559067,
1083
+ "grad_norm": 0.46067437529563904,
1084
+ "learning_rate": 1.6628190899001113e-05,
1085
+ "loss": 0.0297,
1086
+ "step": 1520
1087
+ },
1088
+ {
1089
+ "epoch": 1.6966167498613423,
1090
+ "grad_norm": 0.5661717653274536,
1091
+ "learning_rate": 1.660599334073252e-05,
1092
+ "loss": 0.0372,
1093
+ "step": 1530
1094
+ },
1095
+ {
1096
+ "epoch": 1.7077093732667776,
1097
+ "grad_norm": 0.45938414335250854,
1098
+ "learning_rate": 1.658379578246393e-05,
1099
+ "loss": 0.0371,
1100
+ "step": 1540
1101
+ },
1102
+ {
1103
+ "epoch": 1.718801996672213,
1104
+ "grad_norm": 0.7390128970146179,
1105
+ "learning_rate": 1.656159822419534e-05,
1106
+ "loss": 0.0332,
1107
+ "step": 1550
1108
+ },
1109
+ {
1110
+ "epoch": 1.7298946200776484,
1111
+ "grad_norm": 0.6475571990013123,
1112
+ "learning_rate": 1.653940066592675e-05,
1113
+ "loss": 0.0352,
1114
+ "step": 1560
1115
+ },
1116
+ {
1117
+ "epoch": 1.7409872434830838,
1118
+ "grad_norm": 0.8529049754142761,
1119
+ "learning_rate": 1.651720310765816e-05,
1120
+ "loss": 0.0386,
1121
+ "step": 1570
1122
+ },
1123
+ {
1124
+ "epoch": 1.7520798668885191,
1125
+ "grad_norm": 0.6980950236320496,
1126
+ "learning_rate": 1.649500554938957e-05,
1127
+ "loss": 0.0358,
1128
+ "step": 1580
1129
+ },
1130
+ {
1131
+ "epoch": 1.7631724902939545,
1132
+ "grad_norm": 0.5858293175697327,
1133
+ "learning_rate": 1.6472807991120978e-05,
1134
+ "loss": 0.0418,
1135
+ "step": 1590
1136
+ },
1137
+ {
1138
+ "epoch": 1.7742651136993899,
1139
+ "grad_norm": 0.49686577916145325,
1140
+ "learning_rate": 1.6450610432852388e-05,
1141
+ "loss": 0.0361,
1142
+ "step": 1600
1143
+ },
1144
+ {
1145
+ "epoch": 1.7853577371048253,
1146
+ "grad_norm": 0.3631349205970764,
1147
+ "learning_rate": 1.6428412874583798e-05,
1148
+ "loss": 0.0352,
1149
+ "step": 1610
1150
+ },
1151
+ {
1152
+ "epoch": 1.7964503605102606,
1153
+ "grad_norm": 0.5329940915107727,
1154
+ "learning_rate": 1.6406215316315204e-05,
1155
+ "loss": 0.04,
1156
+ "step": 1620
1157
+ },
1158
+ {
1159
+ "epoch": 1.807542983915696,
1160
+ "grad_norm": 0.5995050668716431,
1161
+ "learning_rate": 1.6384017758046614e-05,
1162
+ "loss": 0.0368,
1163
+ "step": 1630
1164
+ },
1165
+ {
1166
+ "epoch": 1.8186356073211316,
1167
+ "grad_norm": 1.3379504680633545,
1168
+ "learning_rate": 1.6361820199778028e-05,
1169
+ "loss": 0.0397,
1170
+ "step": 1640
1171
+ },
1172
+ {
1173
+ "epoch": 1.829728230726567,
1174
+ "grad_norm": 0.8026002645492554,
1175
+ "learning_rate": 1.6339622641509437e-05,
1176
+ "loss": 0.0318,
1177
+ "step": 1650
1178
+ },
1179
+ {
1180
+ "epoch": 1.8408208541320024,
1181
+ "grad_norm": 0.917950451374054,
1182
+ "learning_rate": 1.6317425083240844e-05,
1183
+ "loss": 0.0352,
1184
+ "step": 1660
1185
+ },
1186
+ {
1187
+ "epoch": 1.8519134775374377,
1188
+ "grad_norm": 0.6763226389884949,
1189
+ "learning_rate": 1.6295227524972254e-05,
1190
+ "loss": 0.0413,
1191
+ "step": 1670
1192
+ },
1193
+ {
1194
+ "epoch": 1.8630061009428731,
1195
+ "grad_norm": 0.5728912949562073,
1196
+ "learning_rate": 1.6273029966703664e-05,
1197
+ "loss": 0.0291,
1198
+ "step": 1680
1199
+ },
1200
+ {
1201
+ "epoch": 1.8740987243483085,
1202
+ "grad_norm": 0.6339443325996399,
1203
+ "learning_rate": 1.6250832408435074e-05,
1204
+ "loss": 0.0409,
1205
+ "step": 1690
1206
+ },
1207
+ {
1208
+ "epoch": 1.8851913477537439,
1209
+ "grad_norm": 0.6930853128433228,
1210
+ "learning_rate": 1.6228634850166483e-05,
1211
+ "loss": 0.0371,
1212
+ "step": 1700
1213
+ },
1214
+ {
1215
+ "epoch": 1.8962839711591792,
1216
+ "grad_norm": 0.37715452909469604,
1217
+ "learning_rate": 1.6206437291897893e-05,
1218
+ "loss": 0.031,
1219
+ "step": 1710
1220
+ },
1221
+ {
1222
+ "epoch": 1.9073765945646146,
1223
+ "grad_norm": 0.5656572580337524,
1224
+ "learning_rate": 1.61842397336293e-05,
1225
+ "loss": 0.0414,
1226
+ "step": 1720
1227
+ },
1228
+ {
1229
+ "epoch": 1.91846921797005,
1230
+ "grad_norm": 0.3563915491104126,
1231
+ "learning_rate": 1.6162042175360713e-05,
1232
+ "loss": 0.0276,
1233
+ "step": 1730
1234
+ },
1235
+ {
1236
+ "epoch": 1.9295618413754854,
1237
+ "grad_norm": 0.8260300159454346,
1238
+ "learning_rate": 1.6139844617092123e-05,
1239
+ "loss": 0.0337,
1240
+ "step": 1740
1241
+ },
1242
+ {
1243
+ "epoch": 1.9406544647809207,
1244
+ "grad_norm": 0.6907551884651184,
1245
+ "learning_rate": 1.6117647058823533e-05,
1246
+ "loss": 0.031,
1247
+ "step": 1750
1248
+ },
1249
+ {
1250
+ "epoch": 1.9517470881863561,
1251
+ "grad_norm": 0.5782826542854309,
1252
+ "learning_rate": 1.609544950055494e-05,
1253
+ "loss": 0.0419,
1254
+ "step": 1760
1255
+ },
1256
+ {
1257
+ "epoch": 1.9628397115917915,
1258
+ "grad_norm": 0.36368125677108765,
1259
+ "learning_rate": 1.607325194228635e-05,
1260
+ "loss": 0.0356,
1261
+ "step": 1770
1262
+ },
1263
+ {
1264
+ "epoch": 1.9739323349972269,
1265
+ "grad_norm": 0.369911789894104,
1266
+ "learning_rate": 1.605105438401776e-05,
1267
+ "loss": 0.0264,
1268
+ "step": 1780
1269
+ },
1270
+ {
1271
+ "epoch": 1.9850249584026622,
1272
+ "grad_norm": 0.6072190999984741,
1273
+ "learning_rate": 1.602885682574917e-05,
1274
+ "loss": 0.0366,
1275
+ "step": 1790
1276
+ },
1277
+ {
1278
+ "epoch": 1.9961175818080976,
1279
+ "grad_norm": 0.9733797311782837,
1280
+ "learning_rate": 1.600665926748058e-05,
1281
+ "loss": 0.0371,
1282
+ "step": 1800
1283
+ },
1284
+ {
1285
+ "epoch": 2.0,
1286
+ "eval_accuracy": 0.8973319824383654,
1287
+ "eval_f1": 0.6872427983539094,
1288
+ "eval_loss": 0.31070804595947266,
1289
+ "eval_precision": 0.8166259168704156,
1290
+ "eval_recall": 0.5932504440497336,
1291
+ "eval_runtime": 2.7114,
1292
+ "eval_samples_per_second": 364.014,
1293
+ "eval_steps_per_second": 11.433,
1294
+ "step": 1804
1295
+ },
1296
+ {
1297
+ "epoch": 2.0066555740432612,
1298
+ "grad_norm": 0.2766992449760437,
1299
+ "learning_rate": 1.598446170921199e-05,
1300
+ "loss": 0.0218,
1301
+ "step": 1810
1302
+ },
1303
+ {
1304
+ "epoch": 2.0177481974486966,
1305
+ "grad_norm": 0.6556938290596008,
1306
+ "learning_rate": 1.59622641509434e-05,
1307
+ "loss": 0.0187,
1308
+ "step": 1820
1309
+ },
1310
+ {
1311
+ "epoch": 2.028840820854132,
1312
+ "grad_norm": 0.24634911119937897,
1313
+ "learning_rate": 1.5940066592674808e-05,
1314
+ "loss": 0.0132,
1315
+ "step": 1830
1316
+ },
1317
+ {
1318
+ "epoch": 2.0399334442595674,
1319
+ "grad_norm": 0.683587372303009,
1320
+ "learning_rate": 1.5917869034406218e-05,
1321
+ "loss": 0.0158,
1322
+ "step": 1840
1323
+ },
1324
+ {
1325
+ "epoch": 2.0510260676650027,
1326
+ "grad_norm": 0.47117751836776733,
1327
+ "learning_rate": 1.5895671476137625e-05,
1328
+ "loss": 0.018,
1329
+ "step": 1850
1330
+ },
1331
+ {
1332
+ "epoch": 2.062118691070438,
1333
+ "grad_norm": 0.4102018475532532,
1334
+ "learning_rate": 1.5873473917869034e-05,
1335
+ "loss": 0.0153,
1336
+ "step": 1860
1337
+ },
1338
+ {
1339
+ "epoch": 2.0732113144758735,
1340
+ "grad_norm": 0.4639064371585846,
1341
+ "learning_rate": 1.5851276359600444e-05,
1342
+ "loss": 0.0171,
1343
+ "step": 1870
1344
+ },
1345
+ {
1346
+ "epoch": 2.084303937881309,
1347
+ "grad_norm": 0.49497199058532715,
1348
+ "learning_rate": 1.5829078801331854e-05,
1349
+ "loss": 0.0168,
1350
+ "step": 1880
1351
+ },
1352
+ {
1353
+ "epoch": 2.0953965612867442,
1354
+ "grad_norm": 0.6302080154418945,
1355
+ "learning_rate": 1.5806881243063264e-05,
1356
+ "loss": 0.0191,
1357
+ "step": 1890
1358
+ },
1359
+ {
1360
+ "epoch": 2.1064891846921796,
1361
+ "grad_norm": 0.3464473783969879,
1362
+ "learning_rate": 1.5784683684794674e-05,
1363
+ "loss": 0.0184,
1364
+ "step": 1900
1365
+ },
1366
+ {
1367
+ "epoch": 2.117581808097615,
1368
+ "grad_norm": 0.6507964730262756,
1369
+ "learning_rate": 1.5762486126526084e-05,
1370
+ "loss": 0.0135,
1371
+ "step": 1910
1372
+ },
1373
+ {
1374
+ "epoch": 2.1286744315030504,
1375
+ "grad_norm": 0.1759006232023239,
1376
+ "learning_rate": 1.5740288568257494e-05,
1377
+ "loss": 0.0118,
1378
+ "step": 1920
1379
+ },
1380
+ {
1381
+ "epoch": 2.1397670549084857,
1382
+ "grad_norm": 0.5952832698822021,
1383
+ "learning_rate": 1.5718091009988904e-05,
1384
+ "loss": 0.0157,
1385
+ "step": 1930
1386
+ },
1387
+ {
1388
+ "epoch": 2.150859678313921,
1389
+ "grad_norm": 0.8094580769538879,
1390
+ "learning_rate": 1.5695893451720313e-05,
1391
+ "loss": 0.0125,
1392
+ "step": 1940
1393
+ },
1394
+ {
1395
+ "epoch": 2.1619523017193565,
1396
+ "grad_norm": 0.6284286379814148,
1397
+ "learning_rate": 1.567369589345172e-05,
1398
+ "loss": 0.0178,
1399
+ "step": 1950
1400
+ },
1401
+ {
1402
+ "epoch": 2.173044925124792,
1403
+ "grad_norm": 0.6017957925796509,
1404
+ "learning_rate": 1.565149833518313e-05,
1405
+ "loss": 0.0136,
1406
+ "step": 1960
1407
+ },
1408
+ {
1409
+ "epoch": 2.1841375485302272,
1410
+ "grad_norm": 1.2836827039718628,
1411
+ "learning_rate": 1.562930077691454e-05,
1412
+ "loss": 0.014,
1413
+ "step": 1970
1414
+ },
1415
+ {
1416
+ "epoch": 2.1952301719356626,
1417
+ "grad_norm": 0.6213756203651428,
1418
+ "learning_rate": 1.560710321864595e-05,
1419
+ "loss": 0.0136,
1420
+ "step": 1980
1421
+ },
1422
+ {
1423
+ "epoch": 2.206322795341098,
1424
+ "grad_norm": 0.48122018575668335,
1425
+ "learning_rate": 1.558490566037736e-05,
1426
+ "loss": 0.0147,
1427
+ "step": 1990
1428
+ },
1429
+ {
1430
+ "epoch": 2.2174154187465334,
1431
+ "grad_norm": 0.7161231637001038,
1432
+ "learning_rate": 1.556270810210877e-05,
1433
+ "loss": 0.0105,
1434
+ "step": 2000
1435
+ },
1436
+ {
1437
+ "epoch": 2.2285080421519687,
1438
+ "grad_norm": 0.6598195433616638,
1439
+ "learning_rate": 1.554051054384018e-05,
1440
+ "loss": 0.0172,
1441
+ "step": 2010
1442
+ },
1443
+ {
1444
+ "epoch": 2.239600665557404,
1445
+ "grad_norm": 0.6077088117599487,
1446
+ "learning_rate": 1.551831298557159e-05,
1447
+ "loss": 0.0144,
1448
+ "step": 2020
1449
+ },
1450
+ {
1451
+ "epoch": 2.2506932889628395,
1452
+ "grad_norm": 0.5622262954711914,
1453
+ "learning_rate": 1.5496115427303e-05,
1454
+ "loss": 0.0175,
1455
+ "step": 2030
1456
+ },
1457
+ {
1458
+ "epoch": 2.261785912368275,
1459
+ "grad_norm": 0.1281885802745819,
1460
+ "learning_rate": 1.547391786903441e-05,
1461
+ "loss": 0.018,
1462
+ "step": 2040
1463
+ },
1464
+ {
1465
+ "epoch": 2.2728785357737102,
1466
+ "grad_norm": 0.408607542514801,
1467
+ "learning_rate": 1.5451720310765815e-05,
1468
+ "loss": 0.0095,
1469
+ "step": 2050
1470
+ },
1471
+ {
1472
+ "epoch": 2.283971159179146,
1473
+ "grad_norm": 0.7786557674407959,
1474
+ "learning_rate": 1.5429522752497225e-05,
1475
+ "loss": 0.0163,
1476
+ "step": 2060
1477
+ },
1478
+ {
1479
+ "epoch": 2.2950637825845814,
1480
+ "grad_norm": 0.22975876927375793,
1481
+ "learning_rate": 1.5407325194228635e-05,
1482
+ "loss": 0.0193,
1483
+ "step": 2070
1484
+ },
1485
+ {
1486
+ "epoch": 2.306156405990017,
1487
+ "grad_norm": 0.4313502311706543,
1488
+ "learning_rate": 1.5385127635960048e-05,
1489
+ "loss": 0.0186,
1490
+ "step": 2080
1491
+ },
1492
+ {
1493
+ "epoch": 2.317249029395452,
1494
+ "grad_norm": 0.794448733329773,
1495
+ "learning_rate": 1.5362930077691455e-05,
1496
+ "loss": 0.0159,
1497
+ "step": 2090
1498
+ },
1499
+ {
1500
+ "epoch": 2.3283416528008876,
1501
+ "grad_norm": 0.4191996157169342,
1502
+ "learning_rate": 1.5340732519422865e-05,
1503
+ "loss": 0.0208,
1504
+ "step": 2100
1505
+ },
1506
+ {
1507
+ "epoch": 2.339434276206323,
1508
+ "grad_norm": 0.5219516754150391,
1509
+ "learning_rate": 1.5318534961154274e-05,
1510
+ "loss": 0.0136,
1511
+ "step": 2110
1512
+ },
1513
+ {
1514
+ "epoch": 2.3505268996117583,
1515
+ "grad_norm": 1.1869897842407227,
1516
+ "learning_rate": 1.5296337402885684e-05,
1517
+ "loss": 0.0177,
1518
+ "step": 2120
1519
+ },
1520
+ {
1521
+ "epoch": 2.3616195230171937,
1522
+ "grad_norm": 0.12048923969268799,
1523
+ "learning_rate": 1.5274139844617094e-05,
1524
+ "loss": 0.0129,
1525
+ "step": 2130
1526
+ },
1527
+ {
1528
+ "epoch": 2.372712146422629,
1529
+ "grad_norm": 0.5213373899459839,
1530
+ "learning_rate": 1.5251942286348502e-05,
1531
+ "loss": 0.0127,
1532
+ "step": 2140
1533
+ },
1534
+ {
1535
+ "epoch": 2.3838047698280644,
1536
+ "grad_norm": 0.43704670667648315,
1537
+ "learning_rate": 1.5229744728079912e-05,
1538
+ "loss": 0.0154,
1539
+ "step": 2150
1540
+ },
1541
+ {
1542
+ "epoch": 2.3948973932335,
1543
+ "grad_norm": 0.567150890827179,
1544
+ "learning_rate": 1.5207547169811324e-05,
1545
+ "loss": 0.0117,
1546
+ "step": 2160
1547
+ },
1548
+ {
1549
+ "epoch": 2.405990016638935,
1550
+ "grad_norm": 0.571408212184906,
1551
+ "learning_rate": 1.5185349611542732e-05,
1552
+ "loss": 0.0152,
1553
+ "step": 2170
1554
+ },
1555
+ {
1556
+ "epoch": 2.4170826400443706,
1557
+ "grad_norm": 0.524834394454956,
1558
+ "learning_rate": 1.5163152053274142e-05,
1559
+ "loss": 0.017,
1560
+ "step": 2180
1561
+ },
1562
+ {
1563
+ "epoch": 2.428175263449806,
1564
+ "grad_norm": 0.5026165246963501,
1565
+ "learning_rate": 1.5140954495005552e-05,
1566
+ "loss": 0.0193,
1567
+ "step": 2190
1568
+ },
1569
+ {
1570
+ "epoch": 2.4392678868552413,
1571
+ "grad_norm": 0.3944782614707947,
1572
+ "learning_rate": 1.511875693673696e-05,
1573
+ "loss": 0.0159,
1574
+ "step": 2200
1575
+ },
1576
+ {
1577
+ "epoch": 2.4503605102606767,
1578
+ "grad_norm": 0.6460635662078857,
1579
+ "learning_rate": 1.509655937846837e-05,
1580
+ "loss": 0.0139,
1581
+ "step": 2210
1582
+ },
1583
+ {
1584
+ "epoch": 2.461453133666112,
1585
+ "grad_norm": 0.5573295950889587,
1586
+ "learning_rate": 1.507436182019978e-05,
1587
+ "loss": 0.0183,
1588
+ "step": 2220
1589
+ },
1590
+ {
1591
+ "epoch": 2.4725457570715474,
1592
+ "grad_norm": 0.5001458525657654,
1593
+ "learning_rate": 1.5052164261931188e-05,
1594
+ "loss": 0.0167,
1595
+ "step": 2230
1596
+ },
1597
+ {
1598
+ "epoch": 2.483638380476983,
1599
+ "grad_norm": 1.0214160680770874,
1600
+ "learning_rate": 1.5029966703662598e-05,
1601
+ "loss": 0.0182,
1602
+ "step": 2240
1603
+ },
1604
+ {
1605
+ "epoch": 2.494731003882418,
1606
+ "grad_norm": 0.5756349563598633,
1607
+ "learning_rate": 1.5007769145394009e-05,
1608
+ "loss": 0.0164,
1609
+ "step": 2250
1610
+ },
1611
+ {
1612
+ "epoch": 2.5058236272878536,
1613
+ "grad_norm": 0.771986722946167,
1614
+ "learning_rate": 1.4985571587125419e-05,
1615
+ "loss": 0.0183,
1616
+ "step": 2260
1617
+ },
1618
+ {
1619
+ "epoch": 2.516916250693289,
1620
+ "grad_norm": 0.5560303926467896,
1621
+ "learning_rate": 1.4963374028856827e-05,
1622
+ "loss": 0.0178,
1623
+ "step": 2270
1624
+ },
1625
+ {
1626
+ "epoch": 2.5280088740987243,
1627
+ "grad_norm": 0.8550804853439331,
1628
+ "learning_rate": 1.4941176470588237e-05,
1629
+ "loss": 0.0118,
1630
+ "step": 2280
1631
+ },
1632
+ {
1633
+ "epoch": 2.5391014975041597,
1634
+ "grad_norm": 0.8129355907440186,
1635
+ "learning_rate": 1.4918978912319645e-05,
1636
+ "loss": 0.0126,
1637
+ "step": 2290
1638
+ },
1639
+ {
1640
+ "epoch": 2.550194120909595,
1641
+ "grad_norm": 1.176377773284912,
1642
+ "learning_rate": 1.4896781354051055e-05,
1643
+ "loss": 0.0126,
1644
+ "step": 2300
1645
+ },
1646
+ {
1647
+ "epoch": 2.5612867443150305,
1648
+ "grad_norm": 0.1419341266155243,
1649
+ "learning_rate": 1.4874583795782465e-05,
1650
+ "loss": 0.0196,
1651
+ "step": 2310
1652
+ },
1653
+ {
1654
+ "epoch": 2.572379367720466,
1655
+ "grad_norm": 1.4311761856079102,
1656
+ "learning_rate": 1.4852386237513873e-05,
1657
+ "loss": 0.0143,
1658
+ "step": 2320
1659
+ },
1660
+ {
1661
+ "epoch": 2.583471991125901,
1662
+ "grad_norm": 0.9924690127372742,
1663
+ "learning_rate": 1.4830188679245283e-05,
1664
+ "loss": 0.0198,
1665
+ "step": 2330
1666
+ },
1667
+ {
1668
+ "epoch": 2.5945646145313366,
1669
+ "grad_norm": 0.6496040225028992,
1670
+ "learning_rate": 1.4807991120976695e-05,
1671
+ "loss": 0.014,
1672
+ "step": 2340
1673
+ },
1674
+ {
1675
+ "epoch": 2.605657237936772,
1676
+ "grad_norm": 0.2335842400789261,
1677
+ "learning_rate": 1.4785793562708104e-05,
1678
+ "loss": 0.0128,
1679
+ "step": 2350
1680
+ },
1681
+ {
1682
+ "epoch": 2.6167498613422073,
1683
+ "grad_norm": 0.6549142003059387,
1684
+ "learning_rate": 1.4763596004439513e-05,
1685
+ "loss": 0.019,
1686
+ "step": 2360
1687
+ },
1688
+ {
1689
+ "epoch": 2.6278424847476427,
1690
+ "grad_norm": 0.3264187276363373,
1691
+ "learning_rate": 1.4741398446170922e-05,
1692
+ "loss": 0.0168,
1693
+ "step": 2370
1694
+ },
1695
+ {
1696
+ "epoch": 2.638935108153078,
1697
+ "grad_norm": 0.4175679683685303,
1698
+ "learning_rate": 1.4719200887902332e-05,
1699
+ "loss": 0.0122,
1700
+ "step": 2380
1701
+ },
1702
+ {
1703
+ "epoch": 2.6500277315585135,
1704
+ "grad_norm": 0.2968366742134094,
1705
+ "learning_rate": 1.469700332963374e-05,
1706
+ "loss": 0.0189,
1707
+ "step": 2390
1708
+ },
1709
+ {
1710
+ "epoch": 2.661120354963949,
1711
+ "grad_norm": 1.5489327907562256,
1712
+ "learning_rate": 1.467480577136515e-05,
1713
+ "loss": 0.0158,
1714
+ "step": 2400
1715
+ },
1716
+ {
1717
+ "epoch": 2.672212978369384,
1718
+ "grad_norm": 0.8017039895057678,
1719
+ "learning_rate": 1.465260821309656e-05,
1720
+ "loss": 0.0178,
1721
+ "step": 2410
1722
+ },
1723
+ {
1724
+ "epoch": 2.6833056017748196,
1725
+ "grad_norm": 0.6136099100112915,
1726
+ "learning_rate": 1.4630410654827972e-05,
1727
+ "loss": 0.0164,
1728
+ "step": 2420
1729
+ },
1730
+ {
1731
+ "epoch": 2.6943982251802554,
1732
+ "grad_norm": 0.5461011528968811,
1733
+ "learning_rate": 1.460821309655938e-05,
1734
+ "loss": 0.0168,
1735
+ "step": 2430
1736
+ },
1737
+ {
1738
+ "epoch": 2.7054908485856908,
1739
+ "grad_norm": 0.2680525779724121,
1740
+ "learning_rate": 1.458601553829079e-05,
1741
+ "loss": 0.0138,
1742
+ "step": 2440
1743
+ },
1744
+ {
1745
+ "epoch": 2.716583471991126,
1746
+ "grad_norm": 0.6884289383888245,
1747
+ "learning_rate": 1.45638179800222e-05,
1748
+ "loss": 0.0145,
1749
+ "step": 2450
1750
+ },
1751
+ {
1752
+ "epoch": 2.7276760953965615,
1753
+ "grad_norm": 1.226928472518921,
1754
+ "learning_rate": 1.4541620421753608e-05,
1755
+ "loss": 0.0145,
1756
+ "step": 2460
1757
+ },
1758
+ {
1759
+ "epoch": 2.738768718801997,
1760
+ "grad_norm": 0.447221040725708,
1761
+ "learning_rate": 1.4519422863485018e-05,
1762
+ "loss": 0.0168,
1763
+ "step": 2470
1764
+ },
1765
+ {
1766
+ "epoch": 2.7498613422074323,
1767
+ "grad_norm": 0.47167718410491943,
1768
+ "learning_rate": 1.4497225305216428e-05,
1769
+ "loss": 0.0124,
1770
+ "step": 2480
1771
+ },
1772
+ {
1773
+ "epoch": 2.7609539656128677,
1774
+ "grad_norm": 1.326282262802124,
1775
+ "learning_rate": 1.4475027746947836e-05,
1776
+ "loss": 0.0183,
1777
+ "step": 2490
1778
+ },
1779
+ {
1780
+ "epoch": 2.772046589018303,
1781
+ "grad_norm": 0.9036744236946106,
1782
+ "learning_rate": 1.4452830188679246e-05,
1783
+ "loss": 0.023,
1784
+ "step": 2500
1785
+ },
1786
+ {
1787
+ "epoch": 2.7831392124237384,
1788
+ "grad_norm": 0.30263999104499817,
1789
+ "learning_rate": 1.4430632630410657e-05,
1790
+ "loss": 0.0163,
1791
+ "step": 2510
1792
+ },
1793
+ {
1794
+ "epoch": 2.794231835829174,
1795
+ "grad_norm": 0.7316049933433533,
1796
+ "learning_rate": 1.4408435072142067e-05,
1797
+ "loss": 0.0156,
1798
+ "step": 2520
1799
+ },
1800
+ {
1801
+ "epoch": 2.805324459234609,
1802
+ "grad_norm": 0.8268038630485535,
1803
+ "learning_rate": 1.4386237513873475e-05,
1804
+ "loss": 0.0169,
1805
+ "step": 2530
1806
+ },
1807
+ {
1808
+ "epoch": 2.8164170826400445,
1809
+ "grad_norm": 0.633816659450531,
1810
+ "learning_rate": 1.4364039955604885e-05,
1811
+ "loss": 0.0147,
1812
+ "step": 2540
1813
+ },
1814
+ {
1815
+ "epoch": 2.82750970604548,
1816
+ "grad_norm": 0.43913397192955017,
1817
+ "learning_rate": 1.4341842397336295e-05,
1818
+ "loss": 0.0122,
1819
+ "step": 2550
1820
+ },
1821
+ {
1822
+ "epoch": 2.8386023294509153,
1823
+ "grad_norm": 0.37594836950302124,
1824
+ "learning_rate": 1.4319644839067703e-05,
1825
+ "loss": 0.0149,
1826
+ "step": 2560
1827
+ },
1828
+ {
1829
+ "epoch": 2.8496949528563507,
1830
+ "grad_norm": 0.5090307593345642,
1831
+ "learning_rate": 1.4297447280799113e-05,
1832
+ "loss": 0.015,
1833
+ "step": 2570
1834
+ },
1835
+ {
1836
+ "epoch": 2.860787576261786,
1837
+ "grad_norm": 0.47723984718322754,
1838
+ "learning_rate": 1.4275249722530521e-05,
1839
+ "loss": 0.0166,
1840
+ "step": 2580
1841
+ },
1842
+ {
1843
+ "epoch": 2.8718801996672214,
1844
+ "grad_norm": 1.2512151002883911,
1845
+ "learning_rate": 1.4253052164261931e-05,
1846
+ "loss": 0.0136,
1847
+ "step": 2590
1848
+ },
1849
+ {
1850
+ "epoch": 2.882972823072657,
1851
+ "grad_norm": 0.676337718963623,
1852
+ "learning_rate": 1.4230854605993343e-05,
1853
+ "loss": 0.0152,
1854
+ "step": 2600
1855
+ },
1856
+ {
1857
+ "epoch": 2.894065446478092,
1858
+ "grad_norm": 0.7379328012466431,
1859
+ "learning_rate": 1.4208657047724752e-05,
1860
+ "loss": 0.0182,
1861
+ "step": 2610
1862
+ },
1863
+ {
1864
+ "epoch": 2.9051580698835275,
1865
+ "grad_norm": 0.8266171813011169,
1866
+ "learning_rate": 1.418645948945616e-05,
1867
+ "loss": 0.0207,
1868
+ "step": 2620
1869
+ },
1870
+ {
1871
+ "epoch": 2.916250693288963,
1872
+ "grad_norm": 0.9925495982170105,
1873
+ "learning_rate": 1.416426193118757e-05,
1874
+ "loss": 0.0157,
1875
+ "step": 2630
1876
+ },
1877
+ {
1878
+ "epoch": 2.9273433166943983,
1879
+ "grad_norm": 1.1120530366897583,
1880
+ "learning_rate": 1.414206437291898e-05,
1881
+ "loss": 0.0164,
1882
+ "step": 2640
1883
+ },
1884
+ {
1885
+ "epoch": 2.9384359400998337,
1886
+ "grad_norm": 0.9687446355819702,
1887
+ "learning_rate": 1.4119866814650389e-05,
1888
+ "loss": 0.0211,
1889
+ "step": 2650
1890
+ },
1891
+ {
1892
+ "epoch": 2.949528563505269,
1893
+ "grad_norm": 0.19811566174030304,
1894
+ "learning_rate": 1.4097669256381798e-05,
1895
+ "loss": 0.0181,
1896
+ "step": 2660
1897
+ },
1898
+ {
1899
+ "epoch": 2.9606211869107044,
1900
+ "grad_norm": 0.9253482818603516,
1901
+ "learning_rate": 1.4075471698113208e-05,
1902
+ "loss": 0.0158,
1903
+ "step": 2670
1904
+ },
1905
+ {
1906
+ "epoch": 2.97171381031614,
1907
+ "grad_norm": 0.677994430065155,
1908
+ "learning_rate": 1.405327413984462e-05,
1909
+ "loss": 0.024,
1910
+ "step": 2680
1911
+ },
1912
+ {
1913
+ "epoch": 2.982806433721575,
1914
+ "grad_norm": 0.3959580361843109,
1915
+ "learning_rate": 1.4031076581576028e-05,
1916
+ "loss": 0.009,
1917
+ "step": 2690
1918
+ },
1919
+ {
1920
+ "epoch": 2.9938990571270105,
1921
+ "grad_norm": 1.0569937229156494,
1922
+ "learning_rate": 1.4008879023307438e-05,
1923
+ "loss": 0.0187,
1924
+ "step": 2700
1925
+ },
1926
+ {
1927
+ "epoch": 3.0,
1928
+ "eval_accuracy": 0.8888888888888888,
1929
+ "eval_f1": 0.7171109200343938,
1930
+ "eval_loss": 0.3638182580471039,
1931
+ "eval_precision": 0.695,
1932
+ "eval_recall": 0.7406749555950266,
1933
+ "eval_runtime": 2.7252,
1934
+ "eval_samples_per_second": 362.177,
1935
+ "eval_steps_per_second": 11.375,
1936
+ "step": 2706
1937
+ },
1938
+ {
1939
+ "epoch": 3.004437049362174,
1940
+ "grad_norm": 0.22475391626358032,
1941
+ "learning_rate": 1.3986681465038848e-05,
1942
+ "loss": 0.0127,
1943
+ "step": 2710
1944
+ },
1945
+ {
1946
+ "epoch": 3.0155296727676095,
1947
+ "grad_norm": 0.42929062247276306,
1948
+ "learning_rate": 1.3964483906770256e-05,
1949
+ "loss": 0.0131,
1950
+ "step": 2720
1951
+ },
1952
+ {
1953
+ "epoch": 3.026622296173045,
1954
+ "grad_norm": 2.398026466369629,
1955
+ "learning_rate": 1.3942286348501666e-05,
1956
+ "loss": 0.0072,
1957
+ "step": 2730
1958
+ },
1959
+ {
1960
+ "epoch": 3.0377149195784803,
1961
+ "grad_norm": 0.14908376336097717,
1962
+ "learning_rate": 1.3920088790233076e-05,
1963
+ "loss": 0.0082,
1964
+ "step": 2740
1965
+ },
1966
+ {
1967
+ "epoch": 3.0488075429839157,
1968
+ "grad_norm": 0.8177769184112549,
1969
+ "learning_rate": 1.3897891231964484e-05,
1970
+ "loss": 0.0088,
1971
+ "step": 2750
1972
+ },
1973
+ {
1974
+ "epoch": 3.059900166389351,
1975
+ "grad_norm": 0.13737396895885468,
1976
+ "learning_rate": 1.3875693673695894e-05,
1977
+ "loss": 0.008,
1978
+ "step": 2760
1979
+ },
1980
+ {
1981
+ "epoch": 3.0709927897947864,
1982
+ "grad_norm": 0.7825105786323547,
1983
+ "learning_rate": 1.3853496115427305e-05,
1984
+ "loss": 0.0034,
1985
+ "step": 2770
1986
+ },
1987
+ {
1988
+ "epoch": 3.082085413200222,
1989
+ "grad_norm": 0.42062872648239136,
1990
+ "learning_rate": 1.3831298557158715e-05,
1991
+ "loss": 0.0062,
1992
+ "step": 2780
1993
+ },
1994
+ {
1995
+ "epoch": 3.093178036605657,
1996
+ "grad_norm": 0.7224815487861633,
1997
+ "learning_rate": 1.3809100998890123e-05,
1998
+ "loss": 0.0055,
1999
+ "step": 2790
2000
+ },
2001
+ {
2002
+ "epoch": 3.1042706600110925,
2003
+ "grad_norm": 0.04094177484512329,
2004
+ "learning_rate": 1.3786903440621533e-05,
2005
+ "loss": 0.002,
2006
+ "step": 2800
2007
+ },
2008
+ {
2009
+ "epoch": 3.115363283416528,
2010
+ "grad_norm": 0.03310002386569977,
2011
+ "learning_rate": 1.3764705882352943e-05,
2012
+ "loss": 0.0055,
2013
+ "step": 2810
2014
+ },
2015
+ {
2016
+ "epoch": 3.1264559068219633,
2017
+ "grad_norm": 0.5118809938430786,
2018
+ "learning_rate": 1.3742508324084351e-05,
2019
+ "loss": 0.0054,
2020
+ "step": 2820
2021
+ },
2022
+ {
2023
+ "epoch": 3.1375485302273987,
2024
+ "grad_norm": 0.5759711861610413,
2025
+ "learning_rate": 1.3720310765815761e-05,
2026
+ "loss": 0.0086,
2027
+ "step": 2830
2028
+ },
2029
+ {
2030
+ "epoch": 3.148641153632834,
2031
+ "grad_norm": 1.1310100555419922,
2032
+ "learning_rate": 1.3698113207547171e-05,
2033
+ "loss": 0.0091,
2034
+ "step": 2840
2035
+ },
2036
+ {
2037
+ "epoch": 3.1597337770382694,
2038
+ "grad_norm": 0.04265379533171654,
2039
+ "learning_rate": 1.3675915649278579e-05,
2040
+ "loss": 0.0061,
2041
+ "step": 2850
2042
+ },
2043
+ {
2044
+ "epoch": 3.170826400443705,
2045
+ "grad_norm": 0.6648514270782471,
2046
+ "learning_rate": 1.365371809100999e-05,
2047
+ "loss": 0.0035,
2048
+ "step": 2860
2049
+ },
2050
+ {
2051
+ "epoch": 3.18191902384914,
2052
+ "grad_norm": 0.02486938051879406,
2053
+ "learning_rate": 1.36315205327414e-05,
2054
+ "loss": 0.0055,
2055
+ "step": 2870
2056
+ },
2057
+ {
2058
+ "epoch": 3.1930116472545755,
2059
+ "grad_norm": 0.5384965538978577,
2060
+ "learning_rate": 1.3609322974472809e-05,
2061
+ "loss": 0.0029,
2062
+ "step": 2880
2063
+ },
2064
+ {
2065
+ "epoch": 3.204104270660011,
2066
+ "grad_norm": 0.6231627464294434,
2067
+ "learning_rate": 1.3587125416204219e-05,
2068
+ "loss": 0.005,
2069
+ "step": 2890
2070
+ },
2071
+ {
2072
+ "epoch": 3.2151968940654463,
2073
+ "grad_norm": 0.5190199613571167,
2074
+ "learning_rate": 1.3564927857935628e-05,
2075
+ "loss": 0.0083,
2076
+ "step": 2900
2077
+ },
2078
+ {
2079
+ "epoch": 3.2262895174708817,
2080
+ "grad_norm": 0.39337947964668274,
2081
+ "learning_rate": 1.3542730299667037e-05,
2082
+ "loss": 0.0081,
2083
+ "step": 2910
2084
+ },
2085
+ {
2086
+ "epoch": 3.237382140876317,
2087
+ "grad_norm": 0.6862775683403015,
2088
+ "learning_rate": 1.3520532741398446e-05,
2089
+ "loss": 0.0074,
2090
+ "step": 2920
2091
+ },
2092
+ {
2093
+ "epoch": 3.2484747642817524,
2094
+ "grad_norm": 1.0862469673156738,
2095
+ "learning_rate": 1.3498335183129856e-05,
2096
+ "loss": 0.0074,
2097
+ "step": 2930
2098
+ },
2099
+ {
2100
+ "epoch": 3.259567387687188,
2101
+ "grad_norm": 1.1846354007720947,
2102
+ "learning_rate": 1.3476137624861268e-05,
2103
+ "loss": 0.0075,
2104
+ "step": 2940
2105
+ },
2106
+ {
2107
+ "epoch": 3.270660011092623,
2108
+ "grad_norm": 1.2481039762496948,
2109
+ "learning_rate": 1.3453940066592676e-05,
2110
+ "loss": 0.003,
2111
+ "step": 2950
2112
+ },
2113
+ {
2114
+ "epoch": 3.281752634498059,
2115
+ "grad_norm": 0.3161996006965637,
2116
+ "learning_rate": 1.3431742508324086e-05,
2117
+ "loss": 0.0034,
2118
+ "step": 2960
2119
+ },
2120
+ {
2121
+ "epoch": 3.2928452579034944,
2122
+ "grad_norm": 0.6262934803962708,
2123
+ "learning_rate": 1.3409544950055496e-05,
2124
+ "loss": 0.0107,
2125
+ "step": 2970
2126
+ },
2127
+ {
2128
+ "epoch": 3.3039378813089297,
2129
+ "grad_norm": 0.9278308153152466,
2130
+ "learning_rate": 1.3387347391786904e-05,
2131
+ "loss": 0.0057,
2132
+ "step": 2980
2133
+ },
2134
+ {
2135
+ "epoch": 3.315030504714365,
2136
+ "grad_norm": 0.40893518924713135,
2137
+ "learning_rate": 1.3365149833518314e-05,
2138
+ "loss": 0.0041,
2139
+ "step": 2990
2140
+ },
2141
+ {
2142
+ "epoch": 3.3261231281198005,
2143
+ "grad_norm": 0.6295328736305237,
2144
+ "learning_rate": 1.3342952275249724e-05,
2145
+ "loss": 0.0032,
2146
+ "step": 3000
2147
+ },
2148
+ {
2149
+ "epoch": 3.337215751525236,
2150
+ "grad_norm": 0.039073534309864044,
2151
+ "learning_rate": 1.3320754716981132e-05,
2152
+ "loss": 0.0067,
2153
+ "step": 3010
2154
+ },
2155
+ {
2156
+ "epoch": 3.3483083749306712,
2157
+ "grad_norm": 0.5157305002212524,
2158
+ "learning_rate": 1.3298557158712542e-05,
2159
+ "loss": 0.0093,
2160
+ "step": 3020
2161
+ },
2162
+ {
2163
+ "epoch": 3.3594009983361066,
2164
+ "grad_norm": 0.9451714158058167,
2165
+ "learning_rate": 1.3276359600443953e-05,
2166
+ "loss": 0.0064,
2167
+ "step": 3030
2168
+ },
2169
+ {
2170
+ "epoch": 3.370493621741542,
2171
+ "grad_norm": 0.043243326246738434,
2172
+ "learning_rate": 1.3254162042175363e-05,
2173
+ "loss": 0.0062,
2174
+ "step": 3040
2175
+ },
2176
+ {
2177
+ "epoch": 3.3815862451469774,
2178
+ "grad_norm": 1.2751350402832031,
2179
+ "learning_rate": 1.3231964483906771e-05,
2180
+ "loss": 0.011,
2181
+ "step": 3050
2182
+ },
2183
+ {
2184
+ "epoch": 3.3926788685524127,
2185
+ "grad_norm": 1.1212610006332397,
2186
+ "learning_rate": 1.3209766925638181e-05,
2187
+ "loss": 0.008,
2188
+ "step": 3060
2189
+ },
2190
+ {
2191
+ "epoch": 3.403771491957848,
2192
+ "grad_norm": 0.3985590934753418,
2193
+ "learning_rate": 1.3187569367369591e-05,
2194
+ "loss": 0.0066,
2195
+ "step": 3070
2196
+ },
2197
+ {
2198
+ "epoch": 3.4148641153632835,
2199
+ "grad_norm": 0.6546558737754822,
2200
+ "learning_rate": 1.3165371809101e-05,
2201
+ "loss": 0.0115,
2202
+ "step": 3080
2203
+ },
2204
+ {
2205
+ "epoch": 3.425956738768719,
2206
+ "grad_norm": 1.0568732023239136,
2207
+ "learning_rate": 1.3143174250832409e-05,
2208
+ "loss": 0.0104,
2209
+ "step": 3090
2210
+ },
2211
+ {
2212
+ "epoch": 3.4370493621741542,
2213
+ "grad_norm": 0.9182056188583374,
2214
+ "learning_rate": 1.3120976692563819e-05,
2215
+ "loss": 0.0079,
2216
+ "step": 3100
2217
+ },
2218
+ {
2219
+ "epoch": 3.4481419855795896,
2220
+ "grad_norm": 1.700160026550293,
2221
+ "learning_rate": 1.3098779134295227e-05,
2222
+ "loss": 0.0061,
2223
+ "step": 3110
2224
+ },
2225
+ {
2226
+ "epoch": 3.459234608985025,
2227
+ "grad_norm": 0.09296636283397675,
2228
+ "learning_rate": 1.3076581576026639e-05,
2229
+ "loss": 0.0106,
2230
+ "step": 3120
2231
+ },
2232
+ {
2233
+ "epoch": 3.4703272323904604,
2234
+ "grad_norm": 0.8721600770950317,
2235
+ "learning_rate": 1.3054384017758049e-05,
2236
+ "loss": 0.0058,
2237
+ "step": 3130
2238
+ },
2239
+ {
2240
+ "epoch": 3.4814198557958957,
2241
+ "grad_norm": 0.5778414607048035,
2242
+ "learning_rate": 1.3032186459489457e-05,
2243
+ "loss": 0.0075,
2244
+ "step": 3140
2245
+ },
2246
+ {
2247
+ "epoch": 3.492512479201331,
2248
+ "grad_norm": 0.632172167301178,
2249
+ "learning_rate": 1.3009988901220867e-05,
2250
+ "loss": 0.0057,
2251
+ "step": 3150
2252
+ },
2253
+ {
2254
+ "epoch": 3.5036051026067665,
2255
+ "grad_norm": 0.49320322275161743,
2256
+ "learning_rate": 1.2987791342952276e-05,
2257
+ "loss": 0.0065,
2258
+ "step": 3160
2259
+ },
2260
+ {
2261
+ "epoch": 3.514697726012202,
2262
+ "grad_norm": 1.2454118728637695,
2263
+ "learning_rate": 1.2965593784683685e-05,
2264
+ "loss": 0.0049,
2265
+ "step": 3170
2266
+ },
2267
+ {
2268
+ "epoch": 3.5257903494176372,
2269
+ "grad_norm": 0.16420379281044006,
2270
+ "learning_rate": 1.2943396226415095e-05,
2271
+ "loss": 0.0042,
2272
+ "step": 3180
2273
+ },
2274
+ {
2275
+ "epoch": 3.5368829728230726,
2276
+ "grad_norm": 0.7556213736534119,
2277
+ "learning_rate": 1.2921198668146504e-05,
2278
+ "loss": 0.0076,
2279
+ "step": 3190
2280
+ },
2281
+ {
2282
+ "epoch": 3.547975596228508,
2283
+ "grad_norm": 0.7741811871528625,
2284
+ "learning_rate": 1.2899001109877916e-05,
2285
+ "loss": 0.0053,
2286
+ "step": 3200
2287
+ },
2288
+ {
2289
+ "epoch": 3.5590682196339434,
2290
+ "grad_norm": 0.3981403708457947,
2291
+ "learning_rate": 1.2876803551609324e-05,
2292
+ "loss": 0.0111,
2293
+ "step": 3210
2294
+ },
2295
+ {
2296
+ "epoch": 3.5701608430393788,
2297
+ "grad_norm": 1.556243658065796,
2298
+ "learning_rate": 1.2854605993340734e-05,
2299
+ "loss": 0.0082,
2300
+ "step": 3220
2301
+ },
2302
+ {
2303
+ "epoch": 3.581253466444814,
2304
+ "grad_norm": 0.24557125568389893,
2305
+ "learning_rate": 1.2832408435072144e-05,
2306
+ "loss": 0.0048,
2307
+ "step": 3230
2308
+ },
2309
+ {
2310
+ "epoch": 3.5923460898502495,
2311
+ "grad_norm": 1.2477893829345703,
2312
+ "learning_rate": 1.2810210876803552e-05,
2313
+ "loss": 0.0023,
2314
+ "step": 3240
2315
+ },
2316
+ {
2317
+ "epoch": 3.603438713255685,
2318
+ "grad_norm": 0.04675845056772232,
2319
+ "learning_rate": 1.2788013318534962e-05,
2320
+ "loss": 0.0041,
2321
+ "step": 3250
2322
+ },
2323
+ {
2324
+ "epoch": 3.6145313366611203,
2325
+ "grad_norm": 0.571173906326294,
2326
+ "learning_rate": 1.2765815760266372e-05,
2327
+ "loss": 0.0055,
2328
+ "step": 3260
2329
+ },
2330
+ {
2331
+ "epoch": 3.6256239600665556,
2332
+ "grad_norm": 0.5565969944000244,
2333
+ "learning_rate": 1.274361820199778e-05,
2334
+ "loss": 0.0092,
2335
+ "step": 3270
2336
+ },
2337
+ {
2338
+ "epoch": 3.636716583471991,
2339
+ "grad_norm": 0.013311430811882019,
2340
+ "learning_rate": 1.272142064372919e-05,
2341
+ "loss": 0.0071,
2342
+ "step": 3280
2343
+ },
2344
+ {
2345
+ "epoch": 3.6478092068774264,
2346
+ "grad_norm": 1.6714210510253906,
2347
+ "learning_rate": 1.2699223085460601e-05,
2348
+ "loss": 0.0069,
2349
+ "step": 3290
2350
+ },
2351
+ {
2352
+ "epoch": 3.6589018302828618,
2353
+ "grad_norm": 0.7213656306266785,
2354
+ "learning_rate": 1.2677025527192011e-05,
2355
+ "loss": 0.0053,
2356
+ "step": 3300
2357
+ },
2358
+ {
2359
+ "epoch": 3.669994453688297,
2360
+ "grad_norm": 0.9073006510734558,
2361
+ "learning_rate": 1.265482796892342e-05,
2362
+ "loss": 0.0101,
2363
+ "step": 3310
2364
+ },
2365
+ {
2366
+ "epoch": 3.6810870770937325,
2367
+ "grad_norm": 0.29895493388175964,
2368
+ "learning_rate": 1.263263041065483e-05,
2369
+ "loss": 0.0052,
2370
+ "step": 3320
2371
+ },
2372
+ {
2373
+ "epoch": 3.6921797004991683,
2374
+ "grad_norm": 1.0842756032943726,
2375
+ "learning_rate": 1.2610432852386239e-05,
2376
+ "loss": 0.008,
2377
+ "step": 3330
2378
+ },
2379
+ {
2380
+ "epoch": 3.7032723239046037,
2381
+ "grad_norm": 0.7172666192054749,
2382
+ "learning_rate": 1.2588235294117647e-05,
2383
+ "loss": 0.01,
2384
+ "step": 3340
2385
+ },
2386
+ {
2387
+ "epoch": 3.714364947310039,
2388
+ "grad_norm": 0.10475369542837143,
2389
+ "learning_rate": 1.2566037735849057e-05,
2390
+ "loss": 0.0063,
2391
+ "step": 3350
2392
+ },
2393
+ {
2394
+ "epoch": 3.7254575707154745,
2395
+ "grad_norm": 0.9639983177185059,
2396
+ "learning_rate": 1.2543840177580467e-05,
2397
+ "loss": 0.0059,
2398
+ "step": 3360
2399
+ },
2400
+ {
2401
+ "epoch": 3.73655019412091,
2402
+ "grad_norm": 0.9579765796661377,
2403
+ "learning_rate": 1.2521642619311875e-05,
2404
+ "loss": 0.0053,
2405
+ "step": 3370
2406
+ },
2407
+ {
2408
+ "epoch": 3.747642817526345,
2409
+ "grad_norm": 0.14046674966812134,
2410
+ "learning_rate": 1.2499445061043287e-05,
2411
+ "loss": 0.0081,
2412
+ "step": 3380
2413
+ },
2414
+ {
2415
+ "epoch": 3.7587354409317806,
2416
+ "grad_norm": 0.4136715233325958,
2417
+ "learning_rate": 1.2477247502774697e-05,
2418
+ "loss": 0.0101,
2419
+ "step": 3390
2420
+ },
2421
+ {
2422
+ "epoch": 3.769828064337216,
2423
+ "grad_norm": 0.30935949087142944,
2424
+ "learning_rate": 1.2455049944506107e-05,
2425
+ "loss": 0.0046,
2426
+ "step": 3400
2427
+ },
2428
+ {
2429
+ "epoch": 3.7809206877426513,
2430
+ "grad_norm": 0.25942277908325195,
2431
+ "learning_rate": 1.2432852386237515e-05,
2432
+ "loss": 0.006,
2433
+ "step": 3410
2434
+ },
2435
+ {
2436
+ "epoch": 3.7920133111480867,
2437
+ "grad_norm": 0.08538083732128143,
2438
+ "learning_rate": 1.2410654827968925e-05,
2439
+ "loss": 0.0038,
2440
+ "step": 3420
2441
+ },
2442
+ {
2443
+ "epoch": 3.803105934553522,
2444
+ "grad_norm": 0.11854628473520279,
2445
+ "learning_rate": 1.2388457269700334e-05,
2446
+ "loss": 0.0038,
2447
+ "step": 3430
2448
+ },
2449
+ {
2450
+ "epoch": 3.8141985579589575,
2451
+ "grad_norm": 0.2808975279331207,
2452
+ "learning_rate": 1.2366259711431743e-05,
2453
+ "loss": 0.0093,
2454
+ "step": 3440
2455
+ },
2456
+ {
2457
+ "epoch": 3.825291181364393,
2458
+ "grad_norm": 1.0285228490829468,
2459
+ "learning_rate": 1.2344062153163152e-05,
2460
+ "loss": 0.0101,
2461
+ "step": 3450
2462
+ },
2463
+ {
2464
+ "epoch": 3.836383804769828,
2465
+ "grad_norm": 0.43529564142227173,
2466
+ "learning_rate": 1.2321864594894564e-05,
2467
+ "loss": 0.0057,
2468
+ "step": 3460
2469
+ },
2470
+ {
2471
+ "epoch": 3.8474764281752636,
2472
+ "grad_norm": 2.0706746578216553,
2473
+ "learning_rate": 1.2299667036625972e-05,
2474
+ "loss": 0.0079,
2475
+ "step": 3470
2476
+ },
2477
+ {
2478
+ "epoch": 3.858569051580699,
2479
+ "grad_norm": 0.9805829524993896,
2480
+ "learning_rate": 1.2277469478357382e-05,
2481
+ "loss": 0.0116,
2482
+ "step": 3480
2483
+ },
2484
+ {
2485
+ "epoch": 3.8696616749861343,
2486
+ "grad_norm": 0.754189133644104,
2487
+ "learning_rate": 1.2255271920088792e-05,
2488
+ "loss": 0.0056,
2489
+ "step": 3490
2490
+ },
2491
+ {
2492
+ "epoch": 3.8807542983915697,
2493
+ "grad_norm": 0.9179038405418396,
2494
+ "learning_rate": 1.22330743618202e-05,
2495
+ "loss": 0.0079,
2496
+ "step": 3500
2497
+ },
2498
+ {
2499
+ "epoch": 3.891846921797005,
2500
+ "grad_norm": 1.5467790365219116,
2501
+ "learning_rate": 1.221087680355161e-05,
2502
+ "loss": 0.009,
2503
+ "step": 3510
2504
+ },
2505
+ {
2506
+ "epoch": 3.9029395452024405,
2507
+ "grad_norm": 0.44232311844825745,
2508
+ "learning_rate": 1.218867924528302e-05,
2509
+ "loss": 0.0072,
2510
+ "step": 3520
2511
+ },
2512
+ {
2513
+ "epoch": 3.914032168607876,
2514
+ "grad_norm": 0.05650203302502632,
2515
+ "learning_rate": 1.2166481687014428e-05,
2516
+ "loss": 0.0053,
2517
+ "step": 3530
2518
+ },
2519
+ {
2520
+ "epoch": 3.925124792013311,
2521
+ "grad_norm": 0.6348217129707336,
2522
+ "learning_rate": 1.2144284128745838e-05,
2523
+ "loss": 0.009,
2524
+ "step": 3540
2525
+ },
2526
+ {
2527
+ "epoch": 3.9362174154187466,
2528
+ "grad_norm": 1.0749729871749878,
2529
+ "learning_rate": 1.212208657047725e-05,
2530
+ "loss": 0.0086,
2531
+ "step": 3550
2532
+ },
2533
+ {
2534
+ "epoch": 3.947310038824182,
2535
+ "grad_norm": 2.320844888687134,
2536
+ "learning_rate": 1.209988901220866e-05,
2537
+ "loss": 0.0096,
2538
+ "step": 3560
2539
+ },
2540
+ {
2541
+ "epoch": 3.9584026622296173,
2542
+ "grad_norm": 0.5694814324378967,
2543
+ "learning_rate": 1.2077691453940067e-05,
2544
+ "loss": 0.008,
2545
+ "step": 3570
2546
+ },
2547
+ {
2548
+ "epoch": 3.9694952856350527,
2549
+ "grad_norm": 0.9794883728027344,
2550
+ "learning_rate": 1.2055493895671477e-05,
2551
+ "loss": 0.0045,
2552
+ "step": 3580
2553
+ },
2554
+ {
2555
+ "epoch": 3.980587909040488,
2556
+ "grad_norm": 0.2133321315050125,
2557
+ "learning_rate": 1.2033296337402887e-05,
2558
+ "loss": 0.0054,
2559
+ "step": 3590
2560
+ },
2561
+ {
2562
+ "epoch": 3.9916805324459235,
2563
+ "grad_norm": 1.423915147781372,
2564
+ "learning_rate": 1.2011098779134295e-05,
2565
+ "loss": 0.0127,
2566
+ "step": 3600
2567
+ },
2568
+ {
2569
+ "epoch": 4.0,
2570
+ "eval_accuracy": 0.8699763593380615,
2571
+ "eval_f1": 0.6728971962616822,
2572
+ "eval_loss": 0.4916483163833618,
2573
+ "eval_precision": 0.6449511400651465,
2574
+ "eval_recall": 0.7033747779751333,
2575
+ "eval_runtime": 2.7388,
2576
+ "eval_samples_per_second": 360.381,
2577
+ "eval_steps_per_second": 11.319,
2578
+ "step": 3608
2579
+ }
2580
+ ],
2581
+ "logging_steps": 10,
2582
+ "max_steps": 9010,
2583
+ "num_input_tokens_seen": 0,
2584
+ "num_train_epochs": 10,
2585
+ "save_steps": 500,
2586
+ "stateful_callbacks": {
2587
+ "TrainerControl": {
2588
+ "args": {
2589
+ "should_epoch_stop": false,
2590
+ "should_evaluate": false,
2591
+ "should_log": false,
2592
+ "should_save": true,
2593
+ "should_training_stop": false
2594
+ },
2595
+ "attributes": {}
2596
+ }
2597
+ },
2598
+ "total_flos": 3.844979161024907e+16,
2599
+ "train_batch_size": 32,
2600
+ "trial_name": null,
2601
+ "trial_params": null
2602
+ }
checkpoint-3608/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e64fe8f815fa80a7291289cde9af3ebb27f62206c40a93e8adf97661cd289de3
3
+ size 5368
checkpoint-4510/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
checkpoint-4510/config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaV2ForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "hidden_act": "gelu",
7
+ "hidden_dropout_prob": 0.1,
8
+ "hidden_size": 1024,
9
+ "id2label": {
10
+ "0": 0,
11
+ "1": 1,
12
+ "2": 2
13
+ },
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 4096,
16
+ "label2id": {
17
+ "0": 0,
18
+ "1": 1,
19
+ "2": 2
20
+ },
21
+ "layer_norm_eps": 1e-07,
22
+ "legacy": true,
23
+ "max_position_embeddings": 512,
24
+ "max_relative_positions": -1,
25
+ "model_type": "deberta-v2",
26
+ "norm_rel_ebd": "layer_norm",
27
+ "num_attention_heads": 16,
28
+ "num_hidden_layers": 24,
29
+ "pad_token_id": 0,
30
+ "pooler_dropout": 0,
31
+ "pooler_hidden_act": "gelu",
32
+ "pooler_hidden_size": 1024,
33
+ "pos_att_type": [
34
+ "p2c",
35
+ "c2p"
36
+ ],
37
+ "position_biased_input": false,
38
+ "position_buckets": 256,
39
+ "problem_type": "multi_label_classification",
40
+ "relative_attention": true,
41
+ "share_att_key": true,
42
+ "torch_dtype": "float32",
43
+ "transformers_version": "4.51.3",
44
+ "type_vocab_size": 0,
45
+ "vocab_size": 128100
46
+ }
checkpoint-4510/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cc67ea10679edd8661e592012efd34b2a9adb2d6184103664ff8555f6a6a7a0
3
+ size 1740308548
checkpoint-4510/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:491a48665da348e407b98370cdea189a32b84971eb958120f5722ba007be11f4
3
+ size 3480848432
checkpoint-4510/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f0bab22d629c79607f7dc65e5d7670659ba328a7271440f8a0f84396181fd1d
3
+ size 14244
checkpoint-4510/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99cc9f37c9526dcec2abfafc6d9c51496f2ca19c3b36c5ec5b9d022975750bba
3
+ size 1064
checkpoint-4510/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[SEP]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": {
9
+ "content": "[UNK]",
10
+ "lstrip": false,
11
+ "normalized": true,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ }
15
+ }
checkpoint-4510/spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
3
+ size 2464616
checkpoint-4510/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-4510/tokenizer_config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[CLS]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[SEP]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[UNK]",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "128000": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "[CLS]",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "[CLS]",
47
+ "do_lower_case": false,
48
+ "eos_token": "[SEP]",
49
+ "extra_special_tokens": {},
50
+ "mask_token": "[MASK]",
51
+ "model_max_length": 1000000000000000019884624838656,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "sp_model_kwargs": {},
55
+ "split_by_punct": false,
56
+ "tokenizer_class": "DebertaV2Tokenizer",
57
+ "unk_token": "[UNK]",
58
+ "vocab_type": "spm"
59
+ }
checkpoint-4510/trainer_state.json ADDED
@@ -0,0 +1,3251 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 5.0,
6
+ "eval_steps": 500,
7
+ "global_step": 4510,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.011092623405435386,
14
+ "grad_norm": 2.009918689727783,
15
+ "learning_rate": 1.998002219755827e-05,
16
+ "loss": 0.5022,
17
+ "step": 10
18
+ },
19
+ {
20
+ "epoch": 0.022185246810870772,
21
+ "grad_norm": 0.49915584921836853,
22
+ "learning_rate": 1.995782463928968e-05,
23
+ "loss": 0.1943,
24
+ "step": 20
25
+ },
26
+ {
27
+ "epoch": 0.033277870216306155,
28
+ "grad_norm": 0.47061625123023987,
29
+ "learning_rate": 1.993562708102109e-05,
30
+ "loss": 0.1436,
31
+ "step": 30
32
+ },
33
+ {
34
+ "epoch": 0.044370493621741544,
35
+ "grad_norm": 0.2260124832391739,
36
+ "learning_rate": 1.99134295227525e-05,
37
+ "loss": 0.1106,
38
+ "step": 40
39
+ },
40
+ {
41
+ "epoch": 0.05546311702717693,
42
+ "grad_norm": 0.31766048073768616,
43
+ "learning_rate": 1.989123196448391e-05,
44
+ "loss": 0.0994,
45
+ "step": 50
46
+ },
47
+ {
48
+ "epoch": 0.06655574043261231,
49
+ "grad_norm": 0.2788391709327698,
50
+ "learning_rate": 1.9869034406215316e-05,
51
+ "loss": 0.111,
52
+ "step": 60
53
+ },
54
+ {
55
+ "epoch": 0.0776483638380477,
56
+ "grad_norm": 0.24376386404037476,
57
+ "learning_rate": 1.9846836847946726e-05,
58
+ "loss": 0.0849,
59
+ "step": 70
60
+ },
61
+ {
62
+ "epoch": 0.08874098724348309,
63
+ "grad_norm": 0.1789788454771042,
64
+ "learning_rate": 1.982463928967814e-05,
65
+ "loss": 0.0971,
66
+ "step": 80
67
+ },
68
+ {
69
+ "epoch": 0.09983361064891846,
70
+ "grad_norm": 0.29698437452316284,
71
+ "learning_rate": 1.9802441731409546e-05,
72
+ "loss": 0.0856,
73
+ "step": 90
74
+ },
75
+ {
76
+ "epoch": 0.11092623405435385,
77
+ "grad_norm": 0.2553412616252899,
78
+ "learning_rate": 1.9780244173140956e-05,
79
+ "loss": 0.0976,
80
+ "step": 100
81
+ },
82
+ {
83
+ "epoch": 0.12201885745978924,
84
+ "grad_norm": 0.20608791708946228,
85
+ "learning_rate": 1.9758046614872365e-05,
86
+ "loss": 0.0947,
87
+ "step": 110
88
+ },
89
+ {
90
+ "epoch": 0.13311148086522462,
91
+ "grad_norm": 0.23551669716835022,
92
+ "learning_rate": 1.9735849056603775e-05,
93
+ "loss": 0.0965,
94
+ "step": 120
95
+ },
96
+ {
97
+ "epoch": 0.14420410427066002,
98
+ "grad_norm": 0.3207148611545563,
99
+ "learning_rate": 1.9713651498335185e-05,
100
+ "loss": 0.0931,
101
+ "step": 130
102
+ },
103
+ {
104
+ "epoch": 0.1552967276760954,
105
+ "grad_norm": 0.34813177585601807,
106
+ "learning_rate": 1.9691453940066595e-05,
107
+ "loss": 0.1,
108
+ "step": 140
109
+ },
110
+ {
111
+ "epoch": 0.16638935108153077,
112
+ "grad_norm": 0.41501104831695557,
113
+ "learning_rate": 1.9669256381798e-05,
114
+ "loss": 0.0922,
115
+ "step": 150
116
+ },
117
+ {
118
+ "epoch": 0.17748197448696618,
119
+ "grad_norm": 0.24550506472587585,
120
+ "learning_rate": 1.964705882352941e-05,
121
+ "loss": 0.0857,
122
+ "step": 160
123
+ },
124
+ {
125
+ "epoch": 0.18857459789240155,
126
+ "grad_norm": 0.4205056130886078,
127
+ "learning_rate": 1.9624861265260825e-05,
128
+ "loss": 0.0831,
129
+ "step": 170
130
+ },
131
+ {
132
+ "epoch": 0.19966722129783693,
133
+ "grad_norm": 0.6002993583679199,
134
+ "learning_rate": 1.9602663706992235e-05,
135
+ "loss": 0.0792,
136
+ "step": 180
137
+ },
138
+ {
139
+ "epoch": 0.21075984470327233,
140
+ "grad_norm": 0.27535638213157654,
141
+ "learning_rate": 1.958046614872364e-05,
142
+ "loss": 0.0786,
143
+ "step": 190
144
+ },
145
+ {
146
+ "epoch": 0.2218524681087077,
147
+ "grad_norm": 0.41602805256843567,
148
+ "learning_rate": 1.955826859045505e-05,
149
+ "loss": 0.0789,
150
+ "step": 200
151
+ },
152
+ {
153
+ "epoch": 0.23294509151414308,
154
+ "grad_norm": 0.41827332973480225,
155
+ "learning_rate": 1.953607103218646e-05,
156
+ "loss": 0.0663,
157
+ "step": 210
158
+ },
159
+ {
160
+ "epoch": 0.24403771491957849,
161
+ "grad_norm": 0.47976112365722656,
162
+ "learning_rate": 1.951387347391787e-05,
163
+ "loss": 0.0727,
164
+ "step": 220
165
+ },
166
+ {
167
+ "epoch": 0.25513033832501386,
168
+ "grad_norm": 0.7771281599998474,
169
+ "learning_rate": 1.949167591564928e-05,
170
+ "loss": 0.0776,
171
+ "step": 230
172
+ },
173
+ {
174
+ "epoch": 0.26622296173044924,
175
+ "grad_norm": 1.0573914051055908,
176
+ "learning_rate": 1.946947835738069e-05,
177
+ "loss": 0.0828,
178
+ "step": 240
179
+ },
180
+ {
181
+ "epoch": 0.2773155851358846,
182
+ "grad_norm": 0.5186040997505188,
183
+ "learning_rate": 1.9447280799112097e-05,
184
+ "loss": 0.0767,
185
+ "step": 250
186
+ },
187
+ {
188
+ "epoch": 0.28840820854132004,
189
+ "grad_norm": 0.5594862699508667,
190
+ "learning_rate": 1.942508324084351e-05,
191
+ "loss": 0.0699,
192
+ "step": 260
193
+ },
194
+ {
195
+ "epoch": 0.2995008319467554,
196
+ "grad_norm": 0.5147818922996521,
197
+ "learning_rate": 1.940288568257492e-05,
198
+ "loss": 0.0591,
199
+ "step": 270
200
+ },
201
+ {
202
+ "epoch": 0.3105934553521908,
203
+ "grad_norm": 0.5510437488555908,
204
+ "learning_rate": 1.938068812430633e-05,
205
+ "loss": 0.0761,
206
+ "step": 280
207
+ },
208
+ {
209
+ "epoch": 0.32168607875762617,
210
+ "grad_norm": 0.4817625880241394,
211
+ "learning_rate": 1.9358490566037736e-05,
212
+ "loss": 0.0688,
213
+ "step": 290
214
+ },
215
+ {
216
+ "epoch": 0.33277870216306155,
217
+ "grad_norm": 0.504362940788269,
218
+ "learning_rate": 1.9336293007769146e-05,
219
+ "loss": 0.0691,
220
+ "step": 300
221
+ },
222
+ {
223
+ "epoch": 0.343871325568497,
224
+ "grad_norm": 0.5732384920120239,
225
+ "learning_rate": 1.9314095449500556e-05,
226
+ "loss": 0.057,
227
+ "step": 310
228
+ },
229
+ {
230
+ "epoch": 0.35496394897393235,
231
+ "grad_norm": 0.5892298221588135,
232
+ "learning_rate": 1.9291897891231966e-05,
233
+ "loss": 0.059,
234
+ "step": 320
235
+ },
236
+ {
237
+ "epoch": 0.36605657237936773,
238
+ "grad_norm": 0.3680512309074402,
239
+ "learning_rate": 1.9269700332963376e-05,
240
+ "loss": 0.0706,
241
+ "step": 330
242
+ },
243
+ {
244
+ "epoch": 0.3771491957848031,
245
+ "grad_norm": 0.41971662640571594,
246
+ "learning_rate": 1.9247502774694786e-05,
247
+ "loss": 0.053,
248
+ "step": 340
249
+ },
250
+ {
251
+ "epoch": 0.3882418191902385,
252
+ "grad_norm": 0.37261390686035156,
253
+ "learning_rate": 1.9225305216426195e-05,
254
+ "loss": 0.0632,
255
+ "step": 350
256
+ },
257
+ {
258
+ "epoch": 0.39933444259567386,
259
+ "grad_norm": 0.48256734013557434,
260
+ "learning_rate": 1.9203107658157605e-05,
261
+ "loss": 0.0651,
262
+ "step": 360
263
+ },
264
+ {
265
+ "epoch": 0.4104270660011093,
266
+ "grad_norm": 0.7914339303970337,
267
+ "learning_rate": 1.9180910099889015e-05,
268
+ "loss": 0.0659,
269
+ "step": 370
270
+ },
271
+ {
272
+ "epoch": 0.42151968940654466,
273
+ "grad_norm": 0.6772429347038269,
274
+ "learning_rate": 1.915871254162042e-05,
275
+ "loss": 0.0655,
276
+ "step": 380
277
+ },
278
+ {
279
+ "epoch": 0.43261231281198004,
280
+ "grad_norm": 0.5621687173843384,
281
+ "learning_rate": 1.913651498335183e-05,
282
+ "loss": 0.0707,
283
+ "step": 390
284
+ },
285
+ {
286
+ "epoch": 0.4437049362174154,
287
+ "grad_norm": 0.30515748262405396,
288
+ "learning_rate": 1.911431742508324e-05,
289
+ "loss": 0.0548,
290
+ "step": 400
291
+ },
292
+ {
293
+ "epoch": 0.4547975596228508,
294
+ "grad_norm": 0.5506859421730042,
295
+ "learning_rate": 1.909211986681465e-05,
296
+ "loss": 0.0554,
297
+ "step": 410
298
+ },
299
+ {
300
+ "epoch": 0.46589018302828616,
301
+ "grad_norm": 0.6387749314308167,
302
+ "learning_rate": 1.906992230854606e-05,
303
+ "loss": 0.0554,
304
+ "step": 420
305
+ },
306
+ {
307
+ "epoch": 0.4769828064337216,
308
+ "grad_norm": 0.18500734865665436,
309
+ "learning_rate": 1.904772475027747e-05,
310
+ "loss": 0.0558,
311
+ "step": 430
312
+ },
313
+ {
314
+ "epoch": 0.48807542983915697,
315
+ "grad_norm": 0.6815407276153564,
316
+ "learning_rate": 1.902552719200888e-05,
317
+ "loss": 0.0537,
318
+ "step": 440
319
+ },
320
+ {
321
+ "epoch": 0.49916805324459235,
322
+ "grad_norm": 0.4826994836330414,
323
+ "learning_rate": 1.900332963374029e-05,
324
+ "loss": 0.0567,
325
+ "step": 450
326
+ },
327
+ {
328
+ "epoch": 0.5102606766500277,
329
+ "grad_norm": 0.24618124961853027,
330
+ "learning_rate": 1.89811320754717e-05,
331
+ "loss": 0.0614,
332
+ "step": 460
333
+ },
334
+ {
335
+ "epoch": 0.5213533000554631,
336
+ "grad_norm": 1.0037415027618408,
337
+ "learning_rate": 1.895893451720311e-05,
338
+ "loss": 0.0604,
339
+ "step": 470
340
+ },
341
+ {
342
+ "epoch": 0.5324459234608985,
343
+ "grad_norm": 0.6000948548316956,
344
+ "learning_rate": 1.8936736958934517e-05,
345
+ "loss": 0.0683,
346
+ "step": 480
347
+ },
348
+ {
349
+ "epoch": 0.5435385468663338,
350
+ "grad_norm": 0.3302474319934845,
351
+ "learning_rate": 1.8914539400665927e-05,
352
+ "loss": 0.0543,
353
+ "step": 490
354
+ },
355
+ {
356
+ "epoch": 0.5546311702717692,
357
+ "grad_norm": 0.5560783743858337,
358
+ "learning_rate": 1.8892341842397337e-05,
359
+ "loss": 0.0569,
360
+ "step": 500
361
+ },
362
+ {
363
+ "epoch": 0.5657237936772047,
364
+ "grad_norm": 0.8041097521781921,
365
+ "learning_rate": 1.8870144284128747e-05,
366
+ "loss": 0.0544,
367
+ "step": 510
368
+ },
369
+ {
370
+ "epoch": 0.5768164170826401,
371
+ "grad_norm": 0.6846103668212891,
372
+ "learning_rate": 1.8847946725860156e-05,
373
+ "loss": 0.058,
374
+ "step": 520
375
+ },
376
+ {
377
+ "epoch": 0.5879090404880755,
378
+ "grad_norm": 0.50434809923172,
379
+ "learning_rate": 1.8825749167591566e-05,
380
+ "loss": 0.0654,
381
+ "step": 530
382
+ },
383
+ {
384
+ "epoch": 0.5990016638935108,
385
+ "grad_norm": 0.54362553358078,
386
+ "learning_rate": 1.8803551609322976e-05,
387
+ "loss": 0.0582,
388
+ "step": 540
389
+ },
390
+ {
391
+ "epoch": 0.6100942872989462,
392
+ "grad_norm": 0.6166839599609375,
393
+ "learning_rate": 1.8781354051054386e-05,
394
+ "loss": 0.0672,
395
+ "step": 550
396
+ },
397
+ {
398
+ "epoch": 0.6211869107043816,
399
+ "grad_norm": 0.4353054165840149,
400
+ "learning_rate": 1.8759156492785796e-05,
401
+ "loss": 0.0532,
402
+ "step": 560
403
+ },
404
+ {
405
+ "epoch": 0.632279534109817,
406
+ "grad_norm": 0.6025580167770386,
407
+ "learning_rate": 1.8736958934517206e-05,
408
+ "loss": 0.0746,
409
+ "step": 570
410
+ },
411
+ {
412
+ "epoch": 0.6433721575152523,
413
+ "grad_norm": 0.7259892225265503,
414
+ "learning_rate": 1.8714761376248612e-05,
415
+ "loss": 0.0642,
416
+ "step": 580
417
+ },
418
+ {
419
+ "epoch": 0.6544647809206877,
420
+ "grad_norm": 0.4940318465232849,
421
+ "learning_rate": 1.8692563817980022e-05,
422
+ "loss": 0.0547,
423
+ "step": 590
424
+ },
425
+ {
426
+ "epoch": 0.6655574043261231,
427
+ "grad_norm": 0.7005699872970581,
428
+ "learning_rate": 1.8670366259711435e-05,
429
+ "loss": 0.0522,
430
+ "step": 600
431
+ },
432
+ {
433
+ "epoch": 0.6766500277315585,
434
+ "grad_norm": 0.4530707895755768,
435
+ "learning_rate": 1.8648168701442845e-05,
436
+ "loss": 0.054,
437
+ "step": 610
438
+ },
439
+ {
440
+ "epoch": 0.687742651136994,
441
+ "grad_norm": 0.9097110629081726,
442
+ "learning_rate": 1.8625971143174252e-05,
443
+ "loss": 0.0622,
444
+ "step": 620
445
+ },
446
+ {
447
+ "epoch": 0.6988352745424293,
448
+ "grad_norm": 0.5374599695205688,
449
+ "learning_rate": 1.860377358490566e-05,
450
+ "loss": 0.0522,
451
+ "step": 630
452
+ },
453
+ {
454
+ "epoch": 0.7099278979478647,
455
+ "grad_norm": 0.3849945664405823,
456
+ "learning_rate": 1.858157602663707e-05,
457
+ "loss": 0.0571,
458
+ "step": 640
459
+ },
460
+ {
461
+ "epoch": 0.7210205213533001,
462
+ "grad_norm": 0.5918008685112,
463
+ "learning_rate": 1.855937846836848e-05,
464
+ "loss": 0.0576,
465
+ "step": 650
466
+ },
467
+ {
468
+ "epoch": 0.7321131447587355,
469
+ "grad_norm": 0.3229956030845642,
470
+ "learning_rate": 1.853718091009989e-05,
471
+ "loss": 0.0537,
472
+ "step": 660
473
+ },
474
+ {
475
+ "epoch": 0.7432057681641708,
476
+ "grad_norm": 0.5264039039611816,
477
+ "learning_rate": 1.85149833518313e-05,
478
+ "loss": 0.06,
479
+ "step": 670
480
+ },
481
+ {
482
+ "epoch": 0.7542983915696062,
483
+ "grad_norm": 0.36795660853385925,
484
+ "learning_rate": 1.8492785793562708e-05,
485
+ "loss": 0.0511,
486
+ "step": 680
487
+ },
488
+ {
489
+ "epoch": 0.7653910149750416,
490
+ "grad_norm": 0.5905130505561829,
491
+ "learning_rate": 1.847058823529412e-05,
492
+ "loss": 0.0636,
493
+ "step": 690
494
+ },
495
+ {
496
+ "epoch": 0.776483638380477,
497
+ "grad_norm": 0.36266571283340454,
498
+ "learning_rate": 1.844839067702553e-05,
499
+ "loss": 0.0598,
500
+ "step": 700
501
+ },
502
+ {
503
+ "epoch": 0.7875762617859123,
504
+ "grad_norm": 0.4978592097759247,
505
+ "learning_rate": 1.8426193118756937e-05,
506
+ "loss": 0.0607,
507
+ "step": 710
508
+ },
509
+ {
510
+ "epoch": 0.7986688851913477,
511
+ "grad_norm": 0.4635021686553955,
512
+ "learning_rate": 1.8403995560488347e-05,
513
+ "loss": 0.0543,
514
+ "step": 720
515
+ },
516
+ {
517
+ "epoch": 0.8097615085967831,
518
+ "grad_norm": 0.44571858644485474,
519
+ "learning_rate": 1.8381798002219757e-05,
520
+ "loss": 0.0678,
521
+ "step": 730
522
+ },
523
+ {
524
+ "epoch": 0.8208541320022186,
525
+ "grad_norm": 0.8265877366065979,
526
+ "learning_rate": 1.8359600443951167e-05,
527
+ "loss": 0.0552,
528
+ "step": 740
529
+ },
530
+ {
531
+ "epoch": 0.831946755407654,
532
+ "grad_norm": 0.5776472091674805,
533
+ "learning_rate": 1.8337402885682577e-05,
534
+ "loss": 0.0505,
535
+ "step": 750
536
+ },
537
+ {
538
+ "epoch": 0.8430393788130893,
539
+ "grad_norm": 0.299274742603302,
540
+ "learning_rate": 1.8315205327413986e-05,
541
+ "loss": 0.0619,
542
+ "step": 760
543
+ },
544
+ {
545
+ "epoch": 0.8541320022185247,
546
+ "grad_norm": 0.587645947933197,
547
+ "learning_rate": 1.8293007769145393e-05,
548
+ "loss": 0.0575,
549
+ "step": 770
550
+ },
551
+ {
552
+ "epoch": 0.8652246256239601,
553
+ "grad_norm": 0.39164137840270996,
554
+ "learning_rate": 1.8270810210876806e-05,
555
+ "loss": 0.0458,
556
+ "step": 780
557
+ },
558
+ {
559
+ "epoch": 0.8763172490293955,
560
+ "grad_norm": 0.4663292169570923,
561
+ "learning_rate": 1.8248612652608216e-05,
562
+ "loss": 0.048,
563
+ "step": 790
564
+ },
565
+ {
566
+ "epoch": 0.8874098724348308,
567
+ "grad_norm": 0.5804581642150879,
568
+ "learning_rate": 1.8226415094339626e-05,
569
+ "loss": 0.0557,
570
+ "step": 800
571
+ },
572
+ {
573
+ "epoch": 0.8985024958402662,
574
+ "grad_norm": 0.4279440939426422,
575
+ "learning_rate": 1.8204217536071032e-05,
576
+ "loss": 0.051,
577
+ "step": 810
578
+ },
579
+ {
580
+ "epoch": 0.9095951192457016,
581
+ "grad_norm": 0.5384302735328674,
582
+ "learning_rate": 1.8182019977802442e-05,
583
+ "loss": 0.0558,
584
+ "step": 820
585
+ },
586
+ {
587
+ "epoch": 0.920687742651137,
588
+ "grad_norm": 0.5049973726272583,
589
+ "learning_rate": 1.8159822419533852e-05,
590
+ "loss": 0.047,
591
+ "step": 830
592
+ },
593
+ {
594
+ "epoch": 0.9317803660565723,
595
+ "grad_norm": 0.8061177134513855,
596
+ "learning_rate": 1.8137624861265262e-05,
597
+ "loss": 0.0572,
598
+ "step": 840
599
+ },
600
+ {
601
+ "epoch": 0.9428729894620078,
602
+ "grad_norm": 0.6056540012359619,
603
+ "learning_rate": 1.8115427302996672e-05,
604
+ "loss": 0.0539,
605
+ "step": 850
606
+ },
607
+ {
608
+ "epoch": 0.9539656128674432,
609
+ "grad_norm": 0.3001384139060974,
610
+ "learning_rate": 1.8093229744728082e-05,
611
+ "loss": 0.053,
612
+ "step": 860
613
+ },
614
+ {
615
+ "epoch": 0.9650582362728786,
616
+ "grad_norm": 0.6709749102592468,
617
+ "learning_rate": 1.807103218645949e-05,
618
+ "loss": 0.0526,
619
+ "step": 870
620
+ },
621
+ {
622
+ "epoch": 0.9761508596783139,
623
+ "grad_norm": 0.8233507871627808,
624
+ "learning_rate": 1.80488346281909e-05,
625
+ "loss": 0.0621,
626
+ "step": 880
627
+ },
628
+ {
629
+ "epoch": 0.9872434830837493,
630
+ "grad_norm": 0.5757150650024414,
631
+ "learning_rate": 1.802663706992231e-05,
632
+ "loss": 0.0546,
633
+ "step": 890
634
+ },
635
+ {
636
+ "epoch": 0.9983361064891847,
637
+ "grad_norm": 0.6834889054298401,
638
+ "learning_rate": 1.800443951165372e-05,
639
+ "loss": 0.0462,
640
+ "step": 900
641
+ },
642
+ {
643
+ "epoch": 1.0,
644
+ "eval_accuracy": 0.8720027017899359,
645
+ "eval_f1": 0.5678449258836944,
646
+ "eval_loss": 0.34414270520210266,
647
+ "eval_precision": 0.7929936305732485,
648
+ "eval_recall": 0.4422735346358792,
649
+ "eval_runtime": 2.727,
650
+ "eval_samples_per_second": 361.937,
651
+ "eval_steps_per_second": 11.368,
652
+ "step": 902
653
+ },
654
+ {
655
+ "epoch": 1.0088740987243483,
656
+ "grad_norm": 0.4411003887653351,
657
+ "learning_rate": 1.7982241953385128e-05,
658
+ "loss": 0.0312,
659
+ "step": 910
660
+ },
661
+ {
662
+ "epoch": 1.0199667221297837,
663
+ "grad_norm": 0.40500083565711975,
664
+ "learning_rate": 1.7960044395116538e-05,
665
+ "loss": 0.0345,
666
+ "step": 920
667
+ },
668
+ {
669
+ "epoch": 1.031059345535219,
670
+ "grad_norm": 0.9287449717521667,
671
+ "learning_rate": 1.7937846836847947e-05,
672
+ "loss": 0.0419,
673
+ "step": 930
674
+ },
675
+ {
676
+ "epoch": 1.0421519689406544,
677
+ "grad_norm": 0.7969145178794861,
678
+ "learning_rate": 1.7915649278579357e-05,
679
+ "loss": 0.042,
680
+ "step": 940
681
+ },
682
+ {
683
+ "epoch": 1.0532445923460898,
684
+ "grad_norm": 0.49280259013175964,
685
+ "learning_rate": 1.7893451720310767e-05,
686
+ "loss": 0.0427,
687
+ "step": 950
688
+ },
689
+ {
690
+ "epoch": 1.0643372157515252,
691
+ "grad_norm": 1.1647624969482422,
692
+ "learning_rate": 1.7871254162042177e-05,
693
+ "loss": 0.0341,
694
+ "step": 960
695
+ },
696
+ {
697
+ "epoch": 1.0754298391569606,
698
+ "grad_norm": 0.2908968925476074,
699
+ "learning_rate": 1.7849056603773587e-05,
700
+ "loss": 0.0389,
701
+ "step": 970
702
+ },
703
+ {
704
+ "epoch": 1.086522462562396,
705
+ "grad_norm": 0.521218478679657,
706
+ "learning_rate": 1.7826859045504997e-05,
707
+ "loss": 0.0396,
708
+ "step": 980
709
+ },
710
+ {
711
+ "epoch": 1.0976150859678313,
712
+ "grad_norm": 0.4156598150730133,
713
+ "learning_rate": 1.7804661487236407e-05,
714
+ "loss": 0.0359,
715
+ "step": 990
716
+ },
717
+ {
718
+ "epoch": 1.1087077093732667,
719
+ "grad_norm": 0.7260242104530334,
720
+ "learning_rate": 1.7782463928967813e-05,
721
+ "loss": 0.0384,
722
+ "step": 1000
723
+ },
724
+ {
725
+ "epoch": 1.119800332778702,
726
+ "grad_norm": 0.43658459186553955,
727
+ "learning_rate": 1.7760266370699223e-05,
728
+ "loss": 0.0383,
729
+ "step": 1010
730
+ },
731
+ {
732
+ "epoch": 1.1308929561841374,
733
+ "grad_norm": 0.3585101068019867,
734
+ "learning_rate": 1.7738068812430633e-05,
735
+ "loss": 0.0401,
736
+ "step": 1020
737
+ },
738
+ {
739
+ "epoch": 1.141985579589573,
740
+ "grad_norm": 0.5010389089584351,
741
+ "learning_rate": 1.7715871254162043e-05,
742
+ "loss": 0.0307,
743
+ "step": 1030
744
+ },
745
+ {
746
+ "epoch": 1.1530782029950084,
747
+ "grad_norm": 0.6415812969207764,
748
+ "learning_rate": 1.7693673695893453e-05,
749
+ "loss": 0.0423,
750
+ "step": 1040
751
+ },
752
+ {
753
+ "epoch": 1.1641708264004438,
754
+ "grad_norm": 0.6393259763717651,
755
+ "learning_rate": 1.7671476137624862e-05,
756
+ "loss": 0.0392,
757
+ "step": 1050
758
+ },
759
+ {
760
+ "epoch": 1.1752634498058792,
761
+ "grad_norm": 0.6339041590690613,
762
+ "learning_rate": 1.7649278579356272e-05,
763
+ "loss": 0.0506,
764
+ "step": 1060
765
+ },
766
+ {
767
+ "epoch": 1.1863560732113145,
768
+ "grad_norm": 0.6086763143539429,
769
+ "learning_rate": 1.7627081021087682e-05,
770
+ "loss": 0.0416,
771
+ "step": 1070
772
+ },
773
+ {
774
+ "epoch": 1.19744869661675,
775
+ "grad_norm": 0.6648682355880737,
776
+ "learning_rate": 1.7604883462819092e-05,
777
+ "loss": 0.0379,
778
+ "step": 1080
779
+ },
780
+ {
781
+ "epoch": 1.2085413200221853,
782
+ "grad_norm": 0.6018221378326416,
783
+ "learning_rate": 1.7582685904550502e-05,
784
+ "loss": 0.0354,
785
+ "step": 1090
786
+ },
787
+ {
788
+ "epoch": 1.2196339434276207,
789
+ "grad_norm": 1.2366654872894287,
790
+ "learning_rate": 1.756048834628191e-05,
791
+ "loss": 0.037,
792
+ "step": 1100
793
+ },
794
+ {
795
+ "epoch": 1.230726566833056,
796
+ "grad_norm": 0.5976310968399048,
797
+ "learning_rate": 1.7538290788013318e-05,
798
+ "loss": 0.0361,
799
+ "step": 1110
800
+ },
801
+ {
802
+ "epoch": 1.2418191902384914,
803
+ "grad_norm": 0.8290308117866516,
804
+ "learning_rate": 1.751609322974473e-05,
805
+ "loss": 0.0383,
806
+ "step": 1120
807
+ },
808
+ {
809
+ "epoch": 1.2529118136439268,
810
+ "grad_norm": 0.382548987865448,
811
+ "learning_rate": 1.749389567147614e-05,
812
+ "loss": 0.0449,
813
+ "step": 1130
814
+ },
815
+ {
816
+ "epoch": 1.2640044370493622,
817
+ "grad_norm": 0.5307976007461548,
818
+ "learning_rate": 1.7471698113207548e-05,
819
+ "loss": 0.0364,
820
+ "step": 1140
821
+ },
822
+ {
823
+ "epoch": 1.2750970604547975,
824
+ "grad_norm": 0.5508521795272827,
825
+ "learning_rate": 1.7449500554938958e-05,
826
+ "loss": 0.0312,
827
+ "step": 1150
828
+ },
829
+ {
830
+ "epoch": 1.286189683860233,
831
+ "grad_norm": 0.37057268619537354,
832
+ "learning_rate": 1.7427302996670368e-05,
833
+ "loss": 0.0397,
834
+ "step": 1160
835
+ },
836
+ {
837
+ "epoch": 1.2972823072656683,
838
+ "grad_norm": 0.4806898236274719,
839
+ "learning_rate": 1.7405105438401777e-05,
840
+ "loss": 0.0391,
841
+ "step": 1170
842
+ },
843
+ {
844
+ "epoch": 1.3083749306711037,
845
+ "grad_norm": 0.759772002696991,
846
+ "learning_rate": 1.7382907880133187e-05,
847
+ "loss": 0.0322,
848
+ "step": 1180
849
+ },
850
+ {
851
+ "epoch": 1.319467554076539,
852
+ "grad_norm": 0.39819085597991943,
853
+ "learning_rate": 1.7360710321864597e-05,
854
+ "loss": 0.0345,
855
+ "step": 1190
856
+ },
857
+ {
858
+ "epoch": 1.3305601774819744,
859
+ "grad_norm": 0.7733897566795349,
860
+ "learning_rate": 1.7338512763596004e-05,
861
+ "loss": 0.0424,
862
+ "step": 1200
863
+ },
864
+ {
865
+ "epoch": 1.3416528008874098,
866
+ "grad_norm": 0.7979075908660889,
867
+ "learning_rate": 1.7316315205327417e-05,
868
+ "loss": 0.0403,
869
+ "step": 1210
870
+ },
871
+ {
872
+ "epoch": 1.3527454242928454,
873
+ "grad_norm": 0.5658752918243408,
874
+ "learning_rate": 1.7294117647058827e-05,
875
+ "loss": 0.0276,
876
+ "step": 1220
877
+ },
878
+ {
879
+ "epoch": 1.3638380476982808,
880
+ "grad_norm": 0.25207844376564026,
881
+ "learning_rate": 1.7271920088790237e-05,
882
+ "loss": 0.0347,
883
+ "step": 1230
884
+ },
885
+ {
886
+ "epoch": 1.3749306711037161,
887
+ "grad_norm": 0.6338945031166077,
888
+ "learning_rate": 1.7249722530521643e-05,
889
+ "loss": 0.0416,
890
+ "step": 1240
891
+ },
892
+ {
893
+ "epoch": 1.3860232945091515,
894
+ "grad_norm": 0.8125913143157959,
895
+ "learning_rate": 1.7227524972253053e-05,
896
+ "loss": 0.0442,
897
+ "step": 1250
898
+ },
899
+ {
900
+ "epoch": 1.397115917914587,
901
+ "grad_norm": 0.5067325234413147,
902
+ "learning_rate": 1.7205327413984463e-05,
903
+ "loss": 0.0358,
904
+ "step": 1260
905
+ },
906
+ {
907
+ "epoch": 1.4082085413200223,
908
+ "grad_norm": 0.40132638812065125,
909
+ "learning_rate": 1.7183129855715873e-05,
910
+ "loss": 0.0303,
911
+ "step": 1270
912
+ },
913
+ {
914
+ "epoch": 1.4193011647254576,
915
+ "grad_norm": 0.6337258815765381,
916
+ "learning_rate": 1.7160932297447283e-05,
917
+ "loss": 0.0471,
918
+ "step": 1280
919
+ },
920
+ {
921
+ "epoch": 1.430393788130893,
922
+ "grad_norm": 0.43456393480300903,
923
+ "learning_rate": 1.713873473917869e-05,
924
+ "loss": 0.0354,
925
+ "step": 1290
926
+ },
927
+ {
928
+ "epoch": 1.4414864115363284,
929
+ "grad_norm": 0.3291069269180298,
930
+ "learning_rate": 1.7116537180910102e-05,
931
+ "loss": 0.0312,
932
+ "step": 1300
933
+ },
934
+ {
935
+ "epoch": 1.4525790349417638,
936
+ "grad_norm": 0.2960388660430908,
937
+ "learning_rate": 1.7094339622641512e-05,
938
+ "loss": 0.0362,
939
+ "step": 1310
940
+ },
941
+ {
942
+ "epoch": 1.4636716583471991,
943
+ "grad_norm": 0.820751428604126,
944
+ "learning_rate": 1.7072142064372922e-05,
945
+ "loss": 0.0396,
946
+ "step": 1320
947
+ },
948
+ {
949
+ "epoch": 1.4747642817526345,
950
+ "grad_norm": 0.7335907816886902,
951
+ "learning_rate": 1.704994450610433e-05,
952
+ "loss": 0.0386,
953
+ "step": 1330
954
+ },
955
+ {
956
+ "epoch": 1.48585690515807,
957
+ "grad_norm": 0.40592002868652344,
958
+ "learning_rate": 1.702774694783574e-05,
959
+ "loss": 0.0358,
960
+ "step": 1340
961
+ },
962
+ {
963
+ "epoch": 1.4969495285635053,
964
+ "grad_norm": 0.6194770336151123,
965
+ "learning_rate": 1.7005549389567148e-05,
966
+ "loss": 0.0361,
967
+ "step": 1350
968
+ },
969
+ {
970
+ "epoch": 1.5080421519689406,
971
+ "grad_norm": 0.44590774178504944,
972
+ "learning_rate": 1.6983351831298558e-05,
973
+ "loss": 0.033,
974
+ "step": 1360
975
+ },
976
+ {
977
+ "epoch": 1.519134775374376,
978
+ "grad_norm": 0.5857370495796204,
979
+ "learning_rate": 1.6961154273029968e-05,
980
+ "loss": 0.0423,
981
+ "step": 1370
982
+ },
983
+ {
984
+ "epoch": 1.5302273987798114,
985
+ "grad_norm": 0.6809953451156616,
986
+ "learning_rate": 1.6938956714761378e-05,
987
+ "loss": 0.0345,
988
+ "step": 1380
989
+ },
990
+ {
991
+ "epoch": 1.5413200221852468,
992
+ "grad_norm": 0.7102778553962708,
993
+ "learning_rate": 1.6916759156492788e-05,
994
+ "loss": 0.0372,
995
+ "step": 1390
996
+ },
997
+ {
998
+ "epoch": 1.5524126455906821,
999
+ "grad_norm": 0.31781554222106934,
1000
+ "learning_rate": 1.6894561598224198e-05,
1001
+ "loss": 0.032,
1002
+ "step": 1400
1003
+ },
1004
+ {
1005
+ "epoch": 1.5635052689961175,
1006
+ "grad_norm": 0.4304943382740021,
1007
+ "learning_rate": 1.6872364039955607e-05,
1008
+ "loss": 0.0375,
1009
+ "step": 1410
1010
+ },
1011
+ {
1012
+ "epoch": 1.574597892401553,
1013
+ "grad_norm": 0.918550968170166,
1014
+ "learning_rate": 1.6850166481687017e-05,
1015
+ "loss": 0.0416,
1016
+ "step": 1420
1017
+ },
1018
+ {
1019
+ "epoch": 1.5856905158069883,
1020
+ "grad_norm": 0.722892701625824,
1021
+ "learning_rate": 1.6827968923418424e-05,
1022
+ "loss": 0.0359,
1023
+ "step": 1430
1024
+ },
1025
+ {
1026
+ "epoch": 1.5967831392124237,
1027
+ "grad_norm": 0.508703351020813,
1028
+ "learning_rate": 1.6805771365149834e-05,
1029
+ "loss": 0.0402,
1030
+ "step": 1440
1031
+ },
1032
+ {
1033
+ "epoch": 1.607875762617859,
1034
+ "grad_norm": 0.4725389778614044,
1035
+ "learning_rate": 1.6783573806881244e-05,
1036
+ "loss": 0.0335,
1037
+ "step": 1450
1038
+ },
1039
+ {
1040
+ "epoch": 1.6189683860232944,
1041
+ "grad_norm": 0.4699971079826355,
1042
+ "learning_rate": 1.6761376248612653e-05,
1043
+ "loss": 0.0277,
1044
+ "step": 1460
1045
+ },
1046
+ {
1047
+ "epoch": 1.6300610094287298,
1048
+ "grad_norm": 0.775764524936676,
1049
+ "learning_rate": 1.6739178690344063e-05,
1050
+ "loss": 0.0333,
1051
+ "step": 1470
1052
+ },
1053
+ {
1054
+ "epoch": 1.6411536328341652,
1055
+ "grad_norm": 0.6115106344223022,
1056
+ "learning_rate": 1.6716981132075473e-05,
1057
+ "loss": 0.0301,
1058
+ "step": 1480
1059
+ },
1060
+ {
1061
+ "epoch": 1.6522462562396005,
1062
+ "grad_norm": 0.743077278137207,
1063
+ "learning_rate": 1.6694783573806883e-05,
1064
+ "loss": 0.035,
1065
+ "step": 1490
1066
+ },
1067
+ {
1068
+ "epoch": 1.663338879645036,
1069
+ "grad_norm": 0.5189201235771179,
1070
+ "learning_rate": 1.6672586015538293e-05,
1071
+ "loss": 0.0475,
1072
+ "step": 1500
1073
+ },
1074
+ {
1075
+ "epoch": 1.6744315030504713,
1076
+ "grad_norm": 0.5089607238769531,
1077
+ "learning_rate": 1.6650388457269703e-05,
1078
+ "loss": 0.0377,
1079
+ "step": 1510
1080
+ },
1081
+ {
1082
+ "epoch": 1.6855241264559067,
1083
+ "grad_norm": 0.46067437529563904,
1084
+ "learning_rate": 1.6628190899001113e-05,
1085
+ "loss": 0.0297,
1086
+ "step": 1520
1087
+ },
1088
+ {
1089
+ "epoch": 1.6966167498613423,
1090
+ "grad_norm": 0.5661717653274536,
1091
+ "learning_rate": 1.660599334073252e-05,
1092
+ "loss": 0.0372,
1093
+ "step": 1530
1094
+ },
1095
+ {
1096
+ "epoch": 1.7077093732667776,
1097
+ "grad_norm": 0.45938414335250854,
1098
+ "learning_rate": 1.658379578246393e-05,
1099
+ "loss": 0.0371,
1100
+ "step": 1540
1101
+ },
1102
+ {
1103
+ "epoch": 1.718801996672213,
1104
+ "grad_norm": 0.7390128970146179,
1105
+ "learning_rate": 1.656159822419534e-05,
1106
+ "loss": 0.0332,
1107
+ "step": 1550
1108
+ },
1109
+ {
1110
+ "epoch": 1.7298946200776484,
1111
+ "grad_norm": 0.6475571990013123,
1112
+ "learning_rate": 1.653940066592675e-05,
1113
+ "loss": 0.0352,
1114
+ "step": 1560
1115
+ },
1116
+ {
1117
+ "epoch": 1.7409872434830838,
1118
+ "grad_norm": 0.8529049754142761,
1119
+ "learning_rate": 1.651720310765816e-05,
1120
+ "loss": 0.0386,
1121
+ "step": 1570
1122
+ },
1123
+ {
1124
+ "epoch": 1.7520798668885191,
1125
+ "grad_norm": 0.6980950236320496,
1126
+ "learning_rate": 1.649500554938957e-05,
1127
+ "loss": 0.0358,
1128
+ "step": 1580
1129
+ },
1130
+ {
1131
+ "epoch": 1.7631724902939545,
1132
+ "grad_norm": 0.5858293175697327,
1133
+ "learning_rate": 1.6472807991120978e-05,
1134
+ "loss": 0.0418,
1135
+ "step": 1590
1136
+ },
1137
+ {
1138
+ "epoch": 1.7742651136993899,
1139
+ "grad_norm": 0.49686577916145325,
1140
+ "learning_rate": 1.6450610432852388e-05,
1141
+ "loss": 0.0361,
1142
+ "step": 1600
1143
+ },
1144
+ {
1145
+ "epoch": 1.7853577371048253,
1146
+ "grad_norm": 0.3631349205970764,
1147
+ "learning_rate": 1.6428412874583798e-05,
1148
+ "loss": 0.0352,
1149
+ "step": 1610
1150
+ },
1151
+ {
1152
+ "epoch": 1.7964503605102606,
1153
+ "grad_norm": 0.5329940915107727,
1154
+ "learning_rate": 1.6406215316315204e-05,
1155
+ "loss": 0.04,
1156
+ "step": 1620
1157
+ },
1158
+ {
1159
+ "epoch": 1.807542983915696,
1160
+ "grad_norm": 0.5995050668716431,
1161
+ "learning_rate": 1.6384017758046614e-05,
1162
+ "loss": 0.0368,
1163
+ "step": 1630
1164
+ },
1165
+ {
1166
+ "epoch": 1.8186356073211316,
1167
+ "grad_norm": 1.3379504680633545,
1168
+ "learning_rate": 1.6361820199778028e-05,
1169
+ "loss": 0.0397,
1170
+ "step": 1640
1171
+ },
1172
+ {
1173
+ "epoch": 1.829728230726567,
1174
+ "grad_norm": 0.8026002645492554,
1175
+ "learning_rate": 1.6339622641509437e-05,
1176
+ "loss": 0.0318,
1177
+ "step": 1650
1178
+ },
1179
+ {
1180
+ "epoch": 1.8408208541320024,
1181
+ "grad_norm": 0.917950451374054,
1182
+ "learning_rate": 1.6317425083240844e-05,
1183
+ "loss": 0.0352,
1184
+ "step": 1660
1185
+ },
1186
+ {
1187
+ "epoch": 1.8519134775374377,
1188
+ "grad_norm": 0.6763226389884949,
1189
+ "learning_rate": 1.6295227524972254e-05,
1190
+ "loss": 0.0413,
1191
+ "step": 1670
1192
+ },
1193
+ {
1194
+ "epoch": 1.8630061009428731,
1195
+ "grad_norm": 0.5728912949562073,
1196
+ "learning_rate": 1.6273029966703664e-05,
1197
+ "loss": 0.0291,
1198
+ "step": 1680
1199
+ },
1200
+ {
1201
+ "epoch": 1.8740987243483085,
1202
+ "grad_norm": 0.6339443325996399,
1203
+ "learning_rate": 1.6250832408435074e-05,
1204
+ "loss": 0.0409,
1205
+ "step": 1690
1206
+ },
1207
+ {
1208
+ "epoch": 1.8851913477537439,
1209
+ "grad_norm": 0.6930853128433228,
1210
+ "learning_rate": 1.6228634850166483e-05,
1211
+ "loss": 0.0371,
1212
+ "step": 1700
1213
+ },
1214
+ {
1215
+ "epoch": 1.8962839711591792,
1216
+ "grad_norm": 0.37715452909469604,
1217
+ "learning_rate": 1.6206437291897893e-05,
1218
+ "loss": 0.031,
1219
+ "step": 1710
1220
+ },
1221
+ {
1222
+ "epoch": 1.9073765945646146,
1223
+ "grad_norm": 0.5656572580337524,
1224
+ "learning_rate": 1.61842397336293e-05,
1225
+ "loss": 0.0414,
1226
+ "step": 1720
1227
+ },
1228
+ {
1229
+ "epoch": 1.91846921797005,
1230
+ "grad_norm": 0.3563915491104126,
1231
+ "learning_rate": 1.6162042175360713e-05,
1232
+ "loss": 0.0276,
1233
+ "step": 1730
1234
+ },
1235
+ {
1236
+ "epoch": 1.9295618413754854,
1237
+ "grad_norm": 0.8260300159454346,
1238
+ "learning_rate": 1.6139844617092123e-05,
1239
+ "loss": 0.0337,
1240
+ "step": 1740
1241
+ },
1242
+ {
1243
+ "epoch": 1.9406544647809207,
1244
+ "grad_norm": 0.6907551884651184,
1245
+ "learning_rate": 1.6117647058823533e-05,
1246
+ "loss": 0.031,
1247
+ "step": 1750
1248
+ },
1249
+ {
1250
+ "epoch": 1.9517470881863561,
1251
+ "grad_norm": 0.5782826542854309,
1252
+ "learning_rate": 1.609544950055494e-05,
1253
+ "loss": 0.0419,
1254
+ "step": 1760
1255
+ },
1256
+ {
1257
+ "epoch": 1.9628397115917915,
1258
+ "grad_norm": 0.36368125677108765,
1259
+ "learning_rate": 1.607325194228635e-05,
1260
+ "loss": 0.0356,
1261
+ "step": 1770
1262
+ },
1263
+ {
1264
+ "epoch": 1.9739323349972269,
1265
+ "grad_norm": 0.369911789894104,
1266
+ "learning_rate": 1.605105438401776e-05,
1267
+ "loss": 0.0264,
1268
+ "step": 1780
1269
+ },
1270
+ {
1271
+ "epoch": 1.9850249584026622,
1272
+ "grad_norm": 0.6072190999984741,
1273
+ "learning_rate": 1.602885682574917e-05,
1274
+ "loss": 0.0366,
1275
+ "step": 1790
1276
+ },
1277
+ {
1278
+ "epoch": 1.9961175818080976,
1279
+ "grad_norm": 0.9733797311782837,
1280
+ "learning_rate": 1.600665926748058e-05,
1281
+ "loss": 0.0371,
1282
+ "step": 1800
1283
+ },
1284
+ {
1285
+ "epoch": 2.0,
1286
+ "eval_accuracy": 0.8973319824383654,
1287
+ "eval_f1": 0.6872427983539094,
1288
+ "eval_loss": 0.31070804595947266,
1289
+ "eval_precision": 0.8166259168704156,
1290
+ "eval_recall": 0.5932504440497336,
1291
+ "eval_runtime": 2.7114,
1292
+ "eval_samples_per_second": 364.014,
1293
+ "eval_steps_per_second": 11.433,
1294
+ "step": 1804
1295
+ },
1296
+ {
1297
+ "epoch": 2.0066555740432612,
1298
+ "grad_norm": 0.2766992449760437,
1299
+ "learning_rate": 1.598446170921199e-05,
1300
+ "loss": 0.0218,
1301
+ "step": 1810
1302
+ },
1303
+ {
1304
+ "epoch": 2.0177481974486966,
1305
+ "grad_norm": 0.6556938290596008,
1306
+ "learning_rate": 1.59622641509434e-05,
1307
+ "loss": 0.0187,
1308
+ "step": 1820
1309
+ },
1310
+ {
1311
+ "epoch": 2.028840820854132,
1312
+ "grad_norm": 0.24634911119937897,
1313
+ "learning_rate": 1.5940066592674808e-05,
1314
+ "loss": 0.0132,
1315
+ "step": 1830
1316
+ },
1317
+ {
1318
+ "epoch": 2.0399334442595674,
1319
+ "grad_norm": 0.683587372303009,
1320
+ "learning_rate": 1.5917869034406218e-05,
1321
+ "loss": 0.0158,
1322
+ "step": 1840
1323
+ },
1324
+ {
1325
+ "epoch": 2.0510260676650027,
1326
+ "grad_norm": 0.47117751836776733,
1327
+ "learning_rate": 1.5895671476137625e-05,
1328
+ "loss": 0.018,
1329
+ "step": 1850
1330
+ },
1331
+ {
1332
+ "epoch": 2.062118691070438,
1333
+ "grad_norm": 0.4102018475532532,
1334
+ "learning_rate": 1.5873473917869034e-05,
1335
+ "loss": 0.0153,
1336
+ "step": 1860
1337
+ },
1338
+ {
1339
+ "epoch": 2.0732113144758735,
1340
+ "grad_norm": 0.4639064371585846,
1341
+ "learning_rate": 1.5851276359600444e-05,
1342
+ "loss": 0.0171,
1343
+ "step": 1870
1344
+ },
1345
+ {
1346
+ "epoch": 2.084303937881309,
1347
+ "grad_norm": 0.49497199058532715,
1348
+ "learning_rate": 1.5829078801331854e-05,
1349
+ "loss": 0.0168,
1350
+ "step": 1880
1351
+ },
1352
+ {
1353
+ "epoch": 2.0953965612867442,
1354
+ "grad_norm": 0.6302080154418945,
1355
+ "learning_rate": 1.5806881243063264e-05,
1356
+ "loss": 0.0191,
1357
+ "step": 1890
1358
+ },
1359
+ {
1360
+ "epoch": 2.1064891846921796,
1361
+ "grad_norm": 0.3464473783969879,
1362
+ "learning_rate": 1.5784683684794674e-05,
1363
+ "loss": 0.0184,
1364
+ "step": 1900
1365
+ },
1366
+ {
1367
+ "epoch": 2.117581808097615,
1368
+ "grad_norm": 0.6507964730262756,
1369
+ "learning_rate": 1.5762486126526084e-05,
1370
+ "loss": 0.0135,
1371
+ "step": 1910
1372
+ },
1373
+ {
1374
+ "epoch": 2.1286744315030504,
1375
+ "grad_norm": 0.1759006232023239,
1376
+ "learning_rate": 1.5740288568257494e-05,
1377
+ "loss": 0.0118,
1378
+ "step": 1920
1379
+ },
1380
+ {
1381
+ "epoch": 2.1397670549084857,
1382
+ "grad_norm": 0.5952832698822021,
1383
+ "learning_rate": 1.5718091009988904e-05,
1384
+ "loss": 0.0157,
1385
+ "step": 1930
1386
+ },
1387
+ {
1388
+ "epoch": 2.150859678313921,
1389
+ "grad_norm": 0.8094580769538879,
1390
+ "learning_rate": 1.5695893451720313e-05,
1391
+ "loss": 0.0125,
1392
+ "step": 1940
1393
+ },
1394
+ {
1395
+ "epoch": 2.1619523017193565,
1396
+ "grad_norm": 0.6284286379814148,
1397
+ "learning_rate": 1.567369589345172e-05,
1398
+ "loss": 0.0178,
1399
+ "step": 1950
1400
+ },
1401
+ {
1402
+ "epoch": 2.173044925124792,
1403
+ "grad_norm": 0.6017957925796509,
1404
+ "learning_rate": 1.565149833518313e-05,
1405
+ "loss": 0.0136,
1406
+ "step": 1960
1407
+ },
1408
+ {
1409
+ "epoch": 2.1841375485302272,
1410
+ "grad_norm": 1.2836827039718628,
1411
+ "learning_rate": 1.562930077691454e-05,
1412
+ "loss": 0.014,
1413
+ "step": 1970
1414
+ },
1415
+ {
1416
+ "epoch": 2.1952301719356626,
1417
+ "grad_norm": 0.6213756203651428,
1418
+ "learning_rate": 1.560710321864595e-05,
1419
+ "loss": 0.0136,
1420
+ "step": 1980
1421
+ },
1422
+ {
1423
+ "epoch": 2.206322795341098,
1424
+ "grad_norm": 0.48122018575668335,
1425
+ "learning_rate": 1.558490566037736e-05,
1426
+ "loss": 0.0147,
1427
+ "step": 1990
1428
+ },
1429
+ {
1430
+ "epoch": 2.2174154187465334,
1431
+ "grad_norm": 0.7161231637001038,
1432
+ "learning_rate": 1.556270810210877e-05,
1433
+ "loss": 0.0105,
1434
+ "step": 2000
1435
+ },
1436
+ {
1437
+ "epoch": 2.2285080421519687,
1438
+ "grad_norm": 0.6598195433616638,
1439
+ "learning_rate": 1.554051054384018e-05,
1440
+ "loss": 0.0172,
1441
+ "step": 2010
1442
+ },
1443
+ {
1444
+ "epoch": 2.239600665557404,
1445
+ "grad_norm": 0.6077088117599487,
1446
+ "learning_rate": 1.551831298557159e-05,
1447
+ "loss": 0.0144,
1448
+ "step": 2020
1449
+ },
1450
+ {
1451
+ "epoch": 2.2506932889628395,
1452
+ "grad_norm": 0.5622262954711914,
1453
+ "learning_rate": 1.5496115427303e-05,
1454
+ "loss": 0.0175,
1455
+ "step": 2030
1456
+ },
1457
+ {
1458
+ "epoch": 2.261785912368275,
1459
+ "grad_norm": 0.1281885802745819,
1460
+ "learning_rate": 1.547391786903441e-05,
1461
+ "loss": 0.018,
1462
+ "step": 2040
1463
+ },
1464
+ {
1465
+ "epoch": 2.2728785357737102,
1466
+ "grad_norm": 0.408607542514801,
1467
+ "learning_rate": 1.5451720310765815e-05,
1468
+ "loss": 0.0095,
1469
+ "step": 2050
1470
+ },
1471
+ {
1472
+ "epoch": 2.283971159179146,
1473
+ "grad_norm": 0.7786557674407959,
1474
+ "learning_rate": 1.5429522752497225e-05,
1475
+ "loss": 0.0163,
1476
+ "step": 2060
1477
+ },
1478
+ {
1479
+ "epoch": 2.2950637825845814,
1480
+ "grad_norm": 0.22975876927375793,
1481
+ "learning_rate": 1.5407325194228635e-05,
1482
+ "loss": 0.0193,
1483
+ "step": 2070
1484
+ },
1485
+ {
1486
+ "epoch": 2.306156405990017,
1487
+ "grad_norm": 0.4313502311706543,
1488
+ "learning_rate": 1.5385127635960048e-05,
1489
+ "loss": 0.0186,
1490
+ "step": 2080
1491
+ },
1492
+ {
1493
+ "epoch": 2.317249029395452,
1494
+ "grad_norm": 0.794448733329773,
1495
+ "learning_rate": 1.5362930077691455e-05,
1496
+ "loss": 0.0159,
1497
+ "step": 2090
1498
+ },
1499
+ {
1500
+ "epoch": 2.3283416528008876,
1501
+ "grad_norm": 0.4191996157169342,
1502
+ "learning_rate": 1.5340732519422865e-05,
1503
+ "loss": 0.0208,
1504
+ "step": 2100
1505
+ },
1506
+ {
1507
+ "epoch": 2.339434276206323,
1508
+ "grad_norm": 0.5219516754150391,
1509
+ "learning_rate": 1.5318534961154274e-05,
1510
+ "loss": 0.0136,
1511
+ "step": 2110
1512
+ },
1513
+ {
1514
+ "epoch": 2.3505268996117583,
1515
+ "grad_norm": 1.1869897842407227,
1516
+ "learning_rate": 1.5296337402885684e-05,
1517
+ "loss": 0.0177,
1518
+ "step": 2120
1519
+ },
1520
+ {
1521
+ "epoch": 2.3616195230171937,
1522
+ "grad_norm": 0.12048923969268799,
1523
+ "learning_rate": 1.5274139844617094e-05,
1524
+ "loss": 0.0129,
1525
+ "step": 2130
1526
+ },
1527
+ {
1528
+ "epoch": 2.372712146422629,
1529
+ "grad_norm": 0.5213373899459839,
1530
+ "learning_rate": 1.5251942286348502e-05,
1531
+ "loss": 0.0127,
1532
+ "step": 2140
1533
+ },
1534
+ {
1535
+ "epoch": 2.3838047698280644,
1536
+ "grad_norm": 0.43704670667648315,
1537
+ "learning_rate": 1.5229744728079912e-05,
1538
+ "loss": 0.0154,
1539
+ "step": 2150
1540
+ },
1541
+ {
1542
+ "epoch": 2.3948973932335,
1543
+ "grad_norm": 0.567150890827179,
1544
+ "learning_rate": 1.5207547169811324e-05,
1545
+ "loss": 0.0117,
1546
+ "step": 2160
1547
+ },
1548
+ {
1549
+ "epoch": 2.405990016638935,
1550
+ "grad_norm": 0.571408212184906,
1551
+ "learning_rate": 1.5185349611542732e-05,
1552
+ "loss": 0.0152,
1553
+ "step": 2170
1554
+ },
1555
+ {
1556
+ "epoch": 2.4170826400443706,
1557
+ "grad_norm": 0.524834394454956,
1558
+ "learning_rate": 1.5163152053274142e-05,
1559
+ "loss": 0.017,
1560
+ "step": 2180
1561
+ },
1562
+ {
1563
+ "epoch": 2.428175263449806,
1564
+ "grad_norm": 0.5026165246963501,
1565
+ "learning_rate": 1.5140954495005552e-05,
1566
+ "loss": 0.0193,
1567
+ "step": 2190
1568
+ },
1569
+ {
1570
+ "epoch": 2.4392678868552413,
1571
+ "grad_norm": 0.3944782614707947,
1572
+ "learning_rate": 1.511875693673696e-05,
1573
+ "loss": 0.0159,
1574
+ "step": 2200
1575
+ },
1576
+ {
1577
+ "epoch": 2.4503605102606767,
1578
+ "grad_norm": 0.6460635662078857,
1579
+ "learning_rate": 1.509655937846837e-05,
1580
+ "loss": 0.0139,
1581
+ "step": 2210
1582
+ },
1583
+ {
1584
+ "epoch": 2.461453133666112,
1585
+ "grad_norm": 0.5573295950889587,
1586
+ "learning_rate": 1.507436182019978e-05,
1587
+ "loss": 0.0183,
1588
+ "step": 2220
1589
+ },
1590
+ {
1591
+ "epoch": 2.4725457570715474,
1592
+ "grad_norm": 0.5001458525657654,
1593
+ "learning_rate": 1.5052164261931188e-05,
1594
+ "loss": 0.0167,
1595
+ "step": 2230
1596
+ },
1597
+ {
1598
+ "epoch": 2.483638380476983,
1599
+ "grad_norm": 1.0214160680770874,
1600
+ "learning_rate": 1.5029966703662598e-05,
1601
+ "loss": 0.0182,
1602
+ "step": 2240
1603
+ },
1604
+ {
1605
+ "epoch": 2.494731003882418,
1606
+ "grad_norm": 0.5756349563598633,
1607
+ "learning_rate": 1.5007769145394009e-05,
1608
+ "loss": 0.0164,
1609
+ "step": 2250
1610
+ },
1611
+ {
1612
+ "epoch": 2.5058236272878536,
1613
+ "grad_norm": 0.771986722946167,
1614
+ "learning_rate": 1.4985571587125419e-05,
1615
+ "loss": 0.0183,
1616
+ "step": 2260
1617
+ },
1618
+ {
1619
+ "epoch": 2.516916250693289,
1620
+ "grad_norm": 0.5560303926467896,
1621
+ "learning_rate": 1.4963374028856827e-05,
1622
+ "loss": 0.0178,
1623
+ "step": 2270
1624
+ },
1625
+ {
1626
+ "epoch": 2.5280088740987243,
1627
+ "grad_norm": 0.8550804853439331,
1628
+ "learning_rate": 1.4941176470588237e-05,
1629
+ "loss": 0.0118,
1630
+ "step": 2280
1631
+ },
1632
+ {
1633
+ "epoch": 2.5391014975041597,
1634
+ "grad_norm": 0.8129355907440186,
1635
+ "learning_rate": 1.4918978912319645e-05,
1636
+ "loss": 0.0126,
1637
+ "step": 2290
1638
+ },
1639
+ {
1640
+ "epoch": 2.550194120909595,
1641
+ "grad_norm": 1.176377773284912,
1642
+ "learning_rate": 1.4896781354051055e-05,
1643
+ "loss": 0.0126,
1644
+ "step": 2300
1645
+ },
1646
+ {
1647
+ "epoch": 2.5612867443150305,
1648
+ "grad_norm": 0.1419341266155243,
1649
+ "learning_rate": 1.4874583795782465e-05,
1650
+ "loss": 0.0196,
1651
+ "step": 2310
1652
+ },
1653
+ {
1654
+ "epoch": 2.572379367720466,
1655
+ "grad_norm": 1.4311761856079102,
1656
+ "learning_rate": 1.4852386237513873e-05,
1657
+ "loss": 0.0143,
1658
+ "step": 2320
1659
+ },
1660
+ {
1661
+ "epoch": 2.583471991125901,
1662
+ "grad_norm": 0.9924690127372742,
1663
+ "learning_rate": 1.4830188679245283e-05,
1664
+ "loss": 0.0198,
1665
+ "step": 2330
1666
+ },
1667
+ {
1668
+ "epoch": 2.5945646145313366,
1669
+ "grad_norm": 0.6496040225028992,
1670
+ "learning_rate": 1.4807991120976695e-05,
1671
+ "loss": 0.014,
1672
+ "step": 2340
1673
+ },
1674
+ {
1675
+ "epoch": 2.605657237936772,
1676
+ "grad_norm": 0.2335842400789261,
1677
+ "learning_rate": 1.4785793562708104e-05,
1678
+ "loss": 0.0128,
1679
+ "step": 2350
1680
+ },
1681
+ {
1682
+ "epoch": 2.6167498613422073,
1683
+ "grad_norm": 0.6549142003059387,
1684
+ "learning_rate": 1.4763596004439513e-05,
1685
+ "loss": 0.019,
1686
+ "step": 2360
1687
+ },
1688
+ {
1689
+ "epoch": 2.6278424847476427,
1690
+ "grad_norm": 0.3264187276363373,
1691
+ "learning_rate": 1.4741398446170922e-05,
1692
+ "loss": 0.0168,
1693
+ "step": 2370
1694
+ },
1695
+ {
1696
+ "epoch": 2.638935108153078,
1697
+ "grad_norm": 0.4175679683685303,
1698
+ "learning_rate": 1.4719200887902332e-05,
1699
+ "loss": 0.0122,
1700
+ "step": 2380
1701
+ },
1702
+ {
1703
+ "epoch": 2.6500277315585135,
1704
+ "grad_norm": 0.2968366742134094,
1705
+ "learning_rate": 1.469700332963374e-05,
1706
+ "loss": 0.0189,
1707
+ "step": 2390
1708
+ },
1709
+ {
1710
+ "epoch": 2.661120354963949,
1711
+ "grad_norm": 1.5489327907562256,
1712
+ "learning_rate": 1.467480577136515e-05,
1713
+ "loss": 0.0158,
1714
+ "step": 2400
1715
+ },
1716
+ {
1717
+ "epoch": 2.672212978369384,
1718
+ "grad_norm": 0.8017039895057678,
1719
+ "learning_rate": 1.465260821309656e-05,
1720
+ "loss": 0.0178,
1721
+ "step": 2410
1722
+ },
1723
+ {
1724
+ "epoch": 2.6833056017748196,
1725
+ "grad_norm": 0.6136099100112915,
1726
+ "learning_rate": 1.4630410654827972e-05,
1727
+ "loss": 0.0164,
1728
+ "step": 2420
1729
+ },
1730
+ {
1731
+ "epoch": 2.6943982251802554,
1732
+ "grad_norm": 0.5461011528968811,
1733
+ "learning_rate": 1.460821309655938e-05,
1734
+ "loss": 0.0168,
1735
+ "step": 2430
1736
+ },
1737
+ {
1738
+ "epoch": 2.7054908485856908,
1739
+ "grad_norm": 0.2680525779724121,
1740
+ "learning_rate": 1.458601553829079e-05,
1741
+ "loss": 0.0138,
1742
+ "step": 2440
1743
+ },
1744
+ {
1745
+ "epoch": 2.716583471991126,
1746
+ "grad_norm": 0.6884289383888245,
1747
+ "learning_rate": 1.45638179800222e-05,
1748
+ "loss": 0.0145,
1749
+ "step": 2450
1750
+ },
1751
+ {
1752
+ "epoch": 2.7276760953965615,
1753
+ "grad_norm": 1.226928472518921,
1754
+ "learning_rate": 1.4541620421753608e-05,
1755
+ "loss": 0.0145,
1756
+ "step": 2460
1757
+ },
1758
+ {
1759
+ "epoch": 2.738768718801997,
1760
+ "grad_norm": 0.447221040725708,
1761
+ "learning_rate": 1.4519422863485018e-05,
1762
+ "loss": 0.0168,
1763
+ "step": 2470
1764
+ },
1765
+ {
1766
+ "epoch": 2.7498613422074323,
1767
+ "grad_norm": 0.47167718410491943,
1768
+ "learning_rate": 1.4497225305216428e-05,
1769
+ "loss": 0.0124,
1770
+ "step": 2480
1771
+ },
1772
+ {
1773
+ "epoch": 2.7609539656128677,
1774
+ "grad_norm": 1.326282262802124,
1775
+ "learning_rate": 1.4475027746947836e-05,
1776
+ "loss": 0.0183,
1777
+ "step": 2490
1778
+ },
1779
+ {
1780
+ "epoch": 2.772046589018303,
1781
+ "grad_norm": 0.9036744236946106,
1782
+ "learning_rate": 1.4452830188679246e-05,
1783
+ "loss": 0.023,
1784
+ "step": 2500
1785
+ },
1786
+ {
1787
+ "epoch": 2.7831392124237384,
1788
+ "grad_norm": 0.30263999104499817,
1789
+ "learning_rate": 1.4430632630410657e-05,
1790
+ "loss": 0.0163,
1791
+ "step": 2510
1792
+ },
1793
+ {
1794
+ "epoch": 2.794231835829174,
1795
+ "grad_norm": 0.7316049933433533,
1796
+ "learning_rate": 1.4408435072142067e-05,
1797
+ "loss": 0.0156,
1798
+ "step": 2520
1799
+ },
1800
+ {
1801
+ "epoch": 2.805324459234609,
1802
+ "grad_norm": 0.8268038630485535,
1803
+ "learning_rate": 1.4386237513873475e-05,
1804
+ "loss": 0.0169,
1805
+ "step": 2530
1806
+ },
1807
+ {
1808
+ "epoch": 2.8164170826400445,
1809
+ "grad_norm": 0.633816659450531,
1810
+ "learning_rate": 1.4364039955604885e-05,
1811
+ "loss": 0.0147,
1812
+ "step": 2540
1813
+ },
1814
+ {
1815
+ "epoch": 2.82750970604548,
1816
+ "grad_norm": 0.43913397192955017,
1817
+ "learning_rate": 1.4341842397336295e-05,
1818
+ "loss": 0.0122,
1819
+ "step": 2550
1820
+ },
1821
+ {
1822
+ "epoch": 2.8386023294509153,
1823
+ "grad_norm": 0.37594836950302124,
1824
+ "learning_rate": 1.4319644839067703e-05,
1825
+ "loss": 0.0149,
1826
+ "step": 2560
1827
+ },
1828
+ {
1829
+ "epoch": 2.8496949528563507,
1830
+ "grad_norm": 0.5090307593345642,
1831
+ "learning_rate": 1.4297447280799113e-05,
1832
+ "loss": 0.015,
1833
+ "step": 2570
1834
+ },
1835
+ {
1836
+ "epoch": 2.860787576261786,
1837
+ "grad_norm": 0.47723984718322754,
1838
+ "learning_rate": 1.4275249722530521e-05,
1839
+ "loss": 0.0166,
1840
+ "step": 2580
1841
+ },
1842
+ {
1843
+ "epoch": 2.8718801996672214,
1844
+ "grad_norm": 1.2512151002883911,
1845
+ "learning_rate": 1.4253052164261931e-05,
1846
+ "loss": 0.0136,
1847
+ "step": 2590
1848
+ },
1849
+ {
1850
+ "epoch": 2.882972823072657,
1851
+ "grad_norm": 0.676337718963623,
1852
+ "learning_rate": 1.4230854605993343e-05,
1853
+ "loss": 0.0152,
1854
+ "step": 2600
1855
+ },
1856
+ {
1857
+ "epoch": 2.894065446478092,
1858
+ "grad_norm": 0.7379328012466431,
1859
+ "learning_rate": 1.4208657047724752e-05,
1860
+ "loss": 0.0182,
1861
+ "step": 2610
1862
+ },
1863
+ {
1864
+ "epoch": 2.9051580698835275,
1865
+ "grad_norm": 0.8266171813011169,
1866
+ "learning_rate": 1.418645948945616e-05,
1867
+ "loss": 0.0207,
1868
+ "step": 2620
1869
+ },
1870
+ {
1871
+ "epoch": 2.916250693288963,
1872
+ "grad_norm": 0.9925495982170105,
1873
+ "learning_rate": 1.416426193118757e-05,
1874
+ "loss": 0.0157,
1875
+ "step": 2630
1876
+ },
1877
+ {
1878
+ "epoch": 2.9273433166943983,
1879
+ "grad_norm": 1.1120530366897583,
1880
+ "learning_rate": 1.414206437291898e-05,
1881
+ "loss": 0.0164,
1882
+ "step": 2640
1883
+ },
1884
+ {
1885
+ "epoch": 2.9384359400998337,
1886
+ "grad_norm": 0.9687446355819702,
1887
+ "learning_rate": 1.4119866814650389e-05,
1888
+ "loss": 0.0211,
1889
+ "step": 2650
1890
+ },
1891
+ {
1892
+ "epoch": 2.949528563505269,
1893
+ "grad_norm": 0.19811566174030304,
1894
+ "learning_rate": 1.4097669256381798e-05,
1895
+ "loss": 0.0181,
1896
+ "step": 2660
1897
+ },
1898
+ {
1899
+ "epoch": 2.9606211869107044,
1900
+ "grad_norm": 0.9253482818603516,
1901
+ "learning_rate": 1.4075471698113208e-05,
1902
+ "loss": 0.0158,
1903
+ "step": 2670
1904
+ },
1905
+ {
1906
+ "epoch": 2.97171381031614,
1907
+ "grad_norm": 0.677994430065155,
1908
+ "learning_rate": 1.405327413984462e-05,
1909
+ "loss": 0.024,
1910
+ "step": 2680
1911
+ },
1912
+ {
1913
+ "epoch": 2.982806433721575,
1914
+ "grad_norm": 0.3959580361843109,
1915
+ "learning_rate": 1.4031076581576028e-05,
1916
+ "loss": 0.009,
1917
+ "step": 2690
1918
+ },
1919
+ {
1920
+ "epoch": 2.9938990571270105,
1921
+ "grad_norm": 1.0569937229156494,
1922
+ "learning_rate": 1.4008879023307438e-05,
1923
+ "loss": 0.0187,
1924
+ "step": 2700
1925
+ },
1926
+ {
1927
+ "epoch": 3.0,
1928
+ "eval_accuracy": 0.8888888888888888,
1929
+ "eval_f1": 0.7171109200343938,
1930
+ "eval_loss": 0.3638182580471039,
1931
+ "eval_precision": 0.695,
1932
+ "eval_recall": 0.7406749555950266,
1933
+ "eval_runtime": 2.7252,
1934
+ "eval_samples_per_second": 362.177,
1935
+ "eval_steps_per_second": 11.375,
1936
+ "step": 2706
1937
+ },
1938
+ {
1939
+ "epoch": 3.004437049362174,
1940
+ "grad_norm": 0.22475391626358032,
1941
+ "learning_rate": 1.3986681465038848e-05,
1942
+ "loss": 0.0127,
1943
+ "step": 2710
1944
+ },
1945
+ {
1946
+ "epoch": 3.0155296727676095,
1947
+ "grad_norm": 0.42929062247276306,
1948
+ "learning_rate": 1.3964483906770256e-05,
1949
+ "loss": 0.0131,
1950
+ "step": 2720
1951
+ },
1952
+ {
1953
+ "epoch": 3.026622296173045,
1954
+ "grad_norm": 2.398026466369629,
1955
+ "learning_rate": 1.3942286348501666e-05,
1956
+ "loss": 0.0072,
1957
+ "step": 2730
1958
+ },
1959
+ {
1960
+ "epoch": 3.0377149195784803,
1961
+ "grad_norm": 0.14908376336097717,
1962
+ "learning_rate": 1.3920088790233076e-05,
1963
+ "loss": 0.0082,
1964
+ "step": 2740
1965
+ },
1966
+ {
1967
+ "epoch": 3.0488075429839157,
1968
+ "grad_norm": 0.8177769184112549,
1969
+ "learning_rate": 1.3897891231964484e-05,
1970
+ "loss": 0.0088,
1971
+ "step": 2750
1972
+ },
1973
+ {
1974
+ "epoch": 3.059900166389351,
1975
+ "grad_norm": 0.13737396895885468,
1976
+ "learning_rate": 1.3875693673695894e-05,
1977
+ "loss": 0.008,
1978
+ "step": 2760
1979
+ },
1980
+ {
1981
+ "epoch": 3.0709927897947864,
1982
+ "grad_norm": 0.7825105786323547,
1983
+ "learning_rate": 1.3853496115427305e-05,
1984
+ "loss": 0.0034,
1985
+ "step": 2770
1986
+ },
1987
+ {
1988
+ "epoch": 3.082085413200222,
1989
+ "grad_norm": 0.42062872648239136,
1990
+ "learning_rate": 1.3831298557158715e-05,
1991
+ "loss": 0.0062,
1992
+ "step": 2780
1993
+ },
1994
+ {
1995
+ "epoch": 3.093178036605657,
1996
+ "grad_norm": 0.7224815487861633,
1997
+ "learning_rate": 1.3809100998890123e-05,
1998
+ "loss": 0.0055,
1999
+ "step": 2790
2000
+ },
2001
+ {
2002
+ "epoch": 3.1042706600110925,
2003
+ "grad_norm": 0.04094177484512329,
2004
+ "learning_rate": 1.3786903440621533e-05,
2005
+ "loss": 0.002,
2006
+ "step": 2800
2007
+ },
2008
+ {
2009
+ "epoch": 3.115363283416528,
2010
+ "grad_norm": 0.03310002386569977,
2011
+ "learning_rate": 1.3764705882352943e-05,
2012
+ "loss": 0.0055,
2013
+ "step": 2810
2014
+ },
2015
+ {
2016
+ "epoch": 3.1264559068219633,
2017
+ "grad_norm": 0.5118809938430786,
2018
+ "learning_rate": 1.3742508324084351e-05,
2019
+ "loss": 0.0054,
2020
+ "step": 2820
2021
+ },
2022
+ {
2023
+ "epoch": 3.1375485302273987,
2024
+ "grad_norm": 0.5759711861610413,
2025
+ "learning_rate": 1.3720310765815761e-05,
2026
+ "loss": 0.0086,
2027
+ "step": 2830
2028
+ },
2029
+ {
2030
+ "epoch": 3.148641153632834,
2031
+ "grad_norm": 1.1310100555419922,
2032
+ "learning_rate": 1.3698113207547171e-05,
2033
+ "loss": 0.0091,
2034
+ "step": 2840
2035
+ },
2036
+ {
2037
+ "epoch": 3.1597337770382694,
2038
+ "grad_norm": 0.04265379533171654,
2039
+ "learning_rate": 1.3675915649278579e-05,
2040
+ "loss": 0.0061,
2041
+ "step": 2850
2042
+ },
2043
+ {
2044
+ "epoch": 3.170826400443705,
2045
+ "grad_norm": 0.6648514270782471,
2046
+ "learning_rate": 1.365371809100999e-05,
2047
+ "loss": 0.0035,
2048
+ "step": 2860
2049
+ },
2050
+ {
2051
+ "epoch": 3.18191902384914,
2052
+ "grad_norm": 0.02486938051879406,
2053
+ "learning_rate": 1.36315205327414e-05,
2054
+ "loss": 0.0055,
2055
+ "step": 2870
2056
+ },
2057
+ {
2058
+ "epoch": 3.1930116472545755,
2059
+ "grad_norm": 0.5384965538978577,
2060
+ "learning_rate": 1.3609322974472809e-05,
2061
+ "loss": 0.0029,
2062
+ "step": 2880
2063
+ },
2064
+ {
2065
+ "epoch": 3.204104270660011,
2066
+ "grad_norm": 0.6231627464294434,
2067
+ "learning_rate": 1.3587125416204219e-05,
2068
+ "loss": 0.005,
2069
+ "step": 2890
2070
+ },
2071
+ {
2072
+ "epoch": 3.2151968940654463,
2073
+ "grad_norm": 0.5190199613571167,
2074
+ "learning_rate": 1.3564927857935628e-05,
2075
+ "loss": 0.0083,
2076
+ "step": 2900
2077
+ },
2078
+ {
2079
+ "epoch": 3.2262895174708817,
2080
+ "grad_norm": 0.39337947964668274,
2081
+ "learning_rate": 1.3542730299667037e-05,
2082
+ "loss": 0.0081,
2083
+ "step": 2910
2084
+ },
2085
+ {
2086
+ "epoch": 3.237382140876317,
2087
+ "grad_norm": 0.6862775683403015,
2088
+ "learning_rate": 1.3520532741398446e-05,
2089
+ "loss": 0.0074,
2090
+ "step": 2920
2091
+ },
2092
+ {
2093
+ "epoch": 3.2484747642817524,
2094
+ "grad_norm": 1.0862469673156738,
2095
+ "learning_rate": 1.3498335183129856e-05,
2096
+ "loss": 0.0074,
2097
+ "step": 2930
2098
+ },
2099
+ {
2100
+ "epoch": 3.259567387687188,
2101
+ "grad_norm": 1.1846354007720947,
2102
+ "learning_rate": 1.3476137624861268e-05,
2103
+ "loss": 0.0075,
2104
+ "step": 2940
2105
+ },
2106
+ {
2107
+ "epoch": 3.270660011092623,
2108
+ "grad_norm": 1.2481039762496948,
2109
+ "learning_rate": 1.3453940066592676e-05,
2110
+ "loss": 0.003,
2111
+ "step": 2950
2112
+ },
2113
+ {
2114
+ "epoch": 3.281752634498059,
2115
+ "grad_norm": 0.3161996006965637,
2116
+ "learning_rate": 1.3431742508324086e-05,
2117
+ "loss": 0.0034,
2118
+ "step": 2960
2119
+ },
2120
+ {
2121
+ "epoch": 3.2928452579034944,
2122
+ "grad_norm": 0.6262934803962708,
2123
+ "learning_rate": 1.3409544950055496e-05,
2124
+ "loss": 0.0107,
2125
+ "step": 2970
2126
+ },
2127
+ {
2128
+ "epoch": 3.3039378813089297,
2129
+ "grad_norm": 0.9278308153152466,
2130
+ "learning_rate": 1.3387347391786904e-05,
2131
+ "loss": 0.0057,
2132
+ "step": 2980
2133
+ },
2134
+ {
2135
+ "epoch": 3.315030504714365,
2136
+ "grad_norm": 0.40893518924713135,
2137
+ "learning_rate": 1.3365149833518314e-05,
2138
+ "loss": 0.0041,
2139
+ "step": 2990
2140
+ },
2141
+ {
2142
+ "epoch": 3.3261231281198005,
2143
+ "grad_norm": 0.6295328736305237,
2144
+ "learning_rate": 1.3342952275249724e-05,
2145
+ "loss": 0.0032,
2146
+ "step": 3000
2147
+ },
2148
+ {
2149
+ "epoch": 3.337215751525236,
2150
+ "grad_norm": 0.039073534309864044,
2151
+ "learning_rate": 1.3320754716981132e-05,
2152
+ "loss": 0.0067,
2153
+ "step": 3010
2154
+ },
2155
+ {
2156
+ "epoch": 3.3483083749306712,
2157
+ "grad_norm": 0.5157305002212524,
2158
+ "learning_rate": 1.3298557158712542e-05,
2159
+ "loss": 0.0093,
2160
+ "step": 3020
2161
+ },
2162
+ {
2163
+ "epoch": 3.3594009983361066,
2164
+ "grad_norm": 0.9451714158058167,
2165
+ "learning_rate": 1.3276359600443953e-05,
2166
+ "loss": 0.0064,
2167
+ "step": 3030
2168
+ },
2169
+ {
2170
+ "epoch": 3.370493621741542,
2171
+ "grad_norm": 0.043243326246738434,
2172
+ "learning_rate": 1.3254162042175363e-05,
2173
+ "loss": 0.0062,
2174
+ "step": 3040
2175
+ },
2176
+ {
2177
+ "epoch": 3.3815862451469774,
2178
+ "grad_norm": 1.2751350402832031,
2179
+ "learning_rate": 1.3231964483906771e-05,
2180
+ "loss": 0.011,
2181
+ "step": 3050
2182
+ },
2183
+ {
2184
+ "epoch": 3.3926788685524127,
2185
+ "grad_norm": 1.1212610006332397,
2186
+ "learning_rate": 1.3209766925638181e-05,
2187
+ "loss": 0.008,
2188
+ "step": 3060
2189
+ },
2190
+ {
2191
+ "epoch": 3.403771491957848,
2192
+ "grad_norm": 0.3985590934753418,
2193
+ "learning_rate": 1.3187569367369591e-05,
2194
+ "loss": 0.0066,
2195
+ "step": 3070
2196
+ },
2197
+ {
2198
+ "epoch": 3.4148641153632835,
2199
+ "grad_norm": 0.6546558737754822,
2200
+ "learning_rate": 1.3165371809101e-05,
2201
+ "loss": 0.0115,
2202
+ "step": 3080
2203
+ },
2204
+ {
2205
+ "epoch": 3.425956738768719,
2206
+ "grad_norm": 1.0568732023239136,
2207
+ "learning_rate": 1.3143174250832409e-05,
2208
+ "loss": 0.0104,
2209
+ "step": 3090
2210
+ },
2211
+ {
2212
+ "epoch": 3.4370493621741542,
2213
+ "grad_norm": 0.9182056188583374,
2214
+ "learning_rate": 1.3120976692563819e-05,
2215
+ "loss": 0.0079,
2216
+ "step": 3100
2217
+ },
2218
+ {
2219
+ "epoch": 3.4481419855795896,
2220
+ "grad_norm": 1.700160026550293,
2221
+ "learning_rate": 1.3098779134295227e-05,
2222
+ "loss": 0.0061,
2223
+ "step": 3110
2224
+ },
2225
+ {
2226
+ "epoch": 3.459234608985025,
2227
+ "grad_norm": 0.09296636283397675,
2228
+ "learning_rate": 1.3076581576026639e-05,
2229
+ "loss": 0.0106,
2230
+ "step": 3120
2231
+ },
2232
+ {
2233
+ "epoch": 3.4703272323904604,
2234
+ "grad_norm": 0.8721600770950317,
2235
+ "learning_rate": 1.3054384017758049e-05,
2236
+ "loss": 0.0058,
2237
+ "step": 3130
2238
+ },
2239
+ {
2240
+ "epoch": 3.4814198557958957,
2241
+ "grad_norm": 0.5778414607048035,
2242
+ "learning_rate": 1.3032186459489457e-05,
2243
+ "loss": 0.0075,
2244
+ "step": 3140
2245
+ },
2246
+ {
2247
+ "epoch": 3.492512479201331,
2248
+ "grad_norm": 0.632172167301178,
2249
+ "learning_rate": 1.3009988901220867e-05,
2250
+ "loss": 0.0057,
2251
+ "step": 3150
2252
+ },
2253
+ {
2254
+ "epoch": 3.5036051026067665,
2255
+ "grad_norm": 0.49320322275161743,
2256
+ "learning_rate": 1.2987791342952276e-05,
2257
+ "loss": 0.0065,
2258
+ "step": 3160
2259
+ },
2260
+ {
2261
+ "epoch": 3.514697726012202,
2262
+ "grad_norm": 1.2454118728637695,
2263
+ "learning_rate": 1.2965593784683685e-05,
2264
+ "loss": 0.0049,
2265
+ "step": 3170
2266
+ },
2267
+ {
2268
+ "epoch": 3.5257903494176372,
2269
+ "grad_norm": 0.16420379281044006,
2270
+ "learning_rate": 1.2943396226415095e-05,
2271
+ "loss": 0.0042,
2272
+ "step": 3180
2273
+ },
2274
+ {
2275
+ "epoch": 3.5368829728230726,
2276
+ "grad_norm": 0.7556213736534119,
2277
+ "learning_rate": 1.2921198668146504e-05,
2278
+ "loss": 0.0076,
2279
+ "step": 3190
2280
+ },
2281
+ {
2282
+ "epoch": 3.547975596228508,
2283
+ "grad_norm": 0.7741811871528625,
2284
+ "learning_rate": 1.2899001109877916e-05,
2285
+ "loss": 0.0053,
2286
+ "step": 3200
2287
+ },
2288
+ {
2289
+ "epoch": 3.5590682196339434,
2290
+ "grad_norm": 0.3981403708457947,
2291
+ "learning_rate": 1.2876803551609324e-05,
2292
+ "loss": 0.0111,
2293
+ "step": 3210
2294
+ },
2295
+ {
2296
+ "epoch": 3.5701608430393788,
2297
+ "grad_norm": 1.556243658065796,
2298
+ "learning_rate": 1.2854605993340734e-05,
2299
+ "loss": 0.0082,
2300
+ "step": 3220
2301
+ },
2302
+ {
2303
+ "epoch": 3.581253466444814,
2304
+ "grad_norm": 0.24557125568389893,
2305
+ "learning_rate": 1.2832408435072144e-05,
2306
+ "loss": 0.0048,
2307
+ "step": 3230
2308
+ },
2309
+ {
2310
+ "epoch": 3.5923460898502495,
2311
+ "grad_norm": 1.2477893829345703,
2312
+ "learning_rate": 1.2810210876803552e-05,
2313
+ "loss": 0.0023,
2314
+ "step": 3240
2315
+ },
2316
+ {
2317
+ "epoch": 3.603438713255685,
2318
+ "grad_norm": 0.04675845056772232,
2319
+ "learning_rate": 1.2788013318534962e-05,
2320
+ "loss": 0.0041,
2321
+ "step": 3250
2322
+ },
2323
+ {
2324
+ "epoch": 3.6145313366611203,
2325
+ "grad_norm": 0.571173906326294,
2326
+ "learning_rate": 1.2765815760266372e-05,
2327
+ "loss": 0.0055,
2328
+ "step": 3260
2329
+ },
2330
+ {
2331
+ "epoch": 3.6256239600665556,
2332
+ "grad_norm": 0.5565969944000244,
2333
+ "learning_rate": 1.274361820199778e-05,
2334
+ "loss": 0.0092,
2335
+ "step": 3270
2336
+ },
2337
+ {
2338
+ "epoch": 3.636716583471991,
2339
+ "grad_norm": 0.013311430811882019,
2340
+ "learning_rate": 1.272142064372919e-05,
2341
+ "loss": 0.0071,
2342
+ "step": 3280
2343
+ },
2344
+ {
2345
+ "epoch": 3.6478092068774264,
2346
+ "grad_norm": 1.6714210510253906,
2347
+ "learning_rate": 1.2699223085460601e-05,
2348
+ "loss": 0.0069,
2349
+ "step": 3290
2350
+ },
2351
+ {
2352
+ "epoch": 3.6589018302828618,
2353
+ "grad_norm": 0.7213656306266785,
2354
+ "learning_rate": 1.2677025527192011e-05,
2355
+ "loss": 0.0053,
2356
+ "step": 3300
2357
+ },
2358
+ {
2359
+ "epoch": 3.669994453688297,
2360
+ "grad_norm": 0.9073006510734558,
2361
+ "learning_rate": 1.265482796892342e-05,
2362
+ "loss": 0.0101,
2363
+ "step": 3310
2364
+ },
2365
+ {
2366
+ "epoch": 3.6810870770937325,
2367
+ "grad_norm": 0.29895493388175964,
2368
+ "learning_rate": 1.263263041065483e-05,
2369
+ "loss": 0.0052,
2370
+ "step": 3320
2371
+ },
2372
+ {
2373
+ "epoch": 3.6921797004991683,
2374
+ "grad_norm": 1.0842756032943726,
2375
+ "learning_rate": 1.2610432852386239e-05,
2376
+ "loss": 0.008,
2377
+ "step": 3330
2378
+ },
2379
+ {
2380
+ "epoch": 3.7032723239046037,
2381
+ "grad_norm": 0.7172666192054749,
2382
+ "learning_rate": 1.2588235294117647e-05,
2383
+ "loss": 0.01,
2384
+ "step": 3340
2385
+ },
2386
+ {
2387
+ "epoch": 3.714364947310039,
2388
+ "grad_norm": 0.10475369542837143,
2389
+ "learning_rate": 1.2566037735849057e-05,
2390
+ "loss": 0.0063,
2391
+ "step": 3350
2392
+ },
2393
+ {
2394
+ "epoch": 3.7254575707154745,
2395
+ "grad_norm": 0.9639983177185059,
2396
+ "learning_rate": 1.2543840177580467e-05,
2397
+ "loss": 0.0059,
2398
+ "step": 3360
2399
+ },
2400
+ {
2401
+ "epoch": 3.73655019412091,
2402
+ "grad_norm": 0.9579765796661377,
2403
+ "learning_rate": 1.2521642619311875e-05,
2404
+ "loss": 0.0053,
2405
+ "step": 3370
2406
+ },
2407
+ {
2408
+ "epoch": 3.747642817526345,
2409
+ "grad_norm": 0.14046674966812134,
2410
+ "learning_rate": 1.2499445061043287e-05,
2411
+ "loss": 0.0081,
2412
+ "step": 3380
2413
+ },
2414
+ {
2415
+ "epoch": 3.7587354409317806,
2416
+ "grad_norm": 0.4136715233325958,
2417
+ "learning_rate": 1.2477247502774697e-05,
2418
+ "loss": 0.0101,
2419
+ "step": 3390
2420
+ },
2421
+ {
2422
+ "epoch": 3.769828064337216,
2423
+ "grad_norm": 0.30935949087142944,
2424
+ "learning_rate": 1.2455049944506107e-05,
2425
+ "loss": 0.0046,
2426
+ "step": 3400
2427
+ },
2428
+ {
2429
+ "epoch": 3.7809206877426513,
2430
+ "grad_norm": 0.25942277908325195,
2431
+ "learning_rate": 1.2432852386237515e-05,
2432
+ "loss": 0.006,
2433
+ "step": 3410
2434
+ },
2435
+ {
2436
+ "epoch": 3.7920133111480867,
2437
+ "grad_norm": 0.08538083732128143,
2438
+ "learning_rate": 1.2410654827968925e-05,
2439
+ "loss": 0.0038,
2440
+ "step": 3420
2441
+ },
2442
+ {
2443
+ "epoch": 3.803105934553522,
2444
+ "grad_norm": 0.11854628473520279,
2445
+ "learning_rate": 1.2388457269700334e-05,
2446
+ "loss": 0.0038,
2447
+ "step": 3430
2448
+ },
2449
+ {
2450
+ "epoch": 3.8141985579589575,
2451
+ "grad_norm": 0.2808975279331207,
2452
+ "learning_rate": 1.2366259711431743e-05,
2453
+ "loss": 0.0093,
2454
+ "step": 3440
2455
+ },
2456
+ {
2457
+ "epoch": 3.825291181364393,
2458
+ "grad_norm": 1.0285228490829468,
2459
+ "learning_rate": 1.2344062153163152e-05,
2460
+ "loss": 0.0101,
2461
+ "step": 3450
2462
+ },
2463
+ {
2464
+ "epoch": 3.836383804769828,
2465
+ "grad_norm": 0.43529564142227173,
2466
+ "learning_rate": 1.2321864594894564e-05,
2467
+ "loss": 0.0057,
2468
+ "step": 3460
2469
+ },
2470
+ {
2471
+ "epoch": 3.8474764281752636,
2472
+ "grad_norm": 2.0706746578216553,
2473
+ "learning_rate": 1.2299667036625972e-05,
2474
+ "loss": 0.0079,
2475
+ "step": 3470
2476
+ },
2477
+ {
2478
+ "epoch": 3.858569051580699,
2479
+ "grad_norm": 0.9805829524993896,
2480
+ "learning_rate": 1.2277469478357382e-05,
2481
+ "loss": 0.0116,
2482
+ "step": 3480
2483
+ },
2484
+ {
2485
+ "epoch": 3.8696616749861343,
2486
+ "grad_norm": 0.754189133644104,
2487
+ "learning_rate": 1.2255271920088792e-05,
2488
+ "loss": 0.0056,
2489
+ "step": 3490
2490
+ },
2491
+ {
2492
+ "epoch": 3.8807542983915697,
2493
+ "grad_norm": 0.9179038405418396,
2494
+ "learning_rate": 1.22330743618202e-05,
2495
+ "loss": 0.0079,
2496
+ "step": 3500
2497
+ },
2498
+ {
2499
+ "epoch": 3.891846921797005,
2500
+ "grad_norm": 1.5467790365219116,
2501
+ "learning_rate": 1.221087680355161e-05,
2502
+ "loss": 0.009,
2503
+ "step": 3510
2504
+ },
2505
+ {
2506
+ "epoch": 3.9029395452024405,
2507
+ "grad_norm": 0.44232311844825745,
2508
+ "learning_rate": 1.218867924528302e-05,
2509
+ "loss": 0.0072,
2510
+ "step": 3520
2511
+ },
2512
+ {
2513
+ "epoch": 3.914032168607876,
2514
+ "grad_norm": 0.05650203302502632,
2515
+ "learning_rate": 1.2166481687014428e-05,
2516
+ "loss": 0.0053,
2517
+ "step": 3530
2518
+ },
2519
+ {
2520
+ "epoch": 3.925124792013311,
2521
+ "grad_norm": 0.6348217129707336,
2522
+ "learning_rate": 1.2144284128745838e-05,
2523
+ "loss": 0.009,
2524
+ "step": 3540
2525
+ },
2526
+ {
2527
+ "epoch": 3.9362174154187466,
2528
+ "grad_norm": 1.0749729871749878,
2529
+ "learning_rate": 1.212208657047725e-05,
2530
+ "loss": 0.0086,
2531
+ "step": 3550
2532
+ },
2533
+ {
2534
+ "epoch": 3.947310038824182,
2535
+ "grad_norm": 2.320844888687134,
2536
+ "learning_rate": 1.209988901220866e-05,
2537
+ "loss": 0.0096,
2538
+ "step": 3560
2539
+ },
2540
+ {
2541
+ "epoch": 3.9584026622296173,
2542
+ "grad_norm": 0.5694814324378967,
2543
+ "learning_rate": 1.2077691453940067e-05,
2544
+ "loss": 0.008,
2545
+ "step": 3570
2546
+ },
2547
+ {
2548
+ "epoch": 3.9694952856350527,
2549
+ "grad_norm": 0.9794883728027344,
2550
+ "learning_rate": 1.2055493895671477e-05,
2551
+ "loss": 0.0045,
2552
+ "step": 3580
2553
+ },
2554
+ {
2555
+ "epoch": 3.980587909040488,
2556
+ "grad_norm": 0.2133321315050125,
2557
+ "learning_rate": 1.2033296337402887e-05,
2558
+ "loss": 0.0054,
2559
+ "step": 3590
2560
+ },
2561
+ {
2562
+ "epoch": 3.9916805324459235,
2563
+ "grad_norm": 1.423915147781372,
2564
+ "learning_rate": 1.2011098779134295e-05,
2565
+ "loss": 0.0127,
2566
+ "step": 3600
2567
+ },
2568
+ {
2569
+ "epoch": 4.0,
2570
+ "eval_accuracy": 0.8699763593380615,
2571
+ "eval_f1": 0.6728971962616822,
2572
+ "eval_loss": 0.4916483163833618,
2573
+ "eval_precision": 0.6449511400651465,
2574
+ "eval_recall": 0.7033747779751333,
2575
+ "eval_runtime": 2.7388,
2576
+ "eval_samples_per_second": 360.381,
2577
+ "eval_steps_per_second": 11.319,
2578
+ "step": 3608
2579
+ },
2580
+ {
2581
+ "epoch": 4.002218524681087,
2582
+ "grad_norm": 0.7580460906028748,
2583
+ "learning_rate": 1.1988901220865705e-05,
2584
+ "loss": 0.0052,
2585
+ "step": 3610
2586
+ },
2587
+ {
2588
+ "epoch": 4.0133111480865225,
2589
+ "grad_norm": 0.13793574273586273,
2590
+ "learning_rate": 1.1966703662597115e-05,
2591
+ "loss": 0.0024,
2592
+ "step": 3620
2593
+ },
2594
+ {
2595
+ "epoch": 4.024403771491958,
2596
+ "grad_norm": 0.41575559973716736,
2597
+ "learning_rate": 1.1944506104328523e-05,
2598
+ "loss": 0.0028,
2599
+ "step": 3630
2600
+ },
2601
+ {
2602
+ "epoch": 4.035496394897393,
2603
+ "grad_norm": 0.268417626619339,
2604
+ "learning_rate": 1.1922308546059935e-05,
2605
+ "loss": 0.0043,
2606
+ "step": 3640
2607
+ },
2608
+ {
2609
+ "epoch": 4.046589018302829,
2610
+ "grad_norm": 0.227146178483963,
2611
+ "learning_rate": 1.1900110987791345e-05,
2612
+ "loss": 0.0014,
2613
+ "step": 3650
2614
+ },
2615
+ {
2616
+ "epoch": 4.057681641708264,
2617
+ "grad_norm": 0.1434301882982254,
2618
+ "learning_rate": 1.1877913429522755e-05,
2619
+ "loss": 0.0012,
2620
+ "step": 3660
2621
+ },
2622
+ {
2623
+ "epoch": 4.068774265113699,
2624
+ "grad_norm": 0.3861040472984314,
2625
+ "learning_rate": 1.1855715871254163e-05,
2626
+ "loss": 0.0027,
2627
+ "step": 3670
2628
+ },
2629
+ {
2630
+ "epoch": 4.079866888519135,
2631
+ "grad_norm": 0.2243143618106842,
2632
+ "learning_rate": 1.1833518312985573e-05,
2633
+ "loss": 0.0062,
2634
+ "step": 3680
2635
+ },
2636
+ {
2637
+ "epoch": 4.09095951192457,
2638
+ "grad_norm": 0.5330334901809692,
2639
+ "learning_rate": 1.1811320754716982e-05,
2640
+ "loss": 0.0074,
2641
+ "step": 3690
2642
+ },
2643
+ {
2644
+ "epoch": 4.1020521353300055,
2645
+ "grad_norm": 0.3948166072368622,
2646
+ "learning_rate": 1.178912319644839e-05,
2647
+ "loss": 0.0047,
2648
+ "step": 3700
2649
+ },
2650
+ {
2651
+ "epoch": 4.113144758735441,
2652
+ "grad_norm": 0.8226463198661804,
2653
+ "learning_rate": 1.17669256381798e-05,
2654
+ "loss": 0.0047,
2655
+ "step": 3710
2656
+ },
2657
+ {
2658
+ "epoch": 4.124237382140876,
2659
+ "grad_norm": 0.5453665852546692,
2660
+ "learning_rate": 1.1744728079911212e-05,
2661
+ "loss": 0.003,
2662
+ "step": 3720
2663
+ },
2664
+ {
2665
+ "epoch": 4.135330005546312,
2666
+ "grad_norm": 0.1433408111333847,
2667
+ "learning_rate": 1.172253052164262e-05,
2668
+ "loss": 0.0029,
2669
+ "step": 3730
2670
+ },
2671
+ {
2672
+ "epoch": 4.146422628951747,
2673
+ "grad_norm": 0.14077675342559814,
2674
+ "learning_rate": 1.170033296337403e-05,
2675
+ "loss": 0.004,
2676
+ "step": 3740
2677
+ },
2678
+ {
2679
+ "epoch": 4.157515252357182,
2680
+ "grad_norm": 0.08359857648611069,
2681
+ "learning_rate": 1.167813540510544e-05,
2682
+ "loss": 0.0034,
2683
+ "step": 3750
2684
+ },
2685
+ {
2686
+ "epoch": 4.168607875762618,
2687
+ "grad_norm": 0.10568249225616455,
2688
+ "learning_rate": 1.1655937846836848e-05,
2689
+ "loss": 0.0019,
2690
+ "step": 3760
2691
+ },
2692
+ {
2693
+ "epoch": 4.179700499168053,
2694
+ "grad_norm": 0.010569937527179718,
2695
+ "learning_rate": 1.1633740288568258e-05,
2696
+ "loss": 0.0039,
2697
+ "step": 3770
2698
+ },
2699
+ {
2700
+ "epoch": 4.1907931225734885,
2701
+ "grad_norm": 0.38508471846580505,
2702
+ "learning_rate": 1.1611542730299668e-05,
2703
+ "loss": 0.0035,
2704
+ "step": 3780
2705
+ },
2706
+ {
2707
+ "epoch": 4.201885745978924,
2708
+ "grad_norm": 0.4583463966846466,
2709
+ "learning_rate": 1.1589345172031076e-05,
2710
+ "loss": 0.0035,
2711
+ "step": 3790
2712
+ },
2713
+ {
2714
+ "epoch": 4.212978369384359,
2715
+ "grad_norm": 0.26573556661605835,
2716
+ "learning_rate": 1.1567147613762486e-05,
2717
+ "loss": 0.0014,
2718
+ "step": 3800
2719
+ },
2720
+ {
2721
+ "epoch": 4.224070992789795,
2722
+ "grad_norm": 0.699878454208374,
2723
+ "learning_rate": 1.1544950055493897e-05,
2724
+ "loss": 0.0034,
2725
+ "step": 3810
2726
+ },
2727
+ {
2728
+ "epoch": 4.23516361619523,
2729
+ "grad_norm": 0.02065327949821949,
2730
+ "learning_rate": 1.1522752497225307e-05,
2731
+ "loss": 0.0028,
2732
+ "step": 3820
2733
+ },
2734
+ {
2735
+ "epoch": 4.246256239600665,
2736
+ "grad_norm": 1.4631462097167969,
2737
+ "learning_rate": 1.1500554938956716e-05,
2738
+ "loss": 0.003,
2739
+ "step": 3830
2740
+ },
2741
+ {
2742
+ "epoch": 4.257348863006101,
2743
+ "grad_norm": 1.0686756372451782,
2744
+ "learning_rate": 1.1478357380688125e-05,
2745
+ "loss": 0.0026,
2746
+ "step": 3840
2747
+ },
2748
+ {
2749
+ "epoch": 4.268441486411536,
2750
+ "grad_norm": 0.15860460698604584,
2751
+ "learning_rate": 1.1456159822419535e-05,
2752
+ "loss": 0.0064,
2753
+ "step": 3850
2754
+ },
2755
+ {
2756
+ "epoch": 4.2795341098169715,
2757
+ "grad_norm": 0.05696827918291092,
2758
+ "learning_rate": 1.1433962264150943e-05,
2759
+ "loss": 0.0033,
2760
+ "step": 3860
2761
+ },
2762
+ {
2763
+ "epoch": 4.290626733222407,
2764
+ "grad_norm": 0.01276635192334652,
2765
+ "learning_rate": 1.1411764705882353e-05,
2766
+ "loss": 0.0059,
2767
+ "step": 3870
2768
+ },
2769
+ {
2770
+ "epoch": 4.301719356627842,
2771
+ "grad_norm": 0.3438052535057068,
2772
+ "learning_rate": 1.1389567147613763e-05,
2773
+ "loss": 0.0022,
2774
+ "step": 3880
2775
+ },
2776
+ {
2777
+ "epoch": 4.312811980033278,
2778
+ "grad_norm": 0.020309099927544594,
2779
+ "learning_rate": 1.1367369589345171e-05,
2780
+ "loss": 0.0029,
2781
+ "step": 3890
2782
+ },
2783
+ {
2784
+ "epoch": 4.323904603438713,
2785
+ "grad_norm": 1.0980569124221802,
2786
+ "learning_rate": 1.1345172031076583e-05,
2787
+ "loss": 0.0012,
2788
+ "step": 3900
2789
+ },
2790
+ {
2791
+ "epoch": 4.334997226844148,
2792
+ "grad_norm": 0.723771333694458,
2793
+ "learning_rate": 1.1322974472807993e-05,
2794
+ "loss": 0.0027,
2795
+ "step": 3910
2796
+ },
2797
+ {
2798
+ "epoch": 4.346089850249584,
2799
+ "grad_norm": 0.009631435386836529,
2800
+ "learning_rate": 1.1300776914539403e-05,
2801
+ "loss": 0.0023,
2802
+ "step": 3920
2803
+ },
2804
+ {
2805
+ "epoch": 4.357182473655019,
2806
+ "grad_norm": 0.29609665274620056,
2807
+ "learning_rate": 1.127857935627081e-05,
2808
+ "loss": 0.0044,
2809
+ "step": 3930
2810
+ },
2811
+ {
2812
+ "epoch": 4.3682750970604545,
2813
+ "grad_norm": 0.06304040551185608,
2814
+ "learning_rate": 1.125638179800222e-05,
2815
+ "loss": 0.0024,
2816
+ "step": 3940
2817
+ },
2818
+ {
2819
+ "epoch": 4.37936772046589,
2820
+ "grad_norm": 0.2765149176120758,
2821
+ "learning_rate": 1.123418423973363e-05,
2822
+ "loss": 0.003,
2823
+ "step": 3950
2824
+ },
2825
+ {
2826
+ "epoch": 4.390460343871325,
2827
+ "grad_norm": 0.628939688205719,
2828
+ "learning_rate": 1.1211986681465039e-05,
2829
+ "loss": 0.0019,
2830
+ "step": 3960
2831
+ },
2832
+ {
2833
+ "epoch": 4.401552967276761,
2834
+ "grad_norm": 0.5144364833831787,
2835
+ "learning_rate": 1.1189789123196449e-05,
2836
+ "loss": 0.0038,
2837
+ "step": 3970
2838
+ },
2839
+ {
2840
+ "epoch": 4.412645590682196,
2841
+ "grad_norm": 0.02215411141514778,
2842
+ "learning_rate": 1.116759156492786e-05,
2843
+ "loss": 0.0052,
2844
+ "step": 3980
2845
+ },
2846
+ {
2847
+ "epoch": 4.423738214087631,
2848
+ "grad_norm": 0.8358492851257324,
2849
+ "learning_rate": 1.114539400665927e-05,
2850
+ "loss": 0.0053,
2851
+ "step": 3990
2852
+ },
2853
+ {
2854
+ "epoch": 4.434830837493067,
2855
+ "grad_norm": 0.49006780982017517,
2856
+ "learning_rate": 1.1123196448390678e-05,
2857
+ "loss": 0.0029,
2858
+ "step": 4000
2859
+ },
2860
+ {
2861
+ "epoch": 4.445923460898502,
2862
+ "grad_norm": 0.08062740415334702,
2863
+ "learning_rate": 1.1100998890122088e-05,
2864
+ "loss": 0.0095,
2865
+ "step": 4010
2866
+ },
2867
+ {
2868
+ "epoch": 4.4570160843039375,
2869
+ "grad_norm": 0.931599497795105,
2870
+ "learning_rate": 1.1078801331853496e-05,
2871
+ "loss": 0.009,
2872
+ "step": 4020
2873
+ },
2874
+ {
2875
+ "epoch": 4.468108707709373,
2876
+ "grad_norm": 1.4779237508773804,
2877
+ "learning_rate": 1.1056603773584906e-05,
2878
+ "loss": 0.0075,
2879
+ "step": 4030
2880
+ },
2881
+ {
2882
+ "epoch": 4.479201331114808,
2883
+ "grad_norm": 0.0675542801618576,
2884
+ "learning_rate": 1.1034406215316316e-05,
2885
+ "loss": 0.0029,
2886
+ "step": 4040
2887
+ },
2888
+ {
2889
+ "epoch": 4.490293954520244,
2890
+ "grad_norm": 1.2957643270492554,
2891
+ "learning_rate": 1.1012208657047724e-05,
2892
+ "loss": 0.0079,
2893
+ "step": 4050
2894
+ },
2895
+ {
2896
+ "epoch": 4.501386577925679,
2897
+ "grad_norm": 1.259100079536438,
2898
+ "learning_rate": 1.0990011098779134e-05,
2899
+ "loss": 0.0045,
2900
+ "step": 4060
2901
+ },
2902
+ {
2903
+ "epoch": 4.512479201331114,
2904
+ "grad_norm": 0.4716441333293915,
2905
+ "learning_rate": 1.0967813540510546e-05,
2906
+ "loss": 0.0022,
2907
+ "step": 4070
2908
+ },
2909
+ {
2910
+ "epoch": 4.52357182473655,
2911
+ "grad_norm": 0.12484975159168243,
2912
+ "learning_rate": 1.0945615982241955e-05,
2913
+ "loss": 0.0031,
2914
+ "step": 4080
2915
+ },
2916
+ {
2917
+ "epoch": 4.534664448141985,
2918
+ "grad_norm": 0.07594279199838638,
2919
+ "learning_rate": 1.0923418423973364e-05,
2920
+ "loss": 0.0027,
2921
+ "step": 4090
2922
+ },
2923
+ {
2924
+ "epoch": 4.5457570715474205,
2925
+ "grad_norm": 0.3258889317512512,
2926
+ "learning_rate": 1.0901220865704773e-05,
2927
+ "loss": 0.0051,
2928
+ "step": 4100
2929
+ },
2930
+ {
2931
+ "epoch": 4.556849694952856,
2932
+ "grad_norm": 0.4499701261520386,
2933
+ "learning_rate": 1.0879023307436183e-05,
2934
+ "loss": 0.0016,
2935
+ "step": 4110
2936
+ },
2937
+ {
2938
+ "epoch": 4.567942318358292,
2939
+ "grad_norm": 0.15266552567481995,
2940
+ "learning_rate": 1.0856825749167591e-05,
2941
+ "loss": 0.0026,
2942
+ "step": 4120
2943
+ },
2944
+ {
2945
+ "epoch": 4.5790349417637275,
2946
+ "grad_norm": 0.07247844338417053,
2947
+ "learning_rate": 1.0834628190899001e-05,
2948
+ "loss": 0.0021,
2949
+ "step": 4130
2950
+ },
2951
+ {
2952
+ "epoch": 4.590127565169163,
2953
+ "grad_norm": 1.395110845565796,
2954
+ "learning_rate": 1.0812430632630411e-05,
2955
+ "loss": 0.0033,
2956
+ "step": 4140
2957
+ },
2958
+ {
2959
+ "epoch": 4.601220188574598,
2960
+ "grad_norm": 1.36636221408844,
2961
+ "learning_rate": 1.079023307436182e-05,
2962
+ "loss": 0.003,
2963
+ "step": 4150
2964
+ },
2965
+ {
2966
+ "epoch": 4.612312811980034,
2967
+ "grad_norm": 1.0960825681686401,
2968
+ "learning_rate": 1.0768035516093231e-05,
2969
+ "loss": 0.0011,
2970
+ "step": 4160
2971
+ },
2972
+ {
2973
+ "epoch": 4.623405435385469,
2974
+ "grad_norm": 1.0867245197296143,
2975
+ "learning_rate": 1.074583795782464e-05,
2976
+ "loss": 0.0045,
2977
+ "step": 4170
2978
+ },
2979
+ {
2980
+ "epoch": 4.634498058790904,
2981
+ "grad_norm": 0.01242793258279562,
2982
+ "learning_rate": 1.072364039955605e-05,
2983
+ "loss": 0.0034,
2984
+ "step": 4180
2985
+ },
2986
+ {
2987
+ "epoch": 4.64559068219634,
2988
+ "grad_norm": 0.0266682468354702,
2989
+ "learning_rate": 1.0701442841287459e-05,
2990
+ "loss": 0.0026,
2991
+ "step": 4190
2992
+ },
2993
+ {
2994
+ "epoch": 4.656683305601775,
2995
+ "grad_norm": 1.711166262626648,
2996
+ "learning_rate": 1.0679245283018869e-05,
2997
+ "loss": 0.0032,
2998
+ "step": 4200
2999
+ },
3000
+ {
3001
+ "epoch": 4.6677759290072105,
3002
+ "grad_norm": 1.225320816040039,
3003
+ "learning_rate": 1.0657047724750279e-05,
3004
+ "loss": 0.0049,
3005
+ "step": 4210
3006
+ },
3007
+ {
3008
+ "epoch": 4.678868552412646,
3009
+ "grad_norm": 0.05706857517361641,
3010
+ "learning_rate": 1.0634850166481687e-05,
3011
+ "loss": 0.0024,
3012
+ "step": 4220
3013
+ },
3014
+ {
3015
+ "epoch": 4.689961175818081,
3016
+ "grad_norm": 0.08420650660991669,
3017
+ "learning_rate": 1.0612652608213097e-05,
3018
+ "loss": 0.006,
3019
+ "step": 4230
3020
+ },
3021
+ {
3022
+ "epoch": 4.701053799223517,
3023
+ "grad_norm": 0.014957012608647346,
3024
+ "learning_rate": 1.0590455049944508e-05,
3025
+ "loss": 0.0033,
3026
+ "step": 4240
3027
+ },
3028
+ {
3029
+ "epoch": 4.712146422628952,
3030
+ "grad_norm": 0.2216530293226242,
3031
+ "learning_rate": 1.0568257491675918e-05,
3032
+ "loss": 0.0039,
3033
+ "step": 4250
3034
+ },
3035
+ {
3036
+ "epoch": 4.723239046034387,
3037
+ "grad_norm": 0.1360142081975937,
3038
+ "learning_rate": 1.0546059933407326e-05,
3039
+ "loss": 0.0018,
3040
+ "step": 4260
3041
+ },
3042
+ {
3043
+ "epoch": 4.734331669439823,
3044
+ "grad_norm": 0.004492219537496567,
3045
+ "learning_rate": 1.0523862375138736e-05,
3046
+ "loss": 0.0016,
3047
+ "step": 4270
3048
+ },
3049
+ {
3050
+ "epoch": 4.745424292845258,
3051
+ "grad_norm": 0.0539228618144989,
3052
+ "learning_rate": 1.0501664816870146e-05,
3053
+ "loss": 0.0022,
3054
+ "step": 4280
3055
+ },
3056
+ {
3057
+ "epoch": 4.7565169162506935,
3058
+ "grad_norm": 0.35766854882240295,
3059
+ "learning_rate": 1.0479467258601554e-05,
3060
+ "loss": 0.0022,
3061
+ "step": 4290
3062
+ },
3063
+ {
3064
+ "epoch": 4.767609539656129,
3065
+ "grad_norm": 0.3328166902065277,
3066
+ "learning_rate": 1.0457269700332964e-05,
3067
+ "loss": 0.0062,
3068
+ "step": 4300
3069
+ },
3070
+ {
3071
+ "epoch": 4.778702163061564,
3072
+ "grad_norm": 0.08985334634780884,
3073
+ "learning_rate": 1.0435072142064372e-05,
3074
+ "loss": 0.0013,
3075
+ "step": 4310
3076
+ },
3077
+ {
3078
+ "epoch": 4.789794786467,
3079
+ "grad_norm": 0.010276420041918755,
3080
+ "learning_rate": 1.0412874583795782e-05,
3081
+ "loss": 0.0031,
3082
+ "step": 4320
3083
+ },
3084
+ {
3085
+ "epoch": 4.800887409872435,
3086
+ "grad_norm": 0.5527135133743286,
3087
+ "learning_rate": 1.0390677025527194e-05,
3088
+ "loss": 0.004,
3089
+ "step": 4330
3090
+ },
3091
+ {
3092
+ "epoch": 4.81198003327787,
3093
+ "grad_norm": 0.21018469333648682,
3094
+ "learning_rate": 1.0368479467258603e-05,
3095
+ "loss": 0.0021,
3096
+ "step": 4340
3097
+ },
3098
+ {
3099
+ "epoch": 4.823072656683306,
3100
+ "grad_norm": 0.10334997624158859,
3101
+ "learning_rate": 1.0346281908990012e-05,
3102
+ "loss": 0.0045,
3103
+ "step": 4350
3104
+ },
3105
+ {
3106
+ "epoch": 4.834165280088741,
3107
+ "grad_norm": 0.09865798056125641,
3108
+ "learning_rate": 1.0324084350721422e-05,
3109
+ "loss": 0.0008,
3110
+ "step": 4360
3111
+ },
3112
+ {
3113
+ "epoch": 4.8452579034941765,
3114
+ "grad_norm": 0.24930772185325623,
3115
+ "learning_rate": 1.0301886792452831e-05,
3116
+ "loss": 0.0021,
3117
+ "step": 4370
3118
+ },
3119
+ {
3120
+ "epoch": 4.856350526899612,
3121
+ "grad_norm": 0.7967151999473572,
3122
+ "learning_rate": 1.027968923418424e-05,
3123
+ "loss": 0.0027,
3124
+ "step": 4380
3125
+ },
3126
+ {
3127
+ "epoch": 4.867443150305047,
3128
+ "grad_norm": 0.2117662876844406,
3129
+ "learning_rate": 1.025749167591565e-05,
3130
+ "loss": 0.0061,
3131
+ "step": 4390
3132
+ },
3133
+ {
3134
+ "epoch": 4.878535773710483,
3135
+ "grad_norm": 0.38198602199554443,
3136
+ "learning_rate": 1.023529411764706e-05,
3137
+ "loss": 0.0032,
3138
+ "step": 4400
3139
+ },
3140
+ {
3141
+ "epoch": 4.889628397115918,
3142
+ "grad_norm": 0.8646931648254395,
3143
+ "learning_rate": 1.0213096559378467e-05,
3144
+ "loss": 0.0074,
3145
+ "step": 4410
3146
+ },
3147
+ {
3148
+ "epoch": 4.900721020521353,
3149
+ "grad_norm": 0.5245551466941833,
3150
+ "learning_rate": 1.0190899001109879e-05,
3151
+ "loss": 0.0029,
3152
+ "step": 4420
3153
+ },
3154
+ {
3155
+ "epoch": 4.911813643926789,
3156
+ "grad_norm": 0.368184894323349,
3157
+ "learning_rate": 1.0168701442841289e-05,
3158
+ "loss": 0.0056,
3159
+ "step": 4430
3160
+ },
3161
+ {
3162
+ "epoch": 4.922906267332224,
3163
+ "grad_norm": 1.3652911186218262,
3164
+ "learning_rate": 1.0146503884572699e-05,
3165
+ "loss": 0.0022,
3166
+ "step": 4440
3167
+ },
3168
+ {
3169
+ "epoch": 4.9339988907376595,
3170
+ "grad_norm": 2.6624059677124023,
3171
+ "learning_rate": 1.0124306326304107e-05,
3172
+ "loss": 0.0059,
3173
+ "step": 4450
3174
+ },
3175
+ {
3176
+ "epoch": 4.945091514143095,
3177
+ "grad_norm": 0.10531987994909286,
3178
+ "learning_rate": 1.0102108768035517e-05,
3179
+ "loss": 0.0017,
3180
+ "step": 4460
3181
+ },
3182
+ {
3183
+ "epoch": 4.95618413754853,
3184
+ "grad_norm": 0.7769778966903687,
3185
+ "learning_rate": 1.0079911209766927e-05,
3186
+ "loss": 0.0054,
3187
+ "step": 4470
3188
+ },
3189
+ {
3190
+ "epoch": 4.967276760953966,
3191
+ "grad_norm": 0.016158653423190117,
3192
+ "learning_rate": 1.0057713651498335e-05,
3193
+ "loss": 0.002,
3194
+ "step": 4480
3195
+ },
3196
+ {
3197
+ "epoch": 4.978369384359401,
3198
+ "grad_norm": 0.09255735576152802,
3199
+ "learning_rate": 1.0035516093229745e-05,
3200
+ "loss": 0.005,
3201
+ "step": 4490
3202
+ },
3203
+ {
3204
+ "epoch": 4.989462007764836,
3205
+ "grad_norm": 0.7796350717544556,
3206
+ "learning_rate": 1.0013318534961156e-05,
3207
+ "loss": 0.0025,
3208
+ "step": 4500
3209
+ },
3210
+ {
3211
+ "epoch": 5.0,
3212
+ "grad_norm": 0.36661309003829956,
3213
+ "learning_rate": 9.991120976692564e-06,
3214
+ "loss": 0.0011,
3215
+ "step": 4510
3216
+ },
3217
+ {
3218
+ "epoch": 5.0,
3219
+ "eval_accuracy": 0.8743667679837892,
3220
+ "eval_f1": 0.7009646302250804,
3221
+ "eval_loss": 0.5751733183860779,
3222
+ "eval_precision": 0.6402349486049926,
3223
+ "eval_recall": 0.7744227353463587,
3224
+ "eval_runtime": 2.6897,
3225
+ "eval_samples_per_second": 366.959,
3226
+ "eval_steps_per_second": 11.526,
3227
+ "step": 4510
3228
+ }
3229
+ ],
3230
+ "logging_steps": 10,
3231
+ "max_steps": 9010,
3232
+ "num_input_tokens_seen": 0,
3233
+ "num_train_epochs": 10,
3234
+ "save_steps": 500,
3235
+ "stateful_callbacks": {
3236
+ "TrainerControl": {
3237
+ "args": {
3238
+ "should_epoch_stop": false,
3239
+ "should_evaluate": false,
3240
+ "should_log": false,
3241
+ "should_save": true,
3242
+ "should_training_stop": false
3243
+ },
3244
+ "attributes": {}
3245
+ }
3246
+ },
3247
+ "total_flos": 4.807363667558798e+16,
3248
+ "train_batch_size": 32,
3249
+ "trial_name": null,
3250
+ "trial_params": null
3251
+ }
checkpoint-4510/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e64fe8f815fa80a7291289cde9af3ebb27f62206c40a93e8adf97661cd289de3
3
+ size 5368
checkpoint-5412/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }