Hugging Face's logo

CaffeineThief
/
cysecbert-ttp-bert-base_data

Safetensors
bert
Model card Files
xet
 Community
CaffeineThief commited on
Commit
db02f52
·
verified ·
1 Parent(s): 2060fee

Upload saved model files

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. checkpoint-109/config.json +130 -0
  2. checkpoint-109/model.safetensors +3 -0
  3. checkpoint-109/optimizer.pt +3 -0
  4. checkpoint-109/rng_state.pth +3 -0
  5. checkpoint-109/scaler.pt +3 -0
  6. checkpoint-109/scheduler.pt +3 -0
  7. checkpoint-109/special_tokens_map.json +37 -0
  8. checkpoint-109/tokenizer.json +0 -0
  9. checkpoint-109/tokenizer_config.json +56 -0
  10. checkpoint-109/trainer_state.json +809 -0
  11. checkpoint-109/training_args.bin +3 -0
  12. checkpoint-109/vocab.txt +0 -0
  13. checkpoint-1090/config.json +130 -0
  14. checkpoint-1090/model.safetensors +3 -0
  15. checkpoint-1090/optimizer.pt +3 -0
  16. checkpoint-1090/rng_state.pth +3 -0
  17. checkpoint-1090/scaler.pt +3 -0
  18. checkpoint-1090/scheduler.pt +3 -0
  19. checkpoint-1090/special_tokens_map.json +37 -0
  20. checkpoint-1090/tokenizer.json +0 -0
  21. checkpoint-1090/tokenizer_config.json +56 -0
  22. checkpoint-1090/trainer_state.json +0 -0
  23. checkpoint-1090/training_args.bin +3 -0
  24. checkpoint-1090/vocab.txt +0 -0
  25. checkpoint-218/config.json +130 -0
  26. checkpoint-218/model.safetensors +3 -0
  27. checkpoint-218/optimizer.pt +3 -0
  28. checkpoint-218/rng_state.pth +3 -0
  29. checkpoint-218/scaler.pt +3 -0
  30. checkpoint-218/scheduler.pt +3 -0
  31. checkpoint-218/special_tokens_map.json +37 -0
  32. checkpoint-218/tokenizer.json +0 -0
  33. checkpoint-218/tokenizer_config.json +56 -0
  34. checkpoint-218/trainer_state.json +1584 -0
  35. checkpoint-218/training_args.bin +3 -0
  36. checkpoint-218/vocab.txt +0 -0
  37. checkpoint-327/config.json +130 -0
  38. checkpoint-327/model.safetensors +3 -0
  39. checkpoint-327/optimizer.pt +3 -0
  40. checkpoint-327/rng_state.pth +3 -0
  41. checkpoint-327/scaler.pt +3 -0
  42. checkpoint-327/scheduler.pt +3 -0
  43. checkpoint-327/special_tokens_map.json +37 -0
  44. checkpoint-327/tokenizer.json +0 -0
  45. checkpoint-327/tokenizer_config.json +56 -0
  46. checkpoint-327/trainer_state.json +2359 -0
  47. checkpoint-327/training_args.bin +3 -0
  48. checkpoint-327/vocab.txt +0 -0
  49. checkpoint-436/config.json +130 -0
  50. checkpoint-436/model.safetensors +3 -0
checkpoint-109/config.json ADDED
@@ -0,0 +1,130 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "T1003.001",
13
+ "1": "T1005",
14
+ "2": "T1012",
15
+ "3": "T1016",
16
+ "4": "T1021.001",
17
+ "5": "T1027",
18
+ "6": "T1033",
19
+ "7": "T1036.005",
20
+ "8": "T1041",
21
+ "9": "T1047",
22
+ "10": "T1053.005",
23
+ "11": "T1055",
24
+ "12": "T1056.001",
25
+ "13": "T1057",
26
+ "14": "T1059.003",
27
+ "15": "T1068",
28
+ "16": "T1070.004",
29
+ "17": "T1071.001",
30
+ "18": "T1072",
31
+ "19": "T1074.001",
32
+ "20": "T1078",
33
+ "21": "T1082",
34
+ "22": "T1083",
35
+ "23": "T1090",
36
+ "24": "T1095",
37
+ "25": "T1105",
38
+ "26": "T1106",
39
+ "27": "T1110",
40
+ "28": "T1112",
41
+ "29": "T1113",
42
+ "30": "T1140",
43
+ "31": "T1190",
44
+ "32": "T1204.002",
45
+ "33": "T1210",
46
+ "34": "T1218.011",
47
+ "35": "T1219",
48
+ "36": "T1484.001",
49
+ "37": "T1518.001",
50
+ "38": "T1543.003",
51
+ "39": "T1547.001",
52
+ "40": "T1548.002",
53
+ "41": "T1552.001",
54
+ "42": "T1557.001",
55
+ "43": "T1562.001",
56
+ "44": "T1564.001",
57
+ "45": "T1566.001",
58
+ "46": "T1569.002",
59
+ "47": "T1570",
60
+ "48": "T1573.001",
61
+ "49": "T1574.002"
62
+ },
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 3072,
65
+ "label2id": {
66
+ "T1003.001": 0,
67
+ "T1005": 1,
68
+ "T1012": 2,
69
+ "T1016": 3,
70
+ "T1021.001": 4,
71
+ "T1027": 5,
72
+ "T1033": 6,
73
+ "T1036.005": 7,
74
+ "T1041": 8,
75
+ "T1047": 9,
76
+ "T1053.005": 10,
77
+ "T1055": 11,
78
+ "T1056.001": 12,
79
+ "T1057": 13,
80
+ "T1059.003": 14,
81
+ "T1068": 15,
82
+ "T1070.004": 16,
83
+ "T1071.001": 17,
84
+ "T1072": 18,
85
+ "T1074.001": 19,
86
+ "T1078": 20,
87
+ "T1082": 21,
88
+ "T1083": 22,
89
+ "T1090": 23,
90
+ "T1095": 24,
91
+ "T1105": 25,
92
+ "T1106": 26,
93
+ "T1110": 27,
94
+ "T1112": 28,
95
+ "T1113": 29,
96
+ "T1140": 30,
97
+ "T1190": 31,
98
+ "T1204.002": 32,
99
+ "T1210": 33,
100
+ "T1218.011": 34,
101
+ "T1219": 35,
102
+ "T1484.001": 36,
103
+ "T1518.001": 37,
104
+ "T1543.003": 38,
105
+ "T1547.001": 39,
106
+ "T1548.002": 40,
107
+ "T1552.001": 41,
108
+ "T1557.001": 42,
109
+ "T1562.001": 43,
110
+ "T1564.001": 44,
111
+ "T1566.001": 45,
112
+ "T1569.002": 46,
113
+ "T1570": 47,
114
+ "T1573.001": 48,
115
+ "T1574.002": 49
116
+ },
117
+ "layer_norm_eps": 1e-12,
118
+ "max_position_embeddings": 512,
119
+ "model_type": "bert",
120
+ "num_attention_heads": 12,
121
+ "num_hidden_layers": 12,
122
+ "pad_token_id": 0,
123
+ "position_embedding_type": "absolute",
124
+ "problem_type": "multi_label_classification",
125
+ "torch_dtype": "float32",
126
+ "transformers_version": "4.55.2",
127
+ "type_vocab_size": 2,
128
+ "use_cache": true,
129
+ "vocab_size": 30522
130
+ }
checkpoint-109/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0770d6fab2e838830d846c89f6bf1bb74a0da9817ec8a036baf42e27102aa3b
3
+ size 438106296
checkpoint-109/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:22b4d9c2e33e6214a2ca0e2229f1beadb8ebe5641cff36a79d3907d702b33e8e
3
+ size 876333626
checkpoint-109/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff6511f45fa379876f237693b23d015862fc6e4358901770ec19f197e41c880a
3
+ size 14244
checkpoint-109/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
3
+ size 988
checkpoint-109/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:02e6d28947baf124a0aeaf60b650ff546068dbe5a5f95f66a1df2d0d92517bab
3
+ size 1064
checkpoint-109/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
checkpoint-109/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-109/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
checkpoint-109/trainer_state.json ADDED
@@ -0,0 +1,809 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 109,
3
+ "best_metric": 0.0,
4
+ "best_model_checkpoint": "./cysecbert-ttp-bert-base_data/checkpoint-109",
5
+ "epoch": 1.0,
6
+ "eval_steps": 500,
7
+ "global_step": 109,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.009174311926605505,
14
+ "grad_norm": 46530.15234375,
15
+ "learning_rate": 0.0,
16
+ "loss": 0.7023,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.01834862385321101,
21
+ "grad_norm": 49571.35546875,
22
+ "learning_rate": 5.000000000000001e-07,
23
+ "loss": 0.7006,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.027522935779816515,
28
+ "grad_norm": 45885.55078125,
29
+ "learning_rate": 1.0000000000000002e-06,
30
+ "loss": 0.7033,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.03669724770642202,
35
+ "grad_norm": 43333.73046875,
36
+ "learning_rate": 1.5e-06,
37
+ "loss": 0.7014,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.045871559633027525,
42
+ "grad_norm": 44340.41796875,
43
+ "learning_rate": 2.0000000000000003e-06,
44
+ "loss": 0.6992,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.05504587155963303,
49
+ "grad_norm": 48282.54296875,
50
+ "learning_rate": 2.5e-06,
51
+ "loss": 0.6924,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.06422018348623854,
56
+ "grad_norm": 48561.56640625,
57
+ "learning_rate": 3e-06,
58
+ "loss": 0.6948,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.07339449541284404,
63
+ "grad_norm": 43580.078125,
64
+ "learning_rate": 3.5000000000000004e-06,
65
+ "loss": 0.6916,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.08256880733944955,
70
+ "grad_norm": 46552.28515625,
71
+ "learning_rate": 4.000000000000001e-06,
72
+ "loss": 0.6854,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.09174311926605505,
77
+ "grad_norm": 49042.15234375,
78
+ "learning_rate": 4.5e-06,
79
+ "loss": 0.6822,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.10091743119266056,
84
+ "grad_norm": 97274.8515625,
85
+ "learning_rate": 5e-06,
86
+ "loss": 0.6788,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.11009174311926606,
91
+ "grad_norm": 55310.375,
92
+ "learning_rate": 5.500000000000001e-06,
93
+ "loss": 0.6781,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.11926605504587157,
98
+ "grad_norm": 70317.6328125,
99
+ "learning_rate": 6e-06,
100
+ "loss": 0.6754,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.12844036697247707,
105
+ "grad_norm": 46825.9765625,
106
+ "learning_rate": 6.5000000000000004e-06,
107
+ "loss": 0.6626,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.13761467889908258,
112
+ "grad_norm": 50745.15625,
113
+ "learning_rate": 7.000000000000001e-06,
114
+ "loss": 0.655,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 0.14678899082568808,
119
+ "grad_norm": 51342.453125,
120
+ "learning_rate": 7.5e-06,
121
+ "loss": 0.6491,
122
+ "step": 16
123
+ },
124
+ {
125
+ "epoch": 0.1559633027522936,
126
+ "grad_norm": 51815.421875,
127
+ "learning_rate": 8.000000000000001e-06,
128
+ "loss": 0.6464,
129
+ "step": 17
130
+ },
131
+ {
132
+ "epoch": 0.1651376146788991,
133
+ "grad_norm": 54610.65234375,
134
+ "learning_rate": 8.500000000000002e-06,
135
+ "loss": 0.6298,
136
+ "step": 18
137
+ },
138
+ {
139
+ "epoch": 0.1743119266055046,
140
+ "grad_norm": 61936.31640625,
141
+ "learning_rate": 9e-06,
142
+ "loss": 0.623,
143
+ "step": 19
144
+ },
145
+ {
146
+ "epoch": 0.1834862385321101,
147
+ "grad_norm": 56021.69921875,
148
+ "learning_rate": 9.5e-06,
149
+ "loss": 0.6156,
150
+ "step": 20
151
+ },
152
+ {
153
+ "epoch": 0.1926605504587156,
154
+ "grad_norm": 60806.10546875,
155
+ "learning_rate": 1e-05,
156
+ "loss": 0.5981,
157
+ "step": 21
158
+ },
159
+ {
160
+ "epoch": 0.2018348623853211,
161
+ "grad_norm": 58582.4140625,
162
+ "learning_rate": 1.05e-05,
163
+ "loss": 0.5869,
164
+ "step": 22
165
+ },
166
+ {
167
+ "epoch": 0.21100917431192662,
168
+ "grad_norm": 59099.16015625,
169
+ "learning_rate": 1.1000000000000001e-05,
170
+ "loss": 0.5732,
171
+ "step": 23
172
+ },
173
+ {
174
+ "epoch": 0.22018348623853212,
175
+ "grad_norm": 52971.81640625,
176
+ "learning_rate": 1.1500000000000002e-05,
177
+ "loss": 0.5587,
178
+ "step": 24
179
+ },
180
+ {
181
+ "epoch": 0.22935779816513763,
182
+ "grad_norm": 51724.43359375,
183
+ "learning_rate": 1.2e-05,
184
+ "loss": 0.5477,
185
+ "step": 25
186
+ },
187
+ {
188
+ "epoch": 0.23853211009174313,
189
+ "grad_norm": 58133.453125,
190
+ "learning_rate": 1.25e-05,
191
+ "loss": 0.5422,
192
+ "step": 26
193
+ },
194
+ {
195
+ "epoch": 0.24770642201834864,
196
+ "grad_norm": 54134.3359375,
197
+ "learning_rate": 1.3000000000000001e-05,
198
+ "loss": 0.5319,
199
+ "step": 27
200
+ },
201
+ {
202
+ "epoch": 0.25688073394495414,
203
+ "grad_norm": 78195.7890625,
204
+ "learning_rate": 1.3500000000000001e-05,
205
+ "loss": 0.5259,
206
+ "step": 28
207
+ },
208
+ {
209
+ "epoch": 0.26605504587155965,
210
+ "grad_norm": 47713.20703125,
211
+ "learning_rate": 1.4000000000000001e-05,
212
+ "loss": 0.5105,
213
+ "step": 29
214
+ },
215
+ {
216
+ "epoch": 0.27522935779816515,
217
+ "grad_norm": 45838.0703125,
218
+ "learning_rate": 1.45e-05,
219
+ "loss": 0.499,
220
+ "step": 30
221
+ },
222
+ {
223
+ "epoch": 0.28440366972477066,
224
+ "grad_norm": 46735.58203125,
225
+ "learning_rate": 1.5e-05,
226
+ "loss": 0.4914,
227
+ "step": 31
228
+ },
229
+ {
230
+ "epoch": 0.29357798165137616,
231
+ "grad_norm": 45432.8828125,
232
+ "learning_rate": 1.55e-05,
233
+ "loss": 0.4783,
234
+ "step": 32
235
+ },
236
+ {
237
+ "epoch": 0.30275229357798167,
238
+ "grad_norm": 46758.58984375,
239
+ "learning_rate": 1.6000000000000003e-05,
240
+ "loss": 0.4758,
241
+ "step": 33
242
+ },
243
+ {
244
+ "epoch": 0.3119266055045872,
245
+ "grad_norm": 45245.82421875,
246
+ "learning_rate": 1.65e-05,
247
+ "loss": 0.4608,
248
+ "step": 34
249
+ },
250
+ {
251
+ "epoch": 0.3211009174311927,
252
+ "grad_norm": 43336.75390625,
253
+ "learning_rate": 1.7000000000000003e-05,
254
+ "loss": 0.4567,
255
+ "step": 35
256
+ },
257
+ {
258
+ "epoch": 0.3302752293577982,
259
+ "grad_norm": 41445.90625,
260
+ "learning_rate": 1.75e-05,
261
+ "loss": 0.4465,
262
+ "step": 36
263
+ },
264
+ {
265
+ "epoch": 0.3394495412844037,
266
+ "grad_norm": 42278.80078125,
267
+ "learning_rate": 1.8e-05,
268
+ "loss": 0.4395,
269
+ "step": 37
270
+ },
271
+ {
272
+ "epoch": 0.3486238532110092,
273
+ "grad_norm": 42039.28515625,
274
+ "learning_rate": 1.85e-05,
275
+ "loss": 0.4289,
276
+ "step": 38
277
+ },
278
+ {
279
+ "epoch": 0.3577981651376147,
280
+ "grad_norm": 41411.61328125,
281
+ "learning_rate": 1.9e-05,
282
+ "loss": 0.423,
283
+ "step": 39
284
+ },
285
+ {
286
+ "epoch": 0.3669724770642202,
287
+ "grad_norm": 40133.84375,
288
+ "learning_rate": 1.9500000000000003e-05,
289
+ "loss": 0.4191,
290
+ "step": 40
291
+ },
292
+ {
293
+ "epoch": 0.3761467889908257,
294
+ "grad_norm": 39758.8828125,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.4071,
297
+ "step": 41
298
+ },
299
+ {
300
+ "epoch": 0.3853211009174312,
301
+ "grad_norm": 38760.36328125,
302
+ "learning_rate": 2.05e-05,
303
+ "loss": 0.3996,
304
+ "step": 42
305
+ },
306
+ {
307
+ "epoch": 0.3944954128440367,
308
+ "grad_norm": 38552.80078125,
309
+ "learning_rate": 2.1e-05,
310
+ "loss": 0.3918,
311
+ "step": 43
312
+ },
313
+ {
314
+ "epoch": 0.4036697247706422,
315
+ "grad_norm": 38561.7578125,
316
+ "learning_rate": 2.15e-05,
317
+ "loss": 0.3865,
318
+ "step": 44
319
+ },
320
+ {
321
+ "epoch": 0.41284403669724773,
322
+ "grad_norm": 37616.85546875,
323
+ "learning_rate": 2.2000000000000003e-05,
324
+ "loss": 0.3794,
325
+ "step": 45
326
+ },
327
+ {
328
+ "epoch": 0.42201834862385323,
329
+ "grad_norm": 37784.78515625,
330
+ "learning_rate": 2.25e-05,
331
+ "loss": 0.3745,
332
+ "step": 46
333
+ },
334
+ {
335
+ "epoch": 0.43119266055045874,
336
+ "grad_norm": 36332.9140625,
337
+ "learning_rate": 2.3000000000000003e-05,
338
+ "loss": 0.3711,
339
+ "step": 47
340
+ },
341
+ {
342
+ "epoch": 0.44036697247706424,
343
+ "grad_norm": 38899.73046875,
344
+ "learning_rate": 2.35e-05,
345
+ "loss": 0.3643,
346
+ "step": 48
347
+ },
348
+ {
349
+ "epoch": 0.44954128440366975,
350
+ "grad_norm": 36174.9765625,
351
+ "learning_rate": 2.4e-05,
352
+ "loss": 0.3575,
353
+ "step": 49
354
+ },
355
+ {
356
+ "epoch": 0.45871559633027525,
357
+ "grad_norm": 35699.83203125,
358
+ "learning_rate": 2.45e-05,
359
+ "loss": 0.347,
360
+ "step": 50
361
+ },
362
+ {
363
+ "epoch": 0.46788990825688076,
364
+ "grad_norm": 35562.3671875,
365
+ "learning_rate": 2.5e-05,
366
+ "loss": 0.3438,
367
+ "step": 51
368
+ },
369
+ {
370
+ "epoch": 0.47706422018348627,
371
+ "grad_norm": 35428.0625,
372
+ "learning_rate": 2.5500000000000003e-05,
373
+ "loss": 0.3325,
374
+ "step": 52
375
+ },
376
+ {
377
+ "epoch": 0.48623853211009177,
378
+ "grad_norm": 34396.80078125,
379
+ "learning_rate": 2.6000000000000002e-05,
380
+ "loss": 0.3302,
381
+ "step": 53
382
+ },
383
+ {
384
+ "epoch": 0.4954128440366973,
385
+ "grad_norm": 38376.41796875,
386
+ "learning_rate": 2.6500000000000004e-05,
387
+ "loss": 0.3218,
388
+ "step": 54
389
+ },
390
+ {
391
+ "epoch": 0.5045871559633027,
392
+ "grad_norm": 33996.23046875,
393
+ "learning_rate": 2.7000000000000002e-05,
394
+ "loss": 0.3131,
395
+ "step": 55
396
+ },
397
+ {
398
+ "epoch": 0.5137614678899083,
399
+ "grad_norm": 33070.55859375,
400
+ "learning_rate": 2.7500000000000004e-05,
401
+ "loss": 0.3129,
402
+ "step": 56
403
+ },
404
+ {
405
+ "epoch": 0.5229357798165137,
406
+ "grad_norm": 33200.44140625,
407
+ "learning_rate": 2.8000000000000003e-05,
408
+ "loss": 0.3003,
409
+ "step": 57
410
+ },
411
+ {
412
+ "epoch": 0.5321100917431193,
413
+ "grad_norm": 33099.1015625,
414
+ "learning_rate": 2.8499999999999998e-05,
415
+ "loss": 0.2904,
416
+ "step": 58
417
+ },
418
+ {
419
+ "epoch": 0.5412844036697247,
420
+ "grad_norm": 31352.298828125,
421
+ "learning_rate": 2.9e-05,
422
+ "loss": 0.2911,
423
+ "step": 59
424
+ },
425
+ {
426
+ "epoch": 0.5504587155963303,
427
+ "grad_norm": 31965.087890625,
428
+ "learning_rate": 2.95e-05,
429
+ "loss": 0.2815,
430
+ "step": 60
431
+ },
432
+ {
433
+ "epoch": 0.5596330275229358,
434
+ "grad_norm": 30810.849609375,
435
+ "learning_rate": 3e-05,
436
+ "loss": 0.2818,
437
+ "step": 61
438
+ },
439
+ {
440
+ "epoch": 0.5688073394495413,
441
+ "grad_norm": 30034.779296875,
442
+ "learning_rate": 3.05e-05,
443
+ "loss": 0.2731,
444
+ "step": 62
445
+ },
446
+ {
447
+ "epoch": 0.5779816513761468,
448
+ "grad_norm": 30009.12109375,
449
+ "learning_rate": 3.1e-05,
450
+ "loss": 0.266,
451
+ "step": 63
452
+ },
453
+ {
454
+ "epoch": 0.5871559633027523,
455
+ "grad_norm": 29305.173828125,
456
+ "learning_rate": 3.15e-05,
457
+ "loss": 0.2609,
458
+ "step": 64
459
+ },
460
+ {
461
+ "epoch": 0.5963302752293578,
462
+ "grad_norm": 29081.853515625,
463
+ "learning_rate": 3.2000000000000005e-05,
464
+ "loss": 0.2543,
465
+ "step": 65
466
+ },
467
+ {
468
+ "epoch": 0.6055045871559633,
469
+ "grad_norm": 28217.021484375,
470
+ "learning_rate": 3.2500000000000004e-05,
471
+ "loss": 0.252,
472
+ "step": 66
473
+ },
474
+ {
475
+ "epoch": 0.6146788990825688,
476
+ "grad_norm": 26448.12890625,
477
+ "learning_rate": 3.3e-05,
478
+ "loss": 0.2565,
479
+ "step": 67
480
+ },
481
+ {
482
+ "epoch": 0.6238532110091743,
483
+ "grad_norm": 27198.80859375,
484
+ "learning_rate": 3.35e-05,
485
+ "loss": 0.2342,
486
+ "step": 68
487
+ },
488
+ {
489
+ "epoch": 0.6330275229357798,
490
+ "grad_norm": 26946.30859375,
491
+ "learning_rate": 3.4000000000000007e-05,
492
+ "loss": 0.2273,
493
+ "step": 69
494
+ },
495
+ {
496
+ "epoch": 0.6422018348623854,
497
+ "grad_norm": 26236.7265625,
498
+ "learning_rate": 3.45e-05,
499
+ "loss": 0.2219,
500
+ "step": 70
501
+ },
502
+ {
503
+ "epoch": 0.6513761467889908,
504
+ "grad_norm": 25535.818359375,
505
+ "learning_rate": 3.5e-05,
506
+ "loss": 0.2197,
507
+ "step": 71
508
+ },
509
+ {
510
+ "epoch": 0.6605504587155964,
511
+ "grad_norm": 24871.8515625,
512
+ "learning_rate": 3.55e-05,
513
+ "loss": 0.2144,
514
+ "step": 72
515
+ },
516
+ {
517
+ "epoch": 0.6697247706422018,
518
+ "grad_norm": 24484.55078125,
519
+ "learning_rate": 3.6e-05,
520
+ "loss": 0.2066,
521
+ "step": 73
522
+ },
523
+ {
524
+ "epoch": 0.6788990825688074,
525
+ "grad_norm": 23655.677734375,
526
+ "learning_rate": 3.65e-05,
527
+ "loss": 0.206,
528
+ "step": 74
529
+ },
530
+ {
531
+ "epoch": 0.6880733944954128,
532
+ "grad_norm": 23129.076171875,
533
+ "learning_rate": 3.7e-05,
534
+ "loss": 0.1979,
535
+ "step": 75
536
+ },
537
+ {
538
+ "epoch": 0.6972477064220184,
539
+ "grad_norm": 21871.90625,
540
+ "learning_rate": 3.7500000000000003e-05,
541
+ "loss": 0.1973,
542
+ "step": 76
543
+ },
544
+ {
545
+ "epoch": 0.7064220183486238,
546
+ "grad_norm": 21561.58984375,
547
+ "learning_rate": 3.8e-05,
548
+ "loss": 0.1935,
549
+ "step": 77
550
+ },
551
+ {
552
+ "epoch": 0.7155963302752294,
553
+ "grad_norm": 22210.29296875,
554
+ "learning_rate": 3.85e-05,
555
+ "loss": 0.1848,
556
+ "step": 78
557
+ },
558
+ {
559
+ "epoch": 0.7247706422018348,
560
+ "grad_norm": 20555.775390625,
561
+ "learning_rate": 3.9000000000000006e-05,
562
+ "loss": 0.1801,
563
+ "step": 79
564
+ },
565
+ {
566
+ "epoch": 0.7339449541284404,
567
+ "grad_norm": 20406.75,
568
+ "learning_rate": 3.9500000000000005e-05,
569
+ "loss": 0.1746,
570
+ "step": 80
571
+ },
572
+ {
573
+ "epoch": 0.7431192660550459,
574
+ "grad_norm": 19150.931640625,
575
+ "learning_rate": 4e-05,
576
+ "loss": 0.1688,
577
+ "step": 81
578
+ },
579
+ {
580
+ "epoch": 0.7522935779816514,
581
+ "grad_norm": 18422.65625,
582
+ "learning_rate": 4.05e-05,
583
+ "loss": 0.1661,
584
+ "step": 82
585
+ },
586
+ {
587
+ "epoch": 0.7614678899082569,
588
+ "grad_norm": 16746.20703125,
589
+ "learning_rate": 4.1e-05,
590
+ "loss": 0.1711,
591
+ "step": 83
592
+ },
593
+ {
594
+ "epoch": 0.7706422018348624,
595
+ "grad_norm": 18229.41015625,
596
+ "learning_rate": 4.15e-05,
597
+ "loss": 0.1646,
598
+ "step": 84
599
+ },
600
+ {
601
+ "epoch": 0.7798165137614679,
602
+ "grad_norm": 17156.267578125,
603
+ "learning_rate": 4.2e-05,
604
+ "loss": 0.1588,
605
+ "step": 85
606
+ },
607
+ {
608
+ "epoch": 0.7889908256880734,
609
+ "grad_norm": 15970.611328125,
610
+ "learning_rate": 4.25e-05,
611
+ "loss": 0.1547,
612
+ "step": 86
613
+ },
614
+ {
615
+ "epoch": 0.7981651376146789,
616
+ "grad_norm": 16997.103515625,
617
+ "learning_rate": 4.3e-05,
618
+ "loss": 0.1493,
619
+ "step": 87
620
+ },
621
+ {
622
+ "epoch": 0.8073394495412844,
623
+ "grad_norm": 15529.58984375,
624
+ "learning_rate": 4.35e-05,
625
+ "loss": 0.1439,
626
+ "step": 88
627
+ },
628
+ {
629
+ "epoch": 0.8165137614678899,
630
+ "grad_norm": 15099.9052734375,
631
+ "learning_rate": 4.4000000000000006e-05,
632
+ "loss": 0.1432,
633
+ "step": 89
634
+ },
635
+ {
636
+ "epoch": 0.8256880733944955,
637
+ "grad_norm": 14261.3232421875,
638
+ "learning_rate": 4.4500000000000004e-05,
639
+ "loss": 0.1396,
640
+ "step": 90
641
+ },
642
+ {
643
+ "epoch": 0.8348623853211009,
644
+ "grad_norm": 13915.87890625,
645
+ "learning_rate": 4.5e-05,
646
+ "loss": 0.1317,
647
+ "step": 91
648
+ },
649
+ {
650
+ "epoch": 0.8440366972477065,
651
+ "grad_norm": 13518.7568359375,
652
+ "learning_rate": 4.55e-05,
653
+ "loss": 0.1289,
654
+ "step": 92
655
+ },
656
+ {
657
+ "epoch": 0.8532110091743119,
658
+ "grad_norm": 12170.1865234375,
659
+ "learning_rate": 4.600000000000001e-05,
660
+ "loss": 0.1371,
661
+ "step": 93
662
+ },
663
+ {
664
+ "epoch": 0.8623853211009175,
665
+ "grad_norm": 11745.9775390625,
666
+ "learning_rate": 4.6500000000000005e-05,
667
+ "loss": 0.1434,
668
+ "step": 94
669
+ },
670
+ {
671
+ "epoch": 0.8715596330275229,
672
+ "grad_norm": 12265.5556640625,
673
+ "learning_rate": 4.7e-05,
674
+ "loss": 0.1343,
675
+ "step": 95
676
+ },
677
+ {
678
+ "epoch": 0.8807339449541285,
679
+ "grad_norm": 11386.3291015625,
680
+ "learning_rate": 4.75e-05,
681
+ "loss": 0.1317,
682
+ "step": 96
683
+ },
684
+ {
685
+ "epoch": 0.8899082568807339,
686
+ "grad_norm": 10869.4599609375,
687
+ "learning_rate": 4.8e-05,
688
+ "loss": 0.1295,
689
+ "step": 97
690
+ },
691
+ {
692
+ "epoch": 0.8990825688073395,
693
+ "grad_norm": 10752.25,
694
+ "learning_rate": 4.85e-05,
695
+ "loss": 0.124,
696
+ "step": 98
697
+ },
698
+ {
699
+ "epoch": 0.908256880733945,
700
+ "grad_norm": 9687.6376953125,
701
+ "learning_rate": 4.9e-05,
702
+ "loss": 0.1271,
703
+ "step": 99
704
+ },
705
+ {
706
+ "epoch": 0.9174311926605505,
707
+ "grad_norm": 9697.15234375,
708
+ "learning_rate": 4.9500000000000004e-05,
709
+ "loss": 0.1194,
710
+ "step": 100
711
+ },
712
+ {
713
+ "epoch": 0.926605504587156,
714
+ "grad_norm": 9423.21875,
715
+ "learning_rate": 5e-05,
716
+ "loss": 0.1176,
717
+ "step": 101
718
+ },
719
+ {
720
+ "epoch": 0.9357798165137615,
721
+ "grad_norm": 10253.8564453125,
722
+ "learning_rate": 4.994949494949495e-05,
723
+ "loss": 0.121,
724
+ "step": 102
725
+ },
726
+ {
727
+ "epoch": 0.944954128440367,
728
+ "grad_norm": 9756.7646484375,
729
+ "learning_rate": 4.98989898989899e-05,
730
+ "loss": 0.1173,
731
+ "step": 103
732
+ },
733
+ {
734
+ "epoch": 0.9541284403669725,
735
+ "grad_norm": 8482.177734375,
736
+ "learning_rate": 4.984848484848485e-05,
737
+ "loss": 0.1233,
738
+ "step": 104
739
+ },
740
+ {
741
+ "epoch": 0.963302752293578,
742
+ "grad_norm": 7755.4892578125,
743
+ "learning_rate": 4.97979797979798e-05,
744
+ "loss": 0.1251,
745
+ "step": 105
746
+ },
747
+ {
748
+ "epoch": 0.9724770642201835,
749
+ "grad_norm": 8553.6689453125,
750
+ "learning_rate": 4.974747474747475e-05,
751
+ "loss": 0.1287,
752
+ "step": 106
753
+ },
754
+ {
755
+ "epoch": 0.981651376146789,
756
+ "grad_norm": 8794.58984375,
757
+ "learning_rate": 4.9696969696969694e-05,
758
+ "loss": 0.1178,
759
+ "step": 107
760
+ },
761
+ {
762
+ "epoch": 0.9908256880733946,
763
+ "grad_norm": 8048.97509765625,
764
+ "learning_rate": 4.964646464646465e-05,
765
+ "loss": 0.1112,
766
+ "step": 108
767
+ },
768
+ {
769
+ "epoch": 1.0,
770
+ "grad_norm": 24349.96484375,
771
+ "learning_rate": 4.9595959595959594e-05,
772
+ "loss": 0.112,
773
+ "step": 109
774
+ },
775
+ {
776
+ "epoch": 1.0,
777
+ "eval_f1_macro": 0.0,
778
+ "eval_f1_micro": 0.0,
779
+ "eval_loss": 0.1133684441447258,
780
+ "eval_precision": 0.0,
781
+ "eval_recall": 0.0,
782
+ "eval_runtime": 0.8578,
783
+ "eval_samples_per_second": 336.892,
784
+ "eval_steps_per_second": 15.154,
785
+ "step": 109
786
+ }
787
+ ],
788
+ "logging_steps": 1,
789
+ "max_steps": 1090,
790
+ "num_input_tokens_seen": 0,
791
+ "num_train_epochs": 10,
792
+ "save_steps": 500,
793
+ "stateful_callbacks": {
794
+ "TrainerControl": {
795
+ "args": {
796
+ "should_epoch_stop": false,
797
+ "should_evaluate": false,
798
+ "should_log": false,
799
+ "should_save": true,
800
+ "should_training_stop": false
801
+ },
802
+ "attributes": {}
803
+ }
804
+ },
805
+ "total_flos": 683067445217280.0,
806
+ "train_batch_size": 24,
807
+ "trial_name": null,
808
+ "trial_params": null
809
+ }
checkpoint-109/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edeef274e762b6821b55d94e571eef919508876a6f05e915f5b9e4cc058a5c62
3
+ size 5368
checkpoint-109/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1090/config.json ADDED
@@ -0,0 +1,130 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "T1003.001",
13
+ "1": "T1005",
14
+ "2": "T1012",
15
+ "3": "T1016",
16
+ "4": "T1021.001",
17
+ "5": "T1027",
18
+ "6": "T1033",
19
+ "7": "T1036.005",
20
+ "8": "T1041",
21
+ "9": "T1047",
22
+ "10": "T1053.005",
23
+ "11": "T1055",
24
+ "12": "T1056.001",
25
+ "13": "T1057",
26
+ "14": "T1059.003",
27
+ "15": "T1068",
28
+ "16": "T1070.004",
29
+ "17": "T1071.001",
30
+ "18": "T1072",
31
+ "19": "T1074.001",
32
+ "20": "T1078",
33
+ "21": "T1082",
34
+ "22": "T1083",
35
+ "23": "T1090",
36
+ "24": "T1095",
37
+ "25": "T1105",
38
+ "26": "T1106",
39
+ "27": "T1110",
40
+ "28": "T1112",
41
+ "29": "T1113",
42
+ "30": "T1140",
43
+ "31": "T1190",
44
+ "32": "T1204.002",
45
+ "33": "T1210",
46
+ "34": "T1218.011",
47
+ "35": "T1219",
48
+ "36": "T1484.001",
49
+ "37": "T1518.001",
50
+ "38": "T1543.003",
51
+ "39": "T1547.001",
52
+ "40": "T1548.002",
53
+ "41": "T1552.001",
54
+ "42": "T1557.001",
55
+ "43": "T1562.001",
56
+ "44": "T1564.001",
57
+ "45": "T1566.001",
58
+ "46": "T1569.002",
59
+ "47": "T1570",
60
+ "48": "T1573.001",
61
+ "49": "T1574.002"
62
+ },
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 3072,
65
+ "label2id": {
66
+ "T1003.001": 0,
67
+ "T1005": 1,
68
+ "T1012": 2,
69
+ "T1016": 3,
70
+ "T1021.001": 4,
71
+ "T1027": 5,
72
+ "T1033": 6,
73
+ "T1036.005": 7,
74
+ "T1041": 8,
75
+ "T1047": 9,
76
+ "T1053.005": 10,
77
+ "T1055": 11,
78
+ "T1056.001": 12,
79
+ "T1057": 13,
80
+ "T1059.003": 14,
81
+ "T1068": 15,
82
+ "T1070.004": 16,
83
+ "T1071.001": 17,
84
+ "T1072": 18,
85
+ "T1074.001": 19,
86
+ "T1078": 20,
87
+ "T1082": 21,
88
+ "T1083": 22,
89
+ "T1090": 23,
90
+ "T1095": 24,
91
+ "T1105": 25,
92
+ "T1106": 26,
93
+ "T1110": 27,
94
+ "T1112": 28,
95
+ "T1113": 29,
96
+ "T1140": 30,
97
+ "T1190": 31,
98
+ "T1204.002": 32,
99
+ "T1210": 33,
100
+ "T1218.011": 34,
101
+ "T1219": 35,
102
+ "T1484.001": 36,
103
+ "T1518.001": 37,
104
+ "T1543.003": 38,
105
+ "T1547.001": 39,
106
+ "T1548.002": 40,
107
+ "T1552.001": 41,
108
+ "T1557.001": 42,
109
+ "T1562.001": 43,
110
+ "T1564.001": 44,
111
+ "T1566.001": 45,
112
+ "T1569.002": 46,
113
+ "T1570": 47,
114
+ "T1573.001": 48,
115
+ "T1574.002": 49
116
+ },
117
+ "layer_norm_eps": 1e-12,
118
+ "max_position_embeddings": 512,
119
+ "model_type": "bert",
120
+ "num_attention_heads": 12,
121
+ "num_hidden_layers": 12,
122
+ "pad_token_id": 0,
123
+ "position_embedding_type": "absolute",
124
+ "problem_type": "multi_label_classification",
125
+ "torch_dtype": "float32",
126
+ "transformers_version": "4.55.2",
127
+ "type_vocab_size": 2,
128
+ "use_cache": true,
129
+ "vocab_size": 30522
130
+ }
checkpoint-1090/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d2c33cb801fb9fd37fba3d062a6f66f9b62765d1d669a6d225fc947114f4fa4
3
+ size 438106296
checkpoint-1090/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34131a9493101283ef6518e9e5b4d588cf3594580bfbc70d14a05a8f80b263ee
3
+ size 876333626
checkpoint-1090/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d167326bd1508b64d72d04e5b71a405062af79101e3c0fd31300b37fc7b8029e
3
+ size 14244
checkpoint-1090/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
3
+ size 988
checkpoint-1090/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:709c0b0268122b52be3e54422e4b5ccc4177c31d098f824bf1f5e01ea9b005da
3
+ size 1064
checkpoint-1090/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
checkpoint-1090/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1090/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
checkpoint-1090/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1090/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edeef274e762b6821b55d94e571eef919508876a6f05e915f5b9e4cc058a5c62
3
+ size 5368
checkpoint-1090/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-218/config.json ADDED
@@ -0,0 +1,130 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "T1003.001",
13
+ "1": "T1005",
14
+ "2": "T1012",
15
+ "3": "T1016",
16
+ "4": "T1021.001",
17
+ "5": "T1027",
18
+ "6": "T1033",
19
+ "7": "T1036.005",
20
+ "8": "T1041",
21
+ "9": "T1047",
22
+ "10": "T1053.005",
23
+ "11": "T1055",
24
+ "12": "T1056.001",
25
+ "13": "T1057",
26
+ "14": "T1059.003",
27
+ "15": "T1068",
28
+ "16": "T1070.004",
29
+ "17": "T1071.001",
30
+ "18": "T1072",
31
+ "19": "T1074.001",
32
+ "20": "T1078",
33
+ "21": "T1082",
34
+ "22": "T1083",
35
+ "23": "T1090",
36
+ "24": "T1095",
37
+ "25": "T1105",
38
+ "26": "T1106",
39
+ "27": "T1110",
40
+ "28": "T1112",
41
+ "29": "T1113",
42
+ "30": "T1140",
43
+ "31": "T1190",
44
+ "32": "T1204.002",
45
+ "33": "T1210",
46
+ "34": "T1218.011",
47
+ "35": "T1219",
48
+ "36": "T1484.001",
49
+ "37": "T1518.001",
50
+ "38": "T1543.003",
51
+ "39": "T1547.001",
52
+ "40": "T1548.002",
53
+ "41": "T1552.001",
54
+ "42": "T1557.001",
55
+ "43": "T1562.001",
56
+ "44": "T1564.001",
57
+ "45": "T1566.001",
58
+ "46": "T1569.002",
59
+ "47": "T1570",
60
+ "48": "T1573.001",
61
+ "49": "T1574.002"
62
+ },
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 3072,
65
+ "label2id": {
66
+ "T1003.001": 0,
67
+ "T1005": 1,
68
+ "T1012": 2,
69
+ "T1016": 3,
70
+ "T1021.001": 4,
71
+ "T1027": 5,
72
+ "T1033": 6,
73
+ "T1036.005": 7,
74
+ "T1041": 8,
75
+ "T1047": 9,
76
+ "T1053.005": 10,
77
+ "T1055": 11,
78
+ "T1056.001": 12,
79
+ "T1057": 13,
80
+ "T1059.003": 14,
81
+ "T1068": 15,
82
+ "T1070.004": 16,
83
+ "T1071.001": 17,
84
+ "T1072": 18,
85
+ "T1074.001": 19,
86
+ "T1078": 20,
87
+ "T1082": 21,
88
+ "T1083": 22,
89
+ "T1090": 23,
90
+ "T1095": 24,
91
+ "T1105": 25,
92
+ "T1106": 26,
93
+ "T1110": 27,
94
+ "T1112": 28,
95
+ "T1113": 29,
96
+ "T1140": 30,
97
+ "T1190": 31,
98
+ "T1204.002": 32,
99
+ "T1210": 33,
100
+ "T1218.011": 34,
101
+ "T1219": 35,
102
+ "T1484.001": 36,
103
+ "T1518.001": 37,
104
+ "T1543.003": 38,
105
+ "T1547.001": 39,
106
+ "T1548.002": 40,
107
+ "T1552.001": 41,
108
+ "T1557.001": 42,
109
+ "T1562.001": 43,
110
+ "T1564.001": 44,
111
+ "T1566.001": 45,
112
+ "T1569.002": 46,
113
+ "T1570": 47,
114
+ "T1573.001": 48,
115
+ "T1574.002": 49
116
+ },
117
+ "layer_norm_eps": 1e-12,
118
+ "max_position_embeddings": 512,
119
+ "model_type": "bert",
120
+ "num_attention_heads": 12,
121
+ "num_hidden_layers": 12,
122
+ "pad_token_id": 0,
123
+ "position_embedding_type": "absolute",
124
+ "problem_type": "multi_label_classification",
125
+ "torch_dtype": "float32",
126
+ "transformers_version": "4.55.2",
127
+ "type_vocab_size": 2,
128
+ "use_cache": true,
129
+ "vocab_size": 30522
130
+ }
checkpoint-218/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04ee5116c279c642a5b0cf6d4474b163866460c3b3086b4973efc1b2970c8c4c
3
+ size 438106296
checkpoint-218/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa8280d3181bec4662e744780aa272fda5cd72af0d15968bc5b2f4cce4e6be7b
3
+ size 876333626
checkpoint-218/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5f04e1d154f31355e187df9c72a9349dbee0e52ad770e42da69a57dafd778e3
3
+ size 14244
checkpoint-218/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
3
+ size 988
checkpoint-218/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:57db727835c163657fe7ba938eaa3c6c5fe3057c67dc8b36f361bb5da91efb88
3
+ size 1064
checkpoint-218/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
checkpoint-218/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-218/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
checkpoint-218/trainer_state.json ADDED
@@ -0,0 +1,1584 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 109,
3
+ "best_metric": 0.0,
4
+ "best_model_checkpoint": "./cysecbert-ttp-bert-base_data/checkpoint-109",
5
+ "epoch": 2.0,
6
+ "eval_steps": 500,
7
+ "global_step": 218,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.009174311926605505,
14
+ "grad_norm": 46530.15234375,
15
+ "learning_rate": 0.0,
16
+ "loss": 0.7023,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.01834862385321101,
21
+ "grad_norm": 49571.35546875,
22
+ "learning_rate": 5.000000000000001e-07,
23
+ "loss": 0.7006,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.027522935779816515,
28
+ "grad_norm": 45885.55078125,
29
+ "learning_rate": 1.0000000000000002e-06,
30
+ "loss": 0.7033,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.03669724770642202,
35
+ "grad_norm": 43333.73046875,
36
+ "learning_rate": 1.5e-06,
37
+ "loss": 0.7014,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.045871559633027525,
42
+ "grad_norm": 44340.41796875,
43
+ "learning_rate": 2.0000000000000003e-06,
44
+ "loss": 0.6992,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.05504587155963303,
49
+ "grad_norm": 48282.54296875,
50
+ "learning_rate": 2.5e-06,
51
+ "loss": 0.6924,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.06422018348623854,
56
+ "grad_norm": 48561.56640625,
57
+ "learning_rate": 3e-06,
58
+ "loss": 0.6948,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.07339449541284404,
63
+ "grad_norm": 43580.078125,
64
+ "learning_rate": 3.5000000000000004e-06,
65
+ "loss": 0.6916,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.08256880733944955,
70
+ "grad_norm": 46552.28515625,
71
+ "learning_rate": 4.000000000000001e-06,
72
+ "loss": 0.6854,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.09174311926605505,
77
+ "grad_norm": 49042.15234375,
78
+ "learning_rate": 4.5e-06,
79
+ "loss": 0.6822,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.10091743119266056,
84
+ "grad_norm": 97274.8515625,
85
+ "learning_rate": 5e-06,
86
+ "loss": 0.6788,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.11009174311926606,
91
+ "grad_norm": 55310.375,
92
+ "learning_rate": 5.500000000000001e-06,
93
+ "loss": 0.6781,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.11926605504587157,
98
+ "grad_norm": 70317.6328125,
99
+ "learning_rate": 6e-06,
100
+ "loss": 0.6754,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.12844036697247707,
105
+ "grad_norm": 46825.9765625,
106
+ "learning_rate": 6.5000000000000004e-06,
107
+ "loss": 0.6626,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.13761467889908258,
112
+ "grad_norm": 50745.15625,
113
+ "learning_rate": 7.000000000000001e-06,
114
+ "loss": 0.655,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 0.14678899082568808,
119
+ "grad_norm": 51342.453125,
120
+ "learning_rate": 7.5e-06,
121
+ "loss": 0.6491,
122
+ "step": 16
123
+ },
124
+ {
125
+ "epoch": 0.1559633027522936,
126
+ "grad_norm": 51815.421875,
127
+ "learning_rate": 8.000000000000001e-06,
128
+ "loss": 0.6464,
129
+ "step": 17
130
+ },
131
+ {
132
+ "epoch": 0.1651376146788991,
133
+ "grad_norm": 54610.65234375,
134
+ "learning_rate": 8.500000000000002e-06,
135
+ "loss": 0.6298,
136
+ "step": 18
137
+ },
138
+ {
139
+ "epoch": 0.1743119266055046,
140
+ "grad_norm": 61936.31640625,
141
+ "learning_rate": 9e-06,
142
+ "loss": 0.623,
143
+ "step": 19
144
+ },
145
+ {
146
+ "epoch": 0.1834862385321101,
147
+ "grad_norm": 56021.69921875,
148
+ "learning_rate": 9.5e-06,
149
+ "loss": 0.6156,
150
+ "step": 20
151
+ },
152
+ {
153
+ "epoch": 0.1926605504587156,
154
+ "grad_norm": 60806.10546875,
155
+ "learning_rate": 1e-05,
156
+ "loss": 0.5981,
157
+ "step": 21
158
+ },
159
+ {
160
+ "epoch": 0.2018348623853211,
161
+ "grad_norm": 58582.4140625,
162
+ "learning_rate": 1.05e-05,
163
+ "loss": 0.5869,
164
+ "step": 22
165
+ },
166
+ {
167
+ "epoch": 0.21100917431192662,
168
+ "grad_norm": 59099.16015625,
169
+ "learning_rate": 1.1000000000000001e-05,
170
+ "loss": 0.5732,
171
+ "step": 23
172
+ },
173
+ {
174
+ "epoch": 0.22018348623853212,
175
+ "grad_norm": 52971.81640625,
176
+ "learning_rate": 1.1500000000000002e-05,
177
+ "loss": 0.5587,
178
+ "step": 24
179
+ },
180
+ {
181
+ "epoch": 0.22935779816513763,
182
+ "grad_norm": 51724.43359375,
183
+ "learning_rate": 1.2e-05,
184
+ "loss": 0.5477,
185
+ "step": 25
186
+ },
187
+ {
188
+ "epoch": 0.23853211009174313,
189
+ "grad_norm": 58133.453125,
190
+ "learning_rate": 1.25e-05,
191
+ "loss": 0.5422,
192
+ "step": 26
193
+ },
194
+ {
195
+ "epoch": 0.24770642201834864,
196
+ "grad_norm": 54134.3359375,
197
+ "learning_rate": 1.3000000000000001e-05,
198
+ "loss": 0.5319,
199
+ "step": 27
200
+ },
201
+ {
202
+ "epoch": 0.25688073394495414,
203
+ "grad_norm": 78195.7890625,
204
+ "learning_rate": 1.3500000000000001e-05,
205
+ "loss": 0.5259,
206
+ "step": 28
207
+ },
208
+ {
209
+ "epoch": 0.26605504587155965,
210
+ "grad_norm": 47713.20703125,
211
+ "learning_rate": 1.4000000000000001e-05,
212
+ "loss": 0.5105,
213
+ "step": 29
214
+ },
215
+ {
216
+ "epoch": 0.27522935779816515,
217
+ "grad_norm": 45838.0703125,
218
+ "learning_rate": 1.45e-05,
219
+ "loss": 0.499,
220
+ "step": 30
221
+ },
222
+ {
223
+ "epoch": 0.28440366972477066,
224
+ "grad_norm": 46735.58203125,
225
+ "learning_rate": 1.5e-05,
226
+ "loss": 0.4914,
227
+ "step": 31
228
+ },
229
+ {
230
+ "epoch": 0.29357798165137616,
231
+ "grad_norm": 45432.8828125,
232
+ "learning_rate": 1.55e-05,
233
+ "loss": 0.4783,
234
+ "step": 32
235
+ },
236
+ {
237
+ "epoch": 0.30275229357798167,
238
+ "grad_norm": 46758.58984375,
239
+ "learning_rate": 1.6000000000000003e-05,
240
+ "loss": 0.4758,
241
+ "step": 33
242
+ },
243
+ {
244
+ "epoch": 0.3119266055045872,
245
+ "grad_norm": 45245.82421875,
246
+ "learning_rate": 1.65e-05,
247
+ "loss": 0.4608,
248
+ "step": 34
249
+ },
250
+ {
251
+ "epoch": 0.3211009174311927,
252
+ "grad_norm": 43336.75390625,
253
+ "learning_rate": 1.7000000000000003e-05,
254
+ "loss": 0.4567,
255
+ "step": 35
256
+ },
257
+ {
258
+ "epoch": 0.3302752293577982,
259
+ "grad_norm": 41445.90625,
260
+ "learning_rate": 1.75e-05,
261
+ "loss": 0.4465,
262
+ "step": 36
263
+ },
264
+ {
265
+ "epoch": 0.3394495412844037,
266
+ "grad_norm": 42278.80078125,
267
+ "learning_rate": 1.8e-05,
268
+ "loss": 0.4395,
269
+ "step": 37
270
+ },
271
+ {
272
+ "epoch": 0.3486238532110092,
273
+ "grad_norm": 42039.28515625,
274
+ "learning_rate": 1.85e-05,
275
+ "loss": 0.4289,
276
+ "step": 38
277
+ },
278
+ {
279
+ "epoch": 0.3577981651376147,
280
+ "grad_norm": 41411.61328125,
281
+ "learning_rate": 1.9e-05,
282
+ "loss": 0.423,
283
+ "step": 39
284
+ },
285
+ {
286
+ "epoch": 0.3669724770642202,
287
+ "grad_norm": 40133.84375,
288
+ "learning_rate": 1.9500000000000003e-05,
289
+ "loss": 0.4191,
290
+ "step": 40
291
+ },
292
+ {
293
+ "epoch": 0.3761467889908257,
294
+ "grad_norm": 39758.8828125,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.4071,
297
+ "step": 41
298
+ },
299
+ {
300
+ "epoch": 0.3853211009174312,
301
+ "grad_norm": 38760.36328125,
302
+ "learning_rate": 2.05e-05,
303
+ "loss": 0.3996,
304
+ "step": 42
305
+ },
306
+ {
307
+ "epoch": 0.3944954128440367,
308
+ "grad_norm": 38552.80078125,
309
+ "learning_rate": 2.1e-05,
310
+ "loss": 0.3918,
311
+ "step": 43
312
+ },
313
+ {
314
+ "epoch": 0.4036697247706422,
315
+ "grad_norm": 38561.7578125,
316
+ "learning_rate": 2.15e-05,
317
+ "loss": 0.3865,
318
+ "step": 44
319
+ },
320
+ {
321
+ "epoch": 0.41284403669724773,
322
+ "grad_norm": 37616.85546875,
323
+ "learning_rate": 2.2000000000000003e-05,
324
+ "loss": 0.3794,
325
+ "step": 45
326
+ },
327
+ {
328
+ "epoch": 0.42201834862385323,
329
+ "grad_norm": 37784.78515625,
330
+ "learning_rate": 2.25e-05,
331
+ "loss": 0.3745,
332
+ "step": 46
333
+ },
334
+ {
335
+ "epoch": 0.43119266055045874,
336
+ "grad_norm": 36332.9140625,
337
+ "learning_rate": 2.3000000000000003e-05,
338
+ "loss": 0.3711,
339
+ "step": 47
340
+ },
341
+ {
342
+ "epoch": 0.44036697247706424,
343
+ "grad_norm": 38899.73046875,
344
+ "learning_rate": 2.35e-05,
345
+ "loss": 0.3643,
346
+ "step": 48
347
+ },
348
+ {
349
+ "epoch": 0.44954128440366975,
350
+ "grad_norm": 36174.9765625,
351
+ "learning_rate": 2.4e-05,
352
+ "loss": 0.3575,
353
+ "step": 49
354
+ },
355
+ {
356
+ "epoch": 0.45871559633027525,
357
+ "grad_norm": 35699.83203125,
358
+ "learning_rate": 2.45e-05,
359
+ "loss": 0.347,
360
+ "step": 50
361
+ },
362
+ {
363
+ "epoch": 0.46788990825688076,
364
+ "grad_norm": 35562.3671875,
365
+ "learning_rate": 2.5e-05,
366
+ "loss": 0.3438,
367
+ "step": 51
368
+ },
369
+ {
370
+ "epoch": 0.47706422018348627,
371
+ "grad_norm": 35428.0625,
372
+ "learning_rate": 2.5500000000000003e-05,
373
+ "loss": 0.3325,
374
+ "step": 52
375
+ },
376
+ {
377
+ "epoch": 0.48623853211009177,
378
+ "grad_norm": 34396.80078125,
379
+ "learning_rate": 2.6000000000000002e-05,
380
+ "loss": 0.3302,
381
+ "step": 53
382
+ },
383
+ {
384
+ "epoch": 0.4954128440366973,
385
+ "grad_norm": 38376.41796875,
386
+ "learning_rate": 2.6500000000000004e-05,
387
+ "loss": 0.3218,
388
+ "step": 54
389
+ },
390
+ {
391
+ "epoch": 0.5045871559633027,
392
+ "grad_norm": 33996.23046875,
393
+ "learning_rate": 2.7000000000000002e-05,
394
+ "loss": 0.3131,
395
+ "step": 55
396
+ },
397
+ {
398
+ "epoch": 0.5137614678899083,
399
+ "grad_norm": 33070.55859375,
400
+ "learning_rate": 2.7500000000000004e-05,
401
+ "loss": 0.3129,
402
+ "step": 56
403
+ },
404
+ {
405
+ "epoch": 0.5229357798165137,
406
+ "grad_norm": 33200.44140625,
407
+ "learning_rate": 2.8000000000000003e-05,
408
+ "loss": 0.3003,
409
+ "step": 57
410
+ },
411
+ {
412
+ "epoch": 0.5321100917431193,
413
+ "grad_norm": 33099.1015625,
414
+ "learning_rate": 2.8499999999999998e-05,
415
+ "loss": 0.2904,
416
+ "step": 58
417
+ },
418
+ {
419
+ "epoch": 0.5412844036697247,
420
+ "grad_norm": 31352.298828125,
421
+ "learning_rate": 2.9e-05,
422
+ "loss": 0.2911,
423
+ "step": 59
424
+ },
425
+ {
426
+ "epoch": 0.5504587155963303,
427
+ "grad_norm": 31965.087890625,
428
+ "learning_rate": 2.95e-05,
429
+ "loss": 0.2815,
430
+ "step": 60
431
+ },
432
+ {
433
+ "epoch": 0.5596330275229358,
434
+ "grad_norm": 30810.849609375,
435
+ "learning_rate": 3e-05,
436
+ "loss": 0.2818,
437
+ "step": 61
438
+ },
439
+ {
440
+ "epoch": 0.5688073394495413,
441
+ "grad_norm": 30034.779296875,
442
+ "learning_rate": 3.05e-05,
443
+ "loss": 0.2731,
444
+ "step": 62
445
+ },
446
+ {
447
+ "epoch": 0.5779816513761468,
448
+ "grad_norm": 30009.12109375,
449
+ "learning_rate": 3.1e-05,
450
+ "loss": 0.266,
451
+ "step": 63
452
+ },
453
+ {
454
+ "epoch": 0.5871559633027523,
455
+ "grad_norm": 29305.173828125,
456
+ "learning_rate": 3.15e-05,
457
+ "loss": 0.2609,
458
+ "step": 64
459
+ },
460
+ {
461
+ "epoch": 0.5963302752293578,
462
+ "grad_norm": 29081.853515625,
463
+ "learning_rate": 3.2000000000000005e-05,
464
+ "loss": 0.2543,
465
+ "step": 65
466
+ },
467
+ {
468
+ "epoch": 0.6055045871559633,
469
+ "grad_norm": 28217.021484375,
470
+ "learning_rate": 3.2500000000000004e-05,
471
+ "loss": 0.252,
472
+ "step": 66
473
+ },
474
+ {
475
+ "epoch": 0.6146788990825688,
476
+ "grad_norm": 26448.12890625,
477
+ "learning_rate": 3.3e-05,
478
+ "loss": 0.2565,
479
+ "step": 67
480
+ },
481
+ {
482
+ "epoch": 0.6238532110091743,
483
+ "grad_norm": 27198.80859375,
484
+ "learning_rate": 3.35e-05,
485
+ "loss": 0.2342,
486
+ "step": 68
487
+ },
488
+ {
489
+ "epoch": 0.6330275229357798,
490
+ "grad_norm": 26946.30859375,
491
+ "learning_rate": 3.4000000000000007e-05,
492
+ "loss": 0.2273,
493
+ "step": 69
494
+ },
495
+ {
496
+ "epoch": 0.6422018348623854,
497
+ "grad_norm": 26236.7265625,
498
+ "learning_rate": 3.45e-05,
499
+ "loss": 0.2219,
500
+ "step": 70
501
+ },
502
+ {
503
+ "epoch": 0.6513761467889908,
504
+ "grad_norm": 25535.818359375,
505
+ "learning_rate": 3.5e-05,
506
+ "loss": 0.2197,
507
+ "step": 71
508
+ },
509
+ {
510
+ "epoch": 0.6605504587155964,
511
+ "grad_norm": 24871.8515625,
512
+ "learning_rate": 3.55e-05,
513
+ "loss": 0.2144,
514
+ "step": 72
515
+ },
516
+ {
517
+ "epoch": 0.6697247706422018,
518
+ "grad_norm": 24484.55078125,
519
+ "learning_rate": 3.6e-05,
520
+ "loss": 0.2066,
521
+ "step": 73
522
+ },
523
+ {
524
+ "epoch": 0.6788990825688074,
525
+ "grad_norm": 23655.677734375,
526
+ "learning_rate": 3.65e-05,
527
+ "loss": 0.206,
528
+ "step": 74
529
+ },
530
+ {
531
+ "epoch": 0.6880733944954128,
532
+ "grad_norm": 23129.076171875,
533
+ "learning_rate": 3.7e-05,
534
+ "loss": 0.1979,
535
+ "step": 75
536
+ },
537
+ {
538
+ "epoch": 0.6972477064220184,
539
+ "grad_norm": 21871.90625,
540
+ "learning_rate": 3.7500000000000003e-05,
541
+ "loss": 0.1973,
542
+ "step": 76
543
+ },
544
+ {
545
+ "epoch": 0.7064220183486238,
546
+ "grad_norm": 21561.58984375,
547
+ "learning_rate": 3.8e-05,
548
+ "loss": 0.1935,
549
+ "step": 77
550
+ },
551
+ {
552
+ "epoch": 0.7155963302752294,
553
+ "grad_norm": 22210.29296875,
554
+ "learning_rate": 3.85e-05,
555
+ "loss": 0.1848,
556
+ "step": 78
557
+ },
558
+ {
559
+ "epoch": 0.7247706422018348,
560
+ "grad_norm": 20555.775390625,
561
+ "learning_rate": 3.9000000000000006e-05,
562
+ "loss": 0.1801,
563
+ "step": 79
564
+ },
565
+ {
566
+ "epoch": 0.7339449541284404,
567
+ "grad_norm": 20406.75,
568
+ "learning_rate": 3.9500000000000005e-05,
569
+ "loss": 0.1746,
570
+ "step": 80
571
+ },
572
+ {
573
+ "epoch": 0.7431192660550459,
574
+ "grad_norm": 19150.931640625,
575
+ "learning_rate": 4e-05,
576
+ "loss": 0.1688,
577
+ "step": 81
578
+ },
579
+ {
580
+ "epoch": 0.7522935779816514,
581
+ "grad_norm": 18422.65625,
582
+ "learning_rate": 4.05e-05,
583
+ "loss": 0.1661,
584
+ "step": 82
585
+ },
586
+ {
587
+ "epoch": 0.7614678899082569,
588
+ "grad_norm": 16746.20703125,
589
+ "learning_rate": 4.1e-05,
590
+ "loss": 0.1711,
591
+ "step": 83
592
+ },
593
+ {
594
+ "epoch": 0.7706422018348624,
595
+ "grad_norm": 18229.41015625,
596
+ "learning_rate": 4.15e-05,
597
+ "loss": 0.1646,
598
+ "step": 84
599
+ },
600
+ {
601
+ "epoch": 0.7798165137614679,
602
+ "grad_norm": 17156.267578125,
603
+ "learning_rate": 4.2e-05,
604
+ "loss": 0.1588,
605
+ "step": 85
606
+ },
607
+ {
608
+ "epoch": 0.7889908256880734,
609
+ "grad_norm": 15970.611328125,
610
+ "learning_rate": 4.25e-05,
611
+ "loss": 0.1547,
612
+ "step": 86
613
+ },
614
+ {
615
+ "epoch": 0.7981651376146789,
616
+ "grad_norm": 16997.103515625,
617
+ "learning_rate": 4.3e-05,
618
+ "loss": 0.1493,
619
+ "step": 87
620
+ },
621
+ {
622
+ "epoch": 0.8073394495412844,
623
+ "grad_norm": 15529.58984375,
624
+ "learning_rate": 4.35e-05,
625
+ "loss": 0.1439,
626
+ "step": 88
627
+ },
628
+ {
629
+ "epoch": 0.8165137614678899,
630
+ "grad_norm": 15099.9052734375,
631
+ "learning_rate": 4.4000000000000006e-05,
632
+ "loss": 0.1432,
633
+ "step": 89
634
+ },
635
+ {
636
+ "epoch": 0.8256880733944955,
637
+ "grad_norm": 14261.3232421875,
638
+ "learning_rate": 4.4500000000000004e-05,
639
+ "loss": 0.1396,
640
+ "step": 90
641
+ },
642
+ {
643
+ "epoch": 0.8348623853211009,
644
+ "grad_norm": 13915.87890625,
645
+ "learning_rate": 4.5e-05,
646
+ "loss": 0.1317,
647
+ "step": 91
648
+ },
649
+ {
650
+ "epoch": 0.8440366972477065,
651
+ "grad_norm": 13518.7568359375,
652
+ "learning_rate": 4.55e-05,
653
+ "loss": 0.1289,
654
+ "step": 92
655
+ },
656
+ {
657
+ "epoch": 0.8532110091743119,
658
+ "grad_norm": 12170.1865234375,
659
+ "learning_rate": 4.600000000000001e-05,
660
+ "loss": 0.1371,
661
+ "step": 93
662
+ },
663
+ {
664
+ "epoch": 0.8623853211009175,
665
+ "grad_norm": 11745.9775390625,
666
+ "learning_rate": 4.6500000000000005e-05,
667
+ "loss": 0.1434,
668
+ "step": 94
669
+ },
670
+ {
671
+ "epoch": 0.8715596330275229,
672
+ "grad_norm": 12265.5556640625,
673
+ "learning_rate": 4.7e-05,
674
+ "loss": 0.1343,
675
+ "step": 95
676
+ },
677
+ {
678
+ "epoch": 0.8807339449541285,
679
+ "grad_norm": 11386.3291015625,
680
+ "learning_rate": 4.75e-05,
681
+ "loss": 0.1317,
682
+ "step": 96
683
+ },
684
+ {
685
+ "epoch": 0.8899082568807339,
686
+ "grad_norm": 10869.4599609375,
687
+ "learning_rate": 4.8e-05,
688
+ "loss": 0.1295,
689
+ "step": 97
690
+ },
691
+ {
692
+ "epoch": 0.8990825688073395,
693
+ "grad_norm": 10752.25,
694
+ "learning_rate": 4.85e-05,
695
+ "loss": 0.124,
696
+ "step": 98
697
+ },
698
+ {
699
+ "epoch": 0.908256880733945,
700
+ "grad_norm": 9687.6376953125,
701
+ "learning_rate": 4.9e-05,
702
+ "loss": 0.1271,
703
+ "step": 99
704
+ },
705
+ {
706
+ "epoch": 0.9174311926605505,
707
+ "grad_norm": 9697.15234375,
708
+ "learning_rate": 4.9500000000000004e-05,
709
+ "loss": 0.1194,
710
+ "step": 100
711
+ },
712
+ {
713
+ "epoch": 0.926605504587156,
714
+ "grad_norm": 9423.21875,
715
+ "learning_rate": 5e-05,
716
+ "loss": 0.1176,
717
+ "step": 101
718
+ },
719
+ {
720
+ "epoch": 0.9357798165137615,
721
+ "grad_norm": 10253.8564453125,
722
+ "learning_rate": 4.994949494949495e-05,
723
+ "loss": 0.121,
724
+ "step": 102
725
+ },
726
+ {
727
+ "epoch": 0.944954128440367,
728
+ "grad_norm": 9756.7646484375,
729
+ "learning_rate": 4.98989898989899e-05,
730
+ "loss": 0.1173,
731
+ "step": 103
732
+ },
733
+ {
734
+ "epoch": 0.9541284403669725,
735
+ "grad_norm": 8482.177734375,
736
+ "learning_rate": 4.984848484848485e-05,
737
+ "loss": 0.1233,
738
+ "step": 104
739
+ },
740
+ {
741
+ "epoch": 0.963302752293578,
742
+ "grad_norm": 7755.4892578125,
743
+ "learning_rate": 4.97979797979798e-05,
744
+ "loss": 0.1251,
745
+ "step": 105
746
+ },
747
+ {
748
+ "epoch": 0.9724770642201835,
749
+ "grad_norm": 8553.6689453125,
750
+ "learning_rate": 4.974747474747475e-05,
751
+ "loss": 0.1287,
752
+ "step": 106
753
+ },
754
+ {
755
+ "epoch": 0.981651376146789,
756
+ "grad_norm": 8794.58984375,
757
+ "learning_rate": 4.9696969696969694e-05,
758
+ "loss": 0.1178,
759
+ "step": 107
760
+ },
761
+ {
762
+ "epoch": 0.9908256880733946,
763
+ "grad_norm": 8048.97509765625,
764
+ "learning_rate": 4.964646464646465e-05,
765
+ "loss": 0.1112,
766
+ "step": 108
767
+ },
768
+ {
769
+ "epoch": 1.0,
770
+ "grad_norm": 24349.96484375,
771
+ "learning_rate": 4.9595959595959594e-05,
772
+ "loss": 0.112,
773
+ "step": 109
774
+ },
775
+ {
776
+ "epoch": 1.0,
777
+ "eval_f1_macro": 0.0,
778
+ "eval_f1_micro": 0.0,
779
+ "eval_loss": 0.1133684441447258,
780
+ "eval_precision": 0.0,
781
+ "eval_recall": 0.0,
782
+ "eval_runtime": 0.8578,
783
+ "eval_samples_per_second": 336.892,
784
+ "eval_steps_per_second": 15.154,
785
+ "step": 109
786
+ },
787
+ {
788
+ "epoch": 1.0091743119266054,
789
+ "grad_norm": 7603.02587890625,
790
+ "learning_rate": 4.9545454545454553e-05,
791
+ "loss": 0.1133,
792
+ "step": 110
793
+ },
794
+ {
795
+ "epoch": 1.018348623853211,
796
+ "grad_norm": 8602.8974609375,
797
+ "learning_rate": 4.94949494949495e-05,
798
+ "loss": 0.1192,
799
+ "step": 111
800
+ },
801
+ {
802
+ "epoch": 1.0275229357798166,
803
+ "grad_norm": 8388.9609375,
804
+ "learning_rate": 4.9444444444444446e-05,
805
+ "loss": 0.1246,
806
+ "step": 112
807
+ },
808
+ {
809
+ "epoch": 1.036697247706422,
810
+ "grad_norm": 9368.150390625,
811
+ "learning_rate": 4.93939393939394e-05,
812
+ "loss": 0.1089,
813
+ "step": 113
814
+ },
815
+ {
816
+ "epoch": 1.0458715596330275,
817
+ "grad_norm": 7053.083984375,
818
+ "learning_rate": 4.9343434343434346e-05,
819
+ "loss": 0.1042,
820
+ "step": 114
821
+ },
822
+ {
823
+ "epoch": 1.0550458715596331,
824
+ "grad_norm": 9539.36328125,
825
+ "learning_rate": 4.92929292929293e-05,
826
+ "loss": 0.1049,
827
+ "step": 115
828
+ },
829
+ {
830
+ "epoch": 1.0642201834862386,
831
+ "grad_norm": 6956.2763671875,
832
+ "learning_rate": 4.9242424242424245e-05,
833
+ "loss": 0.109,
834
+ "step": 116
835
+ },
836
+ {
837
+ "epoch": 1.073394495412844,
838
+ "grad_norm": 6597.9580078125,
839
+ "learning_rate": 4.919191919191919e-05,
840
+ "loss": 0.1124,
841
+ "step": 117
842
+ },
843
+ {
844
+ "epoch": 1.0825688073394495,
845
+ "grad_norm": 8124.76123046875,
846
+ "learning_rate": 4.9141414141414145e-05,
847
+ "loss": 0.1226,
848
+ "step": 118
849
+ },
850
+ {
851
+ "epoch": 1.091743119266055,
852
+ "grad_norm": 7030.619140625,
853
+ "learning_rate": 4.909090909090909e-05,
854
+ "loss": 0.0989,
855
+ "step": 119
856
+ },
857
+ {
858
+ "epoch": 1.1009174311926606,
859
+ "grad_norm": 9377.66796875,
860
+ "learning_rate": 4.9040404040404044e-05,
861
+ "loss": 0.1038,
862
+ "step": 120
863
+ },
864
+ {
865
+ "epoch": 1.110091743119266,
866
+ "grad_norm": 9298.802734375,
867
+ "learning_rate": 4.898989898989899e-05,
868
+ "loss": 0.1216,
869
+ "step": 121
870
+ },
871
+ {
872
+ "epoch": 1.1192660550458715,
873
+ "grad_norm": 7350.28369140625,
874
+ "learning_rate": 4.8939393939393944e-05,
875
+ "loss": 0.0915,
876
+ "step": 122
877
+ },
878
+ {
879
+ "epoch": 1.1284403669724772,
880
+ "grad_norm": 8066.943359375,
881
+ "learning_rate": 4.888888888888889e-05,
882
+ "loss": 0.1291,
883
+ "step": 123
884
+ },
885
+ {
886
+ "epoch": 1.1376146788990826,
887
+ "grad_norm": 8210.3095703125,
888
+ "learning_rate": 4.8838383838383836e-05,
889
+ "loss": 0.1196,
890
+ "step": 124
891
+ },
892
+ {
893
+ "epoch": 1.146788990825688,
894
+ "grad_norm": 7568.0234375,
895
+ "learning_rate": 4.878787878787879e-05,
896
+ "loss": 0.1037,
897
+ "step": 125
898
+ },
899
+ {
900
+ "epoch": 1.1559633027522935,
901
+ "grad_norm": 6394.8896484375,
902
+ "learning_rate": 4.8737373737373736e-05,
903
+ "loss": 0.0961,
904
+ "step": 126
905
+ },
906
+ {
907
+ "epoch": 1.165137614678899,
908
+ "grad_norm": 7549.9951171875,
909
+ "learning_rate": 4.868686868686869e-05,
910
+ "loss": 0.1075,
911
+ "step": 127
912
+ },
913
+ {
914
+ "epoch": 1.1743119266055047,
915
+ "grad_norm": 7983.36865234375,
916
+ "learning_rate": 4.863636363636364e-05,
917
+ "loss": 0.1055,
918
+ "step": 128
919
+ },
920
+ {
921
+ "epoch": 1.18348623853211,
922
+ "grad_norm": 7613.1455078125,
923
+ "learning_rate": 4.858585858585859e-05,
924
+ "loss": 0.1097,
925
+ "step": 129
926
+ },
927
+ {
928
+ "epoch": 1.1926605504587156,
929
+ "grad_norm": 7754.9228515625,
930
+ "learning_rate": 4.853535353535354e-05,
931
+ "loss": 0.1157,
932
+ "step": 130
933
+ },
934
+ {
935
+ "epoch": 1.2018348623853212,
936
+ "grad_norm": 8360.388671875,
937
+ "learning_rate": 4.848484848484849e-05,
938
+ "loss": 0.1019,
939
+ "step": 131
940
+ },
941
+ {
942
+ "epoch": 1.2110091743119267,
943
+ "grad_norm": 8300.9169921875,
944
+ "learning_rate": 4.843434343434344e-05,
945
+ "loss": 0.1098,
946
+ "step": 132
947
+ },
948
+ {
949
+ "epoch": 1.2201834862385321,
950
+ "grad_norm": 7554.8017578125,
951
+ "learning_rate": 4.838383838383839e-05,
952
+ "loss": 0.1075,
953
+ "step": 133
954
+ },
955
+ {
956
+ "epoch": 1.2293577981651376,
957
+ "grad_norm": 6510.427734375,
958
+ "learning_rate": 4.8333333333333334e-05,
959
+ "loss": 0.1041,
960
+ "step": 134
961
+ },
962
+ {
963
+ "epoch": 1.238532110091743,
964
+ "grad_norm": 5649.77978515625,
965
+ "learning_rate": 4.828282828282829e-05,
966
+ "loss": 0.1012,
967
+ "step": 135
968
+ },
969
+ {
970
+ "epoch": 1.2477064220183487,
971
+ "grad_norm": 7662.76513671875,
972
+ "learning_rate": 4.823232323232323e-05,
973
+ "loss": 0.1357,
974
+ "step": 136
975
+ },
976
+ {
977
+ "epoch": 1.2568807339449541,
978
+ "grad_norm": 7261.087890625,
979
+ "learning_rate": 4.8181818181818186e-05,
980
+ "loss": 0.1149,
981
+ "step": 137
982
+ },
983
+ {
984
+ "epoch": 1.2660550458715596,
985
+ "grad_norm": 6818.208984375,
986
+ "learning_rate": 4.813131313131313e-05,
987
+ "loss": 0.1067,
988
+ "step": 138
989
+ },
990
+ {
991
+ "epoch": 1.2752293577981653,
992
+ "grad_norm": 8839.0810546875,
993
+ "learning_rate": 4.808080808080808e-05,
994
+ "loss": 0.089,
995
+ "step": 139
996
+ },
997
+ {
998
+ "epoch": 1.2844036697247707,
999
+ "grad_norm": 8774.0615234375,
1000
+ "learning_rate": 4.803030303030303e-05,
1001
+ "loss": 0.1013,
1002
+ "step": 140
1003
+ },
1004
+ {
1005
+ "epoch": 1.2935779816513762,
1006
+ "grad_norm": 7869.935546875,
1007
+ "learning_rate": 4.797979797979798e-05,
1008
+ "loss": 0.1187,
1009
+ "step": 141
1010
+ },
1011
+ {
1012
+ "epoch": 1.3027522935779816,
1013
+ "grad_norm": 9808.5986328125,
1014
+ "learning_rate": 4.792929292929293e-05,
1015
+ "loss": 0.13,
1016
+ "step": 142
1017
+ },
1018
+ {
1019
+ "epoch": 1.311926605504587,
1020
+ "grad_norm": 7122.400390625,
1021
+ "learning_rate": 4.787878787878788e-05,
1022
+ "loss": 0.1089,
1023
+ "step": 143
1024
+ },
1025
+ {
1026
+ "epoch": 1.3211009174311927,
1027
+ "grad_norm": 6575.39892578125,
1028
+ "learning_rate": 4.782828282828283e-05,
1029
+ "loss": 0.1011,
1030
+ "step": 144
1031
+ },
1032
+ {
1033
+ "epoch": 1.3302752293577982,
1034
+ "grad_norm": 8656.4345703125,
1035
+ "learning_rate": 4.7777777777777784e-05,
1036
+ "loss": 0.0998,
1037
+ "step": 145
1038
+ },
1039
+ {
1040
+ "epoch": 1.3394495412844036,
1041
+ "grad_norm": 7247.53466796875,
1042
+ "learning_rate": 4.772727272727273e-05,
1043
+ "loss": 0.1255,
1044
+ "step": 146
1045
+ },
1046
+ {
1047
+ "epoch": 1.3486238532110093,
1048
+ "grad_norm": 7125.67822265625,
1049
+ "learning_rate": 4.7676767676767684e-05,
1050
+ "loss": 0.1123,
1051
+ "step": 147
1052
+ },
1053
+ {
1054
+ "epoch": 1.3577981651376148,
1055
+ "grad_norm": 7758.84375,
1056
+ "learning_rate": 4.762626262626263e-05,
1057
+ "loss": 0.1138,
1058
+ "step": 148
1059
+ },
1060
+ {
1061
+ "epoch": 1.3669724770642202,
1062
+ "grad_norm": 6958.53173828125,
1063
+ "learning_rate": 4.7575757575757576e-05,
1064
+ "loss": 0.0968,
1065
+ "step": 149
1066
+ },
1067
+ {
1068
+ "epoch": 1.3761467889908257,
1069
+ "grad_norm": 7246.337890625,
1070
+ "learning_rate": 4.752525252525253e-05,
1071
+ "loss": 0.0924,
1072
+ "step": 150
1073
+ },
1074
+ {
1075
+ "epoch": 1.385321100917431,
1076
+ "grad_norm": 7635.5615234375,
1077
+ "learning_rate": 4.7474747474747476e-05,
1078
+ "loss": 0.1152,
1079
+ "step": 151
1080
+ },
1081
+ {
1082
+ "epoch": 1.3944954128440368,
1083
+ "grad_norm": 7913.57080078125,
1084
+ "learning_rate": 4.742424242424243e-05,
1085
+ "loss": 0.0951,
1086
+ "step": 152
1087
+ },
1088
+ {
1089
+ "epoch": 1.4036697247706422,
1090
+ "grad_norm": 8531.388671875,
1091
+ "learning_rate": 4.7373737373737375e-05,
1092
+ "loss": 0.1161,
1093
+ "step": 153
1094
+ },
1095
+ {
1096
+ "epoch": 1.4128440366972477,
1097
+ "grad_norm": 9724.3291015625,
1098
+ "learning_rate": 4.732323232323232e-05,
1099
+ "loss": 0.1253,
1100
+ "step": 154
1101
+ },
1102
+ {
1103
+ "epoch": 1.4220183486238533,
1104
+ "grad_norm": 6726.74365234375,
1105
+ "learning_rate": 4.7272727272727275e-05,
1106
+ "loss": 0.1037,
1107
+ "step": 155
1108
+ },
1109
+ {
1110
+ "epoch": 1.4311926605504588,
1111
+ "grad_norm": 6435.69970703125,
1112
+ "learning_rate": 4.722222222222222e-05,
1113
+ "loss": 0.1036,
1114
+ "step": 156
1115
+ },
1116
+ {
1117
+ "epoch": 1.4403669724770642,
1118
+ "grad_norm": 6851.59814453125,
1119
+ "learning_rate": 4.7171717171717174e-05,
1120
+ "loss": 0.0954,
1121
+ "step": 157
1122
+ },
1123
+ {
1124
+ "epoch": 1.4495412844036697,
1125
+ "grad_norm": 7785.31640625,
1126
+ "learning_rate": 4.712121212121212e-05,
1127
+ "loss": 0.0996,
1128
+ "step": 158
1129
+ },
1130
+ {
1131
+ "epoch": 1.4587155963302751,
1132
+ "grad_norm": 6317.5576171875,
1133
+ "learning_rate": 4.7070707070707074e-05,
1134
+ "loss": 0.0977,
1135
+ "step": 159
1136
+ },
1137
+ {
1138
+ "epoch": 1.4678899082568808,
1139
+ "grad_norm": 8045.7197265625,
1140
+ "learning_rate": 4.702020202020202e-05,
1141
+ "loss": 0.1101,
1142
+ "step": 160
1143
+ },
1144
+ {
1145
+ "epoch": 1.4770642201834863,
1146
+ "grad_norm": 7386.67529296875,
1147
+ "learning_rate": 4.696969696969697e-05,
1148
+ "loss": 0.1137,
1149
+ "step": 161
1150
+ },
1151
+ {
1152
+ "epoch": 1.4862385321100917,
1153
+ "grad_norm": 8017.5791015625,
1154
+ "learning_rate": 4.6919191919191926e-05,
1155
+ "loss": 0.0959,
1156
+ "step": 162
1157
+ },
1158
+ {
1159
+ "epoch": 1.4954128440366974,
1160
+ "grad_norm": 7198.42138671875,
1161
+ "learning_rate": 4.686868686868687e-05,
1162
+ "loss": 0.1005,
1163
+ "step": 163
1164
+ },
1165
+ {
1166
+ "epoch": 1.5045871559633026,
1167
+ "grad_norm": 9139.7900390625,
1168
+ "learning_rate": 4.681818181818182e-05,
1169
+ "loss": 0.0918,
1170
+ "step": 164
1171
+ },
1172
+ {
1173
+ "epoch": 1.5137614678899083,
1174
+ "grad_norm": 6384.1640625,
1175
+ "learning_rate": 4.676767676767677e-05,
1176
+ "loss": 0.0948,
1177
+ "step": 165
1178
+ },
1179
+ {
1180
+ "epoch": 1.5229357798165137,
1181
+ "grad_norm": 5053.224609375,
1182
+ "learning_rate": 4.671717171717172e-05,
1183
+ "loss": 0.1056,
1184
+ "step": 166
1185
+ },
1186
+ {
1187
+ "epoch": 1.5321100917431192,
1188
+ "grad_norm": 9213.654296875,
1189
+ "learning_rate": 4.666666666666667e-05,
1190
+ "loss": 0.0931,
1191
+ "step": 167
1192
+ },
1193
+ {
1194
+ "epoch": 1.5412844036697249,
1195
+ "grad_norm": 7414.76171875,
1196
+ "learning_rate": 4.661616161616162e-05,
1197
+ "loss": 0.1123,
1198
+ "step": 168
1199
+ },
1200
+ {
1201
+ "epoch": 1.5504587155963303,
1202
+ "grad_norm": 6406.48583984375,
1203
+ "learning_rate": 4.656565656565657e-05,
1204
+ "loss": 0.0992,
1205
+ "step": 169
1206
+ },
1207
+ {
1208
+ "epoch": 1.5596330275229358,
1209
+ "grad_norm": 7213.58837890625,
1210
+ "learning_rate": 4.651515151515152e-05,
1211
+ "loss": 0.0952,
1212
+ "step": 170
1213
+ },
1214
+ {
1215
+ "epoch": 1.5688073394495414,
1216
+ "grad_norm": 7537.673828125,
1217
+ "learning_rate": 4.6464646464646464e-05,
1218
+ "loss": 0.0961,
1219
+ "step": 171
1220
+ },
1221
+ {
1222
+ "epoch": 1.5779816513761467,
1223
+ "grad_norm": 7876.294921875,
1224
+ "learning_rate": 4.641414141414142e-05,
1225
+ "loss": 0.1308,
1226
+ "step": 172
1227
+ },
1228
+ {
1229
+ "epoch": 1.5871559633027523,
1230
+ "grad_norm": 9550.125,
1231
+ "learning_rate": 4.636363636363636e-05,
1232
+ "loss": 0.0941,
1233
+ "step": 173
1234
+ },
1235
+ {
1236
+ "epoch": 1.5963302752293578,
1237
+ "grad_norm": 6364.3330078125,
1238
+ "learning_rate": 4.6313131313131316e-05,
1239
+ "loss": 0.0976,
1240
+ "step": 174
1241
+ },
1242
+ {
1243
+ "epoch": 1.6055045871559632,
1244
+ "grad_norm": 6976.2138671875,
1245
+ "learning_rate": 4.626262626262626e-05,
1246
+ "loss": 0.091,
1247
+ "step": 175
1248
+ },
1249
+ {
1250
+ "epoch": 1.614678899082569,
1251
+ "grad_norm": 7259.40234375,
1252
+ "learning_rate": 4.621212121212121e-05,
1253
+ "loss": 0.1096,
1254
+ "step": 176
1255
+ },
1256
+ {
1257
+ "epoch": 1.6238532110091743,
1258
+ "grad_norm": 8176.20849609375,
1259
+ "learning_rate": 4.616161616161616e-05,
1260
+ "loss": 0.1029,
1261
+ "step": 177
1262
+ },
1263
+ {
1264
+ "epoch": 1.6330275229357798,
1265
+ "grad_norm": 5628.04345703125,
1266
+ "learning_rate": 4.6111111111111115e-05,
1267
+ "loss": 0.1008,
1268
+ "step": 178
1269
+ },
1270
+ {
1271
+ "epoch": 1.6422018348623855,
1272
+ "grad_norm": 6802.91064453125,
1273
+ "learning_rate": 4.606060606060607e-05,
1274
+ "loss": 0.1014,
1275
+ "step": 179
1276
+ },
1277
+ {
1278
+ "epoch": 1.6513761467889907,
1279
+ "grad_norm": 14422.4482421875,
1280
+ "learning_rate": 4.6010101010101015e-05,
1281
+ "loss": 0.0912,
1282
+ "step": 180
1283
+ },
1284
+ {
1285
+ "epoch": 1.6605504587155964,
1286
+ "grad_norm": 8855.7744140625,
1287
+ "learning_rate": 4.595959595959596e-05,
1288
+ "loss": 0.0983,
1289
+ "step": 181
1290
+ },
1291
+ {
1292
+ "epoch": 1.6697247706422018,
1293
+ "grad_norm": 7411.93603515625,
1294
+ "learning_rate": 4.5909090909090914e-05,
1295
+ "loss": 0.1099,
1296
+ "step": 182
1297
+ },
1298
+ {
1299
+ "epoch": 1.6788990825688073,
1300
+ "grad_norm": 8138.41748046875,
1301
+ "learning_rate": 4.585858585858586e-05,
1302
+ "loss": 0.1054,
1303
+ "step": 183
1304
+ },
1305
+ {
1306
+ "epoch": 1.688073394495413,
1307
+ "grad_norm": 6157.80908203125,
1308
+ "learning_rate": 4.5808080808080814e-05,
1309
+ "loss": 0.1054,
1310
+ "step": 184
1311
+ },
1312
+ {
1313
+ "epoch": 1.6972477064220184,
1314
+ "grad_norm": 6791.07421875,
1315
+ "learning_rate": 4.575757575757576e-05,
1316
+ "loss": 0.1022,
1317
+ "step": 185
1318
+ },
1319
+ {
1320
+ "epoch": 1.7064220183486238,
1321
+ "grad_norm": 9991.0537109375,
1322
+ "learning_rate": 4.5707070707070706e-05,
1323
+ "loss": 0.1131,
1324
+ "step": 186
1325
+ },
1326
+ {
1327
+ "epoch": 1.7155963302752295,
1328
+ "grad_norm": 7240.28955078125,
1329
+ "learning_rate": 4.565656565656566e-05,
1330
+ "loss": 0.1005,
1331
+ "step": 187
1332
+ },
1333
+ {
1334
+ "epoch": 1.7247706422018347,
1335
+ "grad_norm": 6442.6259765625,
1336
+ "learning_rate": 4.5606060606060606e-05,
1337
+ "loss": 0.1015,
1338
+ "step": 188
1339
+ },
1340
+ {
1341
+ "epoch": 1.7339449541284404,
1342
+ "grad_norm": 8573.3955078125,
1343
+ "learning_rate": 4.555555555555556e-05,
1344
+ "loss": 0.1176,
1345
+ "step": 189
1346
+ },
1347
+ {
1348
+ "epoch": 1.7431192660550459,
1349
+ "grad_norm": 7330.88525390625,
1350
+ "learning_rate": 4.5505050505050505e-05,
1351
+ "loss": 0.0923,
1352
+ "step": 190
1353
+ },
1354
+ {
1355
+ "epoch": 1.7522935779816513,
1356
+ "grad_norm": 7870.916015625,
1357
+ "learning_rate": 4.545454545454546e-05,
1358
+ "loss": 0.1113,
1359
+ "step": 191
1360
+ },
1361
+ {
1362
+ "epoch": 1.761467889908257,
1363
+ "grad_norm": 7789.8564453125,
1364
+ "learning_rate": 4.5404040404040405e-05,
1365
+ "loss": 0.1062,
1366
+ "step": 192
1367
+ },
1368
+ {
1369
+ "epoch": 1.7706422018348624,
1370
+ "grad_norm": 7983.59326171875,
1371
+ "learning_rate": 4.535353535353535e-05,
1372
+ "loss": 0.1078,
1373
+ "step": 193
1374
+ },
1375
+ {
1376
+ "epoch": 1.7798165137614679,
1377
+ "grad_norm": 6533.43994140625,
1378
+ "learning_rate": 4.5303030303030304e-05,
1379
+ "loss": 0.1033,
1380
+ "step": 194
1381
+ },
1382
+ {
1383
+ "epoch": 1.7889908256880735,
1384
+ "grad_norm": 7317.1318359375,
1385
+ "learning_rate": 4.525252525252526e-05,
1386
+ "loss": 0.1106,
1387
+ "step": 195
1388
+ },
1389
+ {
1390
+ "epoch": 1.7981651376146788,
1391
+ "grad_norm": 7829.7880859375,
1392
+ "learning_rate": 4.5202020202020204e-05,
1393
+ "loss": 0.1163,
1394
+ "step": 196
1395
+ },
1396
+ {
1397
+ "epoch": 1.8073394495412844,
1398
+ "grad_norm": 7524.41357421875,
1399
+ "learning_rate": 4.515151515151516e-05,
1400
+ "loss": 0.1043,
1401
+ "step": 197
1402
+ },
1403
+ {
1404
+ "epoch": 1.81651376146789,
1405
+ "grad_norm": 7487.89892578125,
1406
+ "learning_rate": 4.51010101010101e-05,
1407
+ "loss": 0.1091,
1408
+ "step": 198
1409
+ },
1410
+ {
1411
+ "epoch": 1.8256880733944953,
1412
+ "grad_norm": 6964.3154296875,
1413
+ "learning_rate": 4.5050505050505056e-05,
1414
+ "loss": 0.0986,
1415
+ "step": 199
1416
+ },
1417
+ {
1418
+ "epoch": 1.834862385321101,
1419
+ "grad_norm": 7087.05029296875,
1420
+ "learning_rate": 4.5e-05,
1421
+ "loss": 0.0997,
1422
+ "step": 200
1423
+ },
1424
+ {
1425
+ "epoch": 1.8440366972477065,
1426
+ "grad_norm": 7148.7578125,
1427
+ "learning_rate": 4.494949494949495e-05,
1428
+ "loss": 0.0989,
1429
+ "step": 201
1430
+ },
1431
+ {
1432
+ "epoch": 1.853211009174312,
1433
+ "grad_norm": 7850.2470703125,
1434
+ "learning_rate": 4.48989898989899e-05,
1435
+ "loss": 0.0957,
1436
+ "step": 202
1437
+ },
1438
+ {
1439
+ "epoch": 1.8623853211009176,
1440
+ "grad_norm": 7752.841796875,
1441
+ "learning_rate": 4.484848484848485e-05,
1442
+ "loss": 0.1292,
1443
+ "step": 203
1444
+ },
1445
+ {
1446
+ "epoch": 1.8715596330275228,
1447
+ "grad_norm": 6098.35595703125,
1448
+ "learning_rate": 4.47979797979798e-05,
1449
+ "loss": 0.1001,
1450
+ "step": 204
1451
+ },
1452
+ {
1453
+ "epoch": 1.8807339449541285,
1454
+ "grad_norm": 7094.8173828125,
1455
+ "learning_rate": 4.474747474747475e-05,
1456
+ "loss": 0.1286,
1457
+ "step": 205
1458
+ },
1459
+ {
1460
+ "epoch": 1.889908256880734,
1461
+ "grad_norm": 7456.2265625,
1462
+ "learning_rate": 4.46969696969697e-05,
1463
+ "loss": 0.1256,
1464
+ "step": 206
1465
+ },
1466
+ {
1467
+ "epoch": 1.8990825688073394,
1468
+ "grad_norm": 8136.77197265625,
1469
+ "learning_rate": 4.464646464646465e-05,
1470
+ "loss": 0.1014,
1471
+ "step": 207
1472
+ },
1473
+ {
1474
+ "epoch": 1.908256880733945,
1475
+ "grad_norm": 7520.1650390625,
1476
+ "learning_rate": 4.4595959595959594e-05,
1477
+ "loss": 0.0919,
1478
+ "step": 208
1479
+ },
1480
+ {
1481
+ "epoch": 1.9174311926605505,
1482
+ "grad_norm": 8064.79296875,
1483
+ "learning_rate": 4.454545454545455e-05,
1484
+ "loss": 0.0985,
1485
+ "step": 209
1486
+ },
1487
+ {
1488
+ "epoch": 1.926605504587156,
1489
+ "grad_norm": 6153.19677734375,
1490
+ "learning_rate": 4.4494949494949493e-05,
1491
+ "loss": 0.1101,
1492
+ "step": 210
1493
+ },
1494
+ {
1495
+ "epoch": 1.9357798165137616,
1496
+ "grad_norm": 8089.42431640625,
1497
+ "learning_rate": 4.4444444444444447e-05,
1498
+ "loss": 0.0988,
1499
+ "step": 211
1500
+ },
1501
+ {
1502
+ "epoch": 1.9449541284403669,
1503
+ "grad_norm": 8096.140625,
1504
+ "learning_rate": 4.43939393939394e-05,
1505
+ "loss": 0.0992,
1506
+ "step": 212
1507
+ },
1508
+ {
1509
+ "epoch": 1.9541284403669725,
1510
+ "grad_norm": 7865.43798828125,
1511
+ "learning_rate": 4.4343434343434346e-05,
1512
+ "loss": 0.0945,
1513
+ "step": 213
1514
+ },
1515
+ {
1516
+ "epoch": 1.963302752293578,
1517
+ "grad_norm": 6380.67724609375,
1518
+ "learning_rate": 4.42929292929293e-05,
1519
+ "loss": 0.1073,
1520
+ "step": 214
1521
+ },
1522
+ {
1523
+ "epoch": 1.9724770642201834,
1524
+ "grad_norm": 8548.4345703125,
1525
+ "learning_rate": 4.4242424242424246e-05,
1526
+ "loss": 0.1121,
1527
+ "step": 215
1528
+ },
1529
+ {
1530
+ "epoch": 1.981651376146789,
1531
+ "grad_norm": 8465.8466796875,
1532
+ "learning_rate": 4.41919191919192e-05,
1533
+ "loss": 0.1078,
1534
+ "step": 216
1535
+ },
1536
+ {
1537
+ "epoch": 1.9908256880733946,
1538
+ "grad_norm": 8595.419921875,
1539
+ "learning_rate": 4.4141414141414145e-05,
1540
+ "loss": 0.0924,
1541
+ "step": 217
1542
+ },
1543
+ {
1544
+ "epoch": 2.0,
1545
+ "grad_norm": 22229.833984375,
1546
+ "learning_rate": 4.409090909090909e-05,
1547
+ "loss": 0.0785,
1548
+ "step": 218
1549
+ },
1550
+ {
1551
+ "epoch": 2.0,
1552
+ "eval_f1_macro": 0.0,
1553
+ "eval_f1_micro": 0.0,
1554
+ "eval_loss": 0.10520372539758682,
1555
+ "eval_precision": 0.0,
1556
+ "eval_recall": 0.0,
1557
+ "eval_runtime": 0.8127,
1558
+ "eval_samples_per_second": 355.611,
1559
+ "eval_steps_per_second": 15.996,
1560
+ "step": 218
1561
+ }
1562
+ ],
1563
+ "logging_steps": 1,
1564
+ "max_steps": 1090,
1565
+ "num_input_tokens_seen": 0,
1566
+ "num_train_epochs": 10,
1567
+ "save_steps": 500,
1568
+ "stateful_callbacks": {
1569
+ "TrainerControl": {
1570
+ "args": {
1571
+ "should_epoch_stop": false,
1572
+ "should_evaluate": false,
1573
+ "should_log": false,
1574
+ "should_save": true,
1575
+ "should_training_stop": false
1576
+ },
1577
+ "attributes": {}
1578
+ }
1579
+ },
1580
+ "total_flos": 1366134890434560.0,
1581
+ "train_batch_size": 24,
1582
+ "trial_name": null,
1583
+ "trial_params": null
1584
+ }
checkpoint-218/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edeef274e762b6821b55d94e571eef919508876a6f05e915f5b9e4cc058a5c62
3
+ size 5368
checkpoint-218/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-327/config.json ADDED
@@ -0,0 +1,130 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "T1003.001",
13
+ "1": "T1005",
14
+ "2": "T1012",
15
+ "3": "T1016",
16
+ "4": "T1021.001",
17
+ "5": "T1027",
18
+ "6": "T1033",
19
+ "7": "T1036.005",
20
+ "8": "T1041",
21
+ "9": "T1047",
22
+ "10": "T1053.005",
23
+ "11": "T1055",
24
+ "12": "T1056.001",
25
+ "13": "T1057",
26
+ "14": "T1059.003",
27
+ "15": "T1068",
28
+ "16": "T1070.004",
29
+ "17": "T1071.001",
30
+ "18": "T1072",
31
+ "19": "T1074.001",
32
+ "20": "T1078",
33
+ "21": "T1082",
34
+ "22": "T1083",
35
+ "23": "T1090",
36
+ "24": "T1095",
37
+ "25": "T1105",
38
+ "26": "T1106",
39
+ "27": "T1110",
40
+ "28": "T1112",
41
+ "29": "T1113",
42
+ "30": "T1140",
43
+ "31": "T1190",
44
+ "32": "T1204.002",
45
+ "33": "T1210",
46
+ "34": "T1218.011",
47
+ "35": "T1219",
48
+ "36": "T1484.001",
49
+ "37": "T1518.001",
50
+ "38": "T1543.003",
51
+ "39": "T1547.001",
52
+ "40": "T1548.002",
53
+ "41": "T1552.001",
54
+ "42": "T1557.001",
55
+ "43": "T1562.001",
56
+ "44": "T1564.001",
57
+ "45": "T1566.001",
58
+ "46": "T1569.002",
59
+ "47": "T1570",
60
+ "48": "T1573.001",
61
+ "49": "T1574.002"
62
+ },
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 3072,
65
+ "label2id": {
66
+ "T1003.001": 0,
67
+ "T1005": 1,
68
+ "T1012": 2,
69
+ "T1016": 3,
70
+ "T1021.001": 4,
71
+ "T1027": 5,
72
+ "T1033": 6,
73
+ "T1036.005": 7,
74
+ "T1041": 8,
75
+ "T1047": 9,
76
+ "T1053.005": 10,
77
+ "T1055": 11,
78
+ "T1056.001": 12,
79
+ "T1057": 13,
80
+ "T1059.003": 14,
81
+ "T1068": 15,
82
+ "T1070.004": 16,
83
+ "T1071.001": 17,
84
+ "T1072": 18,
85
+ "T1074.001": 19,
86
+ "T1078": 20,
87
+ "T1082": 21,
88
+ "T1083": 22,
89
+ "T1090": 23,
90
+ "T1095": 24,
91
+ "T1105": 25,
92
+ "T1106": 26,
93
+ "T1110": 27,
94
+ "T1112": 28,
95
+ "T1113": 29,
96
+ "T1140": 30,
97
+ "T1190": 31,
98
+ "T1204.002": 32,
99
+ "T1210": 33,
100
+ "T1218.011": 34,
101
+ "T1219": 35,
102
+ "T1484.001": 36,
103
+ "T1518.001": 37,
104
+ "T1543.003": 38,
105
+ "T1547.001": 39,
106
+ "T1548.002": 40,
107
+ "T1552.001": 41,
108
+ "T1557.001": 42,
109
+ "T1562.001": 43,
110
+ "T1564.001": 44,
111
+ "T1566.001": 45,
112
+ "T1569.002": 46,
113
+ "T1570": 47,
114
+ "T1573.001": 48,
115
+ "T1574.002": 49
116
+ },
117
+ "layer_norm_eps": 1e-12,
118
+ "max_position_embeddings": 512,
119
+ "model_type": "bert",
120
+ "num_attention_heads": 12,
121
+ "num_hidden_layers": 12,
122
+ "pad_token_id": 0,
123
+ "position_embedding_type": "absolute",
124
+ "problem_type": "multi_label_classification",
125
+ "torch_dtype": "float32",
126
+ "transformers_version": "4.55.2",
127
+ "type_vocab_size": 2,
128
+ "use_cache": true,
129
+ "vocab_size": 30522
130
+ }
checkpoint-327/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d7d43f4ff02fc807d3e05b421475180af6c1fd105434ec1004e5859936c399b2
3
+ size 438106296
checkpoint-327/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efb024542c786f74f3a4ab72af2e8482074f047daa1d52005953fd5dbba88a39
3
+ size 876333626
checkpoint-327/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b4f996c8f197cbcb67888a3f6abf2a62466a02eaa36a7a5c75c5869267136f4
3
+ size 14244
checkpoint-327/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b30172cf14f5dbe00280d63e36224a9f28dc7a0e8b38a74ceb5eb284e84da363
3
+ size 988
checkpoint-327/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3b18c0bd8f3b3588b2a10fcdcd4d10ec6eefa4ff08603e64e4499dceedf9678
3
+ size 1064
checkpoint-327/special_tokens_map.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": {
3
+ "content": "[CLS]",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "mask_token": {
10
+ "content": "[MASK]",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "[PAD]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "sep_token": {
24
+ "content": "[SEP]",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "unk_token": {
31
+ "content": "[UNK]",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ }
37
+ }
checkpoint-327/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-327/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "BertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
checkpoint-327/trainer_state.json ADDED
@@ -0,0 +1,2359 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 327,
3
+ "best_metric": 0.37575757575757573,
4
+ "best_model_checkpoint": "./cysecbert-ttp-bert-base_data/checkpoint-327",
5
+ "epoch": 3.0,
6
+ "eval_steps": 500,
7
+ "global_step": 327,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.009174311926605505,
14
+ "grad_norm": 46530.15234375,
15
+ "learning_rate": 0.0,
16
+ "loss": 0.7023,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.01834862385321101,
21
+ "grad_norm": 49571.35546875,
22
+ "learning_rate": 5.000000000000001e-07,
23
+ "loss": 0.7006,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.027522935779816515,
28
+ "grad_norm": 45885.55078125,
29
+ "learning_rate": 1.0000000000000002e-06,
30
+ "loss": 0.7033,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.03669724770642202,
35
+ "grad_norm": 43333.73046875,
36
+ "learning_rate": 1.5e-06,
37
+ "loss": 0.7014,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.045871559633027525,
42
+ "grad_norm": 44340.41796875,
43
+ "learning_rate": 2.0000000000000003e-06,
44
+ "loss": 0.6992,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.05504587155963303,
49
+ "grad_norm": 48282.54296875,
50
+ "learning_rate": 2.5e-06,
51
+ "loss": 0.6924,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.06422018348623854,
56
+ "grad_norm": 48561.56640625,
57
+ "learning_rate": 3e-06,
58
+ "loss": 0.6948,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.07339449541284404,
63
+ "grad_norm": 43580.078125,
64
+ "learning_rate": 3.5000000000000004e-06,
65
+ "loss": 0.6916,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.08256880733944955,
70
+ "grad_norm": 46552.28515625,
71
+ "learning_rate": 4.000000000000001e-06,
72
+ "loss": 0.6854,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.09174311926605505,
77
+ "grad_norm": 49042.15234375,
78
+ "learning_rate": 4.5e-06,
79
+ "loss": 0.6822,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.10091743119266056,
84
+ "grad_norm": 97274.8515625,
85
+ "learning_rate": 5e-06,
86
+ "loss": 0.6788,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.11009174311926606,
91
+ "grad_norm": 55310.375,
92
+ "learning_rate": 5.500000000000001e-06,
93
+ "loss": 0.6781,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.11926605504587157,
98
+ "grad_norm": 70317.6328125,
99
+ "learning_rate": 6e-06,
100
+ "loss": 0.6754,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.12844036697247707,
105
+ "grad_norm": 46825.9765625,
106
+ "learning_rate": 6.5000000000000004e-06,
107
+ "loss": 0.6626,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.13761467889908258,
112
+ "grad_norm": 50745.15625,
113
+ "learning_rate": 7.000000000000001e-06,
114
+ "loss": 0.655,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 0.14678899082568808,
119
+ "grad_norm": 51342.453125,
120
+ "learning_rate": 7.5e-06,
121
+ "loss": 0.6491,
122
+ "step": 16
123
+ },
124
+ {
125
+ "epoch": 0.1559633027522936,
126
+ "grad_norm": 51815.421875,
127
+ "learning_rate": 8.000000000000001e-06,
128
+ "loss": 0.6464,
129
+ "step": 17
130
+ },
131
+ {
132
+ "epoch": 0.1651376146788991,
133
+ "grad_norm": 54610.65234375,
134
+ "learning_rate": 8.500000000000002e-06,
135
+ "loss": 0.6298,
136
+ "step": 18
137
+ },
138
+ {
139
+ "epoch": 0.1743119266055046,
140
+ "grad_norm": 61936.31640625,
141
+ "learning_rate": 9e-06,
142
+ "loss": 0.623,
143
+ "step": 19
144
+ },
145
+ {
146
+ "epoch": 0.1834862385321101,
147
+ "grad_norm": 56021.69921875,
148
+ "learning_rate": 9.5e-06,
149
+ "loss": 0.6156,
150
+ "step": 20
151
+ },
152
+ {
153
+ "epoch": 0.1926605504587156,
154
+ "grad_norm": 60806.10546875,
155
+ "learning_rate": 1e-05,
156
+ "loss": 0.5981,
157
+ "step": 21
158
+ },
159
+ {
160
+ "epoch": 0.2018348623853211,
161
+ "grad_norm": 58582.4140625,
162
+ "learning_rate": 1.05e-05,
163
+ "loss": 0.5869,
164
+ "step": 22
165
+ },
166
+ {
167
+ "epoch": 0.21100917431192662,
168
+ "grad_norm": 59099.16015625,
169
+ "learning_rate": 1.1000000000000001e-05,
170
+ "loss": 0.5732,
171
+ "step": 23
172
+ },
173
+ {
174
+ "epoch": 0.22018348623853212,
175
+ "grad_norm": 52971.81640625,
176
+ "learning_rate": 1.1500000000000002e-05,
177
+ "loss": 0.5587,
178
+ "step": 24
179
+ },
180
+ {
181
+ "epoch": 0.22935779816513763,
182
+ "grad_norm": 51724.43359375,
183
+ "learning_rate": 1.2e-05,
184
+ "loss": 0.5477,
185
+ "step": 25
186
+ },
187
+ {
188
+ "epoch": 0.23853211009174313,
189
+ "grad_norm": 58133.453125,
190
+ "learning_rate": 1.25e-05,
191
+ "loss": 0.5422,
192
+ "step": 26
193
+ },
194
+ {
195
+ "epoch": 0.24770642201834864,
196
+ "grad_norm": 54134.3359375,
197
+ "learning_rate": 1.3000000000000001e-05,
198
+ "loss": 0.5319,
199
+ "step": 27
200
+ },
201
+ {
202
+ "epoch": 0.25688073394495414,
203
+ "grad_norm": 78195.7890625,
204
+ "learning_rate": 1.3500000000000001e-05,
205
+ "loss": 0.5259,
206
+ "step": 28
207
+ },
208
+ {
209
+ "epoch": 0.26605504587155965,
210
+ "grad_norm": 47713.20703125,
211
+ "learning_rate": 1.4000000000000001e-05,
212
+ "loss": 0.5105,
213
+ "step": 29
214
+ },
215
+ {
216
+ "epoch": 0.27522935779816515,
217
+ "grad_norm": 45838.0703125,
218
+ "learning_rate": 1.45e-05,
219
+ "loss": 0.499,
220
+ "step": 30
221
+ },
222
+ {
223
+ "epoch": 0.28440366972477066,
224
+ "grad_norm": 46735.58203125,
225
+ "learning_rate": 1.5e-05,
226
+ "loss": 0.4914,
227
+ "step": 31
228
+ },
229
+ {
230
+ "epoch": 0.29357798165137616,
231
+ "grad_norm": 45432.8828125,
232
+ "learning_rate": 1.55e-05,
233
+ "loss": 0.4783,
234
+ "step": 32
235
+ },
236
+ {
237
+ "epoch": 0.30275229357798167,
238
+ "grad_norm": 46758.58984375,
239
+ "learning_rate": 1.6000000000000003e-05,
240
+ "loss": 0.4758,
241
+ "step": 33
242
+ },
243
+ {
244
+ "epoch": 0.3119266055045872,
245
+ "grad_norm": 45245.82421875,
246
+ "learning_rate": 1.65e-05,
247
+ "loss": 0.4608,
248
+ "step": 34
249
+ },
250
+ {
251
+ "epoch": 0.3211009174311927,
252
+ "grad_norm": 43336.75390625,
253
+ "learning_rate": 1.7000000000000003e-05,
254
+ "loss": 0.4567,
255
+ "step": 35
256
+ },
257
+ {
258
+ "epoch": 0.3302752293577982,
259
+ "grad_norm": 41445.90625,
260
+ "learning_rate": 1.75e-05,
261
+ "loss": 0.4465,
262
+ "step": 36
263
+ },
264
+ {
265
+ "epoch": 0.3394495412844037,
266
+ "grad_norm": 42278.80078125,
267
+ "learning_rate": 1.8e-05,
268
+ "loss": 0.4395,
269
+ "step": 37
270
+ },
271
+ {
272
+ "epoch": 0.3486238532110092,
273
+ "grad_norm": 42039.28515625,
274
+ "learning_rate": 1.85e-05,
275
+ "loss": 0.4289,
276
+ "step": 38
277
+ },
278
+ {
279
+ "epoch": 0.3577981651376147,
280
+ "grad_norm": 41411.61328125,
281
+ "learning_rate": 1.9e-05,
282
+ "loss": 0.423,
283
+ "step": 39
284
+ },
285
+ {
286
+ "epoch": 0.3669724770642202,
287
+ "grad_norm": 40133.84375,
288
+ "learning_rate": 1.9500000000000003e-05,
289
+ "loss": 0.4191,
290
+ "step": 40
291
+ },
292
+ {
293
+ "epoch": 0.3761467889908257,
294
+ "grad_norm": 39758.8828125,
295
+ "learning_rate": 2e-05,
296
+ "loss": 0.4071,
297
+ "step": 41
298
+ },
299
+ {
300
+ "epoch": 0.3853211009174312,
301
+ "grad_norm": 38760.36328125,
302
+ "learning_rate": 2.05e-05,
303
+ "loss": 0.3996,
304
+ "step": 42
305
+ },
306
+ {
307
+ "epoch": 0.3944954128440367,
308
+ "grad_norm": 38552.80078125,
309
+ "learning_rate": 2.1e-05,
310
+ "loss": 0.3918,
311
+ "step": 43
312
+ },
313
+ {
314
+ "epoch": 0.4036697247706422,
315
+ "grad_norm": 38561.7578125,
316
+ "learning_rate": 2.15e-05,
317
+ "loss": 0.3865,
318
+ "step": 44
319
+ },
320
+ {
321
+ "epoch": 0.41284403669724773,
322
+ "grad_norm": 37616.85546875,
323
+ "learning_rate": 2.2000000000000003e-05,
324
+ "loss": 0.3794,
325
+ "step": 45
326
+ },
327
+ {
328
+ "epoch": 0.42201834862385323,
329
+ "grad_norm": 37784.78515625,
330
+ "learning_rate": 2.25e-05,
331
+ "loss": 0.3745,
332
+ "step": 46
333
+ },
334
+ {
335
+ "epoch": 0.43119266055045874,
336
+ "grad_norm": 36332.9140625,
337
+ "learning_rate": 2.3000000000000003e-05,
338
+ "loss": 0.3711,
339
+ "step": 47
340
+ },
341
+ {
342
+ "epoch": 0.44036697247706424,
343
+ "grad_norm": 38899.73046875,
344
+ "learning_rate": 2.35e-05,
345
+ "loss": 0.3643,
346
+ "step": 48
347
+ },
348
+ {
349
+ "epoch": 0.44954128440366975,
350
+ "grad_norm": 36174.9765625,
351
+ "learning_rate": 2.4e-05,
352
+ "loss": 0.3575,
353
+ "step": 49
354
+ },
355
+ {
356
+ "epoch": 0.45871559633027525,
357
+ "grad_norm": 35699.83203125,
358
+ "learning_rate": 2.45e-05,
359
+ "loss": 0.347,
360
+ "step": 50
361
+ },
362
+ {
363
+ "epoch": 0.46788990825688076,
364
+ "grad_norm": 35562.3671875,
365
+ "learning_rate": 2.5e-05,
366
+ "loss": 0.3438,
367
+ "step": 51
368
+ },
369
+ {
370
+ "epoch": 0.47706422018348627,
371
+ "grad_norm": 35428.0625,
372
+ "learning_rate": 2.5500000000000003e-05,
373
+ "loss": 0.3325,
374
+ "step": 52
375
+ },
376
+ {
377
+ "epoch": 0.48623853211009177,
378
+ "grad_norm": 34396.80078125,
379
+ "learning_rate": 2.6000000000000002e-05,
380
+ "loss": 0.3302,
381
+ "step": 53
382
+ },
383
+ {
384
+ "epoch": 0.4954128440366973,
385
+ "grad_norm": 38376.41796875,
386
+ "learning_rate": 2.6500000000000004e-05,
387
+ "loss": 0.3218,
388
+ "step": 54
389
+ },
390
+ {
391
+ "epoch": 0.5045871559633027,
392
+ "grad_norm": 33996.23046875,
393
+ "learning_rate": 2.7000000000000002e-05,
394
+ "loss": 0.3131,
395
+ "step": 55
396
+ },
397
+ {
398
+ "epoch": 0.5137614678899083,
399
+ "grad_norm": 33070.55859375,
400
+ "learning_rate": 2.7500000000000004e-05,
401
+ "loss": 0.3129,
402
+ "step": 56
403
+ },
404
+ {
405
+ "epoch": 0.5229357798165137,
406
+ "grad_norm": 33200.44140625,
407
+ "learning_rate": 2.8000000000000003e-05,
408
+ "loss": 0.3003,
409
+ "step": 57
410
+ },
411
+ {
412
+ "epoch": 0.5321100917431193,
413
+ "grad_norm": 33099.1015625,
414
+ "learning_rate": 2.8499999999999998e-05,
415
+ "loss": 0.2904,
416
+ "step": 58
417
+ },
418
+ {
419
+ "epoch": 0.5412844036697247,
420
+ "grad_norm": 31352.298828125,
421
+ "learning_rate": 2.9e-05,
422
+ "loss": 0.2911,
423
+ "step": 59
424
+ },
425
+ {
426
+ "epoch": 0.5504587155963303,
427
+ "grad_norm": 31965.087890625,
428
+ "learning_rate": 2.95e-05,
429
+ "loss": 0.2815,
430
+ "step": 60
431
+ },
432
+ {
433
+ "epoch": 0.5596330275229358,
434
+ "grad_norm": 30810.849609375,
435
+ "learning_rate": 3e-05,
436
+ "loss": 0.2818,
437
+ "step": 61
438
+ },
439
+ {
440
+ "epoch": 0.5688073394495413,
441
+ "grad_norm": 30034.779296875,
442
+ "learning_rate": 3.05e-05,
443
+ "loss": 0.2731,
444
+ "step": 62
445
+ },
446
+ {
447
+ "epoch": 0.5779816513761468,
448
+ "grad_norm": 30009.12109375,
449
+ "learning_rate": 3.1e-05,
450
+ "loss": 0.266,
451
+ "step": 63
452
+ },
453
+ {
454
+ "epoch": 0.5871559633027523,
455
+ "grad_norm": 29305.173828125,
456
+ "learning_rate": 3.15e-05,
457
+ "loss": 0.2609,
458
+ "step": 64
459
+ },
460
+ {
461
+ "epoch": 0.5963302752293578,
462
+ "grad_norm": 29081.853515625,
463
+ "learning_rate": 3.2000000000000005e-05,
464
+ "loss": 0.2543,
465
+ "step": 65
466
+ },
467
+ {
468
+ "epoch": 0.6055045871559633,
469
+ "grad_norm": 28217.021484375,
470
+ "learning_rate": 3.2500000000000004e-05,
471
+ "loss": 0.252,
472
+ "step": 66
473
+ },
474
+ {
475
+ "epoch": 0.6146788990825688,
476
+ "grad_norm": 26448.12890625,
477
+ "learning_rate": 3.3e-05,
478
+ "loss": 0.2565,
479
+ "step": 67
480
+ },
481
+ {
482
+ "epoch": 0.6238532110091743,
483
+ "grad_norm": 27198.80859375,
484
+ "learning_rate": 3.35e-05,
485
+ "loss": 0.2342,
486
+ "step": 68
487
+ },
488
+ {
489
+ "epoch": 0.6330275229357798,
490
+ "grad_norm": 26946.30859375,
491
+ "learning_rate": 3.4000000000000007e-05,
492
+ "loss": 0.2273,
493
+ "step": 69
494
+ },
495
+ {
496
+ "epoch": 0.6422018348623854,
497
+ "grad_norm": 26236.7265625,
498
+ "learning_rate": 3.45e-05,
499
+ "loss": 0.2219,
500
+ "step": 70
501
+ },
502
+ {
503
+ "epoch": 0.6513761467889908,
504
+ "grad_norm": 25535.818359375,
505
+ "learning_rate": 3.5e-05,
506
+ "loss": 0.2197,
507
+ "step": 71
508
+ },
509
+ {
510
+ "epoch": 0.6605504587155964,
511
+ "grad_norm": 24871.8515625,
512
+ "learning_rate": 3.55e-05,
513
+ "loss": 0.2144,
514
+ "step": 72
515
+ },
516
+ {
517
+ "epoch": 0.6697247706422018,
518
+ "grad_norm": 24484.55078125,
519
+ "learning_rate": 3.6e-05,
520
+ "loss": 0.2066,
521
+ "step": 73
522
+ },
523
+ {
524
+ "epoch": 0.6788990825688074,
525
+ "grad_norm": 23655.677734375,
526
+ "learning_rate": 3.65e-05,
527
+ "loss": 0.206,
528
+ "step": 74
529
+ },
530
+ {
531
+ "epoch": 0.6880733944954128,
532
+ "grad_norm": 23129.076171875,
533
+ "learning_rate": 3.7e-05,
534
+ "loss": 0.1979,
535
+ "step": 75
536
+ },
537
+ {
538
+ "epoch": 0.6972477064220184,
539
+ "grad_norm": 21871.90625,
540
+ "learning_rate": 3.7500000000000003e-05,
541
+ "loss": 0.1973,
542
+ "step": 76
543
+ },
544
+ {
545
+ "epoch": 0.7064220183486238,
546
+ "grad_norm": 21561.58984375,
547
+ "learning_rate": 3.8e-05,
548
+ "loss": 0.1935,
549
+ "step": 77
550
+ },
551
+ {
552
+ "epoch": 0.7155963302752294,
553
+ "grad_norm": 22210.29296875,
554
+ "learning_rate": 3.85e-05,
555
+ "loss": 0.1848,
556
+ "step": 78
557
+ },
558
+ {
559
+ "epoch": 0.7247706422018348,
560
+ "grad_norm": 20555.775390625,
561
+ "learning_rate": 3.9000000000000006e-05,
562
+ "loss": 0.1801,
563
+ "step": 79
564
+ },
565
+ {
566
+ "epoch": 0.7339449541284404,
567
+ "grad_norm": 20406.75,
568
+ "learning_rate": 3.9500000000000005e-05,
569
+ "loss": 0.1746,
570
+ "step": 80
571
+ },
572
+ {
573
+ "epoch": 0.7431192660550459,
574
+ "grad_norm": 19150.931640625,
575
+ "learning_rate": 4e-05,
576
+ "loss": 0.1688,
577
+ "step": 81
578
+ },
579
+ {
580
+ "epoch": 0.7522935779816514,
581
+ "grad_norm": 18422.65625,
582
+ "learning_rate": 4.05e-05,
583
+ "loss": 0.1661,
584
+ "step": 82
585
+ },
586
+ {
587
+ "epoch": 0.7614678899082569,
588
+ "grad_norm": 16746.20703125,
589
+ "learning_rate": 4.1e-05,
590
+ "loss": 0.1711,
591
+ "step": 83
592
+ },
593
+ {
594
+ "epoch": 0.7706422018348624,
595
+ "grad_norm": 18229.41015625,
596
+ "learning_rate": 4.15e-05,
597
+ "loss": 0.1646,
598
+ "step": 84
599
+ },
600
+ {
601
+ "epoch": 0.7798165137614679,
602
+ "grad_norm": 17156.267578125,
603
+ "learning_rate": 4.2e-05,
604
+ "loss": 0.1588,
605
+ "step": 85
606
+ },
607
+ {
608
+ "epoch": 0.7889908256880734,
609
+ "grad_norm": 15970.611328125,
610
+ "learning_rate": 4.25e-05,
611
+ "loss": 0.1547,
612
+ "step": 86
613
+ },
614
+ {
615
+ "epoch": 0.7981651376146789,
616
+ "grad_norm": 16997.103515625,
617
+ "learning_rate": 4.3e-05,
618
+ "loss": 0.1493,
619
+ "step": 87
620
+ },
621
+ {
622
+ "epoch": 0.8073394495412844,
623
+ "grad_norm": 15529.58984375,
624
+ "learning_rate": 4.35e-05,
625
+ "loss": 0.1439,
626
+ "step": 88
627
+ },
628
+ {
629
+ "epoch": 0.8165137614678899,
630
+ "grad_norm": 15099.9052734375,
631
+ "learning_rate": 4.4000000000000006e-05,
632
+ "loss": 0.1432,
633
+ "step": 89
634
+ },
635
+ {
636
+ "epoch": 0.8256880733944955,
637
+ "grad_norm": 14261.3232421875,
638
+ "learning_rate": 4.4500000000000004e-05,
639
+ "loss": 0.1396,
640
+ "step": 90
641
+ },
642
+ {
643
+ "epoch": 0.8348623853211009,
644
+ "grad_norm": 13915.87890625,
645
+ "learning_rate": 4.5e-05,
646
+ "loss": 0.1317,
647
+ "step": 91
648
+ },
649
+ {
650
+ "epoch": 0.8440366972477065,
651
+ "grad_norm": 13518.7568359375,
652
+ "learning_rate": 4.55e-05,
653
+ "loss": 0.1289,
654
+ "step": 92
655
+ },
656
+ {
657
+ "epoch": 0.8532110091743119,
658
+ "grad_norm": 12170.1865234375,
659
+ "learning_rate": 4.600000000000001e-05,
660
+ "loss": 0.1371,
661
+ "step": 93
662
+ },
663
+ {
664
+ "epoch": 0.8623853211009175,
665
+ "grad_norm": 11745.9775390625,
666
+ "learning_rate": 4.6500000000000005e-05,
667
+ "loss": 0.1434,
668
+ "step": 94
669
+ },
670
+ {
671
+ "epoch": 0.8715596330275229,
672
+ "grad_norm": 12265.5556640625,
673
+ "learning_rate": 4.7e-05,
674
+ "loss": 0.1343,
675
+ "step": 95
676
+ },
677
+ {
678
+ "epoch": 0.8807339449541285,
679
+ "grad_norm": 11386.3291015625,
680
+ "learning_rate": 4.75e-05,
681
+ "loss": 0.1317,
682
+ "step": 96
683
+ },
684
+ {
685
+ "epoch": 0.8899082568807339,
686
+ "grad_norm": 10869.4599609375,
687
+ "learning_rate": 4.8e-05,
688
+ "loss": 0.1295,
689
+ "step": 97
690
+ },
691
+ {
692
+ "epoch": 0.8990825688073395,
693
+ "grad_norm": 10752.25,
694
+ "learning_rate": 4.85e-05,
695
+ "loss": 0.124,
696
+ "step": 98
697
+ },
698
+ {
699
+ "epoch": 0.908256880733945,
700
+ "grad_norm": 9687.6376953125,
701
+ "learning_rate": 4.9e-05,
702
+ "loss": 0.1271,
703
+ "step": 99
704
+ },
705
+ {
706
+ "epoch": 0.9174311926605505,
707
+ "grad_norm": 9697.15234375,
708
+ "learning_rate": 4.9500000000000004e-05,
709
+ "loss": 0.1194,
710
+ "step": 100
711
+ },
712
+ {
713
+ "epoch": 0.926605504587156,
714
+ "grad_norm": 9423.21875,
715
+ "learning_rate": 5e-05,
716
+ "loss": 0.1176,
717
+ "step": 101
718
+ },
719
+ {
720
+ "epoch": 0.9357798165137615,
721
+ "grad_norm": 10253.8564453125,
722
+ "learning_rate": 4.994949494949495e-05,
723
+ "loss": 0.121,
724
+ "step": 102
725
+ },
726
+ {
727
+ "epoch": 0.944954128440367,
728
+ "grad_norm": 9756.7646484375,
729
+ "learning_rate": 4.98989898989899e-05,
730
+ "loss": 0.1173,
731
+ "step": 103
732
+ },
733
+ {
734
+ "epoch": 0.9541284403669725,
735
+ "grad_norm": 8482.177734375,
736
+ "learning_rate": 4.984848484848485e-05,
737
+ "loss": 0.1233,
738
+ "step": 104
739
+ },
740
+ {
741
+ "epoch": 0.963302752293578,
742
+ "grad_norm": 7755.4892578125,
743
+ "learning_rate": 4.97979797979798e-05,
744
+ "loss": 0.1251,
745
+ "step": 105
746
+ },
747
+ {
748
+ "epoch": 0.9724770642201835,
749
+ "grad_norm": 8553.6689453125,
750
+ "learning_rate": 4.974747474747475e-05,
751
+ "loss": 0.1287,
752
+ "step": 106
753
+ },
754
+ {
755
+ "epoch": 0.981651376146789,
756
+ "grad_norm": 8794.58984375,
757
+ "learning_rate": 4.9696969696969694e-05,
758
+ "loss": 0.1178,
759
+ "step": 107
760
+ },
761
+ {
762
+ "epoch": 0.9908256880733946,
763
+ "grad_norm": 8048.97509765625,
764
+ "learning_rate": 4.964646464646465e-05,
765
+ "loss": 0.1112,
766
+ "step": 108
767
+ },
768
+ {
769
+ "epoch": 1.0,
770
+ "grad_norm": 24349.96484375,
771
+ "learning_rate": 4.9595959595959594e-05,
772
+ "loss": 0.112,
773
+ "step": 109
774
+ },
775
+ {
776
+ "epoch": 1.0,
777
+ "eval_f1_macro": 0.0,
778
+ "eval_f1_micro": 0.0,
779
+ "eval_loss": 0.1133684441447258,
780
+ "eval_precision": 0.0,
781
+ "eval_recall": 0.0,
782
+ "eval_runtime": 0.8578,
783
+ "eval_samples_per_second": 336.892,
784
+ "eval_steps_per_second": 15.154,
785
+ "step": 109
786
+ },
787
+ {
788
+ "epoch": 1.0091743119266054,
789
+ "grad_norm": 7603.02587890625,
790
+ "learning_rate": 4.9545454545454553e-05,
791
+ "loss": 0.1133,
792
+ "step": 110
793
+ },
794
+ {
795
+ "epoch": 1.018348623853211,
796
+ "grad_norm": 8602.8974609375,
797
+ "learning_rate": 4.94949494949495e-05,
798
+ "loss": 0.1192,
799
+ "step": 111
800
+ },
801
+ {
802
+ "epoch": 1.0275229357798166,
803
+ "grad_norm": 8388.9609375,
804
+ "learning_rate": 4.9444444444444446e-05,
805
+ "loss": 0.1246,
806
+ "step": 112
807
+ },
808
+ {
809
+ "epoch": 1.036697247706422,
810
+ "grad_norm": 9368.150390625,
811
+ "learning_rate": 4.93939393939394e-05,
812
+ "loss": 0.1089,
813
+ "step": 113
814
+ },
815
+ {
816
+ "epoch": 1.0458715596330275,
817
+ "grad_norm": 7053.083984375,
818
+ "learning_rate": 4.9343434343434346e-05,
819
+ "loss": 0.1042,
820
+ "step": 114
821
+ },
822
+ {
823
+ "epoch": 1.0550458715596331,
824
+ "grad_norm": 9539.36328125,
825
+ "learning_rate": 4.92929292929293e-05,
826
+ "loss": 0.1049,
827
+ "step": 115
828
+ },
829
+ {
830
+ "epoch": 1.0642201834862386,
831
+ "grad_norm": 6956.2763671875,
832
+ "learning_rate": 4.9242424242424245e-05,
833
+ "loss": 0.109,
834
+ "step": 116
835
+ },
836
+ {
837
+ "epoch": 1.073394495412844,
838
+ "grad_norm": 6597.9580078125,
839
+ "learning_rate": 4.919191919191919e-05,
840
+ "loss": 0.1124,
841
+ "step": 117
842
+ },
843
+ {
844
+ "epoch": 1.0825688073394495,
845
+ "grad_norm": 8124.76123046875,
846
+ "learning_rate": 4.9141414141414145e-05,
847
+ "loss": 0.1226,
848
+ "step": 118
849
+ },
850
+ {
851
+ "epoch": 1.091743119266055,
852
+ "grad_norm": 7030.619140625,
853
+ "learning_rate": 4.909090909090909e-05,
854
+ "loss": 0.0989,
855
+ "step": 119
856
+ },
857
+ {
858
+ "epoch": 1.1009174311926606,
859
+ "grad_norm": 9377.66796875,
860
+ "learning_rate": 4.9040404040404044e-05,
861
+ "loss": 0.1038,
862
+ "step": 120
863
+ },
864
+ {
865
+ "epoch": 1.110091743119266,
866
+ "grad_norm": 9298.802734375,
867
+ "learning_rate": 4.898989898989899e-05,
868
+ "loss": 0.1216,
869
+ "step": 121
870
+ },
871
+ {
872
+ "epoch": 1.1192660550458715,
873
+ "grad_norm": 7350.28369140625,
874
+ "learning_rate": 4.8939393939393944e-05,
875
+ "loss": 0.0915,
876
+ "step": 122
877
+ },
878
+ {
879
+ "epoch": 1.1284403669724772,
880
+ "grad_norm": 8066.943359375,
881
+ "learning_rate": 4.888888888888889e-05,
882
+ "loss": 0.1291,
883
+ "step": 123
884
+ },
885
+ {
886
+ "epoch": 1.1376146788990826,
887
+ "grad_norm": 8210.3095703125,
888
+ "learning_rate": 4.8838383838383836e-05,
889
+ "loss": 0.1196,
890
+ "step": 124
891
+ },
892
+ {
893
+ "epoch": 1.146788990825688,
894
+ "grad_norm": 7568.0234375,
895
+ "learning_rate": 4.878787878787879e-05,
896
+ "loss": 0.1037,
897
+ "step": 125
898
+ },
899
+ {
900
+ "epoch": 1.1559633027522935,
901
+ "grad_norm": 6394.8896484375,
902
+ "learning_rate": 4.8737373737373736e-05,
903
+ "loss": 0.0961,
904
+ "step": 126
905
+ },
906
+ {
907
+ "epoch": 1.165137614678899,
908
+ "grad_norm": 7549.9951171875,
909
+ "learning_rate": 4.868686868686869e-05,
910
+ "loss": 0.1075,
911
+ "step": 127
912
+ },
913
+ {
914
+ "epoch": 1.1743119266055047,
915
+ "grad_norm": 7983.36865234375,
916
+ "learning_rate": 4.863636363636364e-05,
917
+ "loss": 0.1055,
918
+ "step": 128
919
+ },
920
+ {
921
+ "epoch": 1.18348623853211,
922
+ "grad_norm": 7613.1455078125,
923
+ "learning_rate": 4.858585858585859e-05,
924
+ "loss": 0.1097,
925
+ "step": 129
926
+ },
927
+ {
928
+ "epoch": 1.1926605504587156,
929
+ "grad_norm": 7754.9228515625,
930
+ "learning_rate": 4.853535353535354e-05,
931
+ "loss": 0.1157,
932
+ "step": 130
933
+ },
934
+ {
935
+ "epoch": 1.2018348623853212,
936
+ "grad_norm": 8360.388671875,
937
+ "learning_rate": 4.848484848484849e-05,
938
+ "loss": 0.1019,
939
+ "step": 131
940
+ },
941
+ {
942
+ "epoch": 1.2110091743119267,
943
+ "grad_norm": 8300.9169921875,
944
+ "learning_rate": 4.843434343434344e-05,
945
+ "loss": 0.1098,
946
+ "step": 132
947
+ },
948
+ {
949
+ "epoch": 1.2201834862385321,
950
+ "grad_norm": 7554.8017578125,
951
+ "learning_rate": 4.838383838383839e-05,
952
+ "loss": 0.1075,
953
+ "step": 133
954
+ },
955
+ {
956
+ "epoch": 1.2293577981651376,
957
+ "grad_norm": 6510.427734375,
958
+ "learning_rate": 4.8333333333333334e-05,
959
+ "loss": 0.1041,
960
+ "step": 134
961
+ },
962
+ {
963
+ "epoch": 1.238532110091743,
964
+ "grad_norm": 5649.77978515625,
965
+ "learning_rate": 4.828282828282829e-05,
966
+ "loss": 0.1012,
967
+ "step": 135
968
+ },
969
+ {
970
+ "epoch": 1.2477064220183487,
971
+ "grad_norm": 7662.76513671875,
972
+ "learning_rate": 4.823232323232323e-05,
973
+ "loss": 0.1357,
974
+ "step": 136
975
+ },
976
+ {
977
+ "epoch": 1.2568807339449541,
978
+ "grad_norm": 7261.087890625,
979
+ "learning_rate": 4.8181818181818186e-05,
980
+ "loss": 0.1149,
981
+ "step": 137
982
+ },
983
+ {
984
+ "epoch": 1.2660550458715596,
985
+ "grad_norm": 6818.208984375,
986
+ "learning_rate": 4.813131313131313e-05,
987
+ "loss": 0.1067,
988
+ "step": 138
989
+ },
990
+ {
991
+ "epoch": 1.2752293577981653,
992
+ "grad_norm": 8839.0810546875,
993
+ "learning_rate": 4.808080808080808e-05,
994
+ "loss": 0.089,
995
+ "step": 139
996
+ },
997
+ {
998
+ "epoch": 1.2844036697247707,
999
+ "grad_norm": 8774.0615234375,
1000
+ "learning_rate": 4.803030303030303e-05,
1001
+ "loss": 0.1013,
1002
+ "step": 140
1003
+ },
1004
+ {
1005
+ "epoch": 1.2935779816513762,
1006
+ "grad_norm": 7869.935546875,
1007
+ "learning_rate": 4.797979797979798e-05,
1008
+ "loss": 0.1187,
1009
+ "step": 141
1010
+ },
1011
+ {
1012
+ "epoch": 1.3027522935779816,
1013
+ "grad_norm": 9808.5986328125,
1014
+ "learning_rate": 4.792929292929293e-05,
1015
+ "loss": 0.13,
1016
+ "step": 142
1017
+ },
1018
+ {
1019
+ "epoch": 1.311926605504587,
1020
+ "grad_norm": 7122.400390625,
1021
+ "learning_rate": 4.787878787878788e-05,
1022
+ "loss": 0.1089,
1023
+ "step": 143
1024
+ },
1025
+ {
1026
+ "epoch": 1.3211009174311927,
1027
+ "grad_norm": 6575.39892578125,
1028
+ "learning_rate": 4.782828282828283e-05,
1029
+ "loss": 0.1011,
1030
+ "step": 144
1031
+ },
1032
+ {
1033
+ "epoch": 1.3302752293577982,
1034
+ "grad_norm": 8656.4345703125,
1035
+ "learning_rate": 4.7777777777777784e-05,
1036
+ "loss": 0.0998,
1037
+ "step": 145
1038
+ },
1039
+ {
1040
+ "epoch": 1.3394495412844036,
1041
+ "grad_norm": 7247.53466796875,
1042
+ "learning_rate": 4.772727272727273e-05,
1043
+ "loss": 0.1255,
1044
+ "step": 146
1045
+ },
1046
+ {
1047
+ "epoch": 1.3486238532110093,
1048
+ "grad_norm": 7125.67822265625,
1049
+ "learning_rate": 4.7676767676767684e-05,
1050
+ "loss": 0.1123,
1051
+ "step": 147
1052
+ },
1053
+ {
1054
+ "epoch": 1.3577981651376148,
1055
+ "grad_norm": 7758.84375,
1056
+ "learning_rate": 4.762626262626263e-05,
1057
+ "loss": 0.1138,
1058
+ "step": 148
1059
+ },
1060
+ {
1061
+ "epoch": 1.3669724770642202,
1062
+ "grad_norm": 6958.53173828125,
1063
+ "learning_rate": 4.7575757575757576e-05,
1064
+ "loss": 0.0968,
1065
+ "step": 149
1066
+ },
1067
+ {
1068
+ "epoch": 1.3761467889908257,
1069
+ "grad_norm": 7246.337890625,
1070
+ "learning_rate": 4.752525252525253e-05,
1071
+ "loss": 0.0924,
1072
+ "step": 150
1073
+ },
1074
+ {
1075
+ "epoch": 1.385321100917431,
1076
+ "grad_norm": 7635.5615234375,
1077
+ "learning_rate": 4.7474747474747476e-05,
1078
+ "loss": 0.1152,
1079
+ "step": 151
1080
+ },
1081
+ {
1082
+ "epoch": 1.3944954128440368,
1083
+ "grad_norm": 7913.57080078125,
1084
+ "learning_rate": 4.742424242424243e-05,
1085
+ "loss": 0.0951,
1086
+ "step": 152
1087
+ },
1088
+ {
1089
+ "epoch": 1.4036697247706422,
1090
+ "grad_norm": 8531.388671875,
1091
+ "learning_rate": 4.7373737373737375e-05,
1092
+ "loss": 0.1161,
1093
+ "step": 153
1094
+ },
1095
+ {
1096
+ "epoch": 1.4128440366972477,
1097
+ "grad_norm": 9724.3291015625,
1098
+ "learning_rate": 4.732323232323232e-05,
1099
+ "loss": 0.1253,
1100
+ "step": 154
1101
+ },
1102
+ {
1103
+ "epoch": 1.4220183486238533,
1104
+ "grad_norm": 6726.74365234375,
1105
+ "learning_rate": 4.7272727272727275e-05,
1106
+ "loss": 0.1037,
1107
+ "step": 155
1108
+ },
1109
+ {
1110
+ "epoch": 1.4311926605504588,
1111
+ "grad_norm": 6435.69970703125,
1112
+ "learning_rate": 4.722222222222222e-05,
1113
+ "loss": 0.1036,
1114
+ "step": 156
1115
+ },
1116
+ {
1117
+ "epoch": 1.4403669724770642,
1118
+ "grad_norm": 6851.59814453125,
1119
+ "learning_rate": 4.7171717171717174e-05,
1120
+ "loss": 0.0954,
1121
+ "step": 157
1122
+ },
1123
+ {
1124
+ "epoch": 1.4495412844036697,
1125
+ "grad_norm": 7785.31640625,
1126
+ "learning_rate": 4.712121212121212e-05,
1127
+ "loss": 0.0996,
1128
+ "step": 158
1129
+ },
1130
+ {
1131
+ "epoch": 1.4587155963302751,
1132
+ "grad_norm": 6317.5576171875,
1133
+ "learning_rate": 4.7070707070707074e-05,
1134
+ "loss": 0.0977,
1135
+ "step": 159
1136
+ },
1137
+ {
1138
+ "epoch": 1.4678899082568808,
1139
+ "grad_norm": 8045.7197265625,
1140
+ "learning_rate": 4.702020202020202e-05,
1141
+ "loss": 0.1101,
1142
+ "step": 160
1143
+ },
1144
+ {
1145
+ "epoch": 1.4770642201834863,
1146
+ "grad_norm": 7386.67529296875,
1147
+ "learning_rate": 4.696969696969697e-05,
1148
+ "loss": 0.1137,
1149
+ "step": 161
1150
+ },
1151
+ {
1152
+ "epoch": 1.4862385321100917,
1153
+ "grad_norm": 8017.5791015625,
1154
+ "learning_rate": 4.6919191919191926e-05,
1155
+ "loss": 0.0959,
1156
+ "step": 162
1157
+ },
1158
+ {
1159
+ "epoch": 1.4954128440366974,
1160
+ "grad_norm": 7198.42138671875,
1161
+ "learning_rate": 4.686868686868687e-05,
1162
+ "loss": 0.1005,
1163
+ "step": 163
1164
+ },
1165
+ {
1166
+ "epoch": 1.5045871559633026,
1167
+ "grad_norm": 9139.7900390625,
1168
+ "learning_rate": 4.681818181818182e-05,
1169
+ "loss": 0.0918,
1170
+ "step": 164
1171
+ },
1172
+ {
1173
+ "epoch": 1.5137614678899083,
1174
+ "grad_norm": 6384.1640625,
1175
+ "learning_rate": 4.676767676767677e-05,
1176
+ "loss": 0.0948,
1177
+ "step": 165
1178
+ },
1179
+ {
1180
+ "epoch": 1.5229357798165137,
1181
+ "grad_norm": 5053.224609375,
1182
+ "learning_rate": 4.671717171717172e-05,
1183
+ "loss": 0.1056,
1184
+ "step": 166
1185
+ },
1186
+ {
1187
+ "epoch": 1.5321100917431192,
1188
+ "grad_norm": 9213.654296875,
1189
+ "learning_rate": 4.666666666666667e-05,
1190
+ "loss": 0.0931,
1191
+ "step": 167
1192
+ },
1193
+ {
1194
+ "epoch": 1.5412844036697249,
1195
+ "grad_norm": 7414.76171875,
1196
+ "learning_rate": 4.661616161616162e-05,
1197
+ "loss": 0.1123,
1198
+ "step": 168
1199
+ },
1200
+ {
1201
+ "epoch": 1.5504587155963303,
1202
+ "grad_norm": 6406.48583984375,
1203
+ "learning_rate": 4.656565656565657e-05,
1204
+ "loss": 0.0992,
1205
+ "step": 169
1206
+ },
1207
+ {
1208
+ "epoch": 1.5596330275229358,
1209
+ "grad_norm": 7213.58837890625,
1210
+ "learning_rate": 4.651515151515152e-05,
1211
+ "loss": 0.0952,
1212
+ "step": 170
1213
+ },
1214
+ {
1215
+ "epoch": 1.5688073394495414,
1216
+ "grad_norm": 7537.673828125,
1217
+ "learning_rate": 4.6464646464646464e-05,
1218
+ "loss": 0.0961,
1219
+ "step": 171
1220
+ },
1221
+ {
1222
+ "epoch": 1.5779816513761467,
1223
+ "grad_norm": 7876.294921875,
1224
+ "learning_rate": 4.641414141414142e-05,
1225
+ "loss": 0.1308,
1226
+ "step": 172
1227
+ },
1228
+ {
1229
+ "epoch": 1.5871559633027523,
1230
+ "grad_norm": 9550.125,
1231
+ "learning_rate": 4.636363636363636e-05,
1232
+ "loss": 0.0941,
1233
+ "step": 173
1234
+ },
1235
+ {
1236
+ "epoch": 1.5963302752293578,
1237
+ "grad_norm": 6364.3330078125,
1238
+ "learning_rate": 4.6313131313131316e-05,
1239
+ "loss": 0.0976,
1240
+ "step": 174
1241
+ },
1242
+ {
1243
+ "epoch": 1.6055045871559632,
1244
+ "grad_norm": 6976.2138671875,
1245
+ "learning_rate": 4.626262626262626e-05,
1246
+ "loss": 0.091,
1247
+ "step": 175
1248
+ },
1249
+ {
1250
+ "epoch": 1.614678899082569,
1251
+ "grad_norm": 7259.40234375,
1252
+ "learning_rate": 4.621212121212121e-05,
1253
+ "loss": 0.1096,
1254
+ "step": 176
1255
+ },
1256
+ {
1257
+ "epoch": 1.6238532110091743,
1258
+ "grad_norm": 8176.20849609375,
1259
+ "learning_rate": 4.616161616161616e-05,
1260
+ "loss": 0.1029,
1261
+ "step": 177
1262
+ },
1263
+ {
1264
+ "epoch": 1.6330275229357798,
1265
+ "grad_norm": 5628.04345703125,
1266
+ "learning_rate": 4.6111111111111115e-05,
1267
+ "loss": 0.1008,
1268
+ "step": 178
1269
+ },
1270
+ {
1271
+ "epoch": 1.6422018348623855,
1272
+ "grad_norm": 6802.91064453125,
1273
+ "learning_rate": 4.606060606060607e-05,
1274
+ "loss": 0.1014,
1275
+ "step": 179
1276
+ },
1277
+ {
1278
+ "epoch": 1.6513761467889907,
1279
+ "grad_norm": 14422.4482421875,
1280
+ "learning_rate": 4.6010101010101015e-05,
1281
+ "loss": 0.0912,
1282
+ "step": 180
1283
+ },
1284
+ {
1285
+ "epoch": 1.6605504587155964,
1286
+ "grad_norm": 8855.7744140625,
1287
+ "learning_rate": 4.595959595959596e-05,
1288
+ "loss": 0.0983,
1289
+ "step": 181
1290
+ },
1291
+ {
1292
+ "epoch": 1.6697247706422018,
1293
+ "grad_norm": 7411.93603515625,
1294
+ "learning_rate": 4.5909090909090914e-05,
1295
+ "loss": 0.1099,
1296
+ "step": 182
1297
+ },
1298
+ {
1299
+ "epoch": 1.6788990825688073,
1300
+ "grad_norm": 8138.41748046875,
1301
+ "learning_rate": 4.585858585858586e-05,
1302
+ "loss": 0.1054,
1303
+ "step": 183
1304
+ },
1305
+ {
1306
+ "epoch": 1.688073394495413,
1307
+ "grad_norm": 6157.80908203125,
1308
+ "learning_rate": 4.5808080808080814e-05,
1309
+ "loss": 0.1054,
1310
+ "step": 184
1311
+ },
1312
+ {
1313
+ "epoch": 1.6972477064220184,
1314
+ "grad_norm": 6791.07421875,
1315
+ "learning_rate": 4.575757575757576e-05,
1316
+ "loss": 0.1022,
1317
+ "step": 185
1318
+ },
1319
+ {
1320
+ "epoch": 1.7064220183486238,
1321
+ "grad_norm": 9991.0537109375,
1322
+ "learning_rate": 4.5707070707070706e-05,
1323
+ "loss": 0.1131,
1324
+ "step": 186
1325
+ },
1326
+ {
1327
+ "epoch": 1.7155963302752295,
1328
+ "grad_norm": 7240.28955078125,
1329
+ "learning_rate": 4.565656565656566e-05,
1330
+ "loss": 0.1005,
1331
+ "step": 187
1332
+ },
1333
+ {
1334
+ "epoch": 1.7247706422018347,
1335
+ "grad_norm": 6442.6259765625,
1336
+ "learning_rate": 4.5606060606060606e-05,
1337
+ "loss": 0.1015,
1338
+ "step": 188
1339
+ },
1340
+ {
1341
+ "epoch": 1.7339449541284404,
1342
+ "grad_norm": 8573.3955078125,
1343
+ "learning_rate": 4.555555555555556e-05,
1344
+ "loss": 0.1176,
1345
+ "step": 189
1346
+ },
1347
+ {
1348
+ "epoch": 1.7431192660550459,
1349
+ "grad_norm": 7330.88525390625,
1350
+ "learning_rate": 4.5505050505050505e-05,
1351
+ "loss": 0.0923,
1352
+ "step": 190
1353
+ },
1354
+ {
1355
+ "epoch": 1.7522935779816513,
1356
+ "grad_norm": 7870.916015625,
1357
+ "learning_rate": 4.545454545454546e-05,
1358
+ "loss": 0.1113,
1359
+ "step": 191
1360
+ },
1361
+ {
1362
+ "epoch": 1.761467889908257,
1363
+ "grad_norm": 7789.8564453125,
1364
+ "learning_rate": 4.5404040404040405e-05,
1365
+ "loss": 0.1062,
1366
+ "step": 192
1367
+ },
1368
+ {
1369
+ "epoch": 1.7706422018348624,
1370
+ "grad_norm": 7983.59326171875,
1371
+ "learning_rate": 4.535353535353535e-05,
1372
+ "loss": 0.1078,
1373
+ "step": 193
1374
+ },
1375
+ {
1376
+ "epoch": 1.7798165137614679,
1377
+ "grad_norm": 6533.43994140625,
1378
+ "learning_rate": 4.5303030303030304e-05,
1379
+ "loss": 0.1033,
1380
+ "step": 194
1381
+ },
1382
+ {
1383
+ "epoch": 1.7889908256880735,
1384
+ "grad_norm": 7317.1318359375,
1385
+ "learning_rate": 4.525252525252526e-05,
1386
+ "loss": 0.1106,
1387
+ "step": 195
1388
+ },
1389
+ {
1390
+ "epoch": 1.7981651376146788,
1391
+ "grad_norm": 7829.7880859375,
1392
+ "learning_rate": 4.5202020202020204e-05,
1393
+ "loss": 0.1163,
1394
+ "step": 196
1395
+ },
1396
+ {
1397
+ "epoch": 1.8073394495412844,
1398
+ "grad_norm": 7524.41357421875,
1399
+ "learning_rate": 4.515151515151516e-05,
1400
+ "loss": 0.1043,
1401
+ "step": 197
1402
+ },
1403
+ {
1404
+ "epoch": 1.81651376146789,
1405
+ "grad_norm": 7487.89892578125,
1406
+ "learning_rate": 4.51010101010101e-05,
1407
+ "loss": 0.1091,
1408
+ "step": 198
1409
+ },
1410
+ {
1411
+ "epoch": 1.8256880733944953,
1412
+ "grad_norm": 6964.3154296875,
1413
+ "learning_rate": 4.5050505050505056e-05,
1414
+ "loss": 0.0986,
1415
+ "step": 199
1416
+ },
1417
+ {
1418
+ "epoch": 1.834862385321101,
1419
+ "grad_norm": 7087.05029296875,
1420
+ "learning_rate": 4.5e-05,
1421
+ "loss": 0.0997,
1422
+ "step": 200
1423
+ },
1424
+ {
1425
+ "epoch": 1.8440366972477065,
1426
+ "grad_norm": 7148.7578125,
1427
+ "learning_rate": 4.494949494949495e-05,
1428
+ "loss": 0.0989,
1429
+ "step": 201
1430
+ },
1431
+ {
1432
+ "epoch": 1.853211009174312,
1433
+ "grad_norm": 7850.2470703125,
1434
+ "learning_rate": 4.48989898989899e-05,
1435
+ "loss": 0.0957,
1436
+ "step": 202
1437
+ },
1438
+ {
1439
+ "epoch": 1.8623853211009176,
1440
+ "grad_norm": 7752.841796875,
1441
+ "learning_rate": 4.484848484848485e-05,
1442
+ "loss": 0.1292,
1443
+ "step": 203
1444
+ },
1445
+ {
1446
+ "epoch": 1.8715596330275228,
1447
+ "grad_norm": 6098.35595703125,
1448
+ "learning_rate": 4.47979797979798e-05,
1449
+ "loss": 0.1001,
1450
+ "step": 204
1451
+ },
1452
+ {
1453
+ "epoch": 1.8807339449541285,
1454
+ "grad_norm": 7094.8173828125,
1455
+ "learning_rate": 4.474747474747475e-05,
1456
+ "loss": 0.1286,
1457
+ "step": 205
1458
+ },
1459
+ {
1460
+ "epoch": 1.889908256880734,
1461
+ "grad_norm": 7456.2265625,
1462
+ "learning_rate": 4.46969696969697e-05,
1463
+ "loss": 0.1256,
1464
+ "step": 206
1465
+ },
1466
+ {
1467
+ "epoch": 1.8990825688073394,
1468
+ "grad_norm": 8136.77197265625,
1469
+ "learning_rate": 4.464646464646465e-05,
1470
+ "loss": 0.1014,
1471
+ "step": 207
1472
+ },
1473
+ {
1474
+ "epoch": 1.908256880733945,
1475
+ "grad_norm": 7520.1650390625,
1476
+ "learning_rate": 4.4595959595959594e-05,
1477
+ "loss": 0.0919,
1478
+ "step": 208
1479
+ },
1480
+ {
1481
+ "epoch": 1.9174311926605505,
1482
+ "grad_norm": 8064.79296875,
1483
+ "learning_rate": 4.454545454545455e-05,
1484
+ "loss": 0.0985,
1485
+ "step": 209
1486
+ },
1487
+ {
1488
+ "epoch": 1.926605504587156,
1489
+ "grad_norm": 6153.19677734375,
1490
+ "learning_rate": 4.4494949494949493e-05,
1491
+ "loss": 0.1101,
1492
+ "step": 210
1493
+ },
1494
+ {
1495
+ "epoch": 1.9357798165137616,
1496
+ "grad_norm": 8089.42431640625,
1497
+ "learning_rate": 4.4444444444444447e-05,
1498
+ "loss": 0.0988,
1499
+ "step": 211
1500
+ },
1501
+ {
1502
+ "epoch": 1.9449541284403669,
1503
+ "grad_norm": 8096.140625,
1504
+ "learning_rate": 4.43939393939394e-05,
1505
+ "loss": 0.0992,
1506
+ "step": 212
1507
+ },
1508
+ {
1509
+ "epoch": 1.9541284403669725,
1510
+ "grad_norm": 7865.43798828125,
1511
+ "learning_rate": 4.4343434343434346e-05,
1512
+ "loss": 0.0945,
1513
+ "step": 213
1514
+ },
1515
+ {
1516
+ "epoch": 1.963302752293578,
1517
+ "grad_norm": 6380.67724609375,
1518
+ "learning_rate": 4.42929292929293e-05,
1519
+ "loss": 0.1073,
1520
+ "step": 214
1521
+ },
1522
+ {
1523
+ "epoch": 1.9724770642201834,
1524
+ "grad_norm": 8548.4345703125,
1525
+ "learning_rate": 4.4242424242424246e-05,
1526
+ "loss": 0.1121,
1527
+ "step": 215
1528
+ },
1529
+ {
1530
+ "epoch": 1.981651376146789,
1531
+ "grad_norm": 8465.8466796875,
1532
+ "learning_rate": 4.41919191919192e-05,
1533
+ "loss": 0.1078,
1534
+ "step": 216
1535
+ },
1536
+ {
1537
+ "epoch": 1.9908256880733946,
1538
+ "grad_norm": 8595.419921875,
1539
+ "learning_rate": 4.4141414141414145e-05,
1540
+ "loss": 0.0924,
1541
+ "step": 217
1542
+ },
1543
+ {
1544
+ "epoch": 2.0,
1545
+ "grad_norm": 22229.833984375,
1546
+ "learning_rate": 4.409090909090909e-05,
1547
+ "loss": 0.0785,
1548
+ "step": 218
1549
+ },
1550
+ {
1551
+ "epoch": 2.0,
1552
+ "eval_f1_macro": 0.0,
1553
+ "eval_f1_micro": 0.0,
1554
+ "eval_loss": 0.10520372539758682,
1555
+ "eval_precision": 0.0,
1556
+ "eval_recall": 0.0,
1557
+ "eval_runtime": 0.8127,
1558
+ "eval_samples_per_second": 355.611,
1559
+ "eval_steps_per_second": 15.996,
1560
+ "step": 218
1561
+ },
1562
+ {
1563
+ "epoch": 2.0091743119266057,
1564
+ "grad_norm": 8562.740234375,
1565
+ "learning_rate": 4.4040404040404044e-05,
1566
+ "loss": 0.1152,
1567
+ "step": 219
1568
+ },
1569
+ {
1570
+ "epoch": 2.018348623853211,
1571
+ "grad_norm": 7770.8720703125,
1572
+ "learning_rate": 4.398989898989899e-05,
1573
+ "loss": 0.0957,
1574
+ "step": 220
1575
+ },
1576
+ {
1577
+ "epoch": 2.0275229357798166,
1578
+ "grad_norm": 6170.166015625,
1579
+ "learning_rate": 4.3939393939393944e-05,
1580
+ "loss": 0.1027,
1581
+ "step": 221
1582
+ },
1583
+ {
1584
+ "epoch": 2.036697247706422,
1585
+ "grad_norm": 10314.91796875,
1586
+ "learning_rate": 4.388888888888889e-05,
1587
+ "loss": 0.1166,
1588
+ "step": 222
1589
+ },
1590
+ {
1591
+ "epoch": 2.0458715596330275,
1592
+ "grad_norm": 7144.62109375,
1593
+ "learning_rate": 4.383838383838384e-05,
1594
+ "loss": 0.1065,
1595
+ "step": 223
1596
+ },
1597
+ {
1598
+ "epoch": 2.055045871559633,
1599
+ "grad_norm": 7398.2109375,
1600
+ "learning_rate": 4.378787878787879e-05,
1601
+ "loss": 0.1165,
1602
+ "step": 224
1603
+ },
1604
+ {
1605
+ "epoch": 2.0642201834862384,
1606
+ "grad_norm": 7307.375,
1607
+ "learning_rate": 4.3737373737373736e-05,
1608
+ "loss": 0.0941,
1609
+ "step": 225
1610
+ },
1611
+ {
1612
+ "epoch": 2.073394495412844,
1613
+ "grad_norm": 8046.33837890625,
1614
+ "learning_rate": 4.368686868686869e-05,
1615
+ "loss": 0.1092,
1616
+ "step": 226
1617
+ },
1618
+ {
1619
+ "epoch": 2.0825688073394497,
1620
+ "grad_norm": 8320.74609375,
1621
+ "learning_rate": 4.3636363636363636e-05,
1622
+ "loss": 0.0983,
1623
+ "step": 227
1624
+ },
1625
+ {
1626
+ "epoch": 2.091743119266055,
1627
+ "grad_norm": 10485.9384765625,
1628
+ "learning_rate": 4.358585858585859e-05,
1629
+ "loss": 0.0981,
1630
+ "step": 228
1631
+ },
1632
+ {
1633
+ "epoch": 2.1009174311926606,
1634
+ "grad_norm": 7908.02734375,
1635
+ "learning_rate": 4.3535353535353535e-05,
1636
+ "loss": 0.1021,
1637
+ "step": 229
1638
+ },
1639
+ {
1640
+ "epoch": 2.1100917431192663,
1641
+ "grad_norm": 7627.345703125,
1642
+ "learning_rate": 4.348484848484849e-05,
1643
+ "loss": 0.106,
1644
+ "step": 230
1645
+ },
1646
+ {
1647
+ "epoch": 2.1192660550458715,
1648
+ "grad_norm": 7474.5263671875,
1649
+ "learning_rate": 4.343434343434344e-05,
1650
+ "loss": 0.1156,
1651
+ "step": 231
1652
+ },
1653
+ {
1654
+ "epoch": 2.128440366972477,
1655
+ "grad_norm": 6541.48828125,
1656
+ "learning_rate": 4.338383838383839e-05,
1657
+ "loss": 0.0974,
1658
+ "step": 232
1659
+ },
1660
+ {
1661
+ "epoch": 2.1376146788990824,
1662
+ "grad_norm": 10176.2197265625,
1663
+ "learning_rate": 4.3333333333333334e-05,
1664
+ "loss": 0.1172,
1665
+ "step": 233
1666
+ },
1667
+ {
1668
+ "epoch": 2.146788990825688,
1669
+ "grad_norm": 10230.0478515625,
1670
+ "learning_rate": 4.328282828282829e-05,
1671
+ "loss": 0.1089,
1672
+ "step": 234
1673
+ },
1674
+ {
1675
+ "epoch": 2.1559633027522938,
1676
+ "grad_norm": 7409.98583984375,
1677
+ "learning_rate": 4.3232323232323234e-05,
1678
+ "loss": 0.095,
1679
+ "step": 235
1680
+ },
1681
+ {
1682
+ "epoch": 2.165137614678899,
1683
+ "grad_norm": 9656.2216796875,
1684
+ "learning_rate": 4.318181818181819e-05,
1685
+ "loss": 0.0862,
1686
+ "step": 236
1687
+ },
1688
+ {
1689
+ "epoch": 2.1743119266055047,
1690
+ "grad_norm": 10004.51171875,
1691
+ "learning_rate": 4.313131313131313e-05,
1692
+ "loss": 0.1019,
1693
+ "step": 237
1694
+ },
1695
+ {
1696
+ "epoch": 2.18348623853211,
1697
+ "grad_norm": 10041.599609375,
1698
+ "learning_rate": 4.308080808080808e-05,
1699
+ "loss": 0.0828,
1700
+ "step": 238
1701
+ },
1702
+ {
1703
+ "epoch": 2.1926605504587156,
1704
+ "grad_norm": 10838.517578125,
1705
+ "learning_rate": 4.303030303030303e-05,
1706
+ "loss": 0.1093,
1707
+ "step": 239
1708
+ },
1709
+ {
1710
+ "epoch": 2.2018348623853212,
1711
+ "grad_norm": 7099.97265625,
1712
+ "learning_rate": 4.297979797979798e-05,
1713
+ "loss": 0.0962,
1714
+ "step": 240
1715
+ },
1716
+ {
1717
+ "epoch": 2.2110091743119265,
1718
+ "grad_norm": 8438.68359375,
1719
+ "learning_rate": 4.292929292929293e-05,
1720
+ "loss": 0.1095,
1721
+ "step": 241
1722
+ },
1723
+ {
1724
+ "epoch": 2.220183486238532,
1725
+ "grad_norm": 7405.43359375,
1726
+ "learning_rate": 4.287878787878788e-05,
1727
+ "loss": 0.1035,
1728
+ "step": 242
1729
+ },
1730
+ {
1731
+ "epoch": 2.229357798165138,
1732
+ "grad_norm": 9133.9833984375,
1733
+ "learning_rate": 4.282828282828283e-05,
1734
+ "loss": 0.0925,
1735
+ "step": 243
1736
+ },
1737
+ {
1738
+ "epoch": 2.238532110091743,
1739
+ "grad_norm": 7998.142578125,
1740
+ "learning_rate": 4.277777777777778e-05,
1741
+ "loss": 0.0998,
1742
+ "step": 244
1743
+ },
1744
+ {
1745
+ "epoch": 2.2477064220183487,
1746
+ "grad_norm": 9984.765625,
1747
+ "learning_rate": 4.2727272727272724e-05,
1748
+ "loss": 0.1027,
1749
+ "step": 245
1750
+ },
1751
+ {
1752
+ "epoch": 2.2568807339449544,
1753
+ "grad_norm": 8999.0556640625,
1754
+ "learning_rate": 4.267676767676768e-05,
1755
+ "loss": 0.1177,
1756
+ "step": 246
1757
+ },
1758
+ {
1759
+ "epoch": 2.2660550458715596,
1760
+ "grad_norm": 7668.86962890625,
1761
+ "learning_rate": 4.262626262626263e-05,
1762
+ "loss": 0.0946,
1763
+ "step": 247
1764
+ },
1765
+ {
1766
+ "epoch": 2.2752293577981653,
1767
+ "grad_norm": 7349.47998046875,
1768
+ "learning_rate": 4.257575757575758e-05,
1769
+ "loss": 0.1066,
1770
+ "step": 248
1771
+ },
1772
+ {
1773
+ "epoch": 2.2844036697247705,
1774
+ "grad_norm": 8360.69921875,
1775
+ "learning_rate": 4.252525252525253e-05,
1776
+ "loss": 0.1062,
1777
+ "step": 249
1778
+ },
1779
+ {
1780
+ "epoch": 2.293577981651376,
1781
+ "grad_norm": 8717.70703125,
1782
+ "learning_rate": 4.2474747474747476e-05,
1783
+ "loss": 0.0907,
1784
+ "step": 250
1785
+ },
1786
+ {
1787
+ "epoch": 2.302752293577982,
1788
+ "grad_norm": 8706.5791015625,
1789
+ "learning_rate": 4.242424242424243e-05,
1790
+ "loss": 0.0973,
1791
+ "step": 251
1792
+ },
1793
+ {
1794
+ "epoch": 2.311926605504587,
1795
+ "grad_norm": 8500.1220703125,
1796
+ "learning_rate": 4.2373737373737376e-05,
1797
+ "loss": 0.1022,
1798
+ "step": 252
1799
+ },
1800
+ {
1801
+ "epoch": 2.3211009174311927,
1802
+ "grad_norm": 10551.7021484375,
1803
+ "learning_rate": 4.232323232323233e-05,
1804
+ "loss": 0.0979,
1805
+ "step": 253
1806
+ },
1807
+ {
1808
+ "epoch": 2.330275229357798,
1809
+ "grad_norm": 8305.2802734375,
1810
+ "learning_rate": 4.2272727272727275e-05,
1811
+ "loss": 0.0926,
1812
+ "step": 254
1813
+ },
1814
+ {
1815
+ "epoch": 2.3394495412844036,
1816
+ "grad_norm": 10128.9423828125,
1817
+ "learning_rate": 4.222222222222222e-05,
1818
+ "loss": 0.1018,
1819
+ "step": 255
1820
+ },
1821
+ {
1822
+ "epoch": 2.3486238532110093,
1823
+ "grad_norm": 9000.6689453125,
1824
+ "learning_rate": 4.2171717171717175e-05,
1825
+ "loss": 0.091,
1826
+ "step": 256
1827
+ },
1828
+ {
1829
+ "epoch": 2.3577981651376145,
1830
+ "grad_norm": 8849.099609375,
1831
+ "learning_rate": 4.212121212121212e-05,
1832
+ "loss": 0.0993,
1833
+ "step": 257
1834
+ },
1835
+ {
1836
+ "epoch": 2.36697247706422,
1837
+ "grad_norm": 6667.78564453125,
1838
+ "learning_rate": 4.2070707070707074e-05,
1839
+ "loss": 0.1044,
1840
+ "step": 258
1841
+ },
1842
+ {
1843
+ "epoch": 2.376146788990826,
1844
+ "grad_norm": 7288.4970703125,
1845
+ "learning_rate": 4.202020202020202e-05,
1846
+ "loss": 0.0966,
1847
+ "step": 259
1848
+ },
1849
+ {
1850
+ "epoch": 2.385321100917431,
1851
+ "grad_norm": 7155.8310546875,
1852
+ "learning_rate": 4.196969696969697e-05,
1853
+ "loss": 0.1018,
1854
+ "step": 260
1855
+ },
1856
+ {
1857
+ "epoch": 2.3944954128440368,
1858
+ "grad_norm": 6328.56396484375,
1859
+ "learning_rate": 4.191919191919192e-05,
1860
+ "loss": 0.0862,
1861
+ "step": 261
1862
+ },
1863
+ {
1864
+ "epoch": 2.4036697247706424,
1865
+ "grad_norm": 8594.3603515625,
1866
+ "learning_rate": 4.1868686868686866e-05,
1867
+ "loss": 0.1028,
1868
+ "step": 262
1869
+ },
1870
+ {
1871
+ "epoch": 2.4128440366972477,
1872
+ "grad_norm": 8077.544921875,
1873
+ "learning_rate": 4.181818181818182e-05,
1874
+ "loss": 0.1044,
1875
+ "step": 263
1876
+ },
1877
+ {
1878
+ "epoch": 2.4220183486238533,
1879
+ "grad_norm": 6332.14453125,
1880
+ "learning_rate": 4.176767676767677e-05,
1881
+ "loss": 0.0902,
1882
+ "step": 264
1883
+ },
1884
+ {
1885
+ "epoch": 2.4311926605504586,
1886
+ "grad_norm": 7677.5009765625,
1887
+ "learning_rate": 4.171717171717172e-05,
1888
+ "loss": 0.098,
1889
+ "step": 265
1890
+ },
1891
+ {
1892
+ "epoch": 2.4403669724770642,
1893
+ "grad_norm": 7953.89501953125,
1894
+ "learning_rate": 4.166666666666667e-05,
1895
+ "loss": 0.0809,
1896
+ "step": 266
1897
+ },
1898
+ {
1899
+ "epoch": 2.44954128440367,
1900
+ "grad_norm": 6055.19287109375,
1901
+ "learning_rate": 4.161616161616162e-05,
1902
+ "loss": 0.0814,
1903
+ "step": 267
1904
+ },
1905
+ {
1906
+ "epoch": 2.458715596330275,
1907
+ "grad_norm": 6427.98046875,
1908
+ "learning_rate": 4.156565656565657e-05,
1909
+ "loss": 0.0994,
1910
+ "step": 268
1911
+ },
1912
+ {
1913
+ "epoch": 2.467889908256881,
1914
+ "grad_norm": 8880.2197265625,
1915
+ "learning_rate": 4.151515151515152e-05,
1916
+ "loss": 0.0804,
1917
+ "step": 269
1918
+ },
1919
+ {
1920
+ "epoch": 2.477064220183486,
1921
+ "grad_norm": 7500.67578125,
1922
+ "learning_rate": 4.1464646464646464e-05,
1923
+ "loss": 0.0997,
1924
+ "step": 270
1925
+ },
1926
+ {
1927
+ "epoch": 2.4862385321100917,
1928
+ "grad_norm": 7541.759765625,
1929
+ "learning_rate": 4.141414141414142e-05,
1930
+ "loss": 0.0901,
1931
+ "step": 271
1932
+ },
1933
+ {
1934
+ "epoch": 2.4954128440366974,
1935
+ "grad_norm": 6423.1025390625,
1936
+ "learning_rate": 4.1363636363636364e-05,
1937
+ "loss": 0.089,
1938
+ "step": 272
1939
+ },
1940
+ {
1941
+ "epoch": 2.5045871559633026,
1942
+ "grad_norm": 6384.140625,
1943
+ "learning_rate": 4.131313131313132e-05,
1944
+ "loss": 0.0848,
1945
+ "step": 273
1946
+ },
1947
+ {
1948
+ "epoch": 2.5137614678899083,
1949
+ "grad_norm": 6418.369140625,
1950
+ "learning_rate": 4.126262626262626e-05,
1951
+ "loss": 0.0845,
1952
+ "step": 274
1953
+ },
1954
+ {
1955
+ "epoch": 2.522935779816514,
1956
+ "grad_norm": 6441.12060546875,
1957
+ "learning_rate": 4.1212121212121216e-05,
1958
+ "loss": 0.0855,
1959
+ "step": 275
1960
+ },
1961
+ {
1962
+ "epoch": 2.532110091743119,
1963
+ "grad_norm": 6891.9501953125,
1964
+ "learning_rate": 4.116161616161616e-05,
1965
+ "loss": 0.0798,
1966
+ "step": 276
1967
+ },
1968
+ {
1969
+ "epoch": 2.541284403669725,
1970
+ "grad_norm": 8036.39306640625,
1971
+ "learning_rate": 4.111111111111111e-05,
1972
+ "loss": 0.1083,
1973
+ "step": 277
1974
+ },
1975
+ {
1976
+ "epoch": 2.5504587155963305,
1977
+ "grad_norm": 7460.015625,
1978
+ "learning_rate": 4.106060606060606e-05,
1979
+ "loss": 0.0934,
1980
+ "step": 278
1981
+ },
1982
+ {
1983
+ "epoch": 2.5596330275229358,
1984
+ "grad_norm": 7343.33154296875,
1985
+ "learning_rate": 4.101010101010101e-05,
1986
+ "loss": 0.0951,
1987
+ "step": 279
1988
+ },
1989
+ {
1990
+ "epoch": 2.5688073394495414,
1991
+ "grad_norm": 9859.599609375,
1992
+ "learning_rate": 4.095959595959596e-05,
1993
+ "loss": 0.0849,
1994
+ "step": 280
1995
+ },
1996
+ {
1997
+ "epoch": 2.5779816513761467,
1998
+ "grad_norm": 8394.8212890625,
1999
+ "learning_rate": 4.0909090909090915e-05,
2000
+ "loss": 0.079,
2001
+ "step": 281
2002
+ },
2003
+ {
2004
+ "epoch": 2.5871559633027523,
2005
+ "grad_norm": 8307.279296875,
2006
+ "learning_rate": 4.085858585858586e-05,
2007
+ "loss": 0.0985,
2008
+ "step": 282
2009
+ },
2010
+ {
2011
+ "epoch": 2.5963302752293576,
2012
+ "grad_norm": 7505.14794921875,
2013
+ "learning_rate": 4.0808080808080814e-05,
2014
+ "loss": 0.0932,
2015
+ "step": 283
2016
+ },
2017
+ {
2018
+ "epoch": 2.6055045871559632,
2019
+ "grad_norm": 6129.16552734375,
2020
+ "learning_rate": 4.075757575757576e-05,
2021
+ "loss": 0.0797,
2022
+ "step": 284
2023
+ },
2024
+ {
2025
+ "epoch": 2.614678899082569,
2026
+ "grad_norm": 6896.5908203125,
2027
+ "learning_rate": 4.070707070707071e-05,
2028
+ "loss": 0.0891,
2029
+ "step": 285
2030
+ },
2031
+ {
2032
+ "epoch": 2.623853211009174,
2033
+ "grad_norm": 8686.771484375,
2034
+ "learning_rate": 4.065656565656566e-05,
2035
+ "loss": 0.0838,
2036
+ "step": 286
2037
+ },
2038
+ {
2039
+ "epoch": 2.63302752293578,
2040
+ "grad_norm": 6954.4072265625,
2041
+ "learning_rate": 4.0606060606060606e-05,
2042
+ "loss": 0.0976,
2043
+ "step": 287
2044
+ },
2045
+ {
2046
+ "epoch": 2.6422018348623855,
2047
+ "grad_norm": 8336.8720703125,
2048
+ "learning_rate": 4.055555555555556e-05,
2049
+ "loss": 0.1339,
2050
+ "step": 288
2051
+ },
2052
+ {
2053
+ "epoch": 2.6513761467889907,
2054
+ "grad_norm": 6729.21044921875,
2055
+ "learning_rate": 4.0505050505050506e-05,
2056
+ "loss": 0.0897,
2057
+ "step": 289
2058
+ },
2059
+ {
2060
+ "epoch": 2.6605504587155964,
2061
+ "grad_norm": 8650.7490234375,
2062
+ "learning_rate": 4.045454545454546e-05,
2063
+ "loss": 0.1034,
2064
+ "step": 290
2065
+ },
2066
+ {
2067
+ "epoch": 2.669724770642202,
2068
+ "grad_norm": 8123.34228515625,
2069
+ "learning_rate": 4.0404040404040405e-05,
2070
+ "loss": 0.0752,
2071
+ "step": 291
2072
+ },
2073
+ {
2074
+ "epoch": 2.6788990825688073,
2075
+ "grad_norm": 7077.0859375,
2076
+ "learning_rate": 4.035353535353535e-05,
2077
+ "loss": 0.0889,
2078
+ "step": 292
2079
+ },
2080
+ {
2081
+ "epoch": 2.688073394495413,
2082
+ "grad_norm": 8160.107421875,
2083
+ "learning_rate": 4.0303030303030305e-05,
2084
+ "loss": 0.1067,
2085
+ "step": 293
2086
+ },
2087
+ {
2088
+ "epoch": 2.6972477064220186,
2089
+ "grad_norm": 11080.52734375,
2090
+ "learning_rate": 4.025252525252525e-05,
2091
+ "loss": 0.0984,
2092
+ "step": 294
2093
+ },
2094
+ {
2095
+ "epoch": 2.706422018348624,
2096
+ "grad_norm": 8655.8173828125,
2097
+ "learning_rate": 4.0202020202020204e-05,
2098
+ "loss": 0.0977,
2099
+ "step": 295
2100
+ },
2101
+ {
2102
+ "epoch": 2.7155963302752295,
2103
+ "grad_norm": 9417.9384765625,
2104
+ "learning_rate": 4.015151515151515e-05,
2105
+ "loss": 0.0863,
2106
+ "step": 296
2107
+ },
2108
+ {
2109
+ "epoch": 2.7247706422018347,
2110
+ "grad_norm": 7852.45361328125,
2111
+ "learning_rate": 4.01010101010101e-05,
2112
+ "loss": 0.0865,
2113
+ "step": 297
2114
+ },
2115
+ {
2116
+ "epoch": 2.7339449541284404,
2117
+ "grad_norm": 8051.74609375,
2118
+ "learning_rate": 4.005050505050506e-05,
2119
+ "loss": 0.11,
2120
+ "step": 298
2121
+ },
2122
+ {
2123
+ "epoch": 2.7431192660550456,
2124
+ "grad_norm": 7042.64013671875,
2125
+ "learning_rate": 4e-05,
2126
+ "loss": 0.0805,
2127
+ "step": 299
2128
+ },
2129
+ {
2130
+ "epoch": 2.7522935779816513,
2131
+ "grad_norm": 7486.0947265625,
2132
+ "learning_rate": 3.9949494949494956e-05,
2133
+ "loss": 0.0999,
2134
+ "step": 300
2135
+ },
2136
+ {
2137
+ "epoch": 2.761467889908257,
2138
+ "grad_norm": 8937.4580078125,
2139
+ "learning_rate": 3.98989898989899e-05,
2140
+ "loss": 0.0936,
2141
+ "step": 301
2142
+ },
2143
+ {
2144
+ "epoch": 2.770642201834862,
2145
+ "grad_norm": 7193.6708984375,
2146
+ "learning_rate": 3.984848484848485e-05,
2147
+ "loss": 0.0743,
2148
+ "step": 302
2149
+ },
2150
+ {
2151
+ "epoch": 2.779816513761468,
2152
+ "grad_norm": 7543.75341796875,
2153
+ "learning_rate": 3.97979797979798e-05,
2154
+ "loss": 0.0994,
2155
+ "step": 303
2156
+ },
2157
+ {
2158
+ "epoch": 2.7889908256880735,
2159
+ "grad_norm": 8849.693359375,
2160
+ "learning_rate": 3.974747474747475e-05,
2161
+ "loss": 0.0891,
2162
+ "step": 304
2163
+ },
2164
+ {
2165
+ "epoch": 2.7981651376146788,
2166
+ "grad_norm": 6587.58837890625,
2167
+ "learning_rate": 3.96969696969697e-05,
2168
+ "loss": 0.0921,
2169
+ "step": 305
2170
+ },
2171
+ {
2172
+ "epoch": 2.8073394495412844,
2173
+ "grad_norm": 7597.34521484375,
2174
+ "learning_rate": 3.964646464646465e-05,
2175
+ "loss": 0.079,
2176
+ "step": 306
2177
+ },
2178
+ {
2179
+ "epoch": 2.81651376146789,
2180
+ "grad_norm": 8688.705078125,
2181
+ "learning_rate": 3.9595959595959594e-05,
2182
+ "loss": 0.1014,
2183
+ "step": 307
2184
+ },
2185
+ {
2186
+ "epoch": 2.8256880733944953,
2187
+ "grad_norm": 14665.2021484375,
2188
+ "learning_rate": 3.954545454545455e-05,
2189
+ "loss": 0.1354,
2190
+ "step": 308
2191
+ },
2192
+ {
2193
+ "epoch": 2.834862385321101,
2194
+ "grad_norm": 10845.7890625,
2195
+ "learning_rate": 3.9494949494949494e-05,
2196
+ "loss": 0.1076,
2197
+ "step": 309
2198
+ },
2199
+ {
2200
+ "epoch": 2.8440366972477067,
2201
+ "grad_norm": 10935.6669921875,
2202
+ "learning_rate": 3.944444444444445e-05,
2203
+ "loss": 0.1054,
2204
+ "step": 310
2205
+ },
2206
+ {
2207
+ "epoch": 2.853211009174312,
2208
+ "grad_norm": 11541.736328125,
2209
+ "learning_rate": 3.939393939393939e-05,
2210
+ "loss": 0.0943,
2211
+ "step": 311
2212
+ },
2213
+ {
2214
+ "epoch": 2.8623853211009176,
2215
+ "grad_norm": 6934.3125,
2216
+ "learning_rate": 3.9343434343434346e-05,
2217
+ "loss": 0.1018,
2218
+ "step": 312
2219
+ },
2220
+ {
2221
+ "epoch": 2.871559633027523,
2222
+ "grad_norm": 8040.87939453125,
2223
+ "learning_rate": 3.929292929292929e-05,
2224
+ "loss": 0.0926,
2225
+ "step": 313
2226
+ },
2227
+ {
2228
+ "epoch": 2.8807339449541285,
2229
+ "grad_norm": 6876.23876953125,
2230
+ "learning_rate": 3.924242424242424e-05,
2231
+ "loss": 0.0806,
2232
+ "step": 314
2233
+ },
2234
+ {
2235
+ "epoch": 2.8899082568807337,
2236
+ "grad_norm": 8227.6435546875,
2237
+ "learning_rate": 3.91919191919192e-05,
2238
+ "loss": 0.0888,
2239
+ "step": 315
2240
+ },
2241
+ {
2242
+ "epoch": 2.8990825688073394,
2243
+ "grad_norm": 9191.8701171875,
2244
+ "learning_rate": 3.9141414141414145e-05,
2245
+ "loss": 0.0982,
2246
+ "step": 316
2247
+ },
2248
+ {
2249
+ "epoch": 2.908256880733945,
2250
+ "grad_norm": 8231.9013671875,
2251
+ "learning_rate": 3.909090909090909e-05,
2252
+ "loss": 0.0851,
2253
+ "step": 317
2254
+ },
2255
+ {
2256
+ "epoch": 2.9174311926605503,
2257
+ "grad_norm": 8839.322265625,
2258
+ "learning_rate": 3.9040404040404045e-05,
2259
+ "loss": 0.0757,
2260
+ "step": 318
2261
+ },
2262
+ {
2263
+ "epoch": 2.926605504587156,
2264
+ "grad_norm": 8270.8525390625,
2265
+ "learning_rate": 3.898989898989899e-05,
2266
+ "loss": 0.1015,
2267
+ "step": 319
2268
+ },
2269
+ {
2270
+ "epoch": 2.9357798165137616,
2271
+ "grad_norm": 12873.2529296875,
2272
+ "learning_rate": 3.8939393939393944e-05,
2273
+ "loss": 0.0871,
2274
+ "step": 320
2275
+ },
2276
+ {
2277
+ "epoch": 2.944954128440367,
2278
+ "grad_norm": 7598.33837890625,
2279
+ "learning_rate": 3.888888888888889e-05,
2280
+ "loss": 0.0884,
2281
+ "step": 321
2282
+ },
2283
+ {
2284
+ "epoch": 2.9541284403669725,
2285
+ "grad_norm": 8909.1279296875,
2286
+ "learning_rate": 3.8838383838383844e-05,
2287
+ "loss": 0.102,
2288
+ "step": 322
2289
+ },
2290
+ {
2291
+ "epoch": 2.963302752293578,
2292
+ "grad_norm": 9611.2451171875,
2293
+ "learning_rate": 3.878787878787879e-05,
2294
+ "loss": 0.094,
2295
+ "step": 323
2296
+ },
2297
+ {
2298
+ "epoch": 2.9724770642201834,
2299
+ "grad_norm": 8551.978515625,
2300
+ "learning_rate": 3.8737373737373737e-05,
2301
+ "loss": 0.0857,
2302
+ "step": 324
2303
+ },
2304
+ {
2305
+ "epoch": 2.981651376146789,
2306
+ "grad_norm": 7810.45703125,
2307
+ "learning_rate": 3.868686868686869e-05,
2308
+ "loss": 0.0854,
2309
+ "step": 325
2310
+ },
2311
+ {
2312
+ "epoch": 2.9908256880733948,
2313
+ "grad_norm": 8891.1826171875,
2314
+ "learning_rate": 3.8636363636363636e-05,
2315
+ "loss": 0.101,
2316
+ "step": 326
2317
+ },
2318
+ {
2319
+ "epoch": 3.0,
2320
+ "grad_norm": 16918.001953125,
2321
+ "learning_rate": 3.858585858585859e-05,
2322
+ "loss": 0.0625,
2323
+ "step": 327
2324
+ },
2325
+ {
2326
+ "epoch": 3.0,
2327
+ "eval_f1_macro": 0.05806976361668135,
2328
+ "eval_f1_micro": 0.37575757575757573,
2329
+ "eval_loss": 0.0906587690114975,
2330
+ "eval_precision": 0.7380952380952381,
2331
+ "eval_recall": 0.25203252032520324,
2332
+ "eval_runtime": 0.8046,
2333
+ "eval_samples_per_second": 359.182,
2334
+ "eval_steps_per_second": 16.157,
2335
+ "step": 327
2336
+ }
2337
+ ],
2338
+ "logging_steps": 1,
2339
+ "max_steps": 1090,
2340
+ "num_input_tokens_seen": 0,
2341
+ "num_train_epochs": 10,
2342
+ "save_steps": 500,
2343
+ "stateful_callbacks": {
2344
+ "TrainerControl": {
2345
+ "args": {
2346
+ "should_epoch_stop": false,
2347
+ "should_evaluate": false,
2348
+ "should_log": false,
2349
+ "should_save": true,
2350
+ "should_training_stop": false
2351
+ },
2352
+ "attributes": {}
2353
+ }
2354
+ },
2355
+ "total_flos": 2049202335651840.0,
2356
+ "train_batch_size": 24,
2357
+ "trial_name": null,
2358
+ "trial_params": null
2359
+ }
checkpoint-327/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edeef274e762b6821b55d94e571eef919508876a6f05e915f5b9e4cc058a5c62
3
+ size 5368
checkpoint-327/vocab.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-436/config.json ADDED
@@ -0,0 +1,130 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "gradient_checkpointing": false,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.1,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "T1003.001",
13
+ "1": "T1005",
14
+ "2": "T1012",
15
+ "3": "T1016",
16
+ "4": "T1021.001",
17
+ "5": "T1027",
18
+ "6": "T1033",
19
+ "7": "T1036.005",
20
+ "8": "T1041",
21
+ "9": "T1047",
22
+ "10": "T1053.005",
23
+ "11": "T1055",
24
+ "12": "T1056.001",
25
+ "13": "T1057",
26
+ "14": "T1059.003",
27
+ "15": "T1068",
28
+ "16": "T1070.004",
29
+ "17": "T1071.001",
30
+ "18": "T1072",
31
+ "19": "T1074.001",
32
+ "20": "T1078",
33
+ "21": "T1082",
34
+ "22": "T1083",
35
+ "23": "T1090",
36
+ "24": "T1095",
37
+ "25": "T1105",
38
+ "26": "T1106",
39
+ "27": "T1110",
40
+ "28": "T1112",
41
+ "29": "T1113",
42
+ "30": "T1140",
43
+ "31": "T1190",
44
+ "32": "T1204.002",
45
+ "33": "T1210",
46
+ "34": "T1218.011",
47
+ "35": "T1219",
48
+ "36": "T1484.001",
49
+ "37": "T1518.001",
50
+ "38": "T1543.003",
51
+ "39": "T1547.001",
52
+ "40": "T1548.002",
53
+ "41": "T1552.001",
54
+ "42": "T1557.001",
55
+ "43": "T1562.001",
56
+ "44": "T1564.001",
57
+ "45": "T1566.001",
58
+ "46": "T1569.002",
59
+ "47": "T1570",
60
+ "48": "T1573.001",
61
+ "49": "T1574.002"
62
+ },
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 3072,
65
+ "label2id": {
66
+ "T1003.001": 0,
67
+ "T1005": 1,
68
+ "T1012": 2,
69
+ "T1016": 3,
70
+ "T1021.001": 4,
71
+ "T1027": 5,
72
+ "T1033": 6,
73
+ "T1036.005": 7,
74
+ "T1041": 8,
75
+ "T1047": 9,
76
+ "T1053.005": 10,
77
+ "T1055": 11,
78
+ "T1056.001": 12,
79
+ "T1057": 13,
80
+ "T1059.003": 14,
81
+ "T1068": 15,
82
+ "T1070.004": 16,
83
+ "T1071.001": 17,
84
+ "T1072": 18,
85
+ "T1074.001": 19,
86
+ "T1078": 20,
87
+ "T1082": 21,
88
+ "T1083": 22,
89
+ "T1090": 23,
90
+ "T1095": 24,
91
+ "T1105": 25,
92
+ "T1106": 26,
93
+ "T1110": 27,
94
+ "T1112": 28,
95
+ "T1113": 29,
96
+ "T1140": 30,
97
+ "T1190": 31,
98
+ "T1204.002": 32,
99
+ "T1210": 33,
100
+ "T1218.011": 34,
101
+ "T1219": 35,
102
+ "T1484.001": 36,
103
+ "T1518.001": 37,
104
+ "T1543.003": 38,
105
+ "T1547.001": 39,
106
+ "T1548.002": 40,
107
+ "T1552.001": 41,
108
+ "T1557.001": 42,
109
+ "T1562.001": 43,
110
+ "T1564.001": 44,
111
+ "T1566.001": 45,
112
+ "T1569.002": 46,
113
+ "T1570": 47,
114
+ "T1573.001": 48,
115
+ "T1574.002": 49
116
+ },
117
+ "layer_norm_eps": 1e-12,
118
+ "max_position_embeddings": 512,
119
+ "model_type": "bert",
120
+ "num_attention_heads": 12,
121
+ "num_hidden_layers": 12,
122
+ "pad_token_id": 0,
123
+ "position_embedding_type": "absolute",
124
+ "problem_type": "multi_label_classification",
125
+ "torch_dtype": "float32",
126
+ "transformers_version": "4.55.2",
127
+ "type_vocab_size": 2,
128
+ "use_cache": true,
129
+ "vocab_size": 30522
130
+ }
checkpoint-436/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78bdeb866be6f65bd166a90f93414c0266e323c04d8bb23ee426e09eff976a52
3
+ size 438106296