Hugging Face's logo

PhaaNe
/
checkpoint

Safetensors
Model card Files
xet
 Community
PhaaNe commited on
Commit
165710c
·
verified ·
1 Parent(s): f8f62d7

Add files using upload-large-folder tool

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-1309/special_tokens_map.json +51 -0
  2. ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-1309/tokenizer.json +0 -0
  3. ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-1309/tokenizer_config.json +55 -0
  4. ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-1309/trainer_state.json +484 -0
  5. ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748/special_tokens_map.json +51 -0
  6. ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748/tokenizer.json +0 -0
  7. ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748/tokenizer_config.json +55 -0
  8. ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748/trainer_state.json +290 -0
  9. ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-1528/special_tokens_map.json +51 -0
  10. ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-1528/tokenizer.json +0 -0
  11. ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-1528/tokenizer_config.json +55 -0
  12. ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-1528/trainer_state.json +558 -0
  13. ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955/special_tokens_map.json +51 -0
  14. ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955/tokenizer.json +0 -0
  15. ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955/tokenizer_config.json +55 -0
  16. ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955/trainer_state.json +364 -0
  17. ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820/special_tokens_map.json +51 -0
  18. ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820/tokenizer.json +0 -0
  19. ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820/tokenizer_config.json +55 -0
  20. ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820/trainer_state.json +657 -0
  21. ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-2366/special_tokens_map.json +51 -0
  22. ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-2366/tokenizer.json +0 -0
  23. ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-2366/tokenizer_config.json +55 -0
  24. ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2800/special_tokens_map.json +51 -0
  25. ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2800/tokenizer.json +0 -0
  26. ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2800/tokenizer_config.json +55 -0
  27. ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2800/trainer_state.json +1003 -0
  28. ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944/special_tokens_map.json +51 -0
  29. ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944/tokenizer.json +0 -0
  30. ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944/tokenizer_config.json +55 -0
  31. ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944/trainer_state.json +1038 -0
  32. ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-3496/special_tokens_map.json +51 -0
  33. ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-3496/tokenizer.json +0 -0
  34. ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-3496/tokenizer_config.json +55 -0
  35. ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-3496/trainer_state.json +1225 -0
  36. ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379/special_tokens_map.json +51 -0
  37. ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379/tokenizer.json +0 -0
  38. ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379/tokenizer_config.json +55 -0
  39. ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379/trainer_state.json +505 -0
  40. ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1970/special_tokens_map.json +51 -0
  41. ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1970/tokenizer.json +0 -0
  42. ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1970/tokenizer_config.json +55 -0
  43. ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1970/trainer_state.json +699 -0
  44. ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530/special_tokens_map.json +51 -0
  45. ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530/tokenizer.json +0 -0
  46. ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530/tokenizer_config.json +55 -0
  47. ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530/trainer_state.json +569 -0
  48. ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-2040/special_tokens_map.json +51 -0
  49. ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-2040/tokenizer.json +0 -0
  50. ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-2040/tokenizer_config.json +55 -0
ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-1309/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-1309/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-1309/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-1309/trainer_state.json ADDED
@@ -0,0 +1,484 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 748,
3
+ "best_metric": 0.8538576215744639,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748",
5
+ "epoch": 7.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1309,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13368983957219252,
14
+ "grad_norm": 11.328497886657715,
15
+ "learning_rate": 1.2834224598930483e-06,
16
+ "loss": 4.4175,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.26737967914438504,
21
+ "grad_norm": 11.061487197875977,
22
+ "learning_rate": 2.6203208556149735e-06,
23
+ "loss": 4.4074,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.40106951871657753,
28
+ "grad_norm": 70.5512466430664,
29
+ "learning_rate": 3.957219251336899e-06,
30
+ "loss": 4.3799,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5347593582887701,
35
+ "grad_norm": 9.930756568908691,
36
+ "learning_rate": 5.294117647058824e-06,
37
+ "loss": 4.3747,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6684491978609626,
42
+ "grad_norm": 18.983924865722656,
43
+ "learning_rate": 6.631016042780749e-06,
44
+ "loss": 4.3854,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8021390374331551,
49
+ "grad_norm": 10.151260375976562,
50
+ "learning_rate": 7.967914438502674e-06,
51
+ "loss": 4.32,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9358288770053476,
56
+ "grad_norm": 32.483131408691406,
57
+ "learning_rate": 9.3048128342246e-06,
58
+ "loss": 4.2694,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4965034965034965,
64
+ "eval_loss": 1.0118587017059326,
65
+ "eval_macro_f1": 0.4789043499230985,
66
+ "eval_runtime": 0.6453,
67
+ "eval_samples_per_second": 886.35,
68
+ "eval_steps_per_second": 55.784,
69
+ "eval_weighted_f1": 0.48185207309367545,
70
+ "step": 187
71
+ },
72
+ {
73
+ "epoch": 1.0695187165775402,
74
+ "grad_norm": 27.82496452331543,
75
+ "learning_rate": 1.0641711229946525e-05,
76
+ "loss": 4.2217,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2032085561497325,
81
+ "grad_norm": 26.401607513427734,
82
+ "learning_rate": 1.197860962566845e-05,
83
+ "loss": 3.9948,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3368983957219251,
88
+ "grad_norm": 32.15654754638672,
89
+ "learning_rate": 1.3315508021390375e-05,
90
+ "loss": 3.6487,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4705882352941178,
95
+ "grad_norm": 48.68864059448242,
96
+ "learning_rate": 1.4652406417112302e-05,
97
+ "loss": 3.512,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6042780748663101,
102
+ "grad_norm": 53.147613525390625,
103
+ "learning_rate": 1.5989304812834226e-05,
104
+ "loss": 3.1288,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7379679144385025,
109
+ "grad_norm": 116.22325134277344,
110
+ "learning_rate": 1.7326203208556153e-05,
111
+ "loss": 3.1469,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.8716577540106951,
116
+ "grad_norm": 61.6258544921875,
117
+ "learning_rate": 1.8663101604278075e-05,
118
+ "loss": 2.652,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.7465034965034965,
124
+ "eval_loss": 0.5797750949859619,
125
+ "eval_macro_f1": 0.7525661926078041,
126
+ "eval_runtime": 0.6684,
127
+ "eval_samples_per_second": 855.838,
128
+ "eval_steps_per_second": 53.864,
129
+ "eval_weighted_f1": 0.7530983926860606,
130
+ "step": 374
131
+ },
132
+ {
133
+ "epoch": 2.0053475935828877,
134
+ "grad_norm": 517.8780517578125,
135
+ "learning_rate": 2e-05,
136
+ "loss": 2.5975,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.1390374331550803,
141
+ "grad_norm": 61.48631286621094,
142
+ "learning_rate": 1.9851455733808676e-05,
143
+ "loss": 2.3599,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2727272727272725,
148
+ "grad_norm": 43.06319808959961,
149
+ "learning_rate": 1.9702911467617353e-05,
150
+ "loss": 2.1376,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.406417112299465,
155
+ "grad_norm": 44.87933349609375,
156
+ "learning_rate": 1.9554367201426024e-05,
157
+ "loss": 2.111,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.5401069518716577,
162
+ "grad_norm": 30.022062301635742,
163
+ "learning_rate": 1.94058229352347e-05,
164
+ "loss": 2.1151,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.6737967914438503,
169
+ "grad_norm": 28.22661018371582,
170
+ "learning_rate": 1.9257278669043376e-05,
171
+ "loss": 2.0886,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.807486631016043,
176
+ "grad_norm": 56.54759979248047,
177
+ "learning_rate": 1.9108734402852053e-05,
178
+ "loss": 2.0096,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.9411764705882355,
183
+ "grad_norm": 27.68973731994629,
184
+ "learning_rate": 1.8960190136660727e-05,
185
+ "loss": 2.2262,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.7972027972027972,
191
+ "eval_loss": 0.49703487753868103,
192
+ "eval_macro_f1": 0.792668986107612,
193
+ "eval_runtime": 0.6359,
194
+ "eval_samples_per_second": 899.448,
195
+ "eval_steps_per_second": 56.609,
196
+ "eval_weighted_f1": 0.7931116738395309,
197
+ "step": 561
198
+ },
199
+ {
200
+ "epoch": 3.0748663101604277,
201
+ "grad_norm": 46.97283935546875,
202
+ "learning_rate": 1.88116458704694e-05,
203
+ "loss": 1.9467,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2085561497326203,
208
+ "grad_norm": 47.29719161987305,
209
+ "learning_rate": 1.8663101604278075e-05,
210
+ "loss": 1.4003,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.342245989304813,
215
+ "grad_norm": 41.1824951171875,
216
+ "learning_rate": 1.8514557338086753e-05,
217
+ "loss": 1.5192,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.4759358288770055,
222
+ "grad_norm": 68.65145111083984,
223
+ "learning_rate": 1.8366013071895427e-05,
224
+ "loss": 1.649,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6096256684491976,
229
+ "grad_norm": 55.0959358215332,
230
+ "learning_rate": 1.82174688057041e-05,
231
+ "loss": 1.553,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.7433155080213902,
236
+ "grad_norm": 25.891080856323242,
237
+ "learning_rate": 1.8068924539512775e-05,
238
+ "loss": 1.5685,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.877005347593583,
243
+ "grad_norm": 27.374347686767578,
244
+ "learning_rate": 1.7920380273321453e-05,
245
+ "loss": 1.5569,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.8548951048951049,
251
+ "eval_loss": 0.40315213799476624,
252
+ "eval_macro_f1": 0.8538576215744639,
253
+ "eval_runtime": 0.6286,
254
+ "eval_samples_per_second": 910.028,
255
+ "eval_steps_per_second": 57.274,
256
+ "eval_weighted_f1": 0.8536736198460168,
257
+ "step": 748
258
+ },
259
+ {
260
+ "epoch": 4.010695187165775,
261
+ "grad_norm": 56.22517395019531,
262
+ "learning_rate": 1.7771836007130127e-05,
263
+ "loss": 1.4288,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.144385026737968,
268
+ "grad_norm": 59.8333625793457,
269
+ "learning_rate": 1.76232917409388e-05,
270
+ "loss": 1.1554,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.278074866310161,
275
+ "grad_norm": 72.26036071777344,
276
+ "learning_rate": 1.7474747474747475e-05,
277
+ "loss": 0.8808,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.411764705882353,
282
+ "grad_norm": 62.61909866333008,
283
+ "learning_rate": 1.7326203208556153e-05,
284
+ "loss": 0.8415,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.545454545454545,
289
+ "grad_norm": 46.91511917114258,
290
+ "learning_rate": 1.7177658942364827e-05,
291
+ "loss": 1.047,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.6791443850267385,
296
+ "grad_norm": 47.04996871948242,
297
+ "learning_rate": 1.70291146761735e-05,
298
+ "loss": 1.2479,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.81283422459893,
303
+ "grad_norm": 93.73844146728516,
304
+ "learning_rate": 1.6880570409982175e-05,
305
+ "loss": 1.224,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 4.946524064171123,
310
+ "grad_norm": 82.19253540039062,
311
+ "learning_rate": 1.6732026143790852e-05,
312
+ "loss": 1.031,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 5.0,
317
+ "eval_accuracy": 0.8496503496503497,
318
+ "eval_loss": 0.47908440232276917,
319
+ "eval_macro_f1": 0.848181743757675,
320
+ "eval_runtime": 0.6402,
321
+ "eval_samples_per_second": 893.429,
322
+ "eval_steps_per_second": 56.23,
323
+ "eval_weighted_f1": 0.847898149605886,
324
+ "step": 935
325
+ },
326
+ {
327
+ "epoch": 5.080213903743315,
328
+ "grad_norm": 9.348962783813477,
329
+ "learning_rate": 1.6583481877599526e-05,
330
+ "loss": 0.8529,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.213903743315508,
335
+ "grad_norm": 21.49125862121582,
336
+ "learning_rate": 1.64349376114082e-05,
337
+ "loss": 0.8131,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.347593582887701,
342
+ "grad_norm": 64.00634765625,
343
+ "learning_rate": 1.6286393345216875e-05,
344
+ "loss": 0.7135,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.481283422459893,
349
+ "grad_norm": 61.531982421875,
350
+ "learning_rate": 1.6137849079025552e-05,
351
+ "loss": 0.6908,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.614973262032086,
356
+ "grad_norm": 9.150433540344238,
357
+ "learning_rate": 1.5989304812834226e-05,
358
+ "loss": 0.636,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.748663101604278,
363
+ "grad_norm": 42.55214309692383,
364
+ "learning_rate": 1.58407605466429e-05,
365
+ "loss": 0.6091,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.882352941176471,
370
+ "grad_norm": 19.659725189208984,
371
+ "learning_rate": 1.5692216280451574e-05,
372
+ "loss": 0.9111,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 6.0,
377
+ "eval_accuracy": 0.8356643356643356,
378
+ "eval_loss": 0.5444453358650208,
379
+ "eval_macro_f1": 0.8383212074637602,
380
+ "eval_runtime": 0.6421,
381
+ "eval_samples_per_second": 890.775,
382
+ "eval_steps_per_second": 56.063,
383
+ "eval_weighted_f1": 0.8383333857336894,
384
+ "step": 1122
385
+ },
386
+ {
387
+ "epoch": 6.016042780748663,
388
+ "grad_norm": 108.13240051269531,
389
+ "learning_rate": 1.5543672014260252e-05,
390
+ "loss": 0.7784,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.149732620320855,
395
+ "grad_norm": 121.06307220458984,
396
+ "learning_rate": 1.5395127748068926e-05,
397
+ "loss": 0.4299,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.283422459893048,
402
+ "grad_norm": 69.35295104980469,
403
+ "learning_rate": 1.5246583481877602e-05,
404
+ "loss": 0.6042,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.4171122994652405,
409
+ "grad_norm": 94.7804946899414,
410
+ "learning_rate": 1.5098039215686276e-05,
411
+ "loss": 0.694,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.550802139037433,
416
+ "grad_norm": 109.5290756225586,
417
+ "learning_rate": 1.4949494949494952e-05,
418
+ "loss": 0.5509,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.684491978609626,
423
+ "grad_norm": 108.94828796386719,
424
+ "learning_rate": 1.4800950683303626e-05,
425
+ "loss": 0.544,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.818181818181818,
430
+ "grad_norm": 65.55987548828125,
431
+ "learning_rate": 1.4652406417112302e-05,
432
+ "loss": 0.788,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 6.951871657754011,
437
+ "grad_norm": 34.27817916870117,
438
+ "learning_rate": 1.4503862150920976e-05,
439
+ "loss": 0.6882,
440
+ "step": 1300
441
+ },
442
+ {
443
+ "epoch": 7.0,
444
+ "eval_accuracy": 0.8444055944055944,
445
+ "eval_loss": 0.5669668912887573,
446
+ "eval_macro_f1": 0.844275442461584,
447
+ "eval_runtime": 0.6508,
448
+ "eval_samples_per_second": 878.856,
449
+ "eval_steps_per_second": 55.313,
450
+ "eval_weighted_f1": 0.8439799067915141,
451
+ "step": 1309
452
+ }
453
+ ],
454
+ "logging_steps": 25,
455
+ "max_steps": 3740,
456
+ "num_input_tokens_seen": 0,
457
+ "num_train_epochs": 20,
458
+ "save_steps": 500,
459
+ "stateful_callbacks": {
460
+ "EarlyStoppingCallback": {
461
+ "args": {
462
+ "early_stopping_patience": 3,
463
+ "early_stopping_threshold": 0.0
464
+ },
465
+ "attributes": {
466
+ "early_stopping_patience_counter": 3
467
+ }
468
+ },
469
+ "TrainerControl": {
470
+ "args": {
471
+ "should_epoch_stop": false,
472
+ "should_evaluate": false,
473
+ "should_log": false,
474
+ "should_save": true,
475
+ "should_training_stop": true
476
+ },
477
+ "attributes": {}
478
+ }
479
+ },
480
+ "total_flos": 0.0,
481
+ "train_batch_size": 8,
482
+ "trial_name": null,
483
+ "trial_params": null
484
+ }
ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748/trainer_state.json ADDED
@@ -0,0 +1,290 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 748,
3
+ "best_metric": 0.8538576215744639,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-00-s43-20260827T192931Z-3858bb98/trainer/checkpoint-748",
5
+ "epoch": 4.0,
6
+ "eval_steps": 500,
7
+ "global_step": 748,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13368983957219252,
14
+ "grad_norm": 11.328497886657715,
15
+ "learning_rate": 1.2834224598930483e-06,
16
+ "loss": 4.4175,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.26737967914438504,
21
+ "grad_norm": 11.061487197875977,
22
+ "learning_rate": 2.6203208556149735e-06,
23
+ "loss": 4.4074,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.40106951871657753,
28
+ "grad_norm": 70.5512466430664,
29
+ "learning_rate": 3.957219251336899e-06,
30
+ "loss": 4.3799,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5347593582887701,
35
+ "grad_norm": 9.930756568908691,
36
+ "learning_rate": 5.294117647058824e-06,
37
+ "loss": 4.3747,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6684491978609626,
42
+ "grad_norm": 18.983924865722656,
43
+ "learning_rate": 6.631016042780749e-06,
44
+ "loss": 4.3854,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8021390374331551,
49
+ "grad_norm": 10.151260375976562,
50
+ "learning_rate": 7.967914438502674e-06,
51
+ "loss": 4.32,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9358288770053476,
56
+ "grad_norm": 32.483131408691406,
57
+ "learning_rate": 9.3048128342246e-06,
58
+ "loss": 4.2694,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4965034965034965,
64
+ "eval_loss": 1.0118587017059326,
65
+ "eval_macro_f1": 0.4789043499230985,
66
+ "eval_runtime": 0.6453,
67
+ "eval_samples_per_second": 886.35,
68
+ "eval_steps_per_second": 55.784,
69
+ "eval_weighted_f1": 0.48185207309367545,
70
+ "step": 187
71
+ },
72
+ {
73
+ "epoch": 1.0695187165775402,
74
+ "grad_norm": 27.82496452331543,
75
+ "learning_rate": 1.0641711229946525e-05,
76
+ "loss": 4.2217,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2032085561497325,
81
+ "grad_norm": 26.401607513427734,
82
+ "learning_rate": 1.197860962566845e-05,
83
+ "loss": 3.9948,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3368983957219251,
88
+ "grad_norm": 32.15654754638672,
89
+ "learning_rate": 1.3315508021390375e-05,
90
+ "loss": 3.6487,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4705882352941178,
95
+ "grad_norm": 48.68864059448242,
96
+ "learning_rate": 1.4652406417112302e-05,
97
+ "loss": 3.512,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6042780748663101,
102
+ "grad_norm": 53.147613525390625,
103
+ "learning_rate": 1.5989304812834226e-05,
104
+ "loss": 3.1288,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7379679144385025,
109
+ "grad_norm": 116.22325134277344,
110
+ "learning_rate": 1.7326203208556153e-05,
111
+ "loss": 3.1469,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.8716577540106951,
116
+ "grad_norm": 61.6258544921875,
117
+ "learning_rate": 1.8663101604278075e-05,
118
+ "loss": 2.652,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.7465034965034965,
124
+ "eval_loss": 0.5797750949859619,
125
+ "eval_macro_f1": 0.7525661926078041,
126
+ "eval_runtime": 0.6684,
127
+ "eval_samples_per_second": 855.838,
128
+ "eval_steps_per_second": 53.864,
129
+ "eval_weighted_f1": 0.7530983926860606,
130
+ "step": 374
131
+ },
132
+ {
133
+ "epoch": 2.0053475935828877,
134
+ "grad_norm": 517.8780517578125,
135
+ "learning_rate": 2e-05,
136
+ "loss": 2.5975,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.1390374331550803,
141
+ "grad_norm": 61.48631286621094,
142
+ "learning_rate": 1.9851455733808676e-05,
143
+ "loss": 2.3599,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2727272727272725,
148
+ "grad_norm": 43.06319808959961,
149
+ "learning_rate": 1.9702911467617353e-05,
150
+ "loss": 2.1376,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.406417112299465,
155
+ "grad_norm": 44.87933349609375,
156
+ "learning_rate": 1.9554367201426024e-05,
157
+ "loss": 2.111,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.5401069518716577,
162
+ "grad_norm": 30.022062301635742,
163
+ "learning_rate": 1.94058229352347e-05,
164
+ "loss": 2.1151,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.6737967914438503,
169
+ "grad_norm": 28.22661018371582,
170
+ "learning_rate": 1.9257278669043376e-05,
171
+ "loss": 2.0886,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.807486631016043,
176
+ "grad_norm": 56.54759979248047,
177
+ "learning_rate": 1.9108734402852053e-05,
178
+ "loss": 2.0096,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.9411764705882355,
183
+ "grad_norm": 27.68973731994629,
184
+ "learning_rate": 1.8960190136660727e-05,
185
+ "loss": 2.2262,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.7972027972027972,
191
+ "eval_loss": 0.49703487753868103,
192
+ "eval_macro_f1": 0.792668986107612,
193
+ "eval_runtime": 0.6359,
194
+ "eval_samples_per_second": 899.448,
195
+ "eval_steps_per_second": 56.609,
196
+ "eval_weighted_f1": 0.7931116738395309,
197
+ "step": 561
198
+ },
199
+ {
200
+ "epoch": 3.0748663101604277,
201
+ "grad_norm": 46.97283935546875,
202
+ "learning_rate": 1.88116458704694e-05,
203
+ "loss": 1.9467,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2085561497326203,
208
+ "grad_norm": 47.29719161987305,
209
+ "learning_rate": 1.8663101604278075e-05,
210
+ "loss": 1.4003,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.342245989304813,
215
+ "grad_norm": 41.1824951171875,
216
+ "learning_rate": 1.8514557338086753e-05,
217
+ "loss": 1.5192,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.4759358288770055,
222
+ "grad_norm": 68.65145111083984,
223
+ "learning_rate": 1.8366013071895427e-05,
224
+ "loss": 1.649,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6096256684491976,
229
+ "grad_norm": 55.0959358215332,
230
+ "learning_rate": 1.82174688057041e-05,
231
+ "loss": 1.553,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.7433155080213902,
236
+ "grad_norm": 25.891080856323242,
237
+ "learning_rate": 1.8068924539512775e-05,
238
+ "loss": 1.5685,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.877005347593583,
243
+ "grad_norm": 27.374347686767578,
244
+ "learning_rate": 1.7920380273321453e-05,
245
+ "loss": 1.5569,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.8548951048951049,
251
+ "eval_loss": 0.40315213799476624,
252
+ "eval_macro_f1": 0.8538576215744639,
253
+ "eval_runtime": 0.6286,
254
+ "eval_samples_per_second": 910.028,
255
+ "eval_steps_per_second": 57.274,
256
+ "eval_weighted_f1": 0.8536736198460168,
257
+ "step": 748
258
+ }
259
+ ],
260
+ "logging_steps": 25,
261
+ "max_steps": 3740,
262
+ "num_input_tokens_seen": 0,
263
+ "num_train_epochs": 20,
264
+ "save_steps": 500,
265
+ "stateful_callbacks": {
266
+ "EarlyStoppingCallback": {
267
+ "args": {
268
+ "early_stopping_patience": 3,
269
+ "early_stopping_threshold": 0.0
270
+ },
271
+ "attributes": {
272
+ "early_stopping_patience_counter": 0
273
+ }
274
+ },
275
+ "TrainerControl": {
276
+ "args": {
277
+ "should_epoch_stop": false,
278
+ "should_evaluate": false,
279
+ "should_log": false,
280
+ "should_save": true,
281
+ "should_training_stop": false
282
+ },
283
+ "attributes": {}
284
+ }
285
+ },
286
+ "total_flos": 0.0,
287
+ "train_batch_size": 8,
288
+ "trial_name": null,
289
+ "trial_params": null
290
+ }
ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-1528/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-1528/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-1528/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-1528/trainer_state.json ADDED
@@ -0,0 +1,558 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 955,
3
+ "best_metric": 0.8547200320753415,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955",
5
+ "epoch": 8.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1528,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1310615989515072,
14
+ "grad_norm": 14.725127220153809,
15
+ "learning_rate": 1.256544502617801e-06,
16
+ "loss": 4.4405,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2621231979030144,
21
+ "grad_norm": 18.634353637695312,
22
+ "learning_rate": 2.5654450261780104e-06,
23
+ "loss": 4.398,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.3931847968545216,
28
+ "grad_norm": 228.692626953125,
29
+ "learning_rate": 3.874345549738221e-06,
30
+ "loss": 4.4016,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5242463958060288,
35
+ "grad_norm": 8.183308601379395,
36
+ "learning_rate": 5.18324607329843e-06,
37
+ "loss": 4.3876,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6553079947575361,
42
+ "grad_norm": 22.457780838012695,
43
+ "learning_rate": 6.492146596858639e-06,
44
+ "loss": 4.3352,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.7863695937090432,
49
+ "grad_norm": 13.830279350280762,
50
+ "learning_rate": 7.80104712041885e-06,
51
+ "loss": 4.3605,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9174311926605505,
56
+ "grad_norm": 10.861157417297363,
57
+ "learning_rate": 9.109947643979057e-06,
58
+ "loss": 4.2415,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.5017123287671232,
64
+ "eval_loss": 0.989124059677124,
65
+ "eval_macro_f1": 0.4880605104309885,
66
+ "eval_runtime": 0.6919,
67
+ "eval_samples_per_second": 844.113,
68
+ "eval_steps_per_second": 53.48,
69
+ "eval_weighted_f1": 0.48933679234372923,
70
+ "step": 191
71
+ },
72
+ {
73
+ "epoch": 1.0471821756225426,
74
+ "grad_norm": 20.51849365234375,
75
+ "learning_rate": 1.0418848167539269e-05,
76
+ "loss": 3.9673,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.1782437745740497,
81
+ "grad_norm": 30.700273513793945,
82
+ "learning_rate": 1.1727748691099476e-05,
83
+ "loss": 3.7163,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.309305373525557,
88
+ "grad_norm": 51.61935043334961,
89
+ "learning_rate": 1.3036649214659686e-05,
90
+ "loss": 3.2196,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4403669724770642,
95
+ "grad_norm": 60.31619644165039,
96
+ "learning_rate": 1.4345549738219897e-05,
97
+ "loss": 2.7444,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.5714285714285714,
102
+ "grad_norm": 59.442222595214844,
103
+ "learning_rate": 1.5654450261780107e-05,
104
+ "loss": 2.7272,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7024901703800785,
109
+ "grad_norm": 54.38526153564453,
110
+ "learning_rate": 1.6963350785340316e-05,
111
+ "loss": 2.4171,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.833551769331586,
116
+ "grad_norm": 64.13296508789062,
117
+ "learning_rate": 1.8272251308900526e-05,
118
+ "loss": 2.5495,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 1.964613368283093,
123
+ "grad_norm": 35.57573318481445,
124
+ "learning_rate": 1.9581151832460736e-05,
125
+ "loss": 2.5594,
126
+ "step": 375
127
+ },
128
+ {
129
+ "epoch": 2.0,
130
+ "eval_accuracy": 0.803082191780822,
131
+ "eval_loss": 0.4709092378616333,
132
+ "eval_macro_f1": 0.8042426392164436,
133
+ "eval_runtime": 0.6584,
134
+ "eval_samples_per_second": 886.97,
135
+ "eval_steps_per_second": 56.195,
136
+ "eval_weighted_f1": 0.8043397714501659,
137
+ "step": 382
138
+ },
139
+ {
140
+ "epoch": 2.094364351245085,
141
+ "grad_norm": 43.21912384033203,
142
+ "learning_rate": 1.9901105293775453e-05,
143
+ "loss": 1.8763,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2254259501965925,
148
+ "grad_norm": 59.1907844543457,
149
+ "learning_rate": 1.9755671902268764e-05,
150
+ "loss": 1.8154,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.3564875491480994,
155
+ "grad_norm": 56.881744384765625,
156
+ "learning_rate": 1.961023851076207e-05,
157
+ "loss": 1.8681,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.487549148099607,
162
+ "grad_norm": 44.217437744140625,
163
+ "learning_rate": 1.9464805119255382e-05,
164
+ "loss": 1.8112,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.618610747051114,
169
+ "grad_norm": 50.271846771240234,
170
+ "learning_rate": 1.9319371727748693e-05,
171
+ "loss": 1.8015,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.749672346002621,
176
+ "grad_norm": 54.4185791015625,
177
+ "learning_rate": 1.9173938336242e-05,
178
+ "loss": 1.6875,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.8807339449541285,
183
+ "grad_norm": 31.4357852935791,
184
+ "learning_rate": 1.9028504944735315e-05,
185
+ "loss": 1.8809,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.827054794520548,
191
+ "eval_loss": 0.45389169454574585,
192
+ "eval_macro_f1": 0.8272703662421931,
193
+ "eval_runtime": 0.6562,
194
+ "eval_samples_per_second": 889.953,
195
+ "eval_steps_per_second": 56.384,
196
+ "eval_weighted_f1": 0.8274179624478468,
197
+ "step": 573
198
+ },
199
+ {
200
+ "epoch": 3.0104849279161208,
201
+ "grad_norm": 26.563087463378906,
202
+ "learning_rate": 1.8883071553228622e-05,
203
+ "loss": 1.65,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.1415465268676277,
208
+ "grad_norm": 32.46464157104492,
209
+ "learning_rate": 1.8737638161721933e-05,
210
+ "loss": 1.1699,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.272608125819135,
215
+ "grad_norm": 29.514780044555664,
216
+ "learning_rate": 1.8592204770215244e-05,
217
+ "loss": 1.2415,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.4036697247706424,
222
+ "grad_norm": 54.185997009277344,
223
+ "learning_rate": 1.844677137870855e-05,
224
+ "loss": 1.0967,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.5347313237221494,
229
+ "grad_norm": 34.36891555786133,
230
+ "learning_rate": 1.8301337987201865e-05,
231
+ "loss": 1.3803,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.6657929226736568,
236
+ "grad_norm": 38.55197525024414,
237
+ "learning_rate": 1.8155904595695173e-05,
238
+ "loss": 1.5857,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.7968545216251637,
243
+ "grad_norm": 57.356624603271484,
244
+ "learning_rate": 1.8010471204188483e-05,
245
+ "loss": 1.3551,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 3.927916120576671,
250
+ "grad_norm": 68.14253234863281,
251
+ "learning_rate": 1.7865037812681794e-05,
252
+ "loss": 1.3227,
253
+ "step": 750
254
+ },
255
+ {
256
+ "epoch": 4.0,
257
+ "eval_accuracy": 0.8373287671232876,
258
+ "eval_loss": 0.44627076387405396,
259
+ "eval_macro_f1": 0.837965678288259,
260
+ "eval_runtime": 0.6622,
261
+ "eval_samples_per_second": 881.902,
262
+ "eval_steps_per_second": 55.874,
263
+ "eval_weighted_f1": 0.8381923521561171,
264
+ "step": 764
265
+ },
266
+ {
267
+ "epoch": 4.057667103538663,
268
+ "grad_norm": 16.818737030029297,
269
+ "learning_rate": 1.77196044211751e-05,
270
+ "loss": 1.0933,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.18872870249017,
275
+ "grad_norm": 18.714706420898438,
276
+ "learning_rate": 1.7574171029668416e-05,
277
+ "loss": 0.8259,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.319790301441677,
282
+ "grad_norm": 34.60260772705078,
283
+ "learning_rate": 1.7428737638161723e-05,
284
+ "loss": 0.7248,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.450851900393185,
289
+ "grad_norm": 11.71359920501709,
290
+ "learning_rate": 1.7283304246655034e-05,
291
+ "loss": 0.7781,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.581913499344692,
296
+ "grad_norm": 38.82589340209961,
297
+ "learning_rate": 1.7137870855148345e-05,
298
+ "loss": 0.6204,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.712975098296199,
303
+ "grad_norm": 57.82145309448242,
304
+ "learning_rate": 1.6992437463641652e-05,
305
+ "loss": 0.7948,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 4.844036697247707,
310
+ "grad_norm": 27.654638290405273,
311
+ "learning_rate": 1.6847004072134963e-05,
312
+ "loss": 0.7674,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 4.975098296199214,
317
+ "grad_norm": 53.52952575683594,
318
+ "learning_rate": 1.6701570680628274e-05,
319
+ "loss": 0.964,
320
+ "step": 950
321
+ },
322
+ {
323
+ "epoch": 5.0,
324
+ "eval_accuracy": 0.8561643835616438,
325
+ "eval_loss": 0.5143012404441833,
326
+ "eval_macro_f1": 0.8547200320753415,
327
+ "eval_runtime": 0.6642,
328
+ "eval_samples_per_second": 879.289,
329
+ "eval_steps_per_second": 55.708,
330
+ "eval_weighted_f1": 0.8549062794782111,
331
+ "step": 955
332
+ },
333
+ {
334
+ "epoch": 5.104849279161206,
335
+ "grad_norm": 14.367695808410645,
336
+ "learning_rate": 1.6556137289121585e-05,
337
+ "loss": 0.4761,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.235910878112713,
342
+ "grad_norm": 6.829944133758545,
343
+ "learning_rate": 1.6410703897614895e-05,
344
+ "loss": 0.4686,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.36697247706422,
349
+ "grad_norm": 45.58695602416992,
350
+ "learning_rate": 1.6265270506108203e-05,
351
+ "loss": 0.6643,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.498034076015728,
356
+ "grad_norm": 18.722997665405273,
357
+ "learning_rate": 1.6119837114601514e-05,
358
+ "loss": 0.6845,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.6290956749672345,
363
+ "grad_norm": 99.55598449707031,
364
+ "learning_rate": 1.5974403723094824e-05,
365
+ "loss": 0.4751,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.760157273918741,
370
+ "grad_norm": 96.11966705322266,
371
+ "learning_rate": 1.5828970331588135e-05,
372
+ "loss": 0.564,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 5.891218872870249,
377
+ "grad_norm": 55.88045883178711,
378
+ "learning_rate": 1.5683536940081443e-05,
379
+ "loss": 0.6627,
380
+ "step": 1125
381
+ },
382
+ {
383
+ "epoch": 6.0,
384
+ "eval_accuracy": 0.8304794520547946,
385
+ "eval_loss": 0.7207126617431641,
386
+ "eval_macro_f1": 0.8286079414620035,
387
+ "eval_runtime": 0.6721,
388
+ "eval_samples_per_second": 868.902,
389
+ "eval_steps_per_second": 55.05,
390
+ "eval_weighted_f1": 0.8288259582023083,
391
+ "step": 1146
392
+ },
393
+ {
394
+ "epoch": 6.0209698558322415,
395
+ "grad_norm": 101.10545349121094,
396
+ "learning_rate": 1.5538103548574753e-05,
397
+ "loss": 0.5225,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.1520314547837485,
402
+ "grad_norm": 64.68859100341797,
403
+ "learning_rate": 1.5392670157068064e-05,
404
+ "loss": 0.3207,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.283093053735255,
409
+ "grad_norm": 49.65321350097656,
410
+ "learning_rate": 1.5247236765561375e-05,
411
+ "loss": 0.3419,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.414154652686763,
416
+ "grad_norm": 51.250274658203125,
417
+ "learning_rate": 1.5101803374054684e-05,
418
+ "loss": 0.5667,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.54521625163827,
423
+ "grad_norm": 84.01566314697266,
424
+ "learning_rate": 1.4956369982547993e-05,
425
+ "loss": 0.3651,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.676277850589777,
430
+ "grad_norm": 0.7171431183815002,
431
+ "learning_rate": 1.4810936591041306e-05,
432
+ "loss": 0.2575,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 6.807339449541285,
437
+ "grad_norm": 17.243820190429688,
438
+ "learning_rate": 1.4665503199534615e-05,
439
+ "loss": 0.2849,
440
+ "step": 1300
441
+ },
442
+ {
443
+ "epoch": 6.938401048492792,
444
+ "grad_norm": 94.30201721191406,
445
+ "learning_rate": 1.4520069808027924e-05,
446
+ "loss": 0.5828,
447
+ "step": 1325
448
+ },
449
+ {
450
+ "epoch": 7.0,
451
+ "eval_accuracy": 0.8373287671232876,
452
+ "eval_loss": 0.7453076243400574,
453
+ "eval_macro_f1": 0.8346266023720177,
454
+ "eval_runtime": 0.6708,
455
+ "eval_samples_per_second": 870.619,
456
+ "eval_steps_per_second": 55.159,
457
+ "eval_weighted_f1": 0.8349975582591844,
458
+ "step": 1337
459
+ },
460
+ {
461
+ "epoch": 7.068152031454784,
462
+ "grad_norm": 13.990921020507812,
463
+ "learning_rate": 1.4374636416521235e-05,
464
+ "loss": 0.7956,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.199213630406291,
469
+ "grad_norm": 40.4779052734375,
470
+ "learning_rate": 1.4229203025014544e-05,
471
+ "loss": 0.2466,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.330275229357798,
476
+ "grad_norm": 0.42905953526496887,
477
+ "learning_rate": 1.4083769633507855e-05,
478
+ "loss": 0.2836,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.461336828309306,
483
+ "grad_norm": 117.61661529541016,
484
+ "learning_rate": 1.3938336242001166e-05,
485
+ "loss": 0.3839,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.592398427260813,
490
+ "grad_norm": 2.8770437240600586,
491
+ "learning_rate": 1.3792902850494475e-05,
492
+ "loss": 0.2156,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 7.72346002621232,
497
+ "grad_norm": 152.85545349121094,
498
+ "learning_rate": 1.3647469458987785e-05,
499
+ "loss": 0.5404,
500
+ "step": 1475
501
+ },
502
+ {
503
+ "epoch": 7.854521625163827,
504
+ "grad_norm": 61.01215744018555,
505
+ "learning_rate": 1.3502036067481095e-05,
506
+ "loss": 0.403,
507
+ "step": 1500
508
+ },
509
+ {
510
+ "epoch": 7.985583224115334,
511
+ "grad_norm": 113.62518310546875,
512
+ "learning_rate": 1.3356602675974404e-05,
513
+ "loss": 0.3967,
514
+ "step": 1525
515
+ },
516
+ {
517
+ "epoch": 8.0,
518
+ "eval_accuracy": 0.8356164383561644,
519
+ "eval_loss": 0.8855543732643127,
520
+ "eval_macro_f1": 0.836458545694311,
521
+ "eval_runtime": 0.6772,
522
+ "eval_samples_per_second": 862.399,
523
+ "eval_steps_per_second": 54.638,
524
+ "eval_weighted_f1": 0.83654024469477,
525
+ "step": 1528
526
+ }
527
+ ],
528
+ "logging_steps": 25,
529
+ "max_steps": 3820,
530
+ "num_input_tokens_seen": 0,
531
+ "num_train_epochs": 20,
532
+ "save_steps": 500,
533
+ "stateful_callbacks": {
534
+ "EarlyStoppingCallback": {
535
+ "args": {
536
+ "early_stopping_patience": 3,
537
+ "early_stopping_threshold": 0.0
538
+ },
539
+ "attributes": {
540
+ "early_stopping_patience_counter": 3
541
+ }
542
+ },
543
+ "TrainerControl": {
544
+ "args": {
545
+ "should_epoch_stop": false,
546
+ "should_evaluate": false,
547
+ "should_log": false,
548
+ "should_save": true,
549
+ "should_training_stop": true
550
+ },
551
+ "attributes": {}
552
+ }
553
+ },
554
+ "total_flos": 0.0,
555
+ "train_batch_size": 8,
556
+ "trial_name": null,
557
+ "trial_params": null
558
+ }
ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955/trainer_state.json ADDED
@@ -0,0 +1,364 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 955,
3
+ "best_metric": 0.8547200320753415,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-01-s43-20260827T193233Z-3858bb98/trainer/checkpoint-955",
5
+ "epoch": 5.0,
6
+ "eval_steps": 500,
7
+ "global_step": 955,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1310615989515072,
14
+ "grad_norm": 14.725127220153809,
15
+ "learning_rate": 1.256544502617801e-06,
16
+ "loss": 4.4405,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2621231979030144,
21
+ "grad_norm": 18.634353637695312,
22
+ "learning_rate": 2.5654450261780104e-06,
23
+ "loss": 4.398,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.3931847968545216,
28
+ "grad_norm": 228.692626953125,
29
+ "learning_rate": 3.874345549738221e-06,
30
+ "loss": 4.4016,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5242463958060288,
35
+ "grad_norm": 8.183308601379395,
36
+ "learning_rate": 5.18324607329843e-06,
37
+ "loss": 4.3876,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6553079947575361,
42
+ "grad_norm": 22.457780838012695,
43
+ "learning_rate": 6.492146596858639e-06,
44
+ "loss": 4.3352,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.7863695937090432,
49
+ "grad_norm": 13.830279350280762,
50
+ "learning_rate": 7.80104712041885e-06,
51
+ "loss": 4.3605,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9174311926605505,
56
+ "grad_norm": 10.861157417297363,
57
+ "learning_rate": 9.109947643979057e-06,
58
+ "loss": 4.2415,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.5017123287671232,
64
+ "eval_loss": 0.989124059677124,
65
+ "eval_macro_f1": 0.4880605104309885,
66
+ "eval_runtime": 0.6919,
67
+ "eval_samples_per_second": 844.113,
68
+ "eval_steps_per_second": 53.48,
69
+ "eval_weighted_f1": 0.48933679234372923,
70
+ "step": 191
71
+ },
72
+ {
73
+ "epoch": 1.0471821756225426,
74
+ "grad_norm": 20.51849365234375,
75
+ "learning_rate": 1.0418848167539269e-05,
76
+ "loss": 3.9673,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.1782437745740497,
81
+ "grad_norm": 30.700273513793945,
82
+ "learning_rate": 1.1727748691099476e-05,
83
+ "loss": 3.7163,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.309305373525557,
88
+ "grad_norm": 51.61935043334961,
89
+ "learning_rate": 1.3036649214659686e-05,
90
+ "loss": 3.2196,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4403669724770642,
95
+ "grad_norm": 60.31619644165039,
96
+ "learning_rate": 1.4345549738219897e-05,
97
+ "loss": 2.7444,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.5714285714285714,
102
+ "grad_norm": 59.442222595214844,
103
+ "learning_rate": 1.5654450261780107e-05,
104
+ "loss": 2.7272,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7024901703800785,
109
+ "grad_norm": 54.38526153564453,
110
+ "learning_rate": 1.6963350785340316e-05,
111
+ "loss": 2.4171,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.833551769331586,
116
+ "grad_norm": 64.13296508789062,
117
+ "learning_rate": 1.8272251308900526e-05,
118
+ "loss": 2.5495,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 1.964613368283093,
123
+ "grad_norm": 35.57573318481445,
124
+ "learning_rate": 1.9581151832460736e-05,
125
+ "loss": 2.5594,
126
+ "step": 375
127
+ },
128
+ {
129
+ "epoch": 2.0,
130
+ "eval_accuracy": 0.803082191780822,
131
+ "eval_loss": 0.4709092378616333,
132
+ "eval_macro_f1": 0.8042426392164436,
133
+ "eval_runtime": 0.6584,
134
+ "eval_samples_per_second": 886.97,
135
+ "eval_steps_per_second": 56.195,
136
+ "eval_weighted_f1": 0.8043397714501659,
137
+ "step": 382
138
+ },
139
+ {
140
+ "epoch": 2.094364351245085,
141
+ "grad_norm": 43.21912384033203,
142
+ "learning_rate": 1.9901105293775453e-05,
143
+ "loss": 1.8763,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2254259501965925,
148
+ "grad_norm": 59.1907844543457,
149
+ "learning_rate": 1.9755671902268764e-05,
150
+ "loss": 1.8154,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.3564875491480994,
155
+ "grad_norm": 56.881744384765625,
156
+ "learning_rate": 1.961023851076207e-05,
157
+ "loss": 1.8681,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.487549148099607,
162
+ "grad_norm": 44.217437744140625,
163
+ "learning_rate": 1.9464805119255382e-05,
164
+ "loss": 1.8112,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.618610747051114,
169
+ "grad_norm": 50.271846771240234,
170
+ "learning_rate": 1.9319371727748693e-05,
171
+ "loss": 1.8015,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.749672346002621,
176
+ "grad_norm": 54.4185791015625,
177
+ "learning_rate": 1.9173938336242e-05,
178
+ "loss": 1.6875,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.8807339449541285,
183
+ "grad_norm": 31.4357852935791,
184
+ "learning_rate": 1.9028504944735315e-05,
185
+ "loss": 1.8809,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.827054794520548,
191
+ "eval_loss": 0.45389169454574585,
192
+ "eval_macro_f1": 0.8272703662421931,
193
+ "eval_runtime": 0.6562,
194
+ "eval_samples_per_second": 889.953,
195
+ "eval_steps_per_second": 56.384,
196
+ "eval_weighted_f1": 0.8274179624478468,
197
+ "step": 573
198
+ },
199
+ {
200
+ "epoch": 3.0104849279161208,
201
+ "grad_norm": 26.563087463378906,
202
+ "learning_rate": 1.8883071553228622e-05,
203
+ "loss": 1.65,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.1415465268676277,
208
+ "grad_norm": 32.46464157104492,
209
+ "learning_rate": 1.8737638161721933e-05,
210
+ "loss": 1.1699,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.272608125819135,
215
+ "grad_norm": 29.514780044555664,
216
+ "learning_rate": 1.8592204770215244e-05,
217
+ "loss": 1.2415,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.4036697247706424,
222
+ "grad_norm": 54.185997009277344,
223
+ "learning_rate": 1.844677137870855e-05,
224
+ "loss": 1.0967,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.5347313237221494,
229
+ "grad_norm": 34.36891555786133,
230
+ "learning_rate": 1.8301337987201865e-05,
231
+ "loss": 1.3803,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.6657929226736568,
236
+ "grad_norm": 38.55197525024414,
237
+ "learning_rate": 1.8155904595695173e-05,
238
+ "loss": 1.5857,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.7968545216251637,
243
+ "grad_norm": 57.356624603271484,
244
+ "learning_rate": 1.8010471204188483e-05,
245
+ "loss": 1.3551,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 3.927916120576671,
250
+ "grad_norm": 68.14253234863281,
251
+ "learning_rate": 1.7865037812681794e-05,
252
+ "loss": 1.3227,
253
+ "step": 750
254
+ },
255
+ {
256
+ "epoch": 4.0,
257
+ "eval_accuracy": 0.8373287671232876,
258
+ "eval_loss": 0.44627076387405396,
259
+ "eval_macro_f1": 0.837965678288259,
260
+ "eval_runtime": 0.6622,
261
+ "eval_samples_per_second": 881.902,
262
+ "eval_steps_per_second": 55.874,
263
+ "eval_weighted_f1": 0.8381923521561171,
264
+ "step": 764
265
+ },
266
+ {
267
+ "epoch": 4.057667103538663,
268
+ "grad_norm": 16.818737030029297,
269
+ "learning_rate": 1.77196044211751e-05,
270
+ "loss": 1.0933,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.18872870249017,
275
+ "grad_norm": 18.714706420898438,
276
+ "learning_rate": 1.7574171029668416e-05,
277
+ "loss": 0.8259,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.319790301441677,
282
+ "grad_norm": 34.60260772705078,
283
+ "learning_rate": 1.7428737638161723e-05,
284
+ "loss": 0.7248,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.450851900393185,
289
+ "grad_norm": 11.71359920501709,
290
+ "learning_rate": 1.7283304246655034e-05,
291
+ "loss": 0.7781,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.581913499344692,
296
+ "grad_norm": 38.82589340209961,
297
+ "learning_rate": 1.7137870855148345e-05,
298
+ "loss": 0.6204,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.712975098296199,
303
+ "grad_norm": 57.82145309448242,
304
+ "learning_rate": 1.6992437463641652e-05,
305
+ "loss": 0.7948,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 4.844036697247707,
310
+ "grad_norm": 27.654638290405273,
311
+ "learning_rate": 1.6847004072134963e-05,
312
+ "loss": 0.7674,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 4.975098296199214,
317
+ "grad_norm": 53.52952575683594,
318
+ "learning_rate": 1.6701570680628274e-05,
319
+ "loss": 0.964,
320
+ "step": 950
321
+ },
322
+ {
323
+ "epoch": 5.0,
324
+ "eval_accuracy": 0.8561643835616438,
325
+ "eval_loss": 0.5143012404441833,
326
+ "eval_macro_f1": 0.8547200320753415,
327
+ "eval_runtime": 0.6642,
328
+ "eval_samples_per_second": 879.289,
329
+ "eval_steps_per_second": 55.708,
330
+ "eval_weighted_f1": 0.8549062794782111,
331
+ "step": 955
332
+ }
333
+ ],
334
+ "logging_steps": 25,
335
+ "max_steps": 3820,
336
+ "num_input_tokens_seen": 0,
337
+ "num_train_epochs": 20,
338
+ "save_steps": 500,
339
+ "stateful_callbacks": {
340
+ "EarlyStoppingCallback": {
341
+ "args": {
342
+ "early_stopping_patience": 3,
343
+ "early_stopping_threshold": 0.0
344
+ },
345
+ "attributes": {
346
+ "early_stopping_patience_counter": 0
347
+ }
348
+ },
349
+ "TrainerControl": {
350
+ "args": {
351
+ "should_epoch_stop": false,
352
+ "should_evaluate": false,
353
+ "should_log": false,
354
+ "should_save": true,
355
+ "should_training_stop": false
356
+ },
357
+ "attributes": {}
358
+ }
359
+ },
360
+ "total_flos": 0.0,
361
+ "train_batch_size": 8,
362
+ "trial_name": null,
363
+ "trial_params": null
364
+ }
ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820/trainer_state.json ADDED
@@ -0,0 +1,657 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1820,
3
+ "best_metric": 0.869309032967074,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-1820",
5
+ "epoch": 10.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1820,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13793103448275862,
14
+ "grad_norm": 11.58485221862793,
15
+ "learning_rate": 1.3186813186813187e-06,
16
+ "loss": 4.4314,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27586206896551724,
21
+ "grad_norm": 9.425135612487793,
22
+ "learning_rate": 2.6923076923076923e-06,
23
+ "loss": 4.3978,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.41379310344827586,
28
+ "grad_norm": 8.210010528564453,
29
+ "learning_rate": 4.065934065934066e-06,
30
+ "loss": 4.3829,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5517241379310345,
35
+ "grad_norm": 18.993253707885742,
36
+ "learning_rate": 5.43956043956044e-06,
37
+ "loss": 4.3861,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6896551724137931,
42
+ "grad_norm": 19.984159469604492,
43
+ "learning_rate": 6.813186813186814e-06,
44
+ "loss": 4.3753,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8275862068965517,
49
+ "grad_norm": 18.76668357849121,
50
+ "learning_rate": 8.186813186813188e-06,
51
+ "loss": 4.32,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9655172413793104,
56
+ "grad_norm": 19.58595085144043,
57
+ "learning_rate": 9.560439560439562e-06,
58
+ "loss": 4.3283,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.47912524850894633,
64
+ "eval_loss": 1.0147746801376343,
65
+ "eval_macro_f1": 0.4170847419784449,
66
+ "eval_runtime": 0.5825,
67
+ "eval_samples_per_second": 863.45,
68
+ "eval_steps_per_second": 54.931,
69
+ "eval_weighted_f1": 0.4192667321010538,
70
+ "step": 182
71
+ },
72
+ {
73
+ "epoch": 1.0993103448275863,
74
+ "grad_norm": 58.16252517700195,
75
+ "learning_rate": 1.0934065934065933e-05,
76
+ "loss": 3.9772,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.237241379310345,
81
+ "grad_norm": 49.73049545288086,
82
+ "learning_rate": 1.230769230769231e-05,
83
+ "loss": 3.599,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3751724137931034,
88
+ "grad_norm": 58.101070404052734,
89
+ "learning_rate": 1.3681318681318683e-05,
90
+ "loss": 3.1096,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.513103448275862,
95
+ "grad_norm": 45.06196212768555,
96
+ "learning_rate": 1.5054945054945056e-05,
97
+ "loss": 3.0102,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6510344827586207,
102
+ "grad_norm": 41.06315231323242,
103
+ "learning_rate": 1.642857142857143e-05,
104
+ "loss": 2.8282,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7889655172413792,
109
+ "grad_norm": 39.79212951660156,
110
+ "learning_rate": 1.78021978021978e-05,
111
+ "loss": 2.5567,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.926896551724138,
116
+ "grad_norm": 39.06377029418945,
117
+ "learning_rate": 1.9175824175824178e-05,
118
+ "loss": 2.5691,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.7872763419483101,
124
+ "eval_loss": 0.5018846988677979,
125
+ "eval_macro_f1": 0.7884482306920942,
126
+ "eval_runtime": 0.5852,
127
+ "eval_samples_per_second": 859.475,
128
+ "eval_steps_per_second": 54.678,
129
+ "eval_weighted_f1": 0.788913120430031,
130
+ "step": 364
131
+ },
132
+ {
133
+ "epoch": 2.0606896551724136,
134
+ "grad_norm": 42.192054748535156,
135
+ "learning_rate": 1.993894993894994e-05,
136
+ "loss": 2.1688,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.1986206896551725,
141
+ "grad_norm": 43.01150894165039,
142
+ "learning_rate": 1.9786324786324788e-05,
143
+ "loss": 1.7811,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.336551724137931,
148
+ "grad_norm": 27.328853607177734,
149
+ "learning_rate": 1.9633699633699634e-05,
150
+ "loss": 1.8751,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.47448275862069,
155
+ "grad_norm": 53.49696350097656,
156
+ "learning_rate": 1.948107448107448e-05,
157
+ "loss": 1.936,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.6124137931034483,
162
+ "grad_norm": 43.56639099121094,
163
+ "learning_rate": 1.932844932844933e-05,
164
+ "loss": 1.839,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.750344827586207,
169
+ "grad_norm": 59.44014358520508,
170
+ "learning_rate": 1.9175824175824178e-05,
171
+ "loss": 1.9447,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8882758620689657,
176
+ "grad_norm": 32.07001876831055,
177
+ "learning_rate": 1.9023199023199025e-05,
178
+ "loss": 1.7301,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.8330019880715706,
184
+ "eval_loss": 0.4471936821937561,
185
+ "eval_macro_f1": 0.831735853217329,
186
+ "eval_runtime": 0.5916,
187
+ "eval_samples_per_second": 850.291,
188
+ "eval_steps_per_second": 54.094,
189
+ "eval_weighted_f1": 0.8319282926037921,
190
+ "step": 546
191
+ },
192
+ {
193
+ "epoch": 3.0220689655172412,
194
+ "grad_norm": 46.59756088256836,
195
+ "learning_rate": 1.887057387057387e-05,
196
+ "loss": 1.6669,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.16,
201
+ "grad_norm": 66.64631652832031,
202
+ "learning_rate": 1.8717948717948718e-05,
203
+ "loss": 1.2406,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2979310344827586,
208
+ "grad_norm": 43.43894958496094,
209
+ "learning_rate": 1.8565323565323568e-05,
210
+ "loss": 1.1403,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.435862068965517,
215
+ "grad_norm": 42.17364501953125,
216
+ "learning_rate": 1.8412698412698415e-05,
217
+ "loss": 1.3168,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.573793103448276,
222
+ "grad_norm": 50.19221878051758,
223
+ "learning_rate": 1.826007326007326e-05,
224
+ "loss": 1.2885,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.7117241379310344,
229
+ "grad_norm": 53.39130783081055,
230
+ "learning_rate": 1.8107448107448108e-05,
231
+ "loss": 1.2697,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.849655172413793,
236
+ "grad_norm": 57.69573211669922,
237
+ "learning_rate": 1.7954822954822958e-05,
238
+ "loss": 1.1977,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.9875862068965517,
243
+ "grad_norm": 62.88957214355469,
244
+ "learning_rate": 1.78021978021978e-05,
245
+ "loss": 1.2081,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.8469184890656064,
251
+ "eval_loss": 0.4087779223918915,
252
+ "eval_macro_f1": 0.8447465902307477,
253
+ "eval_runtime": 1.5185,
254
+ "eval_samples_per_second": 331.244,
255
+ "eval_steps_per_second": 21.073,
256
+ "eval_weighted_f1": 0.8451006182184996,
257
+ "step": 728
258
+ },
259
+ {
260
+ "epoch": 4.121379310344827,
261
+ "grad_norm": 51.63662338256836,
262
+ "learning_rate": 1.764957264957265e-05,
263
+ "loss": 0.8759,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.259310344827586,
268
+ "grad_norm": 35.93242645263672,
269
+ "learning_rate": 1.74969474969475e-05,
270
+ "loss": 0.8427,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.397241379310345,
275
+ "grad_norm": 54.42876434326172,
276
+ "learning_rate": 1.7344322344322345e-05,
277
+ "loss": 0.7767,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.535172413793103,
282
+ "grad_norm": 44.173789978027344,
283
+ "learning_rate": 1.7191697191697195e-05,
284
+ "loss": 0.9649,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.673103448275862,
289
+ "grad_norm": 29.70044708251953,
290
+ "learning_rate": 1.703907203907204e-05,
291
+ "loss": 0.6833,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.811034482758621,
296
+ "grad_norm": 82.69610595703125,
297
+ "learning_rate": 1.688644688644689e-05,
298
+ "loss": 1.0218,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.94896551724138,
303
+ "grad_norm": 120.68234252929688,
304
+ "learning_rate": 1.6733821733821735e-05,
305
+ "loss": 0.7451,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.8250497017892644,
311
+ "eval_loss": 0.4958866238594055,
312
+ "eval_macro_f1": 0.8235745820260516,
313
+ "eval_runtime": 0.5765,
314
+ "eval_samples_per_second": 872.527,
315
+ "eval_steps_per_second": 55.509,
316
+ "eval_weighted_f1": 0.8239097745772974,
317
+ "step": 910
318
+ },
319
+ {
320
+ "epoch": 5.082758620689655,
321
+ "grad_norm": 36.33766174316406,
322
+ "learning_rate": 1.6581196581196585e-05,
323
+ "loss": 0.605,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.220689655172414,
328
+ "grad_norm": 141.90615844726562,
329
+ "learning_rate": 1.642857142857143e-05,
330
+ "loss": 0.5546,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.358620689655172,
335
+ "grad_norm": 32.26487350463867,
336
+ "learning_rate": 1.627594627594628e-05,
337
+ "loss": 0.5659,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.496551724137931,
342
+ "grad_norm": 42.44941711425781,
343
+ "learning_rate": 1.6123321123321125e-05,
344
+ "loss": 0.5046,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.63448275862069,
349
+ "grad_norm": 57.650203704833984,
350
+ "learning_rate": 1.5970695970695972e-05,
351
+ "loss": 0.5452,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.772413793103448,
356
+ "grad_norm": 15.052305221557617,
357
+ "learning_rate": 1.581807081807082e-05,
358
+ "loss": 0.5173,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.910344827586207,
363
+ "grad_norm": 46.666446685791016,
364
+ "learning_rate": 1.5665445665445665e-05,
365
+ "loss": 0.6587,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.8508946322067594,
371
+ "eval_loss": 0.5746742486953735,
372
+ "eval_macro_f1": 0.8519946935182702,
373
+ "eval_runtime": 0.5866,
374
+ "eval_samples_per_second": 857.472,
375
+ "eval_steps_per_second": 54.551,
376
+ "eval_weighted_f1": 0.8521578610698961,
377
+ "step": 1092
378
+ },
379
+ {
380
+ "epoch": 6.0441379310344825,
381
+ "grad_norm": 81.79331970214844,
382
+ "learning_rate": 1.5512820512820516e-05,
383
+ "loss": 0.5335,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.182068965517241,
388
+ "grad_norm": 114.01929473876953,
389
+ "learning_rate": 1.5360195360195362e-05,
390
+ "loss": 0.4932,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.32,
395
+ "grad_norm": 67.68749237060547,
396
+ "learning_rate": 1.5207570207570207e-05,
397
+ "loss": 0.3986,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.457931034482758,
402
+ "grad_norm": 44.281253814697266,
403
+ "learning_rate": 1.5054945054945056e-05,
404
+ "loss": 0.3298,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.595862068965517,
409
+ "grad_norm": 270.51141357421875,
410
+ "learning_rate": 1.4902319902319904e-05,
411
+ "loss": 0.4579,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.733793103448276,
416
+ "grad_norm": 40.47040557861328,
417
+ "learning_rate": 1.4749694749694752e-05,
418
+ "loss": 0.4534,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.871724137931034,
423
+ "grad_norm": 25.67839813232422,
424
+ "learning_rate": 1.4597069597069597e-05,
425
+ "loss": 0.4817,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 7.0,
430
+ "eval_accuracy": 0.8429423459244533,
431
+ "eval_loss": 0.679874837398529,
432
+ "eval_macro_f1": 0.8419964374428378,
433
+ "eval_runtime": 0.577,
434
+ "eval_samples_per_second": 871.8,
435
+ "eval_steps_per_second": 55.462,
436
+ "eval_weighted_f1": 0.8422470469909669,
437
+ "step": 1274
438
+ },
439
+ {
440
+ "epoch": 7.0055172413793105,
441
+ "grad_norm": 0.925383448600769,
442
+ "learning_rate": 1.4444444444444446e-05,
443
+ "loss": 0.3014,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.1434482758620685,
448
+ "grad_norm": 93.09014892578125,
449
+ "learning_rate": 1.4291819291819293e-05,
450
+ "loss": 0.3997,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.281379310344827,
455
+ "grad_norm": 0.23526118695735931,
456
+ "learning_rate": 1.4139194139194141e-05,
457
+ "loss": 0.37,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.419310344827586,
462
+ "grad_norm": 129.131591796875,
463
+ "learning_rate": 1.3986568986568988e-05,
464
+ "loss": 0.3375,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.557241379310344,
469
+ "grad_norm": 1.749083161354065,
470
+ "learning_rate": 1.3833943833943834e-05,
471
+ "loss": 0.3286,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.695172413793103,
476
+ "grad_norm": 94.52139282226562,
477
+ "learning_rate": 1.3681318681318683e-05,
478
+ "loss": 0.3205,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.833103448275862,
483
+ "grad_norm": 178.98956298828125,
484
+ "learning_rate": 1.3528693528693531e-05,
485
+ "loss": 0.3422,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.971034482758621,
490
+ "grad_norm": 1.7286487817764282,
491
+ "learning_rate": 1.3376068376068376e-05,
492
+ "loss": 0.6482,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 8.0,
497
+ "eval_accuracy": 0.8548707753479126,
498
+ "eval_loss": 0.7450631260871887,
499
+ "eval_macro_f1": 0.8554842491486295,
500
+ "eval_runtime": 0.591,
501
+ "eval_samples_per_second": 851.136,
502
+ "eval_steps_per_second": 54.148,
503
+ "eval_weighted_f1": 0.8557116333768144,
504
+ "step": 1456
505
+ },
506
+ {
507
+ "epoch": 8.104827586206897,
508
+ "grad_norm": 125.99210357666016,
509
+ "learning_rate": 1.3223443223443224e-05,
510
+ "loss": 0.3478,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.242758620689655,
515
+ "grad_norm": 65.56297302246094,
516
+ "learning_rate": 1.3070818070818073e-05,
517
+ "loss": 0.1924,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.380689655172414,
522
+ "grad_norm": 16.62967872619629,
523
+ "learning_rate": 1.291819291819292e-05,
524
+ "loss": 0.2844,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.518620689655172,
529
+ "grad_norm": 0.2591530680656433,
530
+ "learning_rate": 1.2765567765567766e-05,
531
+ "loss": 0.3596,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.65655172413793,
536
+ "grad_norm": 155.3113250732422,
537
+ "learning_rate": 1.2612942612942613e-05,
538
+ "loss": 0.3918,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.79448275862069,
543
+ "grad_norm": 10.316084861755371,
544
+ "learning_rate": 1.2460317460317461e-05,
545
+ "loss": 0.351,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.932413793103448,
550
+ "grad_norm": 0.09080405533313751,
551
+ "learning_rate": 1.230769230769231e-05,
552
+ "loss": 0.1491,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 9.0,
557
+ "eval_accuracy": 0.8608349900596421,
558
+ "eval_loss": 0.7955670952796936,
559
+ "eval_macro_f1": 0.8604545280496444,
560
+ "eval_runtime": 0.5744,
561
+ "eval_samples_per_second": 875.751,
562
+ "eval_steps_per_second": 55.714,
563
+ "eval_weighted_f1": 0.8606573163234017,
564
+ "step": 1638
565
+ },
566
+ {
567
+ "epoch": 9.066206896551725,
568
+ "grad_norm": 35.8708610534668,
569
+ "learning_rate": 1.2155067155067155e-05,
570
+ "loss": 0.1611,
571
+ "step": 1650
572
+ },
573
+ {
574
+ "epoch": 9.204137931034483,
575
+ "grad_norm": 2.7454748153686523,
576
+ "learning_rate": 1.2002442002442003e-05,
577
+ "loss": 0.279,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.34206896551724,
582
+ "grad_norm": 267.3585205078125,
583
+ "learning_rate": 1.1849816849816852e-05,
584
+ "loss": 0.1466,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.48,
589
+ "grad_norm": 195.13316345214844,
590
+ "learning_rate": 1.1697191697191698e-05,
591
+ "loss": 0.3403,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.617931034482758,
596
+ "grad_norm": 0.22231760621070862,
597
+ "learning_rate": 1.1544566544566545e-05,
598
+ "loss": 0.1618,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 9.755862068965516,
603
+ "grad_norm": 0.14777538180351257,
604
+ "learning_rate": 1.1391941391941392e-05,
605
+ "loss": 0.2352,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 9.893793103448276,
610
+ "grad_norm": 27.933271408081055,
611
+ "learning_rate": 1.123931623931624e-05,
612
+ "loss": 0.2912,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 10.0,
617
+ "eval_accuracy": 0.8707753479125249,
618
+ "eval_loss": 0.8268889784812927,
619
+ "eval_macro_f1": 0.869309032967074,
620
+ "eval_runtime": 0.5821,
621
+ "eval_samples_per_second": 864.083,
622
+ "eval_steps_per_second": 54.972,
623
+ "eval_weighted_f1": 0.8693453848174799,
624
+ "step": 1820
625
+ }
626
+ ],
627
+ "logging_steps": 25,
628
+ "max_steps": 3640,
629
+ "num_input_tokens_seen": 0,
630
+ "num_train_epochs": 20,
631
+ "save_steps": 500,
632
+ "stateful_callbacks": {
633
+ "EarlyStoppingCallback": {
634
+ "args": {
635
+ "early_stopping_patience": 3,
636
+ "early_stopping_threshold": 0.0
637
+ },
638
+ "attributes": {
639
+ "early_stopping_patience_counter": 0
640
+ }
641
+ },
642
+ "TrainerControl": {
643
+ "args": {
644
+ "should_epoch_stop": false,
645
+ "should_evaluate": false,
646
+ "should_log": false,
647
+ "should_save": true,
648
+ "should_training_stop": false
649
+ },
650
+ "attributes": {}
651
+ }
652
+ },
653
+ "total_flos": 0.0,
654
+ "train_batch_size": 8,
655
+ "trial_name": null,
656
+ "trial_params": null
657
+ }
ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-2366/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-2366/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-02-s43-20260827T193604Z-3858bb98/trainer/checkpoint-2366/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2800/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2800/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2800/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2800/trainer_state.json ADDED
@@ -0,0 +1,1003 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2275,
3
+ "best_metric": 0.8337519435295319,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-05-s42-20260827T185346Z-6235c159/trainer/checkpoint-2275",
5
+ "epoch": 16.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2800,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1430615164520744,
14
+ "grad_norm": 10.764623641967773,
15
+ "learning_rate": 1.3714285714285717e-06,
16
+ "loss": 4.4589,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2861230329041488,
21
+ "grad_norm": 162.19888305664062,
22
+ "learning_rate": 2.8000000000000003e-06,
23
+ "loss": 4.4266,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4291845493562232,
28
+ "grad_norm": 22.468881607055664,
29
+ "learning_rate": 4.228571428571429e-06,
30
+ "loss": 4.3937,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5722460658082976,
35
+ "grad_norm": 7.81147575378418,
36
+ "learning_rate": 5.6571428571428576e-06,
37
+ "loss": 4.4004,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.7153075822603719,
42
+ "grad_norm": 15.173408508300781,
43
+ "learning_rate": 7.085714285714286e-06,
44
+ "loss": 4.3858,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8583690987124464,
49
+ "grad_norm": 11.736390113830566,
50
+ "learning_rate": 8.514285714285715e-06,
51
+ "loss": 4.337,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "grad_norm": 11.000627517700195,
57
+ "learning_rate": 9.942857142857145e-06,
58
+ "loss": 4.2235,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.38274932614555257,
64
+ "eval_loss": 1.0647900104522705,
65
+ "eval_macro_f1": 0.28862667259291735,
66
+ "eval_runtime": 0.853,
67
+ "eval_samples_per_second": 869.842,
68
+ "eval_steps_per_second": 55.098,
69
+ "eval_weighted_f1": 0.2928804527289634,
70
+ "step": 175
71
+ },
72
+ {
73
+ "epoch": 1.1430615164520743,
74
+ "grad_norm": 14.510539054870605,
75
+ "learning_rate": 1.1371428571428573e-05,
76
+ "loss": 4.2459,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2861230329041489,
81
+ "grad_norm": 13.885258674621582,
82
+ "learning_rate": 1.2800000000000001e-05,
83
+ "loss": 4.1846,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.4291845493562232,
88
+ "grad_norm": 25.825702667236328,
89
+ "learning_rate": 1.422857142857143e-05,
90
+ "loss": 3.7819,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.5722460658082977,
95
+ "grad_norm": 55.849998474121094,
96
+ "learning_rate": 1.5657142857142856e-05,
97
+ "loss": 3.1241,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.715307582260372,
102
+ "grad_norm": 42.44798278808594,
103
+ "learning_rate": 1.708571428571429e-05,
104
+ "loss": 3.0195,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.8583690987124464,
109
+ "grad_norm": 37.7345085144043,
110
+ "learning_rate": 1.8514285714285716e-05,
111
+ "loss": 2.819,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 2.0,
116
+ "grad_norm": 24.260963439941406,
117
+ "learning_rate": 1.9942857142857142e-05,
118
+ "loss": 2.6864,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.7318059299191375,
124
+ "eval_loss": 0.6424435377120972,
125
+ "eval_macro_f1": 0.7346698577409133,
126
+ "eval_runtime": 0.8352,
127
+ "eval_samples_per_second": 888.415,
128
+ "eval_steps_per_second": 56.274,
129
+ "eval_weighted_f1": 0.7344079706632612,
130
+ "step": 350
131
+ },
132
+ {
133
+ "epoch": 2.1430615164520743,
134
+ "grad_norm": 48.79184341430664,
135
+ "learning_rate": 1.984761904761905e-05,
136
+ "loss": 2.3561,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.2861230329041486,
141
+ "grad_norm": 31.460241317749023,
142
+ "learning_rate": 1.968888888888889e-05,
143
+ "loss": 2.0483,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.429184549356223,
148
+ "grad_norm": 41.83769607543945,
149
+ "learning_rate": 1.953015873015873e-05,
150
+ "loss": 2.1106,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.5722460658082977,
155
+ "grad_norm": 40.333953857421875,
156
+ "learning_rate": 1.937142857142857e-05,
157
+ "loss": 2.0525,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.715307582260372,
162
+ "grad_norm": 49.83087158203125,
163
+ "learning_rate": 1.9212698412698414e-05,
164
+ "loss": 1.8606,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.8583690987124464,
169
+ "grad_norm": 61.01762771606445,
170
+ "learning_rate": 1.9053968253968257e-05,
171
+ "loss": 2.0432,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 3.0,
176
+ "grad_norm": 32.610008239746094,
177
+ "learning_rate": 1.8895238095238096e-05,
178
+ "loss": 2.062,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.7735849056603774,
184
+ "eval_loss": 0.5390893220901489,
185
+ "eval_macro_f1": 0.7746363000376896,
186
+ "eval_runtime": 0.8894,
187
+ "eval_samples_per_second": 834.248,
188
+ "eval_steps_per_second": 52.843,
189
+ "eval_weighted_f1": 0.7751295093953015,
190
+ "step": 525
191
+ },
192
+ {
193
+ "epoch": 3.1430615164520743,
194
+ "grad_norm": 42.178749084472656,
195
+ "learning_rate": 1.8736507936507936e-05,
196
+ "loss": 1.4375,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.2861230329041486,
201
+ "grad_norm": 72.75885009765625,
202
+ "learning_rate": 1.857777777777778e-05,
203
+ "loss": 1.5141,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.429184549356223,
208
+ "grad_norm": 60.360504150390625,
209
+ "learning_rate": 1.8419047619047622e-05,
210
+ "loss": 1.5516,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.5722460658082977,
215
+ "grad_norm": 28.8283634185791,
216
+ "learning_rate": 1.8260317460317462e-05,
217
+ "loss": 1.4913,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.715307582260372,
222
+ "grad_norm": 20.996593475341797,
223
+ "learning_rate": 1.81015873015873e-05,
224
+ "loss": 1.1141,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.8583690987124464,
229
+ "grad_norm": 35.394630432128906,
230
+ "learning_rate": 1.7942857142857144e-05,
231
+ "loss": 1.2914,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 4.0,
236
+ "grad_norm": 56.34248733520508,
237
+ "learning_rate": 1.7784126984126987e-05,
238
+ "loss": 1.3023,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 4.0,
243
+ "eval_accuracy": 0.8032345013477089,
244
+ "eval_loss": 0.5447040796279907,
245
+ "eval_macro_f1": 0.8025945866333334,
246
+ "eval_runtime": 0.838,
247
+ "eval_samples_per_second": 885.454,
248
+ "eval_steps_per_second": 56.087,
249
+ "eval_weighted_f1": 0.802648177732895,
250
+ "step": 700
251
+ },
252
+ {
253
+ "epoch": 4.143061516452074,
254
+ "grad_norm": 31.50874900817871,
255
+ "learning_rate": 1.7625396825396827e-05,
256
+ "loss": 0.8751,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.286123032904149,
261
+ "grad_norm": 26.837421417236328,
262
+ "learning_rate": 1.7466666666666667e-05,
263
+ "loss": 0.9729,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.429184549356223,
268
+ "grad_norm": 47.66377639770508,
269
+ "learning_rate": 1.730793650793651e-05,
270
+ "loss": 0.8436,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.572246065808297,
275
+ "grad_norm": 33.50248336791992,
276
+ "learning_rate": 1.7149206349206353e-05,
277
+ "loss": 1.0975,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.715307582260372,
282
+ "grad_norm": 33.4804573059082,
283
+ "learning_rate": 1.6990476190476192e-05,
284
+ "loss": 0.9162,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.858369098712446,
289
+ "grad_norm": 47.16022491455078,
290
+ "learning_rate": 1.6831746031746032e-05,
291
+ "loss": 0.8485,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 5.0,
296
+ "grad_norm": 47.60380554199219,
297
+ "learning_rate": 1.6673015873015875e-05,
298
+ "loss": 0.9271,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 5.0,
303
+ "eval_accuracy": 0.7991913746630728,
304
+ "eval_loss": 0.5958666205406189,
305
+ "eval_macro_f1": 0.7946888806527711,
306
+ "eval_runtime": 0.8419,
307
+ "eval_samples_per_second": 881.29,
308
+ "eval_steps_per_second": 55.823,
309
+ "eval_weighted_f1": 0.7948560005834102,
310
+ "step": 875
311
+ },
312
+ {
313
+ "epoch": 5.143061516452074,
314
+ "grad_norm": 81.07859802246094,
315
+ "learning_rate": 1.6514285714285714e-05,
316
+ "loss": 0.6353,
317
+ "step": 900
318
+ },
319
+ {
320
+ "epoch": 5.286123032904149,
321
+ "grad_norm": 76.7287826538086,
322
+ "learning_rate": 1.6355555555555557e-05,
323
+ "loss": 0.6662,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.429184549356223,
328
+ "grad_norm": 36.765010833740234,
329
+ "learning_rate": 1.6196825396825397e-05,
330
+ "loss": 0.6689,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.572246065808297,
335
+ "grad_norm": 31.146650314331055,
336
+ "learning_rate": 1.603809523809524e-05,
337
+ "loss": 0.475,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.715307582260372,
342
+ "grad_norm": 100.02354431152344,
343
+ "learning_rate": 1.587936507936508e-05,
344
+ "loss": 0.4333,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.858369098712446,
349
+ "grad_norm": 74.33686828613281,
350
+ "learning_rate": 1.5720634920634923e-05,
351
+ "loss": 0.7303,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 6.0,
356
+ "grad_norm": 108.39440155029297,
357
+ "learning_rate": 1.5561904761904762e-05,
358
+ "loss": 0.7531,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 6.0,
363
+ "eval_accuracy": 0.8005390835579514,
364
+ "eval_loss": 0.7989608645439148,
365
+ "eval_macro_f1": 0.7955082870924456,
366
+ "eval_runtime": 0.8314,
367
+ "eval_samples_per_second": 892.452,
368
+ "eval_steps_per_second": 56.53,
369
+ "eval_weighted_f1": 0.7958795167896873,
370
+ "step": 1050
371
+ },
372
+ {
373
+ "epoch": 6.143061516452074,
374
+ "grad_norm": 45.92436981201172,
375
+ "learning_rate": 1.5403174603174605e-05,
376
+ "loss": 0.4501,
377
+ "step": 1075
378
+ },
379
+ {
380
+ "epoch": 6.286123032904149,
381
+ "grad_norm": 50.75591278076172,
382
+ "learning_rate": 1.5244444444444447e-05,
383
+ "loss": 0.4935,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.429184549356223,
388
+ "grad_norm": 102.98908996582031,
389
+ "learning_rate": 1.5085714285714288e-05,
390
+ "loss": 0.4841,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.572246065808297,
395
+ "grad_norm": 71.53900146484375,
396
+ "learning_rate": 1.4926984126984128e-05,
397
+ "loss": 0.4591,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.715307582260372,
402
+ "grad_norm": 200.73916625976562,
403
+ "learning_rate": 1.4768253968253969e-05,
404
+ "loss": 0.3287,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.858369098712446,
409
+ "grad_norm": 143.10678100585938,
410
+ "learning_rate": 1.4609523809523812e-05,
411
+ "loss": 0.4719,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 7.0,
416
+ "grad_norm": 54.07780075073242,
417
+ "learning_rate": 1.4450793650793653e-05,
418
+ "loss": 0.578,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 7.0,
423
+ "eval_accuracy": 0.8126684636118598,
424
+ "eval_loss": 0.888168454170227,
425
+ "eval_macro_f1": 0.8100868380781009,
426
+ "eval_runtime": 0.8353,
427
+ "eval_samples_per_second": 888.287,
428
+ "eval_steps_per_second": 56.266,
429
+ "eval_weighted_f1": 0.8105115755499068,
430
+ "step": 1225
431
+ },
432
+ {
433
+ "epoch": 7.143061516452074,
434
+ "grad_norm": 14.41987419128418,
435
+ "learning_rate": 1.4292063492063493e-05,
436
+ "loss": 0.3751,
437
+ "step": 1250
438
+ },
439
+ {
440
+ "epoch": 7.286123032904149,
441
+ "grad_norm": 136.03530883789062,
442
+ "learning_rate": 1.4133333333333334e-05,
443
+ "loss": 0.587,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.429184549356223,
448
+ "grad_norm": 61.26182174682617,
449
+ "learning_rate": 1.3974603174603177e-05,
450
+ "loss": 0.4754,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.572246065808297,
455
+ "grad_norm": 56.37785720825195,
456
+ "learning_rate": 1.3815873015873018e-05,
457
+ "loss": 0.3748,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.715307582260372,
462
+ "grad_norm": 87.19405364990234,
463
+ "learning_rate": 1.3657142857142858e-05,
464
+ "loss": 0.3728,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.858369098712446,
469
+ "grad_norm": 6.3904876708984375,
470
+ "learning_rate": 1.34984126984127e-05,
471
+ "loss": 0.6402,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 8.0,
476
+ "grad_norm": 0.9372406005859375,
477
+ "learning_rate": 1.333968253968254e-05,
478
+ "loss": 0.3013,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 8.0,
483
+ "eval_accuracy": 0.8140161725067385,
484
+ "eval_loss": 0.9081608653068542,
485
+ "eval_macro_f1": 0.8131613544488082,
486
+ "eval_runtime": 0.848,
487
+ "eval_samples_per_second": 875.019,
488
+ "eval_steps_per_second": 55.426,
489
+ "eval_weighted_f1": 0.8136979617194476,
490
+ "step": 1400
491
+ },
492
+ {
493
+ "epoch": 8.143061516452075,
494
+ "grad_norm": 56.26848220825195,
495
+ "learning_rate": 1.318095238095238e-05,
496
+ "loss": 0.2817,
497
+ "step": 1425
498
+ },
499
+ {
500
+ "epoch": 8.286123032904149,
501
+ "grad_norm": 2.777222156524658,
502
+ "learning_rate": 1.3022222222222223e-05,
503
+ "loss": 0.3345,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.429184549356224,
508
+ "grad_norm": 1.6930785179138184,
509
+ "learning_rate": 1.2863492063492065e-05,
510
+ "loss": 0.427,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.572246065808297,
515
+ "grad_norm": 0.20981667935848236,
516
+ "learning_rate": 1.2704761904761906e-05,
517
+ "loss": 0.1743,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.715307582260372,
522
+ "grad_norm": 0.48982223868370056,
523
+ "learning_rate": 1.2546031746031746e-05,
524
+ "loss": 0.1807,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.858369098712446,
529
+ "grad_norm": 72.02860260009766,
530
+ "learning_rate": 1.2387301587301589e-05,
531
+ "loss": 0.3323,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 9.0,
536
+ "grad_norm": 42.007476806640625,
537
+ "learning_rate": 1.222857142857143e-05,
538
+ "loss": 0.3664,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 9.0,
543
+ "eval_accuracy": 0.8099730458221024,
544
+ "eval_loss": 1.098464012145996,
545
+ "eval_macro_f1": 0.8098916544254772,
546
+ "eval_runtime": 0.851,
547
+ "eval_samples_per_second": 871.869,
548
+ "eval_steps_per_second": 55.226,
549
+ "eval_weighted_f1": 0.8098641707712,
550
+ "step": 1575
551
+ },
552
+ {
553
+ "epoch": 9.143061516452075,
554
+ "grad_norm": 0.31920284032821655,
555
+ "learning_rate": 1.2069841269841271e-05,
556
+ "loss": 0.3034,
557
+ "step": 1600
558
+ },
559
+ {
560
+ "epoch": 9.286123032904149,
561
+ "grad_norm": 2.483795404434204,
562
+ "learning_rate": 1.191111111111111e-05,
563
+ "loss": 0.2451,
564
+ "step": 1625
565
+ },
566
+ {
567
+ "epoch": 9.429184549356224,
568
+ "grad_norm": 5.096146583557129,
569
+ "learning_rate": 1.1752380952380952e-05,
570
+ "loss": 0.2799,
571
+ "step": 1650
572
+ },
573
+ {
574
+ "epoch": 9.572246065808297,
575
+ "grad_norm": 290.0660400390625,
576
+ "learning_rate": 1.1593650793650795e-05,
577
+ "loss": 0.3366,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.715307582260372,
582
+ "grad_norm": 0.3962249159812927,
583
+ "learning_rate": 1.1434920634920636e-05,
584
+ "loss": 0.2094,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.858369098712446,
589
+ "grad_norm": 0.3089562952518463,
590
+ "learning_rate": 1.1276190476190476e-05,
591
+ "loss": 0.2025,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 10.0,
596
+ "grad_norm": 0.048443883657455444,
597
+ "learning_rate": 1.1117460317460317e-05,
598
+ "loss": 0.2143,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 10.0,
603
+ "eval_accuracy": 0.8207547169811321,
604
+ "eval_loss": 1.1146857738494873,
605
+ "eval_macro_f1": 0.8201622337563238,
606
+ "eval_runtime": 0.8438,
607
+ "eval_samples_per_second": 879.403,
608
+ "eval_steps_per_second": 55.703,
609
+ "eval_weighted_f1": 0.8201568140000177,
610
+ "step": 1750
611
+ },
612
+ {
613
+ "epoch": 10.143061516452075,
614
+ "grad_norm": 0.11350910365581512,
615
+ "learning_rate": 1.095873015873016e-05,
616
+ "loss": 0.2029,
617
+ "step": 1775
618
+ },
619
+ {
620
+ "epoch": 10.286123032904149,
621
+ "grad_norm": 215.7249755859375,
622
+ "learning_rate": 1.0800000000000002e-05,
623
+ "loss": 0.0464,
624
+ "step": 1800
625
+ },
626
+ {
627
+ "epoch": 10.429184549356224,
628
+ "grad_norm": 0.207084521651268,
629
+ "learning_rate": 1.0641269841269841e-05,
630
+ "loss": 0.195,
631
+ "step": 1825
632
+ },
633
+ {
634
+ "epoch": 10.572246065808297,
635
+ "grad_norm": 1.5518670082092285,
636
+ "learning_rate": 1.0482539682539683e-05,
637
+ "loss": 0.1188,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.715307582260372,
642
+ "grad_norm": 0.651492714881897,
643
+ "learning_rate": 1.0323809523809526e-05,
644
+ "loss": 0.1733,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.858369098712446,
649
+ "grad_norm": 4.385392189025879,
650
+ "learning_rate": 1.0165079365079367e-05,
651
+ "loss": 0.283,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 11.0,
656
+ "grad_norm": 59.081321716308594,
657
+ "learning_rate": 1.0006349206349206e-05,
658
+ "loss": 0.1867,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 11.0,
663
+ "eval_accuracy": 0.8247978436657682,
664
+ "eval_loss": 1.1053924560546875,
665
+ "eval_macro_f1": 0.8225933246736051,
666
+ "eval_runtime": 0.85,
667
+ "eval_samples_per_second": 872.954,
668
+ "eval_steps_per_second": 55.295,
669
+ "eval_weighted_f1": 0.8229582655673822,
670
+ "step": 1925
671
+ },
672
+ {
673
+ "epoch": 11.143061516452075,
674
+ "grad_norm": 0.31905966997146606,
675
+ "learning_rate": 9.847619047619048e-06,
676
+ "loss": 0.0736,
677
+ "step": 1950
678
+ },
679
+ {
680
+ "epoch": 11.286123032904149,
681
+ "grad_norm": 0.05339258536696434,
682
+ "learning_rate": 9.688888888888889e-06,
683
+ "loss": 0.1638,
684
+ "step": 1975
685
+ },
686
+ {
687
+ "epoch": 11.429184549356224,
688
+ "grad_norm": 0.05733984708786011,
689
+ "learning_rate": 9.53015873015873e-06,
690
+ "loss": 0.1391,
691
+ "step": 2000
692
+ },
693
+ {
694
+ "epoch": 11.572246065808297,
695
+ "grad_norm": 0.04553008824586868,
696
+ "learning_rate": 9.371428571428572e-06,
697
+ "loss": 0.1274,
698
+ "step": 2025
699
+ },
700
+ {
701
+ "epoch": 11.715307582260372,
702
+ "grad_norm": 0.13952229917049408,
703
+ "learning_rate": 9.212698412698413e-06,
704
+ "loss": 0.2145,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 11.858369098712446,
709
+ "grad_norm": 0.042005158960819244,
710
+ "learning_rate": 9.053968253968254e-06,
711
+ "loss": 0.014,
712
+ "step": 2075
713
+ },
714
+ {
715
+ "epoch": 12.0,
716
+ "grad_norm": 7.3003950119018555,
717
+ "learning_rate": 8.895238095238096e-06,
718
+ "loss": 0.09,
719
+ "step": 2100
720
+ },
721
+ {
722
+ "epoch": 12.0,
723
+ "eval_accuracy": 0.8194070080862533,
724
+ "eval_loss": 1.2627427577972412,
725
+ "eval_macro_f1": 0.8175819120708189,
726
+ "eval_runtime": 0.8367,
727
+ "eval_samples_per_second": 886.77,
728
+ "eval_steps_per_second": 56.17,
729
+ "eval_weighted_f1": 0.8178706387836799,
730
+ "step": 2100
731
+ },
732
+ {
733
+ "epoch": 12.143061516452075,
734
+ "grad_norm": 0.04512350261211395,
735
+ "learning_rate": 8.736507936507937e-06,
736
+ "loss": 0.0691,
737
+ "step": 2125
738
+ },
739
+ {
740
+ "epoch": 12.286123032904149,
741
+ "grad_norm": 0.0418599471449852,
742
+ "learning_rate": 8.577777777777778e-06,
743
+ "loss": 0.0586,
744
+ "step": 2150
745
+ },
746
+ {
747
+ "epoch": 12.429184549356224,
748
+ "grad_norm": 0.04520595818758011,
749
+ "learning_rate": 8.41904761904762e-06,
750
+ "loss": 0.082,
751
+ "step": 2175
752
+ },
753
+ {
754
+ "epoch": 12.572246065808297,
755
+ "grad_norm": 30.16748046875,
756
+ "learning_rate": 8.260317460317461e-06,
757
+ "loss": 0.1246,
758
+ "step": 2200
759
+ },
760
+ {
761
+ "epoch": 12.715307582260372,
762
+ "grad_norm": 0.07494717091321945,
763
+ "learning_rate": 8.101587301587302e-06,
764
+ "loss": 0.033,
765
+ "step": 2225
766
+ },
767
+ {
768
+ "epoch": 12.858369098712446,
769
+ "grad_norm": 0.03252115473151207,
770
+ "learning_rate": 7.942857142857144e-06,
771
+ "loss": 0.1144,
772
+ "step": 2250
773
+ },
774
+ {
775
+ "epoch": 13.0,
776
+ "grad_norm": 28.729177474975586,
777
+ "learning_rate": 7.784126984126985e-06,
778
+ "loss": 0.1977,
779
+ "step": 2275
780
+ },
781
+ {
782
+ "epoch": 13.0,
783
+ "eval_accuracy": 0.8342318059299192,
784
+ "eval_loss": 1.1720412969589233,
785
+ "eval_macro_f1": 0.8337519435295319,
786
+ "eval_runtime": 0.8315,
787
+ "eval_samples_per_second": 892.409,
788
+ "eval_steps_per_second": 56.527,
789
+ "eval_weighted_f1": 0.8339796875831188,
790
+ "step": 2275
791
+ },
792
+ {
793
+ "epoch": 13.143061516452075,
794
+ "grad_norm": 5.797855854034424,
795
+ "learning_rate": 7.625396825396826e-06,
796
+ "loss": 0.2229,
797
+ "step": 2300
798
+ },
799
+ {
800
+ "epoch": 13.286123032904149,
801
+ "grad_norm": 0.24729643762111664,
802
+ "learning_rate": 7.4666666666666675e-06,
803
+ "loss": 0.0358,
804
+ "step": 2325
805
+ },
806
+ {
807
+ "epoch": 13.429184549356224,
808
+ "grad_norm": 0.06091878563165665,
809
+ "learning_rate": 7.307936507936508e-06,
810
+ "loss": 0.0894,
811
+ "step": 2350
812
+ },
813
+ {
814
+ "epoch": 13.572246065808297,
815
+ "grad_norm": 0.038106486201286316,
816
+ "learning_rate": 7.14920634920635e-06,
817
+ "loss": 0.0691,
818
+ "step": 2375
819
+ },
820
+ {
821
+ "epoch": 13.715307582260372,
822
+ "grad_norm": 0.058261722326278687,
823
+ "learning_rate": 6.9904761904761905e-06,
824
+ "loss": 0.0793,
825
+ "step": 2400
826
+ },
827
+ {
828
+ "epoch": 13.858369098712446,
829
+ "grad_norm": 0.03747075796127319,
830
+ "learning_rate": 6.831746031746033e-06,
831
+ "loss": 0.0547,
832
+ "step": 2425
833
+ },
834
+ {
835
+ "epoch": 14.0,
836
+ "grad_norm": 1.602468490600586,
837
+ "learning_rate": 6.673015873015873e-06,
838
+ "loss": 0.071,
839
+ "step": 2450
840
+ },
841
+ {
842
+ "epoch": 14.0,
843
+ "eval_accuracy": 0.8247978436657682,
844
+ "eval_loss": 1.2670577764511108,
845
+ "eval_macro_f1": 0.823360740045171,
846
+ "eval_runtime": 0.8477,
847
+ "eval_samples_per_second": 875.268,
848
+ "eval_steps_per_second": 55.442,
849
+ "eval_weighted_f1": 0.823630110741323,
850
+ "step": 2450
851
+ },
852
+ {
853
+ "epoch": 14.143061516452075,
854
+ "grad_norm": 0.02739104814827442,
855
+ "learning_rate": 6.514285714285715e-06,
856
+ "loss": 0.0485,
857
+ "step": 2475
858
+ },
859
+ {
860
+ "epoch": 14.286123032904149,
861
+ "grad_norm": 0.028057582676410675,
862
+ "learning_rate": 6.355555555555556e-06,
863
+ "loss": 0.0502,
864
+ "step": 2500
865
+ },
866
+ {
867
+ "epoch": 14.429184549356224,
868
+ "grad_norm": 279.5613708496094,
869
+ "learning_rate": 6.196825396825398e-06,
870
+ "loss": 0.0431,
871
+ "step": 2525
872
+ },
873
+ {
874
+ "epoch": 14.572246065808297,
875
+ "grad_norm": 6.655642032623291,
876
+ "learning_rate": 6.038095238095238e-06,
877
+ "loss": 0.0505,
878
+ "step": 2550
879
+ },
880
+ {
881
+ "epoch": 14.715307582260372,
882
+ "grad_norm": 293.91485595703125,
883
+ "learning_rate": 5.8793650793650806e-06,
884
+ "loss": 0.0597,
885
+ "step": 2575
886
+ },
887
+ {
888
+ "epoch": 14.858369098712446,
889
+ "grad_norm": 0.10515305399894714,
890
+ "learning_rate": 5.720634920634921e-06,
891
+ "loss": 0.0785,
892
+ "step": 2600
893
+ },
894
+ {
895
+ "epoch": 15.0,
896
+ "grad_norm": 0.43047991394996643,
897
+ "learning_rate": 5.561904761904763e-06,
898
+ "loss": 0.127,
899
+ "step": 2625
900
+ },
901
+ {
902
+ "epoch": 15.0,
903
+ "eval_accuracy": 0.8207547169811321,
904
+ "eval_loss": 1.2633687257766724,
905
+ "eval_macro_f1": 0.8196990300150843,
906
+ "eval_runtime": 0.8644,
907
+ "eval_samples_per_second": 858.378,
908
+ "eval_steps_per_second": 54.372,
909
+ "eval_weighted_f1": 0.8200250401767453,
910
+ "step": 2625
911
+ },
912
+ {
913
+ "epoch": 15.143061516452075,
914
+ "grad_norm": 0.02982182800769806,
915
+ "learning_rate": 5.403174603174604e-06,
916
+ "loss": 0.0322,
917
+ "step": 2650
918
+ },
919
+ {
920
+ "epoch": 15.286123032904149,
921
+ "grad_norm": 0.028712768107652664,
922
+ "learning_rate": 5.244444444444445e-06,
923
+ "loss": 0.0271,
924
+ "step": 2675
925
+ },
926
+ {
927
+ "epoch": 15.429184549356224,
928
+ "grad_norm": 18.056129455566406,
929
+ "learning_rate": 5.085714285714286e-06,
930
+ "loss": 0.0726,
931
+ "step": 2700
932
+ },
933
+ {
934
+ "epoch": 15.572246065808297,
935
+ "grad_norm": 0.05009913071990013,
936
+ "learning_rate": 4.926984126984128e-06,
937
+ "loss": 0.0401,
938
+ "step": 2725
939
+ },
940
+ {
941
+ "epoch": 15.715307582260372,
942
+ "grad_norm": 0.0272875614464283,
943
+ "learning_rate": 4.768253968253969e-06,
944
+ "loss": 0.0817,
945
+ "step": 2750
946
+ },
947
+ {
948
+ "epoch": 15.858369098712446,
949
+ "grad_norm": 0.02538874000310898,
950
+ "learning_rate": 4.60952380952381e-06,
951
+ "loss": 0.0732,
952
+ "step": 2775
953
+ },
954
+ {
955
+ "epoch": 16.0,
956
+ "grad_norm": 0.025316985324025154,
957
+ "learning_rate": 4.4507936507936515e-06,
958
+ "loss": 0.0159,
959
+ "step": 2800
960
+ },
961
+ {
962
+ "epoch": 16.0,
963
+ "eval_accuracy": 0.8099730458221024,
964
+ "eval_loss": 1.3668134212493896,
965
+ "eval_macro_f1": 0.8080128754139545,
966
+ "eval_runtime": 0.8384,
967
+ "eval_samples_per_second": 885.071,
968
+ "eval_steps_per_second": 56.062,
969
+ "eval_weighted_f1": 0.8082965212350719,
970
+ "step": 2800
971
+ }
972
+ ],
973
+ "logging_steps": 25,
974
+ "max_steps": 3500,
975
+ "num_input_tokens_seen": 0,
976
+ "num_train_epochs": 20,
977
+ "save_steps": 500,
978
+ "stateful_callbacks": {
979
+ "EarlyStoppingCallback": {
980
+ "args": {
981
+ "early_stopping_patience": 3,
982
+ "early_stopping_threshold": 0.0
983
+ },
984
+ "attributes": {
985
+ "early_stopping_patience_counter": 3
986
+ }
987
+ },
988
+ "TrainerControl": {
989
+ "args": {
990
+ "should_epoch_stop": false,
991
+ "should_evaluate": false,
992
+ "should_log": false,
993
+ "should_save": true,
994
+ "should_training_stop": true
995
+ },
996
+ "attributes": {}
997
+ }
998
+ },
999
+ "total_flos": 0.0,
1000
+ "train_batch_size": 8,
1001
+ "trial_name": null,
1002
+ "trial_params": null
1003
+ }
ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944/trainer_state.json ADDED
@@ -0,0 +1,1038 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2944,
3
+ "best_metric": 0.8254918793086526,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944",
5
+ "epoch": 16.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2944,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1364256480218281,
14
+ "grad_norm": 13.698607444763184,
15
+ "learning_rate": 1.3043478260869566e-06,
16
+ "loss": 4.4748,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2728512960436562,
21
+ "grad_norm": 8.405525207519531,
22
+ "learning_rate": 2.6630434782608698e-06,
23
+ "loss": 4.4042,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4092769440654843,
28
+ "grad_norm": 14.027036666870117,
29
+ "learning_rate": 4.021739130434783e-06,
30
+ "loss": 4.428,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5457025920873124,
35
+ "grad_norm": 13.634267807006836,
36
+ "learning_rate": 5.380434782608695e-06,
37
+ "loss": 4.3885,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6821282401091405,
42
+ "grad_norm": 14.232049942016602,
43
+ "learning_rate": 6.739130434782609e-06,
44
+ "loss": 4.4061,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8185538881309686,
49
+ "grad_norm": 27.76457405090332,
50
+ "learning_rate": 8.097826086956523e-06,
51
+ "loss": 4.3113,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9549795361527967,
56
+ "grad_norm": 10.04257583618164,
57
+ "learning_rate": 9.456521739130436e-06,
58
+ "loss": 4.3376,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4318181818181818,
64
+ "eval_loss": 1.0600481033325195,
65
+ "eval_macro_f1": 0.33044928653978456,
66
+ "eval_runtime": 0.7955,
67
+ "eval_samples_per_second": 940.271,
68
+ "eval_steps_per_second": 59.081,
69
+ "eval_weighted_f1": 0.33653565188777257,
70
+ "step": 184
71
+ },
72
+ {
73
+ "epoch": 1.08731241473397,
74
+ "grad_norm": 13.584155082702637,
75
+ "learning_rate": 1.0815217391304349e-05,
76
+ "loss": 4.0922,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.223738062755798,
81
+ "grad_norm": 34.05221176147461,
82
+ "learning_rate": 1.2173913043478263e-05,
83
+ "loss": 4.1908,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.360163710777626,
88
+ "grad_norm": 27.138652801513672,
89
+ "learning_rate": 1.3532608695652174e-05,
90
+ "loss": 4.0759,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4965893587994543,
95
+ "grad_norm": 35.364784240722656,
96
+ "learning_rate": 1.4891304347826087e-05,
97
+ "loss": 3.8167,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6330150068212825,
102
+ "grad_norm": 52.16243362426758,
103
+ "learning_rate": 1.6250000000000002e-05,
104
+ "loss": 3.3005,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7694406548431105,
109
+ "grad_norm": 60.48115921020508,
110
+ "learning_rate": 1.7608695652173915e-05,
111
+ "loss": 2.9316,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9058663028649385,
116
+ "grad_norm": 60.63091278076172,
117
+ "learning_rate": 1.8967391304347828e-05,
118
+ "loss": 2.6694,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.7419786096256684,
124
+ "eval_loss": 0.6263679265975952,
125
+ "eval_macro_f1": 0.7350426017747563,
126
+ "eval_runtime": 0.8062,
127
+ "eval_samples_per_second": 927.849,
128
+ "eval_steps_per_second": 58.301,
129
+ "eval_weighted_f1": 0.7364330890420108,
130
+ "step": 368
131
+ },
132
+ {
133
+ "epoch": 2.038199181446112,
134
+ "grad_norm": 26.958227157592773,
135
+ "learning_rate": 1.9963768115942032e-05,
136
+ "loss": 2.5683,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.17462482946794,
141
+ "grad_norm": 53.8985481262207,
142
+ "learning_rate": 1.9812801932367153e-05,
143
+ "loss": 2.2343,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.311050477489768,
148
+ "grad_norm": 61.40669631958008,
149
+ "learning_rate": 1.966183574879227e-05,
150
+ "loss": 2.1879,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.447476125511596,
155
+ "grad_norm": 59.0511474609375,
156
+ "learning_rate": 1.9510869565217393e-05,
157
+ "loss": 2.1735,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.5839017735334244,
162
+ "grad_norm": 37.290348052978516,
163
+ "learning_rate": 1.9359903381642515e-05,
164
+ "loss": 1.9694,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.720327421555252,
169
+ "grad_norm": 36.33408737182617,
170
+ "learning_rate": 1.9208937198067633e-05,
171
+ "loss": 2.1755,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8567530695770804,
176
+ "grad_norm": 41.57215881347656,
177
+ "learning_rate": 1.9057971014492754e-05,
178
+ "loss": 1.8903,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.9931787175989086,
183
+ "grad_norm": 29.599470138549805,
184
+ "learning_rate": 1.8907004830917876e-05,
185
+ "loss": 2.0121,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.8048128342245989,
191
+ "eval_loss": 0.5076773762702942,
192
+ "eval_macro_f1": 0.8023850475862858,
193
+ "eval_runtime": 0.7848,
194
+ "eval_samples_per_second": 953.107,
195
+ "eval_steps_per_second": 59.888,
196
+ "eval_weighted_f1": 0.8030374773044596,
197
+ "step": 552
198
+ },
199
+ {
200
+ "epoch": 3.1255115961800817,
201
+ "grad_norm": 38.99618911743164,
202
+ "learning_rate": 1.8756038647342997e-05,
203
+ "loss": 1.4357,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.26193724420191,
208
+ "grad_norm": 49.06818389892578,
209
+ "learning_rate": 1.8605072463768115e-05,
210
+ "loss": 1.2214,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.398362892223738,
215
+ "grad_norm": 119.49598693847656,
216
+ "learning_rate": 1.8454106280193237e-05,
217
+ "loss": 1.499,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5347885402455663,
222
+ "grad_norm": 39.219112396240234,
223
+ "learning_rate": 1.830314009661836e-05,
224
+ "loss": 1.4728,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6712141882673945,
229
+ "grad_norm": 50.963680267333984,
230
+ "learning_rate": 1.815217391304348e-05,
231
+ "loss": 1.4687,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.8076398362892223,
236
+ "grad_norm": 35.97512435913086,
237
+ "learning_rate": 1.8001207729468598e-05,
238
+ "loss": 1.3411,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.9440654843110505,
243
+ "grad_norm": 101.24478149414062,
244
+ "learning_rate": 1.785024154589372e-05,
245
+ "loss": 1.7263,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.803475935828877,
251
+ "eval_loss": 0.5702747106552124,
252
+ "eval_macro_f1": 0.7982484488032631,
253
+ "eval_runtime": 0.7758,
254
+ "eval_samples_per_second": 964.218,
255
+ "eval_steps_per_second": 60.586,
256
+ "eval_weighted_f1": 0.7989060693412219,
257
+ "step": 736
258
+ },
259
+ {
260
+ "epoch": 4.076398362892224,
261
+ "grad_norm": 10.192824363708496,
262
+ "learning_rate": 1.769927536231884e-05,
263
+ "loss": 1.1625,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.212824010914052,
268
+ "grad_norm": 87.06596374511719,
269
+ "learning_rate": 1.7548309178743963e-05,
270
+ "loss": 0.8203,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.34924965893588,
275
+ "grad_norm": 53.73453903198242,
276
+ "learning_rate": 1.7397342995169084e-05,
277
+ "loss": 1.1154,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.485675306957708,
282
+ "grad_norm": 86.81330108642578,
283
+ "learning_rate": 1.7246376811594206e-05,
284
+ "loss": 0.975,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.622100954979536,
289
+ "grad_norm": 44.46036911010742,
290
+ "learning_rate": 1.7095410628019324e-05,
291
+ "loss": 0.8455,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.758526603001364,
296
+ "grad_norm": 54.869728088378906,
297
+ "learning_rate": 1.6944444444444446e-05,
298
+ "loss": 1.0015,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.894952251023192,
303
+ "grad_norm": 35.585391998291016,
304
+ "learning_rate": 1.6793478260869567e-05,
305
+ "loss": 0.9526,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.7994652406417112,
311
+ "eval_loss": 0.6379645466804504,
312
+ "eval_macro_f1": 0.7996325807700354,
313
+ "eval_runtime": 0.7955,
314
+ "eval_samples_per_second": 940.311,
315
+ "eval_steps_per_second": 59.084,
316
+ "eval_weighted_f1": 0.8001178603315445,
317
+ "step": 920
318
+ },
319
+ {
320
+ "epoch": 5.0272851296043655,
321
+ "grad_norm": 60.07306671142578,
322
+ "learning_rate": 1.664251207729469e-05,
323
+ "loss": 1.0215,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.163710777626194,
328
+ "grad_norm": 50.23310852050781,
329
+ "learning_rate": 1.649154589371981e-05,
330
+ "loss": 0.645,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.300136425648022,
335
+ "grad_norm": 28.637880325317383,
336
+ "learning_rate": 1.634057971014493e-05,
337
+ "loss": 0.5892,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.43656207366985,
342
+ "grad_norm": 25.58955192565918,
343
+ "learning_rate": 1.618961352657005e-05,
344
+ "loss": 0.7481,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.572987721691678,
349
+ "grad_norm": 30.33829116821289,
350
+ "learning_rate": 1.603864734299517e-05,
351
+ "loss": 0.5611,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.709413369713506,
356
+ "grad_norm": 106.69697570800781,
357
+ "learning_rate": 1.5887681159420293e-05,
358
+ "loss": 0.5901,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.845839017735334,
363
+ "grad_norm": 96.58075714111328,
364
+ "learning_rate": 1.5736714975845415e-05,
365
+ "loss": 0.8416,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.982264665757162,
370
+ "grad_norm": 145.4986572265625,
371
+ "learning_rate": 1.5585748792270533e-05,
372
+ "loss": 0.6924,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 6.0,
377
+ "eval_accuracy": 0.8048128342245989,
378
+ "eval_loss": 0.7889442443847656,
379
+ "eval_macro_f1": 0.8061778111906127,
380
+ "eval_runtime": 0.7856,
381
+ "eval_samples_per_second": 952.16,
382
+ "eval_steps_per_second": 59.828,
383
+ "eval_weighted_f1": 0.8066901379149087,
384
+ "step": 1104
385
+ },
386
+ {
387
+ "epoch": 6.114597544338336,
388
+ "grad_norm": 35.43317794799805,
389
+ "learning_rate": 1.5434782608695654e-05,
390
+ "loss": 0.5171,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.251023192360163,
395
+ "grad_norm": 15.85576057434082,
396
+ "learning_rate": 1.5283816425120776e-05,
397
+ "loss": 0.4425,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.387448840381992,
402
+ "grad_norm": 201.84942626953125,
403
+ "learning_rate": 1.5132850241545896e-05,
404
+ "loss": 0.4496,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.52387448840382,
409
+ "grad_norm": 3.254140853881836,
410
+ "learning_rate": 1.4981884057971017e-05,
411
+ "loss": 0.5568,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.660300136425648,
416
+ "grad_norm": 83.1423110961914,
417
+ "learning_rate": 1.4830917874396135e-05,
418
+ "loss": 0.4547,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.796725784447476,
423
+ "grad_norm": 29.0833683013916,
424
+ "learning_rate": 1.4679951690821257e-05,
425
+ "loss": 0.6072,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.933151432469304,
430
+ "grad_norm": 72.84578704833984,
431
+ "learning_rate": 1.4528985507246378e-05,
432
+ "loss": 0.3765,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 7.0,
437
+ "eval_accuracy": 0.8088235294117647,
438
+ "eval_loss": 0.9676269888877869,
439
+ "eval_macro_f1": 0.808268253633415,
440
+ "eval_runtime": 0.7913,
441
+ "eval_samples_per_second": 945.321,
442
+ "eval_steps_per_second": 59.399,
443
+ "eval_weighted_f1": 0.8080799879528013,
444
+ "step": 1288
445
+ },
446
+ {
447
+ "epoch": 7.0654843110504775,
448
+ "grad_norm": 9.068395614624023,
449
+ "learning_rate": 1.43780193236715e-05,
450
+ "loss": 0.5142,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.201909959072306,
455
+ "grad_norm": 104.91226196289062,
456
+ "learning_rate": 1.422705314009662e-05,
457
+ "loss": 0.3286,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.338335607094134,
462
+ "grad_norm": 2.3615524768829346,
463
+ "learning_rate": 1.407608695652174e-05,
464
+ "loss": 0.4441,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.474761255115962,
469
+ "grad_norm": 2.9745988845825195,
470
+ "learning_rate": 1.3925120772946861e-05,
471
+ "loss": 0.3357,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.61118690313779,
476
+ "grad_norm": 0.49000316858291626,
477
+ "learning_rate": 1.3774154589371983e-05,
478
+ "loss": 0.7382,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.747612551159618,
483
+ "grad_norm": 162.61094665527344,
484
+ "learning_rate": 1.3623188405797103e-05,
485
+ "loss": 0.3787,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.884038199181446,
490
+ "grad_norm": 0.5601525902748108,
491
+ "learning_rate": 1.3472222222222224e-05,
492
+ "loss": 0.4359,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 8.0,
497
+ "eval_accuracy": 0.8114973262032086,
498
+ "eval_loss": 1.005163311958313,
499
+ "eval_macro_f1": 0.8100454680654768,
500
+ "eval_runtime": 0.7828,
501
+ "eval_samples_per_second": 955.537,
502
+ "eval_steps_per_second": 60.04,
503
+ "eval_weighted_f1": 0.8108795904417334,
504
+ "step": 1472
505
+ },
506
+ {
507
+ "epoch": 8.016371077762619,
508
+ "grad_norm": 83.41426849365234,
509
+ "learning_rate": 1.3321256038647344e-05,
510
+ "loss": 0.2561,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.152796725784448,
515
+ "grad_norm": 25.318546295166016,
516
+ "learning_rate": 1.3170289855072464e-05,
517
+ "loss": 0.3801,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.289222373806275,
522
+ "grad_norm": 96.02059936523438,
523
+ "learning_rate": 1.3019323671497585e-05,
524
+ "loss": 0.2999,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.425648021828104,
529
+ "grad_norm": 80.63241577148438,
530
+ "learning_rate": 1.2868357487922707e-05,
531
+ "loss": 0.2508,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.562073669849932,
536
+ "grad_norm": 2.628014326095581,
537
+ "learning_rate": 1.2717391304347828e-05,
538
+ "loss": 0.2809,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.69849931787176,
543
+ "grad_norm": 115.1539306640625,
544
+ "learning_rate": 1.2566425120772947e-05,
545
+ "loss": 0.3136,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.834924965893588,
550
+ "grad_norm": 8.638216018676758,
551
+ "learning_rate": 1.2415458937198068e-05,
552
+ "loss": 0.3876,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 8.971350613915416,
557
+ "grad_norm": 21.770883560180664,
558
+ "learning_rate": 1.226449275362319e-05,
559
+ "loss": 0.3336,
560
+ "step": 1650
561
+ },
562
+ {
563
+ "epoch": 9.0,
564
+ "eval_accuracy": 0.8021390374331551,
565
+ "eval_loss": 1.1499035358428955,
566
+ "eval_macro_f1": 0.799858687423883,
567
+ "eval_runtime": 0.7823,
568
+ "eval_samples_per_second": 956.124,
569
+ "eval_steps_per_second": 60.077,
570
+ "eval_weighted_f1": 0.8007311442138713,
571
+ "step": 1656
572
+ },
573
+ {
574
+ "epoch": 9.10368349249659,
575
+ "grad_norm": 42.06435775756836,
576
+ "learning_rate": 1.2113526570048311e-05,
577
+ "loss": 0.1882,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.240109140518417,
582
+ "grad_norm": 0.3936444818973541,
583
+ "learning_rate": 1.1962560386473433e-05,
584
+ "loss": 0.2481,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.376534788540246,
589
+ "grad_norm": 3.401298999786377,
590
+ "learning_rate": 1.181159420289855e-05,
591
+ "loss": 0.269,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.512960436562073,
596
+ "grad_norm": 2.54673171043396,
597
+ "learning_rate": 1.1660628019323672e-05,
598
+ "loss": 0.2902,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 9.649386084583902,
603
+ "grad_norm": 56.028751373291016,
604
+ "learning_rate": 1.1509661835748794e-05,
605
+ "loss": 0.3249,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 9.78581173260573,
610
+ "grad_norm": 6.210036754608154,
611
+ "learning_rate": 1.1358695652173914e-05,
612
+ "loss": 0.1702,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 9.922237380627559,
617
+ "grad_norm": 149.7547149658203,
618
+ "learning_rate": 1.1207729468599035e-05,
619
+ "loss": 0.2522,
620
+ "step": 1825
621
+ },
622
+ {
623
+ "epoch": 10.0,
624
+ "eval_accuracy": 0.8114973262032086,
625
+ "eval_loss": 1.15133798122406,
626
+ "eval_macro_f1": 0.8102329373777906,
627
+ "eval_runtime": 0.7855,
628
+ "eval_samples_per_second": 952.306,
629
+ "eval_steps_per_second": 59.837,
630
+ "eval_weighted_f1": 0.811081420130348,
631
+ "step": 1840
632
+ },
633
+ {
634
+ "epoch": 10.054570259208731,
635
+ "grad_norm": 0.07347074896097183,
636
+ "learning_rate": 1.1056763285024155e-05,
637
+ "loss": 0.1653,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.19099590723056,
642
+ "grad_norm": 5.307486057281494,
643
+ "learning_rate": 1.0905797101449277e-05,
644
+ "loss": 0.1748,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.327421555252387,
649
+ "grad_norm": 15.938469886779785,
650
+ "learning_rate": 1.0754830917874397e-05,
651
+ "loss": 0.118,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.463847203274215,
656
+ "grad_norm": 0.07084821164608002,
657
+ "learning_rate": 1.0603864734299518e-05,
658
+ "loss": 0.1756,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 10.600272851296044,
663
+ "grad_norm": 0.1413358896970749,
664
+ "learning_rate": 1.045289855072464e-05,
665
+ "loss": 0.1429,
666
+ "step": 1950
667
+ },
668
+ {
669
+ "epoch": 10.736698499317871,
670
+ "grad_norm": 3.057255983352661,
671
+ "learning_rate": 1.0301932367149758e-05,
672
+ "loss": 0.1464,
673
+ "step": 1975
674
+ },
675
+ {
676
+ "epoch": 10.8731241473397,
677
+ "grad_norm": 0.1448040008544922,
678
+ "learning_rate": 1.015096618357488e-05,
679
+ "loss": 0.0549,
680
+ "step": 2000
681
+ },
682
+ {
683
+ "epoch": 11.0,
684
+ "eval_accuracy": 0.820855614973262,
685
+ "eval_loss": 1.1292250156402588,
686
+ "eval_macro_f1": 0.8196933576757858,
687
+ "eval_runtime": 0.7845,
688
+ "eval_samples_per_second": 953.483,
689
+ "eval_steps_per_second": 59.911,
690
+ "eval_weighted_f1": 0.8203548439343438,
691
+ "step": 2024
692
+ },
693
+ {
694
+ "epoch": 11.005457025920872,
695
+ "grad_norm": 0.04855697974562645,
696
+ "learning_rate": 1e-05,
697
+ "loss": 0.1918,
698
+ "step": 2025
699
+ },
700
+ {
701
+ "epoch": 11.141882673942701,
702
+ "grad_norm": 0.05384040251374245,
703
+ "learning_rate": 9.849033816425122e-06,
704
+ "loss": 0.086,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 11.278308321964529,
709
+ "grad_norm": 47.7120361328125,
710
+ "learning_rate": 9.698067632850242e-06,
711
+ "loss": 0.1154,
712
+ "step": 2075
713
+ },
714
+ {
715
+ "epoch": 11.414733969986358,
716
+ "grad_norm": 45.42527389526367,
717
+ "learning_rate": 9.547101449275364e-06,
718
+ "loss": 0.1081,
719
+ "step": 2100
720
+ },
721
+ {
722
+ "epoch": 11.551159618008185,
723
+ "grad_norm": 33.62101745605469,
724
+ "learning_rate": 9.396135265700484e-06,
725
+ "loss": 0.1815,
726
+ "step": 2125
727
+ },
728
+ {
729
+ "epoch": 11.687585266030014,
730
+ "grad_norm": 0.4064909815788269,
731
+ "learning_rate": 9.245169082125605e-06,
732
+ "loss": 0.1952,
733
+ "step": 2150
734
+ },
735
+ {
736
+ "epoch": 11.824010914051842,
737
+ "grad_norm": 0.09699592739343643,
738
+ "learning_rate": 9.094202898550725e-06,
739
+ "loss": 0.0889,
740
+ "step": 2175
741
+ },
742
+ {
743
+ "epoch": 11.96043656207367,
744
+ "grad_norm": 0.06755724549293518,
745
+ "learning_rate": 8.943236714975847e-06,
746
+ "loss": 0.1109,
747
+ "step": 2200
748
+ },
749
+ {
750
+ "epoch": 12.0,
751
+ "eval_accuracy": 0.8248663101604278,
752
+ "eval_loss": 1.1814930438995361,
753
+ "eval_macro_f1": 0.8234796920303791,
754
+ "eval_runtime": 0.8306,
755
+ "eval_samples_per_second": 900.551,
756
+ "eval_steps_per_second": 56.585,
757
+ "eval_weighted_f1": 0.8243216010148555,
758
+ "step": 2208
759
+ },
760
+ {
761
+ "epoch": 12.092769440654843,
762
+ "grad_norm": 0.04476558417081833,
763
+ "learning_rate": 8.792270531400966e-06,
764
+ "loss": 0.0878,
765
+ "step": 2225
766
+ },
767
+ {
768
+ "epoch": 12.229195088676672,
769
+ "grad_norm": 0.04227409511804581,
770
+ "learning_rate": 8.641304347826088e-06,
771
+ "loss": 0.0899,
772
+ "step": 2250
773
+ },
774
+ {
775
+ "epoch": 12.3656207366985,
776
+ "grad_norm": 0.03944230452179909,
777
+ "learning_rate": 8.490338164251208e-06,
778
+ "loss": 0.13,
779
+ "step": 2275
780
+ },
781
+ {
782
+ "epoch": 12.502046384720327,
783
+ "grad_norm": 0.03560931980609894,
784
+ "learning_rate": 8.33937198067633e-06,
785
+ "loss": 0.079,
786
+ "step": 2300
787
+ },
788
+ {
789
+ "epoch": 12.638472032742156,
790
+ "grad_norm": 0.032507166266441345,
791
+ "learning_rate": 8.188405797101449e-06,
792
+ "loss": 0.0298,
793
+ "step": 2325
794
+ },
795
+ {
796
+ "epoch": 12.774897680763983,
797
+ "grad_norm": 28.996482849121094,
798
+ "learning_rate": 8.03743961352657e-06,
799
+ "loss": 0.1267,
800
+ "step": 2350
801
+ },
802
+ {
803
+ "epoch": 12.911323328785812,
804
+ "grad_norm": 0.0349334292113781,
805
+ "learning_rate": 7.88647342995169e-06,
806
+ "loss": 0.0799,
807
+ "step": 2375
808
+ },
809
+ {
810
+ "epoch": 13.0,
811
+ "eval_accuracy": 0.8262032085561497,
812
+ "eval_loss": 1.2030959129333496,
813
+ "eval_macro_f1": 0.8249614402043233,
814
+ "eval_runtime": 0.7961,
815
+ "eval_samples_per_second": 939.633,
816
+ "eval_steps_per_second": 59.041,
817
+ "eval_weighted_f1": 0.8256693362217343,
818
+ "step": 2392
819
+ },
820
+ {
821
+ "epoch": 13.043656207366984,
822
+ "grad_norm": 0.05294587090611458,
823
+ "learning_rate": 7.735507246376812e-06,
824
+ "loss": 0.1543,
825
+ "step": 2400
826
+ },
827
+ {
828
+ "epoch": 13.180081855388813,
829
+ "grad_norm": 95.74848175048828,
830
+ "learning_rate": 7.584541062801934e-06,
831
+ "loss": 0.0326,
832
+ "step": 2425
833
+ },
834
+ {
835
+ "epoch": 13.31650750341064,
836
+ "grad_norm": 0.21087993681430817,
837
+ "learning_rate": 7.4335748792270534e-06,
838
+ "loss": 0.0554,
839
+ "step": 2450
840
+ },
841
+ {
842
+ "epoch": 13.45293315143247,
843
+ "grad_norm": 0.03274247795343399,
844
+ "learning_rate": 7.282608695652175e-06,
845
+ "loss": 0.068,
846
+ "step": 2475
847
+ },
848
+ {
849
+ "epoch": 13.589358799454297,
850
+ "grad_norm": 1.2987051010131836,
851
+ "learning_rate": 7.131642512077295e-06,
852
+ "loss": 0.0595,
853
+ "step": 2500
854
+ },
855
+ {
856
+ "epoch": 13.725784447476126,
857
+ "grad_norm": 88.0322036743164,
858
+ "learning_rate": 6.980676328502416e-06,
859
+ "loss": 0.19,
860
+ "step": 2525
861
+ },
862
+ {
863
+ "epoch": 13.862210095497954,
864
+ "grad_norm": 0.03910617530345917,
865
+ "learning_rate": 6.829710144927537e-06,
866
+ "loss": 0.1245,
867
+ "step": 2550
868
+ },
869
+ {
870
+ "epoch": 13.998635743519781,
871
+ "grad_norm": 1.0611357688903809,
872
+ "learning_rate": 6.678743961352657e-06,
873
+ "loss": 0.0397,
874
+ "step": 2575
875
+ },
876
+ {
877
+ "epoch": 14.0,
878
+ "eval_accuracy": 0.8168449197860963,
879
+ "eval_loss": 1.1931055784225464,
880
+ "eval_macro_f1": 0.8144249102142038,
881
+ "eval_runtime": 0.7946,
882
+ "eval_samples_per_second": 941.348,
883
+ "eval_steps_per_second": 59.149,
884
+ "eval_weighted_f1": 0.8152702962092288,
885
+ "step": 2576
886
+ },
887
+ {
888
+ "epoch": 14.130968622100955,
889
+ "grad_norm": 311.4007873535156,
890
+ "learning_rate": 6.5277777777777784e-06,
891
+ "loss": 0.0504,
892
+ "step": 2600
893
+ },
894
+ {
895
+ "epoch": 14.267394270122782,
896
+ "grad_norm": 0.21477359533309937,
897
+ "learning_rate": 6.376811594202898e-06,
898
+ "loss": 0.0531,
899
+ "step": 2625
900
+ },
901
+ {
902
+ "epoch": 14.403819918144611,
903
+ "grad_norm": 0.024863898754119873,
904
+ "learning_rate": 6.22584541062802e-06,
905
+ "loss": 0.1345,
906
+ "step": 2650
907
+ },
908
+ {
909
+ "epoch": 14.540245566166439,
910
+ "grad_norm": 0.032081663608551025,
911
+ "learning_rate": 6.0748792270531405e-06,
912
+ "loss": 0.0562,
913
+ "step": 2675
914
+ },
915
+ {
916
+ "epoch": 14.676671214188268,
917
+ "grad_norm": 0.03855360299348831,
918
+ "learning_rate": 5.923913043478261e-06,
919
+ "loss": 0.1335,
920
+ "step": 2700
921
+ },
922
+ {
923
+ "epoch": 14.813096862210095,
924
+ "grad_norm": 0.027697492390871048,
925
+ "learning_rate": 5.772946859903382e-06,
926
+ "loss": 0.1033,
927
+ "step": 2725
928
+ },
929
+ {
930
+ "epoch": 14.949522510231924,
931
+ "grad_norm": 0.027388738468289375,
932
+ "learning_rate": 5.6219806763285034e-06,
933
+ "loss": 0.0716,
934
+ "step": 2750
935
+ },
936
+ {
937
+ "epoch": 15.0,
938
+ "eval_accuracy": 0.8088235294117647,
939
+ "eval_loss": 1.360243797302246,
940
+ "eval_macro_f1": 0.8076994084347024,
941
+ "eval_runtime": 0.8428,
942
+ "eval_samples_per_second": 887.548,
943
+ "eval_steps_per_second": 55.768,
944
+ "eval_weighted_f1": 0.8083731766424829,
945
+ "step": 2760
946
+ },
947
+ {
948
+ "epoch": 15.081855388813096,
949
+ "grad_norm": 0.025693757459521294,
950
+ "learning_rate": 5.471014492753623e-06,
951
+ "loss": 0.0384,
952
+ "step": 2775
953
+ },
954
+ {
955
+ "epoch": 15.218281036834926,
956
+ "grad_norm": 0.026593701913952827,
957
+ "learning_rate": 5.320048309178745e-06,
958
+ "loss": 0.1151,
959
+ "step": 2800
960
+ },
961
+ {
962
+ "epoch": 15.354706684856753,
963
+ "grad_norm": 0.025015462189912796,
964
+ "learning_rate": 5.169082125603865e-06,
965
+ "loss": 0.0361,
966
+ "step": 2825
967
+ },
968
+ {
969
+ "epoch": 15.491132332878582,
970
+ "grad_norm": 0.2655402421951294,
971
+ "learning_rate": 5.018115942028986e-06,
972
+ "loss": 0.0799,
973
+ "step": 2850
974
+ },
975
+ {
976
+ "epoch": 15.62755798090041,
977
+ "grad_norm": 0.028171056881546974,
978
+ "learning_rate": 4.867149758454107e-06,
979
+ "loss": 0.0516,
980
+ "step": 2875
981
+ },
982
+ {
983
+ "epoch": 15.763983628922237,
984
+ "grad_norm": 0.20584379136562347,
985
+ "learning_rate": 4.716183574879228e-06,
986
+ "loss": 0.0094,
987
+ "step": 2900
988
+ },
989
+ {
990
+ "epoch": 15.900409276944066,
991
+ "grad_norm": 0.022488050162792206,
992
+ "learning_rate": 4.565217391304348e-06,
993
+ "loss": 0.0041,
994
+ "step": 2925
995
+ },
996
+ {
997
+ "epoch": 16.0,
998
+ "eval_accuracy": 0.8262032085561497,
999
+ "eval_loss": 1.2875349521636963,
1000
+ "eval_macro_f1": 0.8254918793086526,
1001
+ "eval_runtime": 0.7797,
1002
+ "eval_samples_per_second": 959.347,
1003
+ "eval_steps_per_second": 60.28,
1004
+ "eval_weighted_f1": 0.8261333678081356,
1005
+ "step": 2944
1006
+ }
1007
+ ],
1008
+ "logging_steps": 25,
1009
+ "max_steps": 3680,
1010
+ "num_input_tokens_seen": 0,
1011
+ "num_train_epochs": 20,
1012
+ "save_steps": 500,
1013
+ "stateful_callbacks": {
1014
+ "EarlyStoppingCallback": {
1015
+ "args": {
1016
+ "early_stopping_patience": 3,
1017
+ "early_stopping_threshold": 0.0
1018
+ },
1019
+ "attributes": {
1020
+ "early_stopping_patience_counter": 0
1021
+ }
1022
+ },
1023
+ "TrainerControl": {
1024
+ "args": {
1025
+ "should_epoch_stop": false,
1026
+ "should_evaluate": false,
1027
+ "should_log": false,
1028
+ "should_save": true,
1029
+ "should_training_stop": false
1030
+ },
1031
+ "attributes": {}
1032
+ }
1033
+ },
1034
+ "total_flos": 0.0,
1035
+ "train_batch_size": 8,
1036
+ "trial_name": null,
1037
+ "trial_params": null
1038
+ }
ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-3496/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-3496/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-3496/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-3496/trainer_state.json ADDED
@@ -0,0 +1,1225 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2944,
3
+ "best_metric": 0.8254918793086526,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-06-s42-20260827T190009Z-6235c159/trainer/checkpoint-2944",
5
+ "epoch": 19.0,
6
+ "eval_steps": 500,
7
+ "global_step": 3496,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1364256480218281,
14
+ "grad_norm": 13.698607444763184,
15
+ "learning_rate": 1.3043478260869566e-06,
16
+ "loss": 4.4748,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2728512960436562,
21
+ "grad_norm": 8.405525207519531,
22
+ "learning_rate": 2.6630434782608698e-06,
23
+ "loss": 4.4042,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4092769440654843,
28
+ "grad_norm": 14.027036666870117,
29
+ "learning_rate": 4.021739130434783e-06,
30
+ "loss": 4.428,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5457025920873124,
35
+ "grad_norm": 13.634267807006836,
36
+ "learning_rate": 5.380434782608695e-06,
37
+ "loss": 4.3885,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6821282401091405,
42
+ "grad_norm": 14.232049942016602,
43
+ "learning_rate": 6.739130434782609e-06,
44
+ "loss": 4.4061,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8185538881309686,
49
+ "grad_norm": 27.76457405090332,
50
+ "learning_rate": 8.097826086956523e-06,
51
+ "loss": 4.3113,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9549795361527967,
56
+ "grad_norm": 10.04257583618164,
57
+ "learning_rate": 9.456521739130436e-06,
58
+ "loss": 4.3376,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4318181818181818,
64
+ "eval_loss": 1.0600481033325195,
65
+ "eval_macro_f1": 0.33044928653978456,
66
+ "eval_runtime": 0.7955,
67
+ "eval_samples_per_second": 940.271,
68
+ "eval_steps_per_second": 59.081,
69
+ "eval_weighted_f1": 0.33653565188777257,
70
+ "step": 184
71
+ },
72
+ {
73
+ "epoch": 1.08731241473397,
74
+ "grad_norm": 13.584155082702637,
75
+ "learning_rate": 1.0815217391304349e-05,
76
+ "loss": 4.0922,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.223738062755798,
81
+ "grad_norm": 34.05221176147461,
82
+ "learning_rate": 1.2173913043478263e-05,
83
+ "loss": 4.1908,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.360163710777626,
88
+ "grad_norm": 27.138652801513672,
89
+ "learning_rate": 1.3532608695652174e-05,
90
+ "loss": 4.0759,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4965893587994543,
95
+ "grad_norm": 35.364784240722656,
96
+ "learning_rate": 1.4891304347826087e-05,
97
+ "loss": 3.8167,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6330150068212825,
102
+ "grad_norm": 52.16243362426758,
103
+ "learning_rate": 1.6250000000000002e-05,
104
+ "loss": 3.3005,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7694406548431105,
109
+ "grad_norm": 60.48115921020508,
110
+ "learning_rate": 1.7608695652173915e-05,
111
+ "loss": 2.9316,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9058663028649385,
116
+ "grad_norm": 60.63091278076172,
117
+ "learning_rate": 1.8967391304347828e-05,
118
+ "loss": 2.6694,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.7419786096256684,
124
+ "eval_loss": 0.6263679265975952,
125
+ "eval_macro_f1": 0.7350426017747563,
126
+ "eval_runtime": 0.8062,
127
+ "eval_samples_per_second": 927.849,
128
+ "eval_steps_per_second": 58.301,
129
+ "eval_weighted_f1": 0.7364330890420108,
130
+ "step": 368
131
+ },
132
+ {
133
+ "epoch": 2.038199181446112,
134
+ "grad_norm": 26.958227157592773,
135
+ "learning_rate": 1.9963768115942032e-05,
136
+ "loss": 2.5683,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.17462482946794,
141
+ "grad_norm": 53.8985481262207,
142
+ "learning_rate": 1.9812801932367153e-05,
143
+ "loss": 2.2343,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.311050477489768,
148
+ "grad_norm": 61.40669631958008,
149
+ "learning_rate": 1.966183574879227e-05,
150
+ "loss": 2.1879,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.447476125511596,
155
+ "grad_norm": 59.0511474609375,
156
+ "learning_rate": 1.9510869565217393e-05,
157
+ "loss": 2.1735,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.5839017735334244,
162
+ "grad_norm": 37.290348052978516,
163
+ "learning_rate": 1.9359903381642515e-05,
164
+ "loss": 1.9694,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.720327421555252,
169
+ "grad_norm": 36.33408737182617,
170
+ "learning_rate": 1.9208937198067633e-05,
171
+ "loss": 2.1755,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8567530695770804,
176
+ "grad_norm": 41.57215881347656,
177
+ "learning_rate": 1.9057971014492754e-05,
178
+ "loss": 1.8903,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.9931787175989086,
183
+ "grad_norm": 29.599470138549805,
184
+ "learning_rate": 1.8907004830917876e-05,
185
+ "loss": 2.0121,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.8048128342245989,
191
+ "eval_loss": 0.5076773762702942,
192
+ "eval_macro_f1": 0.8023850475862858,
193
+ "eval_runtime": 0.7848,
194
+ "eval_samples_per_second": 953.107,
195
+ "eval_steps_per_second": 59.888,
196
+ "eval_weighted_f1": 0.8030374773044596,
197
+ "step": 552
198
+ },
199
+ {
200
+ "epoch": 3.1255115961800817,
201
+ "grad_norm": 38.99618911743164,
202
+ "learning_rate": 1.8756038647342997e-05,
203
+ "loss": 1.4357,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.26193724420191,
208
+ "grad_norm": 49.06818389892578,
209
+ "learning_rate": 1.8605072463768115e-05,
210
+ "loss": 1.2214,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.398362892223738,
215
+ "grad_norm": 119.49598693847656,
216
+ "learning_rate": 1.8454106280193237e-05,
217
+ "loss": 1.499,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5347885402455663,
222
+ "grad_norm": 39.219112396240234,
223
+ "learning_rate": 1.830314009661836e-05,
224
+ "loss": 1.4728,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6712141882673945,
229
+ "grad_norm": 50.963680267333984,
230
+ "learning_rate": 1.815217391304348e-05,
231
+ "loss": 1.4687,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.8076398362892223,
236
+ "grad_norm": 35.97512435913086,
237
+ "learning_rate": 1.8001207729468598e-05,
238
+ "loss": 1.3411,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.9440654843110505,
243
+ "grad_norm": 101.24478149414062,
244
+ "learning_rate": 1.785024154589372e-05,
245
+ "loss": 1.7263,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.803475935828877,
251
+ "eval_loss": 0.5702747106552124,
252
+ "eval_macro_f1": 0.7982484488032631,
253
+ "eval_runtime": 0.7758,
254
+ "eval_samples_per_second": 964.218,
255
+ "eval_steps_per_second": 60.586,
256
+ "eval_weighted_f1": 0.7989060693412219,
257
+ "step": 736
258
+ },
259
+ {
260
+ "epoch": 4.076398362892224,
261
+ "grad_norm": 10.192824363708496,
262
+ "learning_rate": 1.769927536231884e-05,
263
+ "loss": 1.1625,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.212824010914052,
268
+ "grad_norm": 87.06596374511719,
269
+ "learning_rate": 1.7548309178743963e-05,
270
+ "loss": 0.8203,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.34924965893588,
275
+ "grad_norm": 53.73453903198242,
276
+ "learning_rate": 1.7397342995169084e-05,
277
+ "loss": 1.1154,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.485675306957708,
282
+ "grad_norm": 86.81330108642578,
283
+ "learning_rate": 1.7246376811594206e-05,
284
+ "loss": 0.975,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.622100954979536,
289
+ "grad_norm": 44.46036911010742,
290
+ "learning_rate": 1.7095410628019324e-05,
291
+ "loss": 0.8455,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.758526603001364,
296
+ "grad_norm": 54.869728088378906,
297
+ "learning_rate": 1.6944444444444446e-05,
298
+ "loss": 1.0015,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.894952251023192,
303
+ "grad_norm": 35.585391998291016,
304
+ "learning_rate": 1.6793478260869567e-05,
305
+ "loss": 0.9526,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.7994652406417112,
311
+ "eval_loss": 0.6379645466804504,
312
+ "eval_macro_f1": 0.7996325807700354,
313
+ "eval_runtime": 0.7955,
314
+ "eval_samples_per_second": 940.311,
315
+ "eval_steps_per_second": 59.084,
316
+ "eval_weighted_f1": 0.8001178603315445,
317
+ "step": 920
318
+ },
319
+ {
320
+ "epoch": 5.0272851296043655,
321
+ "grad_norm": 60.07306671142578,
322
+ "learning_rate": 1.664251207729469e-05,
323
+ "loss": 1.0215,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.163710777626194,
328
+ "grad_norm": 50.23310852050781,
329
+ "learning_rate": 1.649154589371981e-05,
330
+ "loss": 0.645,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.300136425648022,
335
+ "grad_norm": 28.637880325317383,
336
+ "learning_rate": 1.634057971014493e-05,
337
+ "loss": 0.5892,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.43656207366985,
342
+ "grad_norm": 25.58955192565918,
343
+ "learning_rate": 1.618961352657005e-05,
344
+ "loss": 0.7481,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.572987721691678,
349
+ "grad_norm": 30.33829116821289,
350
+ "learning_rate": 1.603864734299517e-05,
351
+ "loss": 0.5611,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.709413369713506,
356
+ "grad_norm": 106.69697570800781,
357
+ "learning_rate": 1.5887681159420293e-05,
358
+ "loss": 0.5901,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.845839017735334,
363
+ "grad_norm": 96.58075714111328,
364
+ "learning_rate": 1.5736714975845415e-05,
365
+ "loss": 0.8416,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.982264665757162,
370
+ "grad_norm": 145.4986572265625,
371
+ "learning_rate": 1.5585748792270533e-05,
372
+ "loss": 0.6924,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 6.0,
377
+ "eval_accuracy": 0.8048128342245989,
378
+ "eval_loss": 0.7889442443847656,
379
+ "eval_macro_f1": 0.8061778111906127,
380
+ "eval_runtime": 0.7856,
381
+ "eval_samples_per_second": 952.16,
382
+ "eval_steps_per_second": 59.828,
383
+ "eval_weighted_f1": 0.8066901379149087,
384
+ "step": 1104
385
+ },
386
+ {
387
+ "epoch": 6.114597544338336,
388
+ "grad_norm": 35.43317794799805,
389
+ "learning_rate": 1.5434782608695654e-05,
390
+ "loss": 0.5171,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.251023192360163,
395
+ "grad_norm": 15.85576057434082,
396
+ "learning_rate": 1.5283816425120776e-05,
397
+ "loss": 0.4425,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.387448840381992,
402
+ "grad_norm": 201.84942626953125,
403
+ "learning_rate": 1.5132850241545896e-05,
404
+ "loss": 0.4496,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.52387448840382,
409
+ "grad_norm": 3.254140853881836,
410
+ "learning_rate": 1.4981884057971017e-05,
411
+ "loss": 0.5568,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.660300136425648,
416
+ "grad_norm": 83.1423110961914,
417
+ "learning_rate": 1.4830917874396135e-05,
418
+ "loss": 0.4547,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.796725784447476,
423
+ "grad_norm": 29.0833683013916,
424
+ "learning_rate": 1.4679951690821257e-05,
425
+ "loss": 0.6072,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.933151432469304,
430
+ "grad_norm": 72.84578704833984,
431
+ "learning_rate": 1.4528985507246378e-05,
432
+ "loss": 0.3765,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 7.0,
437
+ "eval_accuracy": 0.8088235294117647,
438
+ "eval_loss": 0.9676269888877869,
439
+ "eval_macro_f1": 0.808268253633415,
440
+ "eval_runtime": 0.7913,
441
+ "eval_samples_per_second": 945.321,
442
+ "eval_steps_per_second": 59.399,
443
+ "eval_weighted_f1": 0.8080799879528013,
444
+ "step": 1288
445
+ },
446
+ {
447
+ "epoch": 7.0654843110504775,
448
+ "grad_norm": 9.068395614624023,
449
+ "learning_rate": 1.43780193236715e-05,
450
+ "loss": 0.5142,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.201909959072306,
455
+ "grad_norm": 104.91226196289062,
456
+ "learning_rate": 1.422705314009662e-05,
457
+ "loss": 0.3286,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.338335607094134,
462
+ "grad_norm": 2.3615524768829346,
463
+ "learning_rate": 1.407608695652174e-05,
464
+ "loss": 0.4441,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.474761255115962,
469
+ "grad_norm": 2.9745988845825195,
470
+ "learning_rate": 1.3925120772946861e-05,
471
+ "loss": 0.3357,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.61118690313779,
476
+ "grad_norm": 0.49000316858291626,
477
+ "learning_rate": 1.3774154589371983e-05,
478
+ "loss": 0.7382,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.747612551159618,
483
+ "grad_norm": 162.61094665527344,
484
+ "learning_rate": 1.3623188405797103e-05,
485
+ "loss": 0.3787,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.884038199181446,
490
+ "grad_norm": 0.5601525902748108,
491
+ "learning_rate": 1.3472222222222224e-05,
492
+ "loss": 0.4359,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 8.0,
497
+ "eval_accuracy": 0.8114973262032086,
498
+ "eval_loss": 1.005163311958313,
499
+ "eval_macro_f1": 0.8100454680654768,
500
+ "eval_runtime": 0.7828,
501
+ "eval_samples_per_second": 955.537,
502
+ "eval_steps_per_second": 60.04,
503
+ "eval_weighted_f1": 0.8108795904417334,
504
+ "step": 1472
505
+ },
506
+ {
507
+ "epoch": 8.016371077762619,
508
+ "grad_norm": 83.41426849365234,
509
+ "learning_rate": 1.3321256038647344e-05,
510
+ "loss": 0.2561,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.152796725784448,
515
+ "grad_norm": 25.318546295166016,
516
+ "learning_rate": 1.3170289855072464e-05,
517
+ "loss": 0.3801,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.289222373806275,
522
+ "grad_norm": 96.02059936523438,
523
+ "learning_rate": 1.3019323671497585e-05,
524
+ "loss": 0.2999,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.425648021828104,
529
+ "grad_norm": 80.63241577148438,
530
+ "learning_rate": 1.2868357487922707e-05,
531
+ "loss": 0.2508,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.562073669849932,
536
+ "grad_norm": 2.628014326095581,
537
+ "learning_rate": 1.2717391304347828e-05,
538
+ "loss": 0.2809,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.69849931787176,
543
+ "grad_norm": 115.1539306640625,
544
+ "learning_rate": 1.2566425120772947e-05,
545
+ "loss": 0.3136,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.834924965893588,
550
+ "grad_norm": 8.638216018676758,
551
+ "learning_rate": 1.2415458937198068e-05,
552
+ "loss": 0.3876,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 8.971350613915416,
557
+ "grad_norm": 21.770883560180664,
558
+ "learning_rate": 1.226449275362319e-05,
559
+ "loss": 0.3336,
560
+ "step": 1650
561
+ },
562
+ {
563
+ "epoch": 9.0,
564
+ "eval_accuracy": 0.8021390374331551,
565
+ "eval_loss": 1.1499035358428955,
566
+ "eval_macro_f1": 0.799858687423883,
567
+ "eval_runtime": 0.7823,
568
+ "eval_samples_per_second": 956.124,
569
+ "eval_steps_per_second": 60.077,
570
+ "eval_weighted_f1": 0.8007311442138713,
571
+ "step": 1656
572
+ },
573
+ {
574
+ "epoch": 9.10368349249659,
575
+ "grad_norm": 42.06435775756836,
576
+ "learning_rate": 1.2113526570048311e-05,
577
+ "loss": 0.1882,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.240109140518417,
582
+ "grad_norm": 0.3936444818973541,
583
+ "learning_rate": 1.1962560386473433e-05,
584
+ "loss": 0.2481,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.376534788540246,
589
+ "grad_norm": 3.401298999786377,
590
+ "learning_rate": 1.181159420289855e-05,
591
+ "loss": 0.269,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.512960436562073,
596
+ "grad_norm": 2.54673171043396,
597
+ "learning_rate": 1.1660628019323672e-05,
598
+ "loss": 0.2902,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 9.649386084583902,
603
+ "grad_norm": 56.028751373291016,
604
+ "learning_rate": 1.1509661835748794e-05,
605
+ "loss": 0.3249,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 9.78581173260573,
610
+ "grad_norm": 6.210036754608154,
611
+ "learning_rate": 1.1358695652173914e-05,
612
+ "loss": 0.1702,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 9.922237380627559,
617
+ "grad_norm": 149.7547149658203,
618
+ "learning_rate": 1.1207729468599035e-05,
619
+ "loss": 0.2522,
620
+ "step": 1825
621
+ },
622
+ {
623
+ "epoch": 10.0,
624
+ "eval_accuracy": 0.8114973262032086,
625
+ "eval_loss": 1.15133798122406,
626
+ "eval_macro_f1": 0.8102329373777906,
627
+ "eval_runtime": 0.7855,
628
+ "eval_samples_per_second": 952.306,
629
+ "eval_steps_per_second": 59.837,
630
+ "eval_weighted_f1": 0.811081420130348,
631
+ "step": 1840
632
+ },
633
+ {
634
+ "epoch": 10.054570259208731,
635
+ "grad_norm": 0.07347074896097183,
636
+ "learning_rate": 1.1056763285024155e-05,
637
+ "loss": 0.1653,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.19099590723056,
642
+ "grad_norm": 5.307486057281494,
643
+ "learning_rate": 1.0905797101449277e-05,
644
+ "loss": 0.1748,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.327421555252387,
649
+ "grad_norm": 15.938469886779785,
650
+ "learning_rate": 1.0754830917874397e-05,
651
+ "loss": 0.118,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.463847203274215,
656
+ "grad_norm": 0.07084821164608002,
657
+ "learning_rate": 1.0603864734299518e-05,
658
+ "loss": 0.1756,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 10.600272851296044,
663
+ "grad_norm": 0.1413358896970749,
664
+ "learning_rate": 1.045289855072464e-05,
665
+ "loss": 0.1429,
666
+ "step": 1950
667
+ },
668
+ {
669
+ "epoch": 10.736698499317871,
670
+ "grad_norm": 3.057255983352661,
671
+ "learning_rate": 1.0301932367149758e-05,
672
+ "loss": 0.1464,
673
+ "step": 1975
674
+ },
675
+ {
676
+ "epoch": 10.8731241473397,
677
+ "grad_norm": 0.1448040008544922,
678
+ "learning_rate": 1.015096618357488e-05,
679
+ "loss": 0.0549,
680
+ "step": 2000
681
+ },
682
+ {
683
+ "epoch": 11.0,
684
+ "eval_accuracy": 0.820855614973262,
685
+ "eval_loss": 1.1292250156402588,
686
+ "eval_macro_f1": 0.8196933576757858,
687
+ "eval_runtime": 0.7845,
688
+ "eval_samples_per_second": 953.483,
689
+ "eval_steps_per_second": 59.911,
690
+ "eval_weighted_f1": 0.8203548439343438,
691
+ "step": 2024
692
+ },
693
+ {
694
+ "epoch": 11.005457025920872,
695
+ "grad_norm": 0.04855697974562645,
696
+ "learning_rate": 1e-05,
697
+ "loss": 0.1918,
698
+ "step": 2025
699
+ },
700
+ {
701
+ "epoch": 11.141882673942701,
702
+ "grad_norm": 0.05384040251374245,
703
+ "learning_rate": 9.849033816425122e-06,
704
+ "loss": 0.086,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 11.278308321964529,
709
+ "grad_norm": 47.7120361328125,
710
+ "learning_rate": 9.698067632850242e-06,
711
+ "loss": 0.1154,
712
+ "step": 2075
713
+ },
714
+ {
715
+ "epoch": 11.414733969986358,
716
+ "grad_norm": 45.42527389526367,
717
+ "learning_rate": 9.547101449275364e-06,
718
+ "loss": 0.1081,
719
+ "step": 2100
720
+ },
721
+ {
722
+ "epoch": 11.551159618008185,
723
+ "grad_norm": 33.62101745605469,
724
+ "learning_rate": 9.396135265700484e-06,
725
+ "loss": 0.1815,
726
+ "step": 2125
727
+ },
728
+ {
729
+ "epoch": 11.687585266030014,
730
+ "grad_norm": 0.4064909815788269,
731
+ "learning_rate": 9.245169082125605e-06,
732
+ "loss": 0.1952,
733
+ "step": 2150
734
+ },
735
+ {
736
+ "epoch": 11.824010914051842,
737
+ "grad_norm": 0.09699592739343643,
738
+ "learning_rate": 9.094202898550725e-06,
739
+ "loss": 0.0889,
740
+ "step": 2175
741
+ },
742
+ {
743
+ "epoch": 11.96043656207367,
744
+ "grad_norm": 0.06755724549293518,
745
+ "learning_rate": 8.943236714975847e-06,
746
+ "loss": 0.1109,
747
+ "step": 2200
748
+ },
749
+ {
750
+ "epoch": 12.0,
751
+ "eval_accuracy": 0.8248663101604278,
752
+ "eval_loss": 1.1814930438995361,
753
+ "eval_macro_f1": 0.8234796920303791,
754
+ "eval_runtime": 0.8306,
755
+ "eval_samples_per_second": 900.551,
756
+ "eval_steps_per_second": 56.585,
757
+ "eval_weighted_f1": 0.8243216010148555,
758
+ "step": 2208
759
+ },
760
+ {
761
+ "epoch": 12.092769440654843,
762
+ "grad_norm": 0.04476558417081833,
763
+ "learning_rate": 8.792270531400966e-06,
764
+ "loss": 0.0878,
765
+ "step": 2225
766
+ },
767
+ {
768
+ "epoch": 12.229195088676672,
769
+ "grad_norm": 0.04227409511804581,
770
+ "learning_rate": 8.641304347826088e-06,
771
+ "loss": 0.0899,
772
+ "step": 2250
773
+ },
774
+ {
775
+ "epoch": 12.3656207366985,
776
+ "grad_norm": 0.03944230452179909,
777
+ "learning_rate": 8.490338164251208e-06,
778
+ "loss": 0.13,
779
+ "step": 2275
780
+ },
781
+ {
782
+ "epoch": 12.502046384720327,
783
+ "grad_norm": 0.03560931980609894,
784
+ "learning_rate": 8.33937198067633e-06,
785
+ "loss": 0.079,
786
+ "step": 2300
787
+ },
788
+ {
789
+ "epoch": 12.638472032742156,
790
+ "grad_norm": 0.032507166266441345,
791
+ "learning_rate": 8.188405797101449e-06,
792
+ "loss": 0.0298,
793
+ "step": 2325
794
+ },
795
+ {
796
+ "epoch": 12.774897680763983,
797
+ "grad_norm": 28.996482849121094,
798
+ "learning_rate": 8.03743961352657e-06,
799
+ "loss": 0.1267,
800
+ "step": 2350
801
+ },
802
+ {
803
+ "epoch": 12.911323328785812,
804
+ "grad_norm": 0.0349334292113781,
805
+ "learning_rate": 7.88647342995169e-06,
806
+ "loss": 0.0799,
807
+ "step": 2375
808
+ },
809
+ {
810
+ "epoch": 13.0,
811
+ "eval_accuracy": 0.8262032085561497,
812
+ "eval_loss": 1.2030959129333496,
813
+ "eval_macro_f1": 0.8249614402043233,
814
+ "eval_runtime": 0.7961,
815
+ "eval_samples_per_second": 939.633,
816
+ "eval_steps_per_second": 59.041,
817
+ "eval_weighted_f1": 0.8256693362217343,
818
+ "step": 2392
819
+ },
820
+ {
821
+ "epoch": 13.043656207366984,
822
+ "grad_norm": 0.05294587090611458,
823
+ "learning_rate": 7.735507246376812e-06,
824
+ "loss": 0.1543,
825
+ "step": 2400
826
+ },
827
+ {
828
+ "epoch": 13.180081855388813,
829
+ "grad_norm": 95.74848175048828,
830
+ "learning_rate": 7.584541062801934e-06,
831
+ "loss": 0.0326,
832
+ "step": 2425
833
+ },
834
+ {
835
+ "epoch": 13.31650750341064,
836
+ "grad_norm": 0.21087993681430817,
837
+ "learning_rate": 7.4335748792270534e-06,
838
+ "loss": 0.0554,
839
+ "step": 2450
840
+ },
841
+ {
842
+ "epoch": 13.45293315143247,
843
+ "grad_norm": 0.03274247795343399,
844
+ "learning_rate": 7.282608695652175e-06,
845
+ "loss": 0.068,
846
+ "step": 2475
847
+ },
848
+ {
849
+ "epoch": 13.589358799454297,
850
+ "grad_norm": 1.2987051010131836,
851
+ "learning_rate": 7.131642512077295e-06,
852
+ "loss": 0.0595,
853
+ "step": 2500
854
+ },
855
+ {
856
+ "epoch": 13.725784447476126,
857
+ "grad_norm": 88.0322036743164,
858
+ "learning_rate": 6.980676328502416e-06,
859
+ "loss": 0.19,
860
+ "step": 2525
861
+ },
862
+ {
863
+ "epoch": 13.862210095497954,
864
+ "grad_norm": 0.03910617530345917,
865
+ "learning_rate": 6.829710144927537e-06,
866
+ "loss": 0.1245,
867
+ "step": 2550
868
+ },
869
+ {
870
+ "epoch": 13.998635743519781,
871
+ "grad_norm": 1.0611357688903809,
872
+ "learning_rate": 6.678743961352657e-06,
873
+ "loss": 0.0397,
874
+ "step": 2575
875
+ },
876
+ {
877
+ "epoch": 14.0,
878
+ "eval_accuracy": 0.8168449197860963,
879
+ "eval_loss": 1.1931055784225464,
880
+ "eval_macro_f1": 0.8144249102142038,
881
+ "eval_runtime": 0.7946,
882
+ "eval_samples_per_second": 941.348,
883
+ "eval_steps_per_second": 59.149,
884
+ "eval_weighted_f1": 0.8152702962092288,
885
+ "step": 2576
886
+ },
887
+ {
888
+ "epoch": 14.130968622100955,
889
+ "grad_norm": 311.4007873535156,
890
+ "learning_rate": 6.5277777777777784e-06,
891
+ "loss": 0.0504,
892
+ "step": 2600
893
+ },
894
+ {
895
+ "epoch": 14.267394270122782,
896
+ "grad_norm": 0.21477359533309937,
897
+ "learning_rate": 6.376811594202898e-06,
898
+ "loss": 0.0531,
899
+ "step": 2625
900
+ },
901
+ {
902
+ "epoch": 14.403819918144611,
903
+ "grad_norm": 0.024863898754119873,
904
+ "learning_rate": 6.22584541062802e-06,
905
+ "loss": 0.1345,
906
+ "step": 2650
907
+ },
908
+ {
909
+ "epoch": 14.540245566166439,
910
+ "grad_norm": 0.032081663608551025,
911
+ "learning_rate": 6.0748792270531405e-06,
912
+ "loss": 0.0562,
913
+ "step": 2675
914
+ },
915
+ {
916
+ "epoch": 14.676671214188268,
917
+ "grad_norm": 0.03855360299348831,
918
+ "learning_rate": 5.923913043478261e-06,
919
+ "loss": 0.1335,
920
+ "step": 2700
921
+ },
922
+ {
923
+ "epoch": 14.813096862210095,
924
+ "grad_norm": 0.027697492390871048,
925
+ "learning_rate": 5.772946859903382e-06,
926
+ "loss": 0.1033,
927
+ "step": 2725
928
+ },
929
+ {
930
+ "epoch": 14.949522510231924,
931
+ "grad_norm": 0.027388738468289375,
932
+ "learning_rate": 5.6219806763285034e-06,
933
+ "loss": 0.0716,
934
+ "step": 2750
935
+ },
936
+ {
937
+ "epoch": 15.0,
938
+ "eval_accuracy": 0.8088235294117647,
939
+ "eval_loss": 1.360243797302246,
940
+ "eval_macro_f1": 0.8076994084347024,
941
+ "eval_runtime": 0.8428,
942
+ "eval_samples_per_second": 887.548,
943
+ "eval_steps_per_second": 55.768,
944
+ "eval_weighted_f1": 0.8083731766424829,
945
+ "step": 2760
946
+ },
947
+ {
948
+ "epoch": 15.081855388813096,
949
+ "grad_norm": 0.025693757459521294,
950
+ "learning_rate": 5.471014492753623e-06,
951
+ "loss": 0.0384,
952
+ "step": 2775
953
+ },
954
+ {
955
+ "epoch": 15.218281036834926,
956
+ "grad_norm": 0.026593701913952827,
957
+ "learning_rate": 5.320048309178745e-06,
958
+ "loss": 0.1151,
959
+ "step": 2800
960
+ },
961
+ {
962
+ "epoch": 15.354706684856753,
963
+ "grad_norm": 0.025015462189912796,
964
+ "learning_rate": 5.169082125603865e-06,
965
+ "loss": 0.0361,
966
+ "step": 2825
967
+ },
968
+ {
969
+ "epoch": 15.491132332878582,
970
+ "grad_norm": 0.2655402421951294,
971
+ "learning_rate": 5.018115942028986e-06,
972
+ "loss": 0.0799,
973
+ "step": 2850
974
+ },
975
+ {
976
+ "epoch": 15.62755798090041,
977
+ "grad_norm": 0.028171056881546974,
978
+ "learning_rate": 4.867149758454107e-06,
979
+ "loss": 0.0516,
980
+ "step": 2875
981
+ },
982
+ {
983
+ "epoch": 15.763983628922237,
984
+ "grad_norm": 0.20584379136562347,
985
+ "learning_rate": 4.716183574879228e-06,
986
+ "loss": 0.0094,
987
+ "step": 2900
988
+ },
989
+ {
990
+ "epoch": 15.900409276944066,
991
+ "grad_norm": 0.022488050162792206,
992
+ "learning_rate": 4.565217391304348e-06,
993
+ "loss": 0.0041,
994
+ "step": 2925
995
+ },
996
+ {
997
+ "epoch": 16.0,
998
+ "eval_accuracy": 0.8262032085561497,
999
+ "eval_loss": 1.2875349521636963,
1000
+ "eval_macro_f1": 0.8254918793086526,
1001
+ "eval_runtime": 0.7797,
1002
+ "eval_samples_per_second": 959.347,
1003
+ "eval_steps_per_second": 60.28,
1004
+ "eval_weighted_f1": 0.8261333678081356,
1005
+ "step": 2944
1006
+ },
1007
+ {
1008
+ "epoch": 16.032742155525238,
1009
+ "grad_norm": 0.029698899015784264,
1010
+ "learning_rate": 4.414251207729469e-06,
1011
+ "loss": 0.0033,
1012
+ "step": 2950
1013
+ },
1014
+ {
1015
+ "epoch": 16.169167803547065,
1016
+ "grad_norm": 0.0360022634267807,
1017
+ "learning_rate": 4.26328502415459e-06,
1018
+ "loss": 0.0026,
1019
+ "step": 2975
1020
+ },
1021
+ {
1022
+ "epoch": 16.305593451568896,
1023
+ "grad_norm": 0.025346431881189346,
1024
+ "learning_rate": 4.11231884057971e-06,
1025
+ "loss": 0.0174,
1026
+ "step": 3000
1027
+ },
1028
+ {
1029
+ "epoch": 16.442019099590723,
1030
+ "grad_norm": 0.02235783450305462,
1031
+ "learning_rate": 3.961352657004831e-06,
1032
+ "loss": 0.1101,
1033
+ "step": 3025
1034
+ },
1035
+ {
1036
+ "epoch": 16.57844474761255,
1037
+ "grad_norm": 0.5413618683815002,
1038
+ "learning_rate": 3.810386473429952e-06,
1039
+ "loss": 0.0223,
1040
+ "step": 3050
1041
+ },
1042
+ {
1043
+ "epoch": 16.714870395634378,
1044
+ "grad_norm": 0.027648868039250374,
1045
+ "learning_rate": 3.659420289855073e-06,
1046
+ "loss": 0.0598,
1047
+ "step": 3075
1048
+ },
1049
+ {
1050
+ "epoch": 16.85129604365621,
1051
+ "grad_norm": 0.02309337444603443,
1052
+ "learning_rate": 3.5084541062801935e-06,
1053
+ "loss": 0.0534,
1054
+ "step": 3100
1055
+ },
1056
+ {
1057
+ "epoch": 16.987721691678036,
1058
+ "grad_norm": 0.030747761949896812,
1059
+ "learning_rate": 3.3574879227053142e-06,
1060
+ "loss": 0.0027,
1061
+ "step": 3125
1062
+ },
1063
+ {
1064
+ "epoch": 17.0,
1065
+ "eval_accuracy": 0.8088235294117647,
1066
+ "eval_loss": 1.4057111740112305,
1067
+ "eval_macro_f1": 0.8078065144581422,
1068
+ "eval_runtime": 0.7826,
1069
+ "eval_samples_per_second": 955.778,
1070
+ "eval_steps_per_second": 60.056,
1071
+ "eval_weighted_f1": 0.808302638369542,
1072
+ "step": 3128
1073
+ },
1074
+ {
1075
+ "epoch": 17.12005457025921,
1076
+ "grad_norm": 0.020325487479567528,
1077
+ "learning_rate": 3.206521739130435e-06,
1078
+ "loss": 0.0631,
1079
+ "step": 3150
1080
+ },
1081
+ {
1082
+ "epoch": 17.256480218281038,
1083
+ "grad_norm": 0.021230516955256462,
1084
+ "learning_rate": 3.055555555555556e-06,
1085
+ "loss": 0.0059,
1086
+ "step": 3175
1087
+ },
1088
+ {
1089
+ "epoch": 17.392905866302865,
1090
+ "grad_norm": 3.355180501937866,
1091
+ "learning_rate": 2.9045893719806767e-06,
1092
+ "loss": 0.0422,
1093
+ "step": 3200
1094
+ },
1095
+ {
1096
+ "epoch": 17.529331514324692,
1097
+ "grad_norm": 0.022968197241425514,
1098
+ "learning_rate": 2.7536231884057974e-06,
1099
+ "loss": 0.0481,
1100
+ "step": 3225
1101
+ },
1102
+ {
1103
+ "epoch": 17.66575716234652,
1104
+ "grad_norm": 574.4872436523438,
1105
+ "learning_rate": 2.602657004830918e-06,
1106
+ "loss": 0.0159,
1107
+ "step": 3250
1108
+ },
1109
+ {
1110
+ "epoch": 17.80218281036835,
1111
+ "grad_norm": 0.049941010773181915,
1112
+ "learning_rate": 2.451690821256039e-06,
1113
+ "loss": 0.0738,
1114
+ "step": 3275
1115
+ },
1116
+ {
1117
+ "epoch": 17.938608458390178,
1118
+ "grad_norm": 0.020618176087737083,
1119
+ "learning_rate": 2.3007246376811595e-06,
1120
+ "loss": 0.0022,
1121
+ "step": 3300
1122
+ },
1123
+ {
1124
+ "epoch": 18.0,
1125
+ "eval_accuracy": 0.820855614973262,
1126
+ "eval_loss": 1.370205044746399,
1127
+ "eval_macro_f1": 0.8203022242829886,
1128
+ "eval_runtime": 0.7937,
1129
+ "eval_samples_per_second": 942.381,
1130
+ "eval_steps_per_second": 59.214,
1131
+ "eval_weighted_f1": 0.8209840718915221,
1132
+ "step": 3312
1133
+ },
1134
+ {
1135
+ "epoch": 18.07094133697135,
1136
+ "grad_norm": 0.02301851101219654,
1137
+ "learning_rate": 2.1497584541062806e-06,
1138
+ "loss": 0.0021,
1139
+ "step": 3325
1140
+ },
1141
+ {
1142
+ "epoch": 18.20736698499318,
1143
+ "grad_norm": 0.019857747480273247,
1144
+ "learning_rate": 1.9987922705314013e-06,
1145
+ "loss": 0.0079,
1146
+ "step": 3350
1147
+ },
1148
+ {
1149
+ "epoch": 18.343792633015006,
1150
+ "grad_norm": 0.045158278197050095,
1151
+ "learning_rate": 1.8478260869565218e-06,
1152
+ "loss": 0.1128,
1153
+ "step": 3375
1154
+ },
1155
+ {
1156
+ "epoch": 18.480218281036834,
1157
+ "grad_norm": 0.01906937174499035,
1158
+ "learning_rate": 1.6968599033816427e-06,
1159
+ "loss": 0.0619,
1160
+ "step": 3400
1161
+ },
1162
+ {
1163
+ "epoch": 18.616643929058664,
1164
+ "grad_norm": 0.021298298612236977,
1165
+ "learning_rate": 1.5458937198067634e-06,
1166
+ "loss": 0.0401,
1167
+ "step": 3425
1168
+ },
1169
+ {
1170
+ "epoch": 18.753069577080492,
1171
+ "grad_norm": 0.020442141219973564,
1172
+ "learning_rate": 1.3949275362318843e-06,
1173
+ "loss": 0.0115,
1174
+ "step": 3450
1175
+ },
1176
+ {
1177
+ "epoch": 18.88949522510232,
1178
+ "grad_norm": 0.019422555342316628,
1179
+ "learning_rate": 1.243961352657005e-06,
1180
+ "loss": 0.0751,
1181
+ "step": 3475
1182
+ },
1183
+ {
1184
+ "epoch": 19.0,
1185
+ "eval_accuracy": 0.8235294117647058,
1186
+ "eval_loss": 1.3653334379196167,
1187
+ "eval_macro_f1": 0.8229055488197744,
1188
+ "eval_runtime": 0.7794,
1189
+ "eval_samples_per_second": 959.686,
1190
+ "eval_steps_per_second": 60.301,
1191
+ "eval_weighted_f1": 0.8236105828584015,
1192
+ "step": 3496
1193
+ }
1194
+ ],
1195
+ "logging_steps": 25,
1196
+ "max_steps": 3680,
1197
+ "num_input_tokens_seen": 0,
1198
+ "num_train_epochs": 20,
1199
+ "save_steps": 500,
1200
+ "stateful_callbacks": {
1201
+ "EarlyStoppingCallback": {
1202
+ "args": {
1203
+ "early_stopping_patience": 3,
1204
+ "early_stopping_threshold": 0.0
1205
+ },
1206
+ "attributes": {
1207
+ "early_stopping_patience_counter": 3
1208
+ }
1209
+ },
1210
+ "TrainerControl": {
1211
+ "args": {
1212
+ "should_epoch_stop": false,
1213
+ "should_evaluate": false,
1214
+ "should_log": false,
1215
+ "should_save": true,
1216
+ "should_training_stop": true
1217
+ },
1218
+ "attributes": {}
1219
+ }
1220
+ },
1221
+ "total_flos": 0.0,
1222
+ "train_batch_size": 8,
1223
+ "trial_name": null,
1224
+ "trial_params": null
1225
+ }
ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379/trainer_state.json ADDED
@@ -0,0 +1,505 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1379,
3
+ "best_metric": 0.8525588574943566,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379",
5
+ "epoch": 7.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1379,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.12706480304955528,
14
+ "grad_norm": 15.085540771484375,
15
+ "learning_rate": 1.218274111675127e-06,
16
+ "loss": 4.4563,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.25412960609911056,
21
+ "grad_norm": 8.763714790344238,
22
+ "learning_rate": 2.487309644670051e-06,
23
+ "loss": 4.4293,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.3811944091486658,
28
+ "grad_norm": 15.892329216003418,
29
+ "learning_rate": 3.756345177664975e-06,
30
+ "loss": 4.4088,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5082592121982211,
35
+ "grad_norm": 13.681346893310547,
36
+ "learning_rate": 5.025380710659899e-06,
37
+ "loss": 4.4234,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6353240152477764,
42
+ "grad_norm": 62.75432205200195,
43
+ "learning_rate": 6.2944162436548225e-06,
44
+ "loss": 4.3818,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.7623888182973316,
49
+ "grad_norm": 14.515042304992676,
50
+ "learning_rate": 7.563451776649747e-06,
51
+ "loss": 4.3619,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.8894536213468869,
56
+ "grad_norm": 13.431984901428223,
57
+ "learning_rate": 8.832487309644671e-06,
58
+ "loss": 4.2928,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.5635276532137519,
64
+ "eval_loss": 0.9356052875518799,
65
+ "eval_macro_f1": 0.559522096781545,
66
+ "eval_runtime": 0.7174,
67
+ "eval_samples_per_second": 932.481,
68
+ "eval_steps_per_second": 58.541,
69
+ "eval_weighted_f1": 0.5597942808083388,
70
+ "step": 197
71
+ },
72
+ {
73
+ "epoch": 1.0152477763659467,
74
+ "grad_norm": 24.51892852783203,
75
+ "learning_rate": 1.0101522842639596e-05,
76
+ "loss": 4.0043,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.1423125794155018,
81
+ "grad_norm": 42.52670669555664,
82
+ "learning_rate": 1.1370558375634519e-05,
83
+ "loss": 3.4663,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.2693773824650572,
88
+ "grad_norm": 34.29641342163086,
89
+ "learning_rate": 1.2639593908629442e-05,
90
+ "loss": 3.1677,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.3964421855146125,
95
+ "grad_norm": 61.411476135253906,
96
+ "learning_rate": 1.3908629441624366e-05,
97
+ "loss": 2.8312,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.5235069885641677,
102
+ "grad_norm": 44.43330383300781,
103
+ "learning_rate": 1.5177664974619291e-05,
104
+ "loss": 2.7258,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.650571791613723,
109
+ "grad_norm": 54.96050262451172,
110
+ "learning_rate": 1.6446700507614214e-05,
111
+ "loss": 2.6636,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.7776365946632784,
116
+ "grad_norm": 48.811378479003906,
117
+ "learning_rate": 1.771573604060914e-05,
118
+ "loss": 2.358,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 1.9047013977128335,
123
+ "grad_norm": 52.216590881347656,
124
+ "learning_rate": 1.8984771573604063e-05,
125
+ "loss": 2.2766,
126
+ "step": 375
127
+ },
128
+ {
129
+ "epoch": 2.0,
130
+ "eval_accuracy": 0.7862481315396114,
131
+ "eval_loss": 0.5088972449302673,
132
+ "eval_macro_f1": 0.786588698920894,
133
+ "eval_runtime": 0.7298,
134
+ "eval_samples_per_second": 916.714,
135
+ "eval_steps_per_second": 57.552,
136
+ "eval_weighted_f1": 0.7870716935072392,
137
+ "step": 394
138
+ },
139
+ {
140
+ "epoch": 2.0304955527318933,
141
+ "grad_norm": 61.090213775634766,
142
+ "learning_rate": 1.997179921037789e-05,
143
+ "loss": 2.1162,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.1575603557814484,
148
+ "grad_norm": 54.048797607421875,
149
+ "learning_rate": 1.9830795262267346e-05,
150
+ "loss": 1.6129,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.2846251588310036,
155
+ "grad_norm": 45.04215621948242,
156
+ "learning_rate": 1.96897913141568e-05,
157
+ "loss": 1.8085,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.411689961880559,
162
+ "grad_norm": 40.17599105834961,
163
+ "learning_rate": 1.954878736604625e-05,
164
+ "loss": 1.7853,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.5387547649301143,
169
+ "grad_norm": 28.08652114868164,
170
+ "learning_rate": 1.9407783417935703e-05,
171
+ "loss": 1.9402,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.6658195679796695,
176
+ "grad_norm": 23.983867645263672,
177
+ "learning_rate": 1.9266779469825156e-05,
178
+ "loss": 1.8668,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.792884371029225,
183
+ "grad_norm": 48.15681076049805,
184
+ "learning_rate": 1.912577552171461e-05,
185
+ "loss": 1.5519,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 2.91994917407878,
190
+ "grad_norm": 26.708223342895508,
191
+ "learning_rate": 1.8984771573604063e-05,
192
+ "loss": 1.9312,
193
+ "step": 575
194
+ },
195
+ {
196
+ "epoch": 3.0,
197
+ "eval_accuracy": 0.8295964125560538,
198
+ "eval_loss": 0.4801219701766968,
199
+ "eval_macro_f1": 0.8313914528110596,
200
+ "eval_runtime": 0.7294,
201
+ "eval_samples_per_second": 917.182,
202
+ "eval_steps_per_second": 57.581,
203
+ "eval_weighted_f1": 0.8313654015323609,
204
+ "step": 591
205
+ },
206
+ {
207
+ "epoch": 3.04574332909784,
208
+ "grad_norm": 33.821617126464844,
209
+ "learning_rate": 1.8843767625493517e-05,
210
+ "loss": 1.5808,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.172808132147395,
215
+ "grad_norm": 41.84688949584961,
216
+ "learning_rate": 1.870276367738297e-05,
217
+ "loss": 1.1358,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.2998729351969507,
222
+ "grad_norm": 44.807376861572266,
223
+ "learning_rate": 1.856175972927242e-05,
224
+ "loss": 1.2669,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.426937738246506,
229
+ "grad_norm": 272.6358947753906,
230
+ "learning_rate": 1.8420755781161873e-05,
231
+ "loss": 1.1596,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.554002541296061,
236
+ "grad_norm": 37.20458984375,
237
+ "learning_rate": 1.8279751833051327e-05,
238
+ "loss": 1.0346,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.681067344345616,
243
+ "grad_norm": 38.68766403198242,
244
+ "learning_rate": 1.813874788494078e-05,
245
+ "loss": 1.2677,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 3.8081321473951717,
250
+ "grad_norm": 45.18739700317383,
251
+ "learning_rate": 1.7997743936830234e-05,
252
+ "loss": 1.2086,
253
+ "step": 750
254
+ },
255
+ {
256
+ "epoch": 3.935196950444727,
257
+ "grad_norm": 45.09466552734375,
258
+ "learning_rate": 1.7856739988719684e-05,
259
+ "loss": 1.1931,
260
+ "step": 775
261
+ },
262
+ {
263
+ "epoch": 4.0,
264
+ "eval_accuracy": 0.8490284005979073,
265
+ "eval_loss": 0.45191213488578796,
266
+ "eval_macro_f1": 0.8491685735308848,
267
+ "eval_runtime": 0.7238,
268
+ "eval_samples_per_second": 924.282,
269
+ "eval_steps_per_second": 58.027,
270
+ "eval_weighted_f1": 0.8491846182767624,
271
+ "step": 788
272
+ },
273
+ {
274
+ "epoch": 4.060991105463787,
275
+ "grad_norm": 45.72401428222656,
276
+ "learning_rate": 1.771573604060914e-05,
277
+ "loss": 1.117,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.188055908513342,
282
+ "grad_norm": 43.56619644165039,
283
+ "learning_rate": 1.7574732092498594e-05,
284
+ "loss": 0.7974,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.315120711562897,
289
+ "grad_norm": 22.15397071838379,
290
+ "learning_rate": 1.7433728144388044e-05,
291
+ "loss": 0.7436,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.4421855146124525,
296
+ "grad_norm": 38.34001159667969,
297
+ "learning_rate": 1.7292724196277497e-05,
298
+ "loss": 0.7424,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.569250317662007,
303
+ "grad_norm": 15.297261238098145,
304
+ "learning_rate": 1.715172024816695e-05,
305
+ "loss": 0.8808,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 4.696315120711563,
310
+ "grad_norm": 47.723731994628906,
311
+ "learning_rate": 1.7010716300056404e-05,
312
+ "loss": 0.8489,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 4.823379923761118,
317
+ "grad_norm": 37.75666427612305,
318
+ "learning_rate": 1.6869712351945854e-05,
319
+ "loss": 0.898,
320
+ "step": 950
321
+ },
322
+ {
323
+ "epoch": 4.950444726810673,
324
+ "grad_norm": 37.142539978027344,
325
+ "learning_rate": 1.672870840383531e-05,
326
+ "loss": 0.8993,
327
+ "step": 975
328
+ },
329
+ {
330
+ "epoch": 5.0,
331
+ "eval_accuracy": 0.8460388639760837,
332
+ "eval_loss": 0.48817121982574463,
333
+ "eval_macro_f1": 0.8460164400467091,
334
+ "eval_runtime": 0.7247,
335
+ "eval_samples_per_second": 923.149,
336
+ "eval_steps_per_second": 57.956,
337
+ "eval_weighted_f1": 0.8460335099896523,
338
+ "step": 985
339
+ },
340
+ {
341
+ "epoch": 5.076238881829733,
342
+ "grad_norm": 46.55158996582031,
343
+ "learning_rate": 1.6587704455724764e-05,
344
+ "loss": 0.6348,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.203303684879288,
349
+ "grad_norm": 89.48847961425781,
350
+ "learning_rate": 1.6446700507614214e-05,
351
+ "loss": 0.4828,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.330368487928844,
356
+ "grad_norm": 75.65311431884766,
357
+ "learning_rate": 1.6305696559503667e-05,
358
+ "loss": 0.4371,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.457433290978399,
363
+ "grad_norm": 88.13261413574219,
364
+ "learning_rate": 1.616469261139312e-05,
365
+ "loss": 0.5345,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.584498094027954,
370
+ "grad_norm": 20.740280151367188,
371
+ "learning_rate": 1.6023688663282574e-05,
372
+ "loss": 0.5955,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 5.71156289707751,
377
+ "grad_norm": 187.943359375,
378
+ "learning_rate": 1.5882684715172024e-05,
379
+ "loss": 0.5359,
380
+ "step": 1125
381
+ },
382
+ {
383
+ "epoch": 5.838627700127065,
384
+ "grad_norm": 41.54878234863281,
385
+ "learning_rate": 1.5741680767061477e-05,
386
+ "loss": 0.6068,
387
+ "step": 1150
388
+ },
389
+ {
390
+ "epoch": 5.96569250317662,
391
+ "grad_norm": 29.867746353149414,
392
+ "learning_rate": 1.5600676818950934e-05,
393
+ "loss": 0.5915,
394
+ "step": 1175
395
+ },
396
+ {
397
+ "epoch": 6.0,
398
+ "eval_accuracy": 0.8445440956651719,
399
+ "eval_loss": 0.622381329536438,
400
+ "eval_macro_f1": 0.8431216401219178,
401
+ "eval_runtime": 0.7351,
402
+ "eval_samples_per_second": 910.028,
403
+ "eval_steps_per_second": 57.132,
404
+ "eval_weighted_f1": 0.8431094777236665,
405
+ "step": 1182
406
+ },
407
+ {
408
+ "epoch": 6.09148665819568,
409
+ "grad_norm": 51.3915901184082,
410
+ "learning_rate": 1.5459672870840384e-05,
411
+ "loss": 0.3325,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.2185514612452355,
416
+ "grad_norm": 0.9594596028327942,
417
+ "learning_rate": 1.5318668922729838e-05,
418
+ "loss": 0.3621,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.34561626429479,
423
+ "grad_norm": 0.6977389454841614,
424
+ "learning_rate": 1.5177664974619291e-05,
425
+ "loss": 0.3612,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.472681067344346,
430
+ "grad_norm": 6.876659393310547,
431
+ "learning_rate": 1.5036661026508743e-05,
432
+ "loss": 0.4863,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 6.599745870393901,
437
+ "grad_norm": 55.13117218017578,
438
+ "learning_rate": 1.4895657078398196e-05,
439
+ "loss": 0.445,
440
+ "step": 1300
441
+ },
442
+ {
443
+ "epoch": 6.726810673443456,
444
+ "grad_norm": 1.8285411596298218,
445
+ "learning_rate": 1.4754653130287648e-05,
446
+ "loss": 0.2787,
447
+ "step": 1325
448
+ },
449
+ {
450
+ "epoch": 6.853875476493012,
451
+ "grad_norm": 82.604248046875,
452
+ "learning_rate": 1.4613649182177103e-05,
453
+ "loss": 0.5279,
454
+ "step": 1350
455
+ },
456
+ {
457
+ "epoch": 6.980940279542566,
458
+ "grad_norm": 52.5632209777832,
459
+ "learning_rate": 1.4472645234066556e-05,
460
+ "loss": 0.4638,
461
+ "step": 1375
462
+ },
463
+ {
464
+ "epoch": 7.0,
465
+ "eval_accuracy": 0.8535127055306427,
466
+ "eval_loss": 0.7533432245254517,
467
+ "eval_macro_f1": 0.8525588574943566,
468
+ "eval_runtime": 0.7556,
469
+ "eval_samples_per_second": 885.361,
470
+ "eval_steps_per_second": 55.583,
471
+ "eval_weighted_f1": 0.8524908732633673,
472
+ "step": 1379
473
+ }
474
+ ],
475
+ "logging_steps": 25,
476
+ "max_steps": 3940,
477
+ "num_input_tokens_seen": 0,
478
+ "num_train_epochs": 20,
479
+ "save_steps": 500,
480
+ "stateful_callbacks": {
481
+ "EarlyStoppingCallback": {
482
+ "args": {
483
+ "early_stopping_patience": 3,
484
+ "early_stopping_threshold": 0.0
485
+ },
486
+ "attributes": {
487
+ "early_stopping_patience_counter": 0
488
+ }
489
+ },
490
+ "TrainerControl": {
491
+ "args": {
492
+ "should_epoch_stop": false,
493
+ "should_evaluate": false,
494
+ "should_log": false,
495
+ "should_save": true,
496
+ "should_training_stop": false
497
+ },
498
+ "attributes": {}
499
+ }
500
+ },
501
+ "total_flos": 0.0,
502
+ "train_batch_size": 8,
503
+ "trial_name": null,
504
+ "trial_params": null
505
+ }
ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1970/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1970/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1970/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1970/trainer_state.json ADDED
@@ -0,0 +1,699 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1379,
3
+ "best_metric": 0.8525588574943566,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-07-s42-20260827T190759Z-6235c159/trainer/checkpoint-1379",
5
+ "epoch": 10.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1970,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.12706480304955528,
14
+ "grad_norm": 15.085540771484375,
15
+ "learning_rate": 1.218274111675127e-06,
16
+ "loss": 4.4563,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.25412960609911056,
21
+ "grad_norm": 8.763714790344238,
22
+ "learning_rate": 2.487309644670051e-06,
23
+ "loss": 4.4293,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.3811944091486658,
28
+ "grad_norm": 15.892329216003418,
29
+ "learning_rate": 3.756345177664975e-06,
30
+ "loss": 4.4088,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5082592121982211,
35
+ "grad_norm": 13.681346893310547,
36
+ "learning_rate": 5.025380710659899e-06,
37
+ "loss": 4.4234,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6353240152477764,
42
+ "grad_norm": 62.75432205200195,
43
+ "learning_rate": 6.2944162436548225e-06,
44
+ "loss": 4.3818,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.7623888182973316,
49
+ "grad_norm": 14.515042304992676,
50
+ "learning_rate": 7.563451776649747e-06,
51
+ "loss": 4.3619,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.8894536213468869,
56
+ "grad_norm": 13.431984901428223,
57
+ "learning_rate": 8.832487309644671e-06,
58
+ "loss": 4.2928,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.5635276532137519,
64
+ "eval_loss": 0.9356052875518799,
65
+ "eval_macro_f1": 0.559522096781545,
66
+ "eval_runtime": 0.7174,
67
+ "eval_samples_per_second": 932.481,
68
+ "eval_steps_per_second": 58.541,
69
+ "eval_weighted_f1": 0.5597942808083388,
70
+ "step": 197
71
+ },
72
+ {
73
+ "epoch": 1.0152477763659467,
74
+ "grad_norm": 24.51892852783203,
75
+ "learning_rate": 1.0101522842639596e-05,
76
+ "loss": 4.0043,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.1423125794155018,
81
+ "grad_norm": 42.52670669555664,
82
+ "learning_rate": 1.1370558375634519e-05,
83
+ "loss": 3.4663,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.2693773824650572,
88
+ "grad_norm": 34.29641342163086,
89
+ "learning_rate": 1.2639593908629442e-05,
90
+ "loss": 3.1677,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.3964421855146125,
95
+ "grad_norm": 61.411476135253906,
96
+ "learning_rate": 1.3908629441624366e-05,
97
+ "loss": 2.8312,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.5235069885641677,
102
+ "grad_norm": 44.43330383300781,
103
+ "learning_rate": 1.5177664974619291e-05,
104
+ "loss": 2.7258,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.650571791613723,
109
+ "grad_norm": 54.96050262451172,
110
+ "learning_rate": 1.6446700507614214e-05,
111
+ "loss": 2.6636,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.7776365946632784,
116
+ "grad_norm": 48.811378479003906,
117
+ "learning_rate": 1.771573604060914e-05,
118
+ "loss": 2.358,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 1.9047013977128335,
123
+ "grad_norm": 52.216590881347656,
124
+ "learning_rate": 1.8984771573604063e-05,
125
+ "loss": 2.2766,
126
+ "step": 375
127
+ },
128
+ {
129
+ "epoch": 2.0,
130
+ "eval_accuracy": 0.7862481315396114,
131
+ "eval_loss": 0.5088972449302673,
132
+ "eval_macro_f1": 0.786588698920894,
133
+ "eval_runtime": 0.7298,
134
+ "eval_samples_per_second": 916.714,
135
+ "eval_steps_per_second": 57.552,
136
+ "eval_weighted_f1": 0.7870716935072392,
137
+ "step": 394
138
+ },
139
+ {
140
+ "epoch": 2.0304955527318933,
141
+ "grad_norm": 61.090213775634766,
142
+ "learning_rate": 1.997179921037789e-05,
143
+ "loss": 2.1162,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.1575603557814484,
148
+ "grad_norm": 54.048797607421875,
149
+ "learning_rate": 1.9830795262267346e-05,
150
+ "loss": 1.6129,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.2846251588310036,
155
+ "grad_norm": 45.04215621948242,
156
+ "learning_rate": 1.96897913141568e-05,
157
+ "loss": 1.8085,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.411689961880559,
162
+ "grad_norm": 40.17599105834961,
163
+ "learning_rate": 1.954878736604625e-05,
164
+ "loss": 1.7853,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.5387547649301143,
169
+ "grad_norm": 28.08652114868164,
170
+ "learning_rate": 1.9407783417935703e-05,
171
+ "loss": 1.9402,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.6658195679796695,
176
+ "grad_norm": 23.983867645263672,
177
+ "learning_rate": 1.9266779469825156e-05,
178
+ "loss": 1.8668,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.792884371029225,
183
+ "grad_norm": 48.15681076049805,
184
+ "learning_rate": 1.912577552171461e-05,
185
+ "loss": 1.5519,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 2.91994917407878,
190
+ "grad_norm": 26.708223342895508,
191
+ "learning_rate": 1.8984771573604063e-05,
192
+ "loss": 1.9312,
193
+ "step": 575
194
+ },
195
+ {
196
+ "epoch": 3.0,
197
+ "eval_accuracy": 0.8295964125560538,
198
+ "eval_loss": 0.4801219701766968,
199
+ "eval_macro_f1": 0.8313914528110596,
200
+ "eval_runtime": 0.7294,
201
+ "eval_samples_per_second": 917.182,
202
+ "eval_steps_per_second": 57.581,
203
+ "eval_weighted_f1": 0.8313654015323609,
204
+ "step": 591
205
+ },
206
+ {
207
+ "epoch": 3.04574332909784,
208
+ "grad_norm": 33.821617126464844,
209
+ "learning_rate": 1.8843767625493517e-05,
210
+ "loss": 1.5808,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.172808132147395,
215
+ "grad_norm": 41.84688949584961,
216
+ "learning_rate": 1.870276367738297e-05,
217
+ "loss": 1.1358,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.2998729351969507,
222
+ "grad_norm": 44.807376861572266,
223
+ "learning_rate": 1.856175972927242e-05,
224
+ "loss": 1.2669,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.426937738246506,
229
+ "grad_norm": 272.6358947753906,
230
+ "learning_rate": 1.8420755781161873e-05,
231
+ "loss": 1.1596,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.554002541296061,
236
+ "grad_norm": 37.20458984375,
237
+ "learning_rate": 1.8279751833051327e-05,
238
+ "loss": 1.0346,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.681067344345616,
243
+ "grad_norm": 38.68766403198242,
244
+ "learning_rate": 1.813874788494078e-05,
245
+ "loss": 1.2677,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 3.8081321473951717,
250
+ "grad_norm": 45.18739700317383,
251
+ "learning_rate": 1.7997743936830234e-05,
252
+ "loss": 1.2086,
253
+ "step": 750
254
+ },
255
+ {
256
+ "epoch": 3.935196950444727,
257
+ "grad_norm": 45.09466552734375,
258
+ "learning_rate": 1.7856739988719684e-05,
259
+ "loss": 1.1931,
260
+ "step": 775
261
+ },
262
+ {
263
+ "epoch": 4.0,
264
+ "eval_accuracy": 0.8490284005979073,
265
+ "eval_loss": 0.45191213488578796,
266
+ "eval_macro_f1": 0.8491685735308848,
267
+ "eval_runtime": 0.7238,
268
+ "eval_samples_per_second": 924.282,
269
+ "eval_steps_per_second": 58.027,
270
+ "eval_weighted_f1": 0.8491846182767624,
271
+ "step": 788
272
+ },
273
+ {
274
+ "epoch": 4.060991105463787,
275
+ "grad_norm": 45.72401428222656,
276
+ "learning_rate": 1.771573604060914e-05,
277
+ "loss": 1.117,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.188055908513342,
282
+ "grad_norm": 43.56619644165039,
283
+ "learning_rate": 1.7574732092498594e-05,
284
+ "loss": 0.7974,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.315120711562897,
289
+ "grad_norm": 22.15397071838379,
290
+ "learning_rate": 1.7433728144388044e-05,
291
+ "loss": 0.7436,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.4421855146124525,
296
+ "grad_norm": 38.34001159667969,
297
+ "learning_rate": 1.7292724196277497e-05,
298
+ "loss": 0.7424,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.569250317662007,
303
+ "grad_norm": 15.297261238098145,
304
+ "learning_rate": 1.715172024816695e-05,
305
+ "loss": 0.8808,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 4.696315120711563,
310
+ "grad_norm": 47.723731994628906,
311
+ "learning_rate": 1.7010716300056404e-05,
312
+ "loss": 0.8489,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 4.823379923761118,
317
+ "grad_norm": 37.75666427612305,
318
+ "learning_rate": 1.6869712351945854e-05,
319
+ "loss": 0.898,
320
+ "step": 950
321
+ },
322
+ {
323
+ "epoch": 4.950444726810673,
324
+ "grad_norm": 37.142539978027344,
325
+ "learning_rate": 1.672870840383531e-05,
326
+ "loss": 0.8993,
327
+ "step": 975
328
+ },
329
+ {
330
+ "epoch": 5.0,
331
+ "eval_accuracy": 0.8460388639760837,
332
+ "eval_loss": 0.48817121982574463,
333
+ "eval_macro_f1": 0.8460164400467091,
334
+ "eval_runtime": 0.7247,
335
+ "eval_samples_per_second": 923.149,
336
+ "eval_steps_per_second": 57.956,
337
+ "eval_weighted_f1": 0.8460335099896523,
338
+ "step": 985
339
+ },
340
+ {
341
+ "epoch": 5.076238881829733,
342
+ "grad_norm": 46.55158996582031,
343
+ "learning_rate": 1.6587704455724764e-05,
344
+ "loss": 0.6348,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.203303684879288,
349
+ "grad_norm": 89.48847961425781,
350
+ "learning_rate": 1.6446700507614214e-05,
351
+ "loss": 0.4828,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.330368487928844,
356
+ "grad_norm": 75.65311431884766,
357
+ "learning_rate": 1.6305696559503667e-05,
358
+ "loss": 0.4371,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.457433290978399,
363
+ "grad_norm": 88.13261413574219,
364
+ "learning_rate": 1.616469261139312e-05,
365
+ "loss": 0.5345,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.584498094027954,
370
+ "grad_norm": 20.740280151367188,
371
+ "learning_rate": 1.6023688663282574e-05,
372
+ "loss": 0.5955,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 5.71156289707751,
377
+ "grad_norm": 187.943359375,
378
+ "learning_rate": 1.5882684715172024e-05,
379
+ "loss": 0.5359,
380
+ "step": 1125
381
+ },
382
+ {
383
+ "epoch": 5.838627700127065,
384
+ "grad_norm": 41.54878234863281,
385
+ "learning_rate": 1.5741680767061477e-05,
386
+ "loss": 0.6068,
387
+ "step": 1150
388
+ },
389
+ {
390
+ "epoch": 5.96569250317662,
391
+ "grad_norm": 29.867746353149414,
392
+ "learning_rate": 1.5600676818950934e-05,
393
+ "loss": 0.5915,
394
+ "step": 1175
395
+ },
396
+ {
397
+ "epoch": 6.0,
398
+ "eval_accuracy": 0.8445440956651719,
399
+ "eval_loss": 0.622381329536438,
400
+ "eval_macro_f1": 0.8431216401219178,
401
+ "eval_runtime": 0.7351,
402
+ "eval_samples_per_second": 910.028,
403
+ "eval_steps_per_second": 57.132,
404
+ "eval_weighted_f1": 0.8431094777236665,
405
+ "step": 1182
406
+ },
407
+ {
408
+ "epoch": 6.09148665819568,
409
+ "grad_norm": 51.3915901184082,
410
+ "learning_rate": 1.5459672870840384e-05,
411
+ "loss": 0.3325,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.2185514612452355,
416
+ "grad_norm": 0.9594596028327942,
417
+ "learning_rate": 1.5318668922729838e-05,
418
+ "loss": 0.3621,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.34561626429479,
423
+ "grad_norm": 0.6977389454841614,
424
+ "learning_rate": 1.5177664974619291e-05,
425
+ "loss": 0.3612,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.472681067344346,
430
+ "grad_norm": 6.876659393310547,
431
+ "learning_rate": 1.5036661026508743e-05,
432
+ "loss": 0.4863,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 6.599745870393901,
437
+ "grad_norm": 55.13117218017578,
438
+ "learning_rate": 1.4895657078398196e-05,
439
+ "loss": 0.445,
440
+ "step": 1300
441
+ },
442
+ {
443
+ "epoch": 6.726810673443456,
444
+ "grad_norm": 1.8285411596298218,
445
+ "learning_rate": 1.4754653130287648e-05,
446
+ "loss": 0.2787,
447
+ "step": 1325
448
+ },
449
+ {
450
+ "epoch": 6.853875476493012,
451
+ "grad_norm": 82.604248046875,
452
+ "learning_rate": 1.4613649182177103e-05,
453
+ "loss": 0.5279,
454
+ "step": 1350
455
+ },
456
+ {
457
+ "epoch": 6.980940279542566,
458
+ "grad_norm": 52.5632209777832,
459
+ "learning_rate": 1.4472645234066556e-05,
460
+ "loss": 0.4638,
461
+ "step": 1375
462
+ },
463
+ {
464
+ "epoch": 7.0,
465
+ "eval_accuracy": 0.8535127055306427,
466
+ "eval_loss": 0.7533432245254517,
467
+ "eval_macro_f1": 0.8525588574943566,
468
+ "eval_runtime": 0.7556,
469
+ "eval_samples_per_second": 885.361,
470
+ "eval_steps_per_second": 55.583,
471
+ "eval_weighted_f1": 0.8524908732633673,
472
+ "step": 1379
473
+ },
474
+ {
475
+ "epoch": 7.106734434561626,
476
+ "grad_norm": 55.73310089111328,
477
+ "learning_rate": 1.4331641285956008e-05,
478
+ "loss": 0.2568,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.233799237611182,
483
+ "grad_norm": 5.577467918395996,
484
+ "learning_rate": 1.4190637337845461e-05,
485
+ "loss": 0.3423,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.360864040660737,
490
+ "grad_norm": 126.92430877685547,
491
+ "learning_rate": 1.4049633389734913e-05,
492
+ "loss": 0.3348,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 7.487928843710292,
497
+ "grad_norm": 22.87789535522461,
498
+ "learning_rate": 1.3908629441624366e-05,
499
+ "loss": 0.5321,
500
+ "step": 1475
501
+ },
502
+ {
503
+ "epoch": 7.614993646759848,
504
+ "grad_norm": 134.18499755859375,
505
+ "learning_rate": 1.3767625493513818e-05,
506
+ "loss": 0.3859,
507
+ "step": 1500
508
+ },
509
+ {
510
+ "epoch": 7.742058449809403,
511
+ "grad_norm": 176.86053466796875,
512
+ "learning_rate": 1.3626621545403271e-05,
513
+ "loss": 0.2192,
514
+ "step": 1525
515
+ },
516
+ {
517
+ "epoch": 7.869123252858958,
518
+ "grad_norm": 114.168212890625,
519
+ "learning_rate": 1.3485617597292726e-05,
520
+ "loss": 0.4917,
521
+ "step": 1550
522
+ },
523
+ {
524
+ "epoch": 7.9961880559085134,
525
+ "grad_norm": 1.5546883344650269,
526
+ "learning_rate": 1.3344613649182178e-05,
527
+ "loss": 0.3951,
528
+ "step": 1575
529
+ },
530
+ {
531
+ "epoch": 8.0,
532
+ "eval_accuracy": 0.8400597907324364,
533
+ "eval_loss": 0.7629181146621704,
534
+ "eval_macro_f1": 0.839451102006168,
535
+ "eval_runtime": 0.7124,
536
+ "eval_samples_per_second": 939.104,
537
+ "eval_steps_per_second": 58.957,
538
+ "eval_weighted_f1": 0.8395208802707667,
539
+ "step": 1576
540
+ },
541
+ {
542
+ "epoch": 8.121982210927573,
543
+ "grad_norm": 96.23695373535156,
544
+ "learning_rate": 1.3203609701071631e-05,
545
+ "loss": 0.1721,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.249047013977128,
550
+ "grad_norm": 116.43880462646484,
551
+ "learning_rate": 1.3062605752961083e-05,
552
+ "loss": 0.3282,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 8.376111817026684,
557
+ "grad_norm": 99.85157775878906,
558
+ "learning_rate": 1.2921601804850537e-05,
559
+ "loss": 0.4269,
560
+ "step": 1650
561
+ },
562
+ {
563
+ "epoch": 8.503176620076239,
564
+ "grad_norm": 2.0939526557922363,
565
+ "learning_rate": 1.278059785673999e-05,
566
+ "loss": 0.3576,
567
+ "step": 1675
568
+ },
569
+ {
570
+ "epoch": 8.630241423125794,
571
+ "grad_norm": 55.3084716796875,
572
+ "learning_rate": 1.2639593908629442e-05,
573
+ "loss": 0.2196,
574
+ "step": 1700
575
+ },
576
+ {
577
+ "epoch": 8.757306226175348,
578
+ "grad_norm": 47.04745864868164,
579
+ "learning_rate": 1.2498589960518897e-05,
580
+ "loss": 0.2041,
581
+ "step": 1725
582
+ },
583
+ {
584
+ "epoch": 8.884371029224905,
585
+ "grad_norm": 228.26966857910156,
586
+ "learning_rate": 1.2357586012408348e-05,
587
+ "loss": 0.3237,
588
+ "step": 1750
589
+ },
590
+ {
591
+ "epoch": 9.0,
592
+ "eval_accuracy": 0.8430493273542601,
593
+ "eval_loss": 0.8689716458320618,
594
+ "eval_macro_f1": 0.843683658203653,
595
+ "eval_runtime": 0.7123,
596
+ "eval_samples_per_second": 939.219,
597
+ "eval_steps_per_second": 58.964,
598
+ "eval_weighted_f1": 0.8437893757333649,
599
+ "step": 1773
600
+ },
601
+ {
602
+ "epoch": 9.010165184243965,
603
+ "grad_norm": 6.63241720199585,
604
+ "learning_rate": 1.2216582064297802e-05,
605
+ "loss": 0.568,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 9.13722998729352,
610
+ "grad_norm": 0.07384615391492844,
611
+ "learning_rate": 1.2075578116187253e-05,
612
+ "loss": 0.1182,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 9.264294790343074,
617
+ "grad_norm": 101.28473663330078,
618
+ "learning_rate": 1.1934574168076707e-05,
619
+ "loss": 0.169,
620
+ "step": 1825
621
+ },
622
+ {
623
+ "epoch": 9.39135959339263,
624
+ "grad_norm": 0.07773292064666748,
625
+ "learning_rate": 1.179357021996616e-05,
626
+ "loss": 0.2225,
627
+ "step": 1850
628
+ },
629
+ {
630
+ "epoch": 9.518424396442185,
631
+ "grad_norm": 0.11557679623365402,
632
+ "learning_rate": 1.1652566271855612e-05,
633
+ "loss": 0.0956,
634
+ "step": 1875
635
+ },
636
+ {
637
+ "epoch": 9.64548919949174,
638
+ "grad_norm": 0.10225258022546768,
639
+ "learning_rate": 1.1511562323745067e-05,
640
+ "loss": 0.1632,
641
+ "step": 1900
642
+ },
643
+ {
644
+ "epoch": 9.772554002541296,
645
+ "grad_norm": 0.07881728559732437,
646
+ "learning_rate": 1.1370558375634519e-05,
647
+ "loss": 0.2586,
648
+ "step": 1925
649
+ },
650
+ {
651
+ "epoch": 9.899618805590851,
652
+ "grad_norm": 17.513851165771484,
653
+ "learning_rate": 1.1229554427523972e-05,
654
+ "loss": 0.336,
655
+ "step": 1950
656
+ },
657
+ {
658
+ "epoch": 10.0,
659
+ "eval_accuracy": 0.8445440956651719,
660
+ "eval_loss": 0.9900490045547485,
661
+ "eval_macro_f1": 0.8444863976249267,
662
+ "eval_runtime": 0.7232,
663
+ "eval_samples_per_second": 925.04,
664
+ "eval_steps_per_second": 58.074,
665
+ "eval_weighted_f1": 0.8443585287806236,
666
+ "step": 1970
667
+ }
668
+ ],
669
+ "logging_steps": 25,
670
+ "max_steps": 3940,
671
+ "num_input_tokens_seen": 0,
672
+ "num_train_epochs": 20,
673
+ "save_steps": 500,
674
+ "stateful_callbacks": {
675
+ "EarlyStoppingCallback": {
676
+ "args": {
677
+ "early_stopping_patience": 3,
678
+ "early_stopping_threshold": 0.0
679
+ },
680
+ "attributes": {
681
+ "early_stopping_patience_counter": 3
682
+ }
683
+ },
684
+ "TrainerControl": {
685
+ "args": {
686
+ "should_epoch_stop": false,
687
+ "should_evaluate": false,
688
+ "should_log": false,
689
+ "should_save": true,
690
+ "should_training_stop": true
691
+ },
692
+ "attributes": {}
693
+ }
694
+ },
695
+ "total_flos": 0.0,
696
+ "train_batch_size": 8,
697
+ "trial_name": null,
698
+ "trial_params": null
699
+ }
ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }
ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530/trainer_state.json ADDED
@@ -0,0 +1,569 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1530,
3
+ "best_metric": 0.8243176180408366,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-1530",
5
+ "epoch": 9.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1530,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.14727540500736377,
14
+ "grad_norm": 8.618729591369629,
15
+ "learning_rate": 1.4117647058823531e-06,
16
+ "loss": 4.4598,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.29455081001472755,
21
+ "grad_norm": 12.339950561523438,
22
+ "learning_rate": 2.8823529411764704e-06,
23
+ "loss": 4.4121,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4418262150220913,
28
+ "grad_norm": 12.81640911102295,
29
+ "learning_rate": 4.352941176470588e-06,
30
+ "loss": 4.4231,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5891016200294551,
35
+ "grad_norm": 10.555315971374512,
36
+ "learning_rate": 5.823529411764707e-06,
37
+ "loss": 4.3918,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.7363770250368189,
42
+ "grad_norm": 9.827973365783691,
43
+ "learning_rate": 7.294117647058823e-06,
44
+ "loss": 4.3722,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8836524300441826,
49
+ "grad_norm": 205.26437377929688,
50
+ "learning_rate": 8.764705882352942e-06,
51
+ "loss": 4.2838,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "eval_accuracy": 0.44714038128249567,
57
+ "eval_loss": 1.0794451236724854,
58
+ "eval_macro_f1": 0.3696054141515373,
59
+ "eval_runtime": 0.6665,
60
+ "eval_samples_per_second": 865.696,
61
+ "eval_steps_per_second": 55.513,
62
+ "eval_weighted_f1": 0.37278531363140255,
63
+ "step": 170
64
+ },
65
+ {
66
+ "epoch": 1.0294550810014726,
67
+ "grad_norm": 13.06892204284668,
68
+ "learning_rate": 1.023529411764706e-05,
69
+ "loss": 4.2496,
70
+ "step": 175
71
+ },
72
+ {
73
+ "epoch": 1.1767304860088366,
74
+ "grad_norm": 30.303871154785156,
75
+ "learning_rate": 1.1705882352941178e-05,
76
+ "loss": 4.1939,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.3240058910162003,
81
+ "grad_norm": 36.40523147583008,
82
+ "learning_rate": 1.3176470588235294e-05,
83
+ "loss": 3.8883,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.4712812960235642,
88
+ "grad_norm": 35.80412673950195,
89
+ "learning_rate": 1.4647058823529412e-05,
90
+ "loss": 3.2581,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.6185567010309279,
95
+ "grad_norm": 50.22141647338867,
96
+ "learning_rate": 1.6117647058823533e-05,
97
+ "loss": 3.0102,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.7658321060382915,
102
+ "grad_norm": 87.98308563232422,
103
+ "learning_rate": 1.7588235294117647e-05,
104
+ "loss": 3.0492,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.9131075110456552,
109
+ "grad_norm": 35.41440200805664,
110
+ "learning_rate": 1.9058823529411764e-05,
111
+ "loss": 2.8344,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 2.0,
116
+ "eval_accuracy": 0.7590987868284229,
117
+ "eval_loss": 0.5970668196678162,
118
+ "eval_macro_f1": 0.7560835096046364,
119
+ "eval_runtime": 0.6544,
120
+ "eval_samples_per_second": 881.686,
121
+ "eval_steps_per_second": 56.538,
122
+ "eval_weighted_f1": 0.7563858465193687,
123
+ "step": 340
124
+ },
125
+ {
126
+ "epoch": 2.0589101620029453,
127
+ "grad_norm": 51.65792465209961,
128
+ "learning_rate": 1.9941176470588238e-05,
129
+ "loss": 2.4335,
130
+ "step": 350
131
+ },
132
+ {
133
+ "epoch": 2.2061855670103094,
134
+ "grad_norm": 41.59065246582031,
135
+ "learning_rate": 1.977777777777778e-05,
136
+ "loss": 2.3281,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.353460972017673,
141
+ "grad_norm": 34.73653030395508,
142
+ "learning_rate": 1.961437908496732e-05,
143
+ "loss": 2.0699,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.500736377025037,
148
+ "grad_norm": 74.14517974853516,
149
+ "learning_rate": 1.9450980392156864e-05,
150
+ "loss": 2.1578,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.6480117820324005,
155
+ "grad_norm": 63.165306091308594,
156
+ "learning_rate": 1.9287581699346406e-05,
157
+ "loss": 1.913,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.795287187039764,
162
+ "grad_norm": 47.502498626708984,
163
+ "learning_rate": 1.912418300653595e-05,
164
+ "loss": 1.7691,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.9425625920471283,
169
+ "grad_norm": 39.62498474121094,
170
+ "learning_rate": 1.8960784313725492e-05,
171
+ "loss": 2.0327,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 3.0,
176
+ "eval_accuracy": 0.7989601386481803,
177
+ "eval_loss": 0.5072414875030518,
178
+ "eval_macro_f1": 0.793744570286909,
179
+ "eval_runtime": 0.6419,
180
+ "eval_samples_per_second": 898.936,
181
+ "eval_steps_per_second": 57.644,
182
+ "eval_weighted_f1": 0.7941990861861373,
183
+ "step": 510
184
+ },
185
+ {
186
+ "epoch": 3.0883652430044184,
187
+ "grad_norm": 49.47594451904297,
188
+ "learning_rate": 1.8797385620915035e-05,
189
+ "loss": 1.5598,
190
+ "step": 525
191
+ },
192
+ {
193
+ "epoch": 3.235640648011782,
194
+ "grad_norm": 61.620418548583984,
195
+ "learning_rate": 1.8633986928104578e-05,
196
+ "loss": 1.2781,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.3829160530191458,
201
+ "grad_norm": 61.680816650390625,
202
+ "learning_rate": 1.847058823529412e-05,
203
+ "loss": 1.2836,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.5301914580265095,
208
+ "grad_norm": 25.939306259155273,
209
+ "learning_rate": 1.830718954248366e-05,
210
+ "loss": 1.3021,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.6774668630338736,
215
+ "grad_norm": 61.35898971557617,
216
+ "learning_rate": 1.8143790849673203e-05,
217
+ "loss": 1.292,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.824742268041237,
222
+ "grad_norm": 47.5286979675293,
223
+ "learning_rate": 1.7980392156862746e-05,
224
+ "loss": 1.2698,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.972017673048601,
229
+ "grad_norm": 42.40422439575195,
230
+ "learning_rate": 1.781699346405229e-05,
231
+ "loss": 1.4948,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 4.0,
236
+ "eval_accuracy": 0.8128249566724437,
237
+ "eval_loss": 0.508634626865387,
238
+ "eval_macro_f1": 0.8109021014998965,
239
+ "eval_runtime": 0.6429,
240
+ "eval_samples_per_second": 897.559,
241
+ "eval_steps_per_second": 57.556,
242
+ "eval_weighted_f1": 0.8110331206976222,
243
+ "step": 680
244
+ },
245
+ {
246
+ "epoch": 4.117820324005891,
247
+ "grad_norm": 13.027673721313477,
248
+ "learning_rate": 1.7653594771241832e-05,
249
+ "loss": 1.0072,
250
+ "step": 700
251
+ },
252
+ {
253
+ "epoch": 4.265095729013255,
254
+ "grad_norm": 25.903133392333984,
255
+ "learning_rate": 1.7490196078431375e-05,
256
+ "loss": 0.8442,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.412371134020619,
261
+ "grad_norm": 74.11421966552734,
262
+ "learning_rate": 1.7326797385620918e-05,
263
+ "loss": 0.7711,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.559646539027982,
268
+ "grad_norm": 16.307165145874023,
269
+ "learning_rate": 1.7163398692810457e-05,
270
+ "loss": 0.8034,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.706921944035346,
275
+ "grad_norm": 30.609487533569336,
276
+ "learning_rate": 1.7e-05,
277
+ "loss": 0.8812,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.8541973490427095,
282
+ "grad_norm": 73.01841735839844,
283
+ "learning_rate": 1.6836601307189543e-05,
284
+ "loss": 0.897,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 5.0,
289
+ "grad_norm": 43.02462387084961,
290
+ "learning_rate": 1.6673202614379086e-05,
291
+ "loss": 1.0138,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 5.0,
296
+ "eval_accuracy": 0.8058925476603119,
297
+ "eval_loss": 0.621121346950531,
298
+ "eval_macro_f1": 0.8021918078032467,
299
+ "eval_runtime": 0.656,
300
+ "eval_samples_per_second": 879.529,
301
+ "eval_steps_per_second": 56.4,
302
+ "eval_weighted_f1": 0.8023839758528064,
303
+ "step": 850
304
+ },
305
+ {
306
+ "epoch": 5.147275405007364,
307
+ "grad_norm": 80.87950134277344,
308
+ "learning_rate": 1.650980392156863e-05,
309
+ "loss": 0.5408,
310
+ "step": 875
311
+ },
312
+ {
313
+ "epoch": 5.294550810014727,
314
+ "grad_norm": 51.10795211791992,
315
+ "learning_rate": 1.634640522875817e-05,
316
+ "loss": 0.5015,
317
+ "step": 900
318
+ },
319
+ {
320
+ "epoch": 5.4418262150220915,
321
+ "grad_norm": 106.66352081298828,
322
+ "learning_rate": 1.6183006535947714e-05,
323
+ "loss": 0.4298,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.589101620029455,
328
+ "grad_norm": 46.167118072509766,
329
+ "learning_rate": 1.6019607843137257e-05,
330
+ "loss": 0.7623,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.736377025036819,
335
+ "grad_norm": 8.59582805633545,
336
+ "learning_rate": 1.5856209150326797e-05,
337
+ "loss": 0.7861,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.883652430044183,
342
+ "grad_norm": 38.83647918701172,
343
+ "learning_rate": 1.569281045751634e-05,
344
+ "loss": 0.7567,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 6.0,
349
+ "eval_accuracy": 0.7972270363951474,
350
+ "eval_loss": 0.7285588979721069,
351
+ "eval_macro_f1": 0.7971036150082808,
352
+ "eval_runtime": 0.6486,
353
+ "eval_samples_per_second": 889.553,
354
+ "eval_steps_per_second": 57.042,
355
+ "eval_weighted_f1": 0.7969802816000557,
356
+ "step": 1020
357
+ },
358
+ {
359
+ "epoch": 6.029455081001473,
360
+ "grad_norm": 40.64036178588867,
361
+ "learning_rate": 1.5529411764705882e-05,
362
+ "loss": 0.4507,
363
+ "step": 1025
364
+ },
365
+ {
366
+ "epoch": 6.176730486008837,
367
+ "grad_norm": 12.018685340881348,
368
+ "learning_rate": 1.536601307189543e-05,
369
+ "loss": 0.4766,
370
+ "step": 1050
371
+ },
372
+ {
373
+ "epoch": 6.3240058910162,
374
+ "grad_norm": 7.10188102722168,
375
+ "learning_rate": 1.5202614379084968e-05,
376
+ "loss": 0.4288,
377
+ "step": 1075
378
+ },
379
+ {
380
+ "epoch": 6.471281296023564,
381
+ "grad_norm": 50.2728157043457,
382
+ "learning_rate": 1.5039215686274511e-05,
383
+ "loss": 0.4884,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.618556701030927,
388
+ "grad_norm": 153.4954376220703,
389
+ "learning_rate": 1.4875816993464054e-05,
390
+ "loss": 0.5846,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.7658321060382915,
395
+ "grad_norm": 1.2279057502746582,
396
+ "learning_rate": 1.4712418300653597e-05,
397
+ "loss": 0.5291,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.913107511045656,
402
+ "grad_norm": 226.41575622558594,
403
+ "learning_rate": 1.4549019607843138e-05,
404
+ "loss": 0.3512,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 7.0,
409
+ "eval_accuracy": 0.8128249566724437,
410
+ "eval_loss": 0.8911365270614624,
411
+ "eval_macro_f1": 0.8116135452466854,
412
+ "eval_runtime": 0.6497,
413
+ "eval_samples_per_second": 888.108,
414
+ "eval_steps_per_second": 56.95,
415
+ "eval_weighted_f1": 0.8116623140981447,
416
+ "step": 1190
417
+ },
418
+ {
419
+ "epoch": 7.058910162002945,
420
+ "grad_norm": 167.6545867919922,
421
+ "learning_rate": 1.438562091503268e-05,
422
+ "loss": 0.5701,
423
+ "step": 1200
424
+ },
425
+ {
426
+ "epoch": 7.206185567010309,
427
+ "grad_norm": 10.70135498046875,
428
+ "learning_rate": 1.4222222222222224e-05,
429
+ "loss": 0.3486,
430
+ "step": 1225
431
+ },
432
+ {
433
+ "epoch": 7.353460972017673,
434
+ "grad_norm": 0.9517020583152771,
435
+ "learning_rate": 1.4058823529411765e-05,
436
+ "loss": 0.2325,
437
+ "step": 1250
438
+ },
439
+ {
440
+ "epoch": 7.500736377025037,
441
+ "grad_norm": 18.72394561767578,
442
+ "learning_rate": 1.3895424836601308e-05,
443
+ "loss": 0.4756,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.648011782032401,
448
+ "grad_norm": 13.212563514709473,
449
+ "learning_rate": 1.373202614379085e-05,
450
+ "loss": 0.3734,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.795287187039764,
455
+ "grad_norm": 20.52510643005371,
456
+ "learning_rate": 1.3568627450980393e-05,
457
+ "loss": 0.3536,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.942562592047128,
462
+ "grad_norm": 135.24163818359375,
463
+ "learning_rate": 1.3405228758169935e-05,
464
+ "loss": 0.3952,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 8.0,
469
+ "eval_accuracy": 0.8058925476603119,
470
+ "eval_loss": 1.0216491222381592,
471
+ "eval_macro_f1": 0.8080280301047013,
472
+ "eval_runtime": 0.651,
473
+ "eval_samples_per_second": 886.303,
474
+ "eval_steps_per_second": 56.834,
475
+ "eval_weighted_f1": 0.8080734525321493,
476
+ "step": 1360
477
+ },
478
+ {
479
+ "epoch": 8.088365243004418,
480
+ "grad_norm": 129.58322143554688,
481
+ "learning_rate": 1.3241830065359478e-05,
482
+ "loss": 0.3004,
483
+ "step": 1375
484
+ },
485
+ {
486
+ "epoch": 8.235640648011781,
487
+ "grad_norm": 260.3665771484375,
488
+ "learning_rate": 1.307843137254902e-05,
489
+ "loss": 0.2231,
490
+ "step": 1400
491
+ },
492
+ {
493
+ "epoch": 8.382916053019146,
494
+ "grad_norm": 93.43733978271484,
495
+ "learning_rate": 1.2915032679738563e-05,
496
+ "loss": 0.2782,
497
+ "step": 1425
498
+ },
499
+ {
500
+ "epoch": 8.53019145802651,
501
+ "grad_norm": 5.474648475646973,
502
+ "learning_rate": 1.2751633986928104e-05,
503
+ "loss": 0.203,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.677466863033873,
508
+ "grad_norm": 0.6706470251083374,
509
+ "learning_rate": 1.2588235294117647e-05,
510
+ "loss": 0.1767,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.824742268041238,
515
+ "grad_norm": 1.8230680227279663,
516
+ "learning_rate": 1.242483660130719e-05,
517
+ "loss": 0.2426,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.972017673048601,
522
+ "grad_norm": 0.09084439277648926,
523
+ "learning_rate": 1.2261437908496733e-05,
524
+ "loss": 0.355,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 9.0,
529
+ "eval_accuracy": 0.8249566724436742,
530
+ "eval_loss": 1.0626007318496704,
531
+ "eval_macro_f1": 0.8243176180408366,
532
+ "eval_runtime": 0.6656,
533
+ "eval_samples_per_second": 866.845,
534
+ "eval_steps_per_second": 55.586,
535
+ "eval_weighted_f1": 0.8243971445103069,
536
+ "step": 1530
537
+ }
538
+ ],
539
+ "logging_steps": 25,
540
+ "max_steps": 3400,
541
+ "num_input_tokens_seen": 0,
542
+ "num_train_epochs": 20,
543
+ "save_steps": 500,
544
+ "stateful_callbacks": {
545
+ "EarlyStoppingCallback": {
546
+ "args": {
547
+ "early_stopping_patience": 3,
548
+ "early_stopping_threshold": 0.0
549
+ },
550
+ "attributes": {
551
+ "early_stopping_patience_counter": 0
552
+ }
553
+ },
554
+ "TrainerControl": {
555
+ "args": {
556
+ "should_epoch_stop": false,
557
+ "should_evaluate": false,
558
+ "should_log": false,
559
+ "should_save": true,
560
+ "should_training_stop": false
561
+ },
562
+ "attributes": {}
563
+ }
564
+ },
565
+ "total_flos": 0.0,
566
+ "train_batch_size": 8,
567
+ "trial_name": null,
568
+ "trial_params": null
569
+ }
ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-2040/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-2040/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-bamibert-topic-domain-08-s42-20260827T191225Z-6235c159/trainer/checkpoint-2040/tokenizer_config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "<unk>",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "20480": {
36
+ "content": "<mask>",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 2048,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "tokenizer_class": "XLMRobertaTokenizerFast",
54
+ "unk_token": "<unk>"
55
+ }