PhaaNe commited on
Commit
31deea9
·
verified ·
1 Parent(s): 6554329

Add files using upload-large-folder tool

Browse files
Files changed (50) hide show
  1. ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-1368/special_tokens_map.json +15 -0
  2. ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-1368/tokenizer.json +0 -0
  3. ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-1368/tokenizer_config.json +56 -0
  4. ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-1368/trainer_state.json +509 -0
  5. ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-855/trainer_state.json +336 -0
  6. ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629/special_tokens_map.json +15 -0
  7. ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629/tokenizer.json +0 -0
  8. ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629/tokenizer_config.json +56 -0
  9. ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629/trainer_state.json +597 -0
  10. ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-2172/special_tokens_map.json +15 -0
  11. ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-2172/tokenizer.json +0 -0
  12. ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-2172/tokenizer_config.json +56 -0
  13. ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-2172/trainer_state.json +777 -0
  14. ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464/special_tokens_map.json +15 -0
  15. ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464/tokenizer.json +0 -0
  16. ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464/tokenizer_config.json +56 -0
  17. ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464/trainer_state.json +537 -0
  18. ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-2013/special_tokens_map.json +15 -0
  19. ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-2013/tokenizer.json +0 -0
  20. ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-2013/tokenizer_config.json +56 -0
  21. ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-2013/trainer_state.json +724 -0
  22. ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288/special_tokens_map.json +15 -0
  23. ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288/tokenizer.json +0 -0
  24. ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288/tokenizer_config.json +56 -0
  25. ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288/trainer_state.json +823 -0
  26. ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2816/special_tokens_map.json +15 -0
  27. ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2816/tokenizer.json +0 -0
  28. ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2816/tokenizer_config.json +56 -0
  29. ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2816/trainer_state.json +1003 -0
  30. ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620/special_tokens_map.json +15 -0
  31. ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620/tokenizer.json +0 -0
  32. ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620/tokenizer_config.json +56 -0
  33. ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620/trainer_state.json +590 -0
  34. ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-2160/special_tokens_map.json +15 -0
  35. ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-2160/tokenizer.json +0 -0
  36. ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-2160/tokenizer_config.json +56 -0
  37. ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-2160/trainer_state.json +777 -0
  38. ood-vifactcheck-visobert-source-source-05-s44-20260831T033516Z-59e46264/trainer/checkpoint-1080/trainer_state.json +410 -0
  39. ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035/special_tokens_map.json +15 -0
  40. ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035/tokenizer.json +0 -0
  41. ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035/tokenizer_config.json +56 -0
  42. ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035/trainer_state.json +731 -0
  43. ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110/special_tokens_map.json +15 -0
  44. ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110/tokenizer.json +0 -0
  45. ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110/tokenizer_config.json +56 -0
  46. ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110/trainer_state.json +417 -0
  47. ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1665/special_tokens_map.json +15 -0
  48. ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1665/tokenizer.json +0 -0
  49. ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1665/tokenizer_config.json +56 -0
  50. ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1665/trainer_state.json +604 -0
ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-1368/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-1368/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-1368/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-1368/trainer_state.json ADDED
@@ -0,0 +1,509 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 855,
3
+ "best_metric": 0.5695085866794131,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-855",
5
+ "epoch": 8.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1368,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.14684287812041116,
14
+ "grad_norm": 21.10831069946289,
15
+ "learning_rate": 1.4035087719298246e-06,
16
+ "loss": 4.9373,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2936857562408223,
21
+ "grad_norm": 26.937105178833008,
22
+ "learning_rate": 2.865497076023392e-06,
23
+ "loss": 4.5007,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.44052863436123346,
28
+ "grad_norm": 30.653785705566406,
29
+ "learning_rate": 4.3274853801169596e-06,
30
+ "loss": 4.5134,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5873715124816447,
35
+ "grad_norm": 19.527666091918945,
36
+ "learning_rate": 5.789473684210527e-06,
37
+ "loss": 4.4307,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.7342143906020558,
42
+ "grad_norm": 18.099462509155273,
43
+ "learning_rate": 7.251461988304094e-06,
44
+ "loss": 4.415,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8810572687224669,
49
+ "grad_norm": 20.5585994720459,
50
+ "learning_rate": 8.713450292397661e-06,
51
+ "loss": 4.3751,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "eval_accuracy": 0.42170818505338076,
57
+ "eval_loss": 1.0713446140289307,
58
+ "eval_macro_f1": 0.4171546546546547,
59
+ "eval_runtime": 1.2554,
60
+ "eval_samples_per_second": 447.67,
61
+ "eval_steps_per_second": 28.676,
62
+ "eval_weighted_f1": 0.4164686928921804,
63
+ "step": 171
64
+ },
65
+ {
66
+ "epoch": 1.0234948604992657,
67
+ "grad_norm": 26.58049774169922,
68
+ "learning_rate": 1.017543859649123e-05,
69
+ "loss": 4.1254,
70
+ "step": 175
71
+ },
72
+ {
73
+ "epoch": 1.170337738619677,
74
+ "grad_norm": 21.180662155151367,
75
+ "learning_rate": 1.1637426900584797e-05,
76
+ "loss": 4.2122,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.3171806167400881,
81
+ "grad_norm": 25.780235290527344,
82
+ "learning_rate": 1.3099415204678362e-05,
83
+ "loss": 4.1942,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.4640234948604993,
88
+ "grad_norm": 30.095077514648438,
89
+ "learning_rate": 1.4561403508771931e-05,
90
+ "loss": 4.2033,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.6108663729809103,
95
+ "grad_norm": 18.672189712524414,
96
+ "learning_rate": 1.60233918128655e-05,
97
+ "loss": 4.2413,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.7577092511013217,
102
+ "grad_norm": 26.892139434814453,
103
+ "learning_rate": 1.7485380116959064e-05,
104
+ "loss": 4.2269,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.9045521292217327,
109
+ "grad_norm": 36.147430419921875,
110
+ "learning_rate": 1.894736842105263e-05,
111
+ "loss": 4.2321,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 2.0,
116
+ "eval_accuracy": 0.4786476868327402,
117
+ "eval_loss": 1.019220232963562,
118
+ "eval_macro_f1": 0.4709104823363945,
119
+ "eval_runtime": 1.219,
120
+ "eval_samples_per_second": 461.038,
121
+ "eval_steps_per_second": 29.533,
122
+ "eval_weighted_f1": 0.47169902989800816,
123
+ "step": 342
124
+ },
125
+ {
126
+ "epoch": 2.0469897209985315,
127
+ "grad_norm": 46.64394760131836,
128
+ "learning_rate": 1.99545159194282e-05,
129
+ "loss": 3.8873,
130
+ "step": 350
131
+ },
132
+ {
133
+ "epoch": 2.193832599118943,
134
+ "grad_norm": 26.428043365478516,
135
+ "learning_rate": 1.9792072774528916e-05,
136
+ "loss": 3.784,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.340675477239354,
141
+ "grad_norm": 50.76394271850586,
142
+ "learning_rate": 1.962962962962963e-05,
143
+ "loss": 3.7823,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.4875183553597653,
148
+ "grad_norm": 34.32876968383789,
149
+ "learning_rate": 1.9467186484730346e-05,
150
+ "loss": 3.7576,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.6343612334801763,
155
+ "grad_norm": 23.77153205871582,
156
+ "learning_rate": 1.930474333983106e-05,
157
+ "loss": 3.6831,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.7812041116005872,
162
+ "grad_norm": 30.293790817260742,
163
+ "learning_rate": 1.9142300194931776e-05,
164
+ "loss": 3.5789,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.9280469897209986,
169
+ "grad_norm": 34.51399230957031,
170
+ "learning_rate": 1.897985705003249e-05,
171
+ "loss": 3.545,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 3.0,
176
+ "eval_accuracy": 0.5249110320284698,
177
+ "eval_loss": 0.9542731642723083,
178
+ "eval_macro_f1": 0.5298419937367601,
179
+ "eval_runtime": 1.3075,
180
+ "eval_samples_per_second": 429.825,
181
+ "eval_steps_per_second": 27.533,
182
+ "eval_weighted_f1": 0.5290611044899665,
183
+ "step": 513
184
+ },
185
+ {
186
+ "epoch": 3.0704845814977975,
187
+ "grad_norm": 42.1556396484375,
188
+ "learning_rate": 1.8817413905133203e-05,
189
+ "loss": 3.1233,
190
+ "step": 525
191
+ },
192
+ {
193
+ "epoch": 3.2173274596182084,
194
+ "grad_norm": 31.78765296936035,
195
+ "learning_rate": 1.8654970760233918e-05,
196
+ "loss": 2.8295,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.36417033773862,
201
+ "grad_norm": 48.715728759765625,
202
+ "learning_rate": 1.8492527615334633e-05,
203
+ "loss": 2.7569,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.511013215859031,
208
+ "grad_norm": 49.466880798339844,
209
+ "learning_rate": 1.833008447043535e-05,
210
+ "loss": 2.9454,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.657856093979442,
215
+ "grad_norm": 24.45052719116211,
216
+ "learning_rate": 1.8167641325536064e-05,
217
+ "loss": 2.653,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.804698972099853,
222
+ "grad_norm": 56.912906646728516,
223
+ "learning_rate": 1.800519818063678e-05,
224
+ "loss": 2.73,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.951541850220264,
229
+ "grad_norm": 36.87748336791992,
230
+ "learning_rate": 1.7842755035737494e-05,
231
+ "loss": 2.8009,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 4.0,
236
+ "eval_accuracy": 0.5480427046263345,
237
+ "eval_loss": 1.0056161880493164,
238
+ "eval_macro_f1": 0.5520629043093742,
239
+ "eval_runtime": 1.1849,
240
+ "eval_samples_per_second": 474.283,
241
+ "eval_steps_per_second": 30.381,
242
+ "eval_weighted_f1": 0.5530168205091344,
243
+ "step": 684
244
+ },
245
+ {
246
+ "epoch": 4.093979441997063,
247
+ "grad_norm": 50.74189758300781,
248
+ "learning_rate": 1.768031189083821e-05,
249
+ "loss": 2.0904,
250
+ "step": 700
251
+ },
252
+ {
253
+ "epoch": 4.240822320117474,
254
+ "grad_norm": 90.03434753417969,
255
+ "learning_rate": 1.751786874593892e-05,
256
+ "loss": 1.8455,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.387665198237886,
261
+ "grad_norm": 29.36235237121582,
262
+ "learning_rate": 1.7355425601039636e-05,
263
+ "loss": 1.7616,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.534508076358296,
268
+ "grad_norm": 60.075355529785156,
269
+ "learning_rate": 1.719298245614035e-05,
270
+ "loss": 1.8333,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.681350954478708,
275
+ "grad_norm": 49.219120025634766,
276
+ "learning_rate": 1.7030539311241066e-05,
277
+ "loss": 1.7533,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.828193832599119,
282
+ "grad_norm": 62.47343063354492,
283
+ "learning_rate": 1.686809616634178e-05,
284
+ "loss": 1.7121,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.975036710719531,
289
+ "grad_norm": 66.96786499023438,
290
+ "learning_rate": 1.6705653021442497e-05,
291
+ "loss": 1.8352,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 5.0,
296
+ "eval_accuracy": 0.5729537366548043,
297
+ "eval_loss": 1.1687474250793457,
298
+ "eval_macro_f1": 0.5695085866794131,
299
+ "eval_runtime": 1.151,
300
+ "eval_samples_per_second": 488.253,
301
+ "eval_steps_per_second": 31.276,
302
+ "eval_weighted_f1": 0.5699134927369511,
303
+ "step": 855
304
+ },
305
+ {
306
+ "epoch": 5.117474302496329,
307
+ "grad_norm": 59.37667465209961,
308
+ "learning_rate": 1.654320987654321e-05,
309
+ "loss": 1.0574,
310
+ "step": 875
311
+ },
312
+ {
313
+ "epoch": 5.26431718061674,
314
+ "grad_norm": 38.62677001953125,
315
+ "learning_rate": 1.6380766731643927e-05,
316
+ "loss": 0.8771,
317
+ "step": 900
318
+ },
319
+ {
320
+ "epoch": 5.411160058737151,
321
+ "grad_norm": 33.850303649902344,
322
+ "learning_rate": 1.621832358674464e-05,
323
+ "loss": 0.9906,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.558002936857562,
328
+ "grad_norm": 34.10219955444336,
329
+ "learning_rate": 1.6055880441845354e-05,
330
+ "loss": 0.8834,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.704845814977974,
335
+ "grad_norm": 24.249958038330078,
336
+ "learning_rate": 1.589343729694607e-05,
337
+ "loss": 1.029,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.851688693098385,
342
+ "grad_norm": 53.64544677734375,
343
+ "learning_rate": 1.5730994152046787e-05,
344
+ "loss": 1.0961,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.998531571218796,
349
+ "grad_norm": 44.56748962402344,
350
+ "learning_rate": 1.55685510071475e-05,
351
+ "loss": 1.064,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 6.0,
356
+ "eval_accuracy": 0.5604982206405694,
357
+ "eval_loss": 1.5051192045211792,
358
+ "eval_macro_f1": 0.5515477372676975,
359
+ "eval_runtime": 1.1275,
360
+ "eval_samples_per_second": 498.443,
361
+ "eval_steps_per_second": 31.929,
362
+ "eval_weighted_f1": 0.5527902647910987,
363
+ "step": 1026
364
+ },
365
+ {
366
+ "epoch": 6.140969162995595,
367
+ "grad_norm": 31.487876892089844,
368
+ "learning_rate": 1.5406107862248214e-05,
369
+ "loss": 0.4447,
370
+ "step": 1050
371
+ },
372
+ {
373
+ "epoch": 6.2878120411160054,
374
+ "grad_norm": 52.43296432495117,
375
+ "learning_rate": 1.524366471734893e-05,
376
+ "loss": 0.4975,
377
+ "step": 1075
378
+ },
379
+ {
380
+ "epoch": 6.434654919236417,
381
+ "grad_norm": 38.7028694152832,
382
+ "learning_rate": 1.5081221572449645e-05,
383
+ "loss": 0.7424,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.581497797356828,
388
+ "grad_norm": 3.461725950241089,
389
+ "learning_rate": 1.4918778427550358e-05,
390
+ "loss": 0.6284,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.72834067547724,
395
+ "grad_norm": 48.99492645263672,
396
+ "learning_rate": 1.4756335282651073e-05,
397
+ "loss": 0.5094,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.87518355359765,
402
+ "grad_norm": 40.62295913696289,
403
+ "learning_rate": 1.4593892137751788e-05,
404
+ "loss": 0.799,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 7.0,
409
+ "eval_accuracy": 0.5622775800711743,
410
+ "eval_loss": 1.751611590385437,
411
+ "eval_macro_f1": 0.5669466031024418,
412
+ "eval_runtime": 1.1852,
413
+ "eval_samples_per_second": 474.174,
414
+ "eval_steps_per_second": 30.374,
415
+ "eval_weighted_f1": 0.5671642276524538,
416
+ "step": 1197
417
+ },
418
+ {
419
+ "epoch": 7.0176211453744495,
420
+ "grad_norm": 26.79924774169922,
421
+ "learning_rate": 1.4431448992852503e-05,
422
+ "loss": 0.8194,
423
+ "step": 1200
424
+ },
425
+ {
426
+ "epoch": 7.164464023494861,
427
+ "grad_norm": 21.90039825439453,
428
+ "learning_rate": 1.4269005847953217e-05,
429
+ "loss": 0.2912,
430
+ "step": 1225
431
+ },
432
+ {
433
+ "epoch": 7.311306901615271,
434
+ "grad_norm": 15.028008460998535,
435
+ "learning_rate": 1.4106562703053932e-05,
436
+ "loss": 0.2157,
437
+ "step": 1250
438
+ },
439
+ {
440
+ "epoch": 7.458149779735683,
441
+ "grad_norm": 95.66842651367188,
442
+ "learning_rate": 1.3944119558154647e-05,
443
+ "loss": 0.4676,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.604992657856094,
448
+ "grad_norm": 106.09507751464844,
449
+ "learning_rate": 1.3781676413255362e-05,
450
+ "loss": 0.3346,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.751835535976505,
455
+ "grad_norm": 4.569489002227783,
456
+ "learning_rate": 1.3619233268356076e-05,
457
+ "loss": 0.2991,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.898678414096916,
462
+ "grad_norm": 48.604896545410156,
463
+ "learning_rate": 1.345679012345679e-05,
464
+ "loss": 0.4393,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 8.0,
469
+ "eval_accuracy": 0.5569395017793595,
470
+ "eval_loss": 2.273167848587036,
471
+ "eval_macro_f1": 0.5568613709975643,
472
+ "eval_runtime": 1.2119,
473
+ "eval_samples_per_second": 463.734,
474
+ "eval_steps_per_second": 29.705,
475
+ "eval_weighted_f1": 0.5573499184084568,
476
+ "step": 1368
477
+ }
478
+ ],
479
+ "logging_steps": 25,
480
+ "max_steps": 3420,
481
+ "num_input_tokens_seen": 0,
482
+ "num_train_epochs": 20,
483
+ "save_steps": 500,
484
+ "stateful_callbacks": {
485
+ "EarlyStoppingCallback": {
486
+ "args": {
487
+ "early_stopping_patience": 3,
488
+ "early_stopping_threshold": 0.0
489
+ },
490
+ "attributes": {
491
+ "early_stopping_patience_counter": 3
492
+ }
493
+ },
494
+ "TrainerControl": {
495
+ "args": {
496
+ "should_epoch_stop": false,
497
+ "should_evaluate": false,
498
+ "should_log": false,
499
+ "should_save": true,
500
+ "should_training_stop": true
501
+ },
502
+ "attributes": {}
503
+ }
504
+ },
505
+ "total_flos": 0.0,
506
+ "train_batch_size": 8,
507
+ "trial_name": null,
508
+ "trial_params": null
509
+ }
ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-855/trainer_state.json ADDED
@@ -0,0 +1,336 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 855,
3
+ "best_metric": 0.5695085866794131,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-855",
5
+ "epoch": 5.0,
6
+ "eval_steps": 500,
7
+ "global_step": 855,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.14684287812041116,
14
+ "grad_norm": 21.10831069946289,
15
+ "learning_rate": 1.4035087719298246e-06,
16
+ "loss": 4.9373,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2936857562408223,
21
+ "grad_norm": 26.937105178833008,
22
+ "learning_rate": 2.865497076023392e-06,
23
+ "loss": 4.5007,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.44052863436123346,
28
+ "grad_norm": 30.653785705566406,
29
+ "learning_rate": 4.3274853801169596e-06,
30
+ "loss": 4.5134,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5873715124816447,
35
+ "grad_norm": 19.527666091918945,
36
+ "learning_rate": 5.789473684210527e-06,
37
+ "loss": 4.4307,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.7342143906020558,
42
+ "grad_norm": 18.099462509155273,
43
+ "learning_rate": 7.251461988304094e-06,
44
+ "loss": 4.415,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8810572687224669,
49
+ "grad_norm": 20.5585994720459,
50
+ "learning_rate": 8.713450292397661e-06,
51
+ "loss": 4.3751,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "eval_accuracy": 0.42170818505338076,
57
+ "eval_loss": 1.0713446140289307,
58
+ "eval_macro_f1": 0.4171546546546547,
59
+ "eval_runtime": 1.2554,
60
+ "eval_samples_per_second": 447.67,
61
+ "eval_steps_per_second": 28.676,
62
+ "eval_weighted_f1": 0.4164686928921804,
63
+ "step": 171
64
+ },
65
+ {
66
+ "epoch": 1.0234948604992657,
67
+ "grad_norm": 26.58049774169922,
68
+ "learning_rate": 1.017543859649123e-05,
69
+ "loss": 4.1254,
70
+ "step": 175
71
+ },
72
+ {
73
+ "epoch": 1.170337738619677,
74
+ "grad_norm": 21.180662155151367,
75
+ "learning_rate": 1.1637426900584797e-05,
76
+ "loss": 4.2122,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.3171806167400881,
81
+ "grad_norm": 25.780235290527344,
82
+ "learning_rate": 1.3099415204678362e-05,
83
+ "loss": 4.1942,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.4640234948604993,
88
+ "grad_norm": 30.095077514648438,
89
+ "learning_rate": 1.4561403508771931e-05,
90
+ "loss": 4.2033,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.6108663729809103,
95
+ "grad_norm": 18.672189712524414,
96
+ "learning_rate": 1.60233918128655e-05,
97
+ "loss": 4.2413,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.7577092511013217,
102
+ "grad_norm": 26.892139434814453,
103
+ "learning_rate": 1.7485380116959064e-05,
104
+ "loss": 4.2269,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.9045521292217327,
109
+ "grad_norm": 36.147430419921875,
110
+ "learning_rate": 1.894736842105263e-05,
111
+ "loss": 4.2321,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 2.0,
116
+ "eval_accuracy": 0.4786476868327402,
117
+ "eval_loss": 1.019220232963562,
118
+ "eval_macro_f1": 0.4709104823363945,
119
+ "eval_runtime": 1.219,
120
+ "eval_samples_per_second": 461.038,
121
+ "eval_steps_per_second": 29.533,
122
+ "eval_weighted_f1": 0.47169902989800816,
123
+ "step": 342
124
+ },
125
+ {
126
+ "epoch": 2.0469897209985315,
127
+ "grad_norm": 46.64394760131836,
128
+ "learning_rate": 1.99545159194282e-05,
129
+ "loss": 3.8873,
130
+ "step": 350
131
+ },
132
+ {
133
+ "epoch": 2.193832599118943,
134
+ "grad_norm": 26.428043365478516,
135
+ "learning_rate": 1.9792072774528916e-05,
136
+ "loss": 3.784,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.340675477239354,
141
+ "grad_norm": 50.76394271850586,
142
+ "learning_rate": 1.962962962962963e-05,
143
+ "loss": 3.7823,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.4875183553597653,
148
+ "grad_norm": 34.32876968383789,
149
+ "learning_rate": 1.9467186484730346e-05,
150
+ "loss": 3.7576,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.6343612334801763,
155
+ "grad_norm": 23.77153205871582,
156
+ "learning_rate": 1.930474333983106e-05,
157
+ "loss": 3.6831,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.7812041116005872,
162
+ "grad_norm": 30.293790817260742,
163
+ "learning_rate": 1.9142300194931776e-05,
164
+ "loss": 3.5789,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.9280469897209986,
169
+ "grad_norm": 34.51399230957031,
170
+ "learning_rate": 1.897985705003249e-05,
171
+ "loss": 3.545,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 3.0,
176
+ "eval_accuracy": 0.5249110320284698,
177
+ "eval_loss": 0.9542731642723083,
178
+ "eval_macro_f1": 0.5298419937367601,
179
+ "eval_runtime": 1.3075,
180
+ "eval_samples_per_second": 429.825,
181
+ "eval_steps_per_second": 27.533,
182
+ "eval_weighted_f1": 0.5290611044899665,
183
+ "step": 513
184
+ },
185
+ {
186
+ "epoch": 3.0704845814977975,
187
+ "grad_norm": 42.1556396484375,
188
+ "learning_rate": 1.8817413905133203e-05,
189
+ "loss": 3.1233,
190
+ "step": 525
191
+ },
192
+ {
193
+ "epoch": 3.2173274596182084,
194
+ "grad_norm": 31.78765296936035,
195
+ "learning_rate": 1.8654970760233918e-05,
196
+ "loss": 2.8295,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.36417033773862,
201
+ "grad_norm": 48.715728759765625,
202
+ "learning_rate": 1.8492527615334633e-05,
203
+ "loss": 2.7569,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.511013215859031,
208
+ "grad_norm": 49.466880798339844,
209
+ "learning_rate": 1.833008447043535e-05,
210
+ "loss": 2.9454,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.657856093979442,
215
+ "grad_norm": 24.45052719116211,
216
+ "learning_rate": 1.8167641325536064e-05,
217
+ "loss": 2.653,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.804698972099853,
222
+ "grad_norm": 56.912906646728516,
223
+ "learning_rate": 1.800519818063678e-05,
224
+ "loss": 2.73,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.951541850220264,
229
+ "grad_norm": 36.87748336791992,
230
+ "learning_rate": 1.7842755035737494e-05,
231
+ "loss": 2.8009,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 4.0,
236
+ "eval_accuracy": 0.5480427046263345,
237
+ "eval_loss": 1.0056161880493164,
238
+ "eval_macro_f1": 0.5520629043093742,
239
+ "eval_runtime": 1.1849,
240
+ "eval_samples_per_second": 474.283,
241
+ "eval_steps_per_second": 30.381,
242
+ "eval_weighted_f1": 0.5530168205091344,
243
+ "step": 684
244
+ },
245
+ {
246
+ "epoch": 4.093979441997063,
247
+ "grad_norm": 50.74189758300781,
248
+ "learning_rate": 1.768031189083821e-05,
249
+ "loss": 2.0904,
250
+ "step": 700
251
+ },
252
+ {
253
+ "epoch": 4.240822320117474,
254
+ "grad_norm": 90.03434753417969,
255
+ "learning_rate": 1.751786874593892e-05,
256
+ "loss": 1.8455,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.387665198237886,
261
+ "grad_norm": 29.36235237121582,
262
+ "learning_rate": 1.7355425601039636e-05,
263
+ "loss": 1.7616,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.534508076358296,
268
+ "grad_norm": 60.075355529785156,
269
+ "learning_rate": 1.719298245614035e-05,
270
+ "loss": 1.8333,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.681350954478708,
275
+ "grad_norm": 49.219120025634766,
276
+ "learning_rate": 1.7030539311241066e-05,
277
+ "loss": 1.7533,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.828193832599119,
282
+ "grad_norm": 62.47343063354492,
283
+ "learning_rate": 1.686809616634178e-05,
284
+ "loss": 1.7121,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.975036710719531,
289
+ "grad_norm": 66.96786499023438,
290
+ "learning_rate": 1.6705653021442497e-05,
291
+ "loss": 1.8352,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 5.0,
296
+ "eval_accuracy": 0.5729537366548043,
297
+ "eval_loss": 1.1687474250793457,
298
+ "eval_macro_f1": 0.5695085866794131,
299
+ "eval_runtime": 1.151,
300
+ "eval_samples_per_second": 488.253,
301
+ "eval_steps_per_second": 31.276,
302
+ "eval_weighted_f1": 0.5699134927369511,
303
+ "step": 855
304
+ }
305
+ ],
306
+ "logging_steps": 25,
307
+ "max_steps": 3420,
308
+ "num_input_tokens_seen": 0,
309
+ "num_train_epochs": 20,
310
+ "save_steps": 500,
311
+ "stateful_callbacks": {
312
+ "EarlyStoppingCallback": {
313
+ "args": {
314
+ "early_stopping_patience": 3,
315
+ "early_stopping_threshold": 0.0
316
+ },
317
+ "attributes": {
318
+ "early_stopping_patience_counter": 0
319
+ }
320
+ },
321
+ "TrainerControl": {
322
+ "args": {
323
+ "should_epoch_stop": false,
324
+ "should_evaluate": false,
325
+ "should_log": false,
326
+ "should_save": true,
327
+ "should_training_stop": false
328
+ },
329
+ "attributes": {}
330
+ }
331
+ },
332
+ "total_flos": 0.0,
333
+ "train_batch_size": 8,
334
+ "trial_name": null,
335
+ "trial_params": null
336
+ }
ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629/trainer_state.json ADDED
@@ -0,0 +1,597 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1629,
3
+ "best_metric": 0.5891044915695222,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629",
5
+ "epoch": 9.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1629,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13812154696132597,
14
+ "grad_norm": 27.335662841796875,
15
+ "learning_rate": 1.3259668508287293e-06,
16
+ "loss": 4.9035,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27624309392265195,
21
+ "grad_norm": 21.327495574951172,
22
+ "learning_rate": 2.707182320441989e-06,
23
+ "loss": 4.541,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4143646408839779,
28
+ "grad_norm": 22.216156005859375,
29
+ "learning_rate": 4.088397790055249e-06,
30
+ "loss": 4.4314,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5524861878453039,
35
+ "grad_norm": 27.660001754760742,
36
+ "learning_rate": 5.469613259668509e-06,
37
+ "loss": 4.4338,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6906077348066298,
42
+ "grad_norm": 25.91851806640625,
43
+ "learning_rate": 6.850828729281769e-06,
44
+ "loss": 4.4312,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8287292817679558,
49
+ "grad_norm": 31.54214859008789,
50
+ "learning_rate": 8.232044198895029e-06,
51
+ "loss": 4.3709,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9668508287292817,
56
+ "grad_norm": 34.3691520690918,
57
+ "learning_rate": 9.613259668508288e-06,
58
+ "loss": 4.3934,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4381443298969072,
64
+ "eval_loss": 1.053492784500122,
65
+ "eval_macro_f1": 0.39091448290736547,
66
+ "eval_runtime": 1.2234,
67
+ "eval_samples_per_second": 475.721,
68
+ "eval_steps_per_second": 30.243,
69
+ "eval_weighted_f1": 0.39632724315466367,
70
+ "step": 181
71
+ },
72
+ {
73
+ "epoch": 1.1049723756906078,
74
+ "grad_norm": 25.648290634155273,
75
+ "learning_rate": 1.0994475138121547e-05,
76
+ "loss": 4.224,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2430939226519337,
81
+ "grad_norm": 19.31011390686035,
82
+ "learning_rate": 1.2375690607734806e-05,
83
+ "loss": 4.18,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3812154696132597,
88
+ "grad_norm": 38.72648620605469,
89
+ "learning_rate": 1.3756906077348067e-05,
90
+ "loss": 4.2667,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.5193370165745856,
95
+ "grad_norm": 26.38353157043457,
96
+ "learning_rate": 1.5138121546961326e-05,
97
+ "loss": 4.2076,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6574585635359116,
102
+ "grad_norm": 57.3071403503418,
103
+ "learning_rate": 1.6519337016574586e-05,
104
+ "loss": 4.3571,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7955801104972375,
109
+ "grad_norm": 18.260107040405273,
110
+ "learning_rate": 1.7900552486187846e-05,
111
+ "loss": 4.2109,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9337016574585635,
116
+ "grad_norm": 14.118206024169922,
117
+ "learning_rate": 1.9281767955801107e-05,
118
+ "loss": 4.3317,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.47594501718213056,
124
+ "eval_loss": 1.0273001194000244,
125
+ "eval_macro_f1": 0.44513648245031207,
126
+ "eval_runtime": 1.2291,
127
+ "eval_samples_per_second": 473.504,
128
+ "eval_steps_per_second": 30.102,
129
+ "eval_weighted_f1": 0.44956769058920576,
130
+ "step": 362
131
+ },
132
+ {
133
+ "epoch": 2.0718232044198897,
134
+ "grad_norm": 18.690710067749023,
135
+ "learning_rate": 1.9926335174953962e-05,
136
+ "loss": 4.0973,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.2099447513812156,
141
+ "grad_norm": 24.73369598388672,
142
+ "learning_rate": 1.9772866789441377e-05,
143
+ "loss": 3.8954,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.3480662983425415,
148
+ "grad_norm": 34.313846588134766,
149
+ "learning_rate": 1.9619398403928794e-05,
150
+ "loss": 3.9883,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.4861878453038675,
155
+ "grad_norm": 21.537168502807617,
156
+ "learning_rate": 1.946593001841621e-05,
157
+ "loss": 3.7673,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.6243093922651934,
162
+ "grad_norm": 21.02507209777832,
163
+ "learning_rate": 1.9312461632903623e-05,
164
+ "loss": 3.5875,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7624309392265194,
169
+ "grad_norm": 30.997478485107422,
170
+ "learning_rate": 1.915899324739104e-05,
171
+ "loss": 3.5076,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.9005524861878453,
176
+ "grad_norm": 26.355712890625,
177
+ "learning_rate": 1.9005524861878455e-05,
178
+ "loss": 3.5718,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.49140893470790376,
184
+ "eval_loss": 1.0402227640151978,
185
+ "eval_macro_f1": 0.4766026445849854,
186
+ "eval_runtime": 1.2221,
187
+ "eval_samples_per_second": 476.225,
188
+ "eval_steps_per_second": 30.276,
189
+ "eval_weighted_f1": 0.47945146096021224,
190
+ "step": 543
191
+ },
192
+ {
193
+ "epoch": 3.0386740331491713,
194
+ "grad_norm": 27.272193908691406,
195
+ "learning_rate": 1.885205647636587e-05,
196
+ "loss": 3.5849,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.1767955801104972,
201
+ "grad_norm": 27.247163772583008,
202
+ "learning_rate": 1.8698588090853287e-05,
203
+ "loss": 2.9377,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.314917127071823,
208
+ "grad_norm": 43.23213195800781,
209
+ "learning_rate": 1.85451197053407e-05,
210
+ "loss": 3.0916,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.453038674033149,
215
+ "grad_norm": 43.39284896850586,
216
+ "learning_rate": 1.8391651319828116e-05,
217
+ "loss": 2.7921,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.591160220994475,
222
+ "grad_norm": 41.5806999206543,
223
+ "learning_rate": 1.823818293431553e-05,
224
+ "loss": 2.8743,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.729281767955801,
229
+ "grad_norm": 37.16095733642578,
230
+ "learning_rate": 1.8084714548802948e-05,
231
+ "loss": 2.8339,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.867403314917127,
236
+ "grad_norm": 74.77664184570312,
237
+ "learning_rate": 1.7931246163290362e-05,
238
+ "loss": 2.8439,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 4.0,
243
+ "eval_accuracy": 0.5584192439862543,
244
+ "eval_loss": 1.016396403312683,
245
+ "eval_macro_f1": 0.5380304714259044,
246
+ "eval_runtime": 1.2131,
247
+ "eval_samples_per_second": 479.762,
248
+ "eval_steps_per_second": 30.5,
249
+ "eval_weighted_f1": 0.5391344861520974,
250
+ "step": 724
251
+ },
252
+ {
253
+ "epoch": 4.005524861878453,
254
+ "grad_norm": 27.540451049804688,
255
+ "learning_rate": 1.7777777777777777e-05,
256
+ "loss": 2.8797,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.143646408839779,
261
+ "grad_norm": 39.97393035888672,
262
+ "learning_rate": 1.7624309392265194e-05,
263
+ "loss": 1.9942,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.281767955801105,
268
+ "grad_norm": 41.98676681518555,
269
+ "learning_rate": 1.747084100675261e-05,
270
+ "loss": 1.9718,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.419889502762431,
275
+ "grad_norm": 28.36094093322754,
276
+ "learning_rate": 1.7317372621240026e-05,
277
+ "loss": 1.8035,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.558011049723757,
282
+ "grad_norm": 39.36656188964844,
283
+ "learning_rate": 1.716390423572744e-05,
284
+ "loss": 1.897,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.696132596685083,
289
+ "grad_norm": 49.739192962646484,
290
+ "learning_rate": 1.701043585021486e-05,
291
+ "loss": 2.1822,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.834254143646409,
296
+ "grad_norm": 29.114208221435547,
297
+ "learning_rate": 1.6856967464702273e-05,
298
+ "loss": 1.947,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.972375690607735,
303
+ "grad_norm": 40.23611831665039,
304
+ "learning_rate": 1.670349907918969e-05,
305
+ "loss": 1.8745,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.5721649484536082,
311
+ "eval_loss": 1.1483445167541504,
312
+ "eval_macro_f1": 0.560929301297984,
313
+ "eval_runtime": 1.2323,
314
+ "eval_samples_per_second": 472.27,
315
+ "eval_steps_per_second": 30.024,
316
+ "eval_weighted_f1": 0.5569620726176342,
317
+ "step": 905
318
+ },
319
+ {
320
+ "epoch": 5.110497237569061,
321
+ "grad_norm": 27.892236709594727,
322
+ "learning_rate": 1.6550030693677105e-05,
323
+ "loss": 1.2467,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.248618784530387,
328
+ "grad_norm": 77.62355041503906,
329
+ "learning_rate": 1.639656230816452e-05,
330
+ "loss": 1.0786,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.386740331491713,
335
+ "grad_norm": 41.20305252075195,
336
+ "learning_rate": 1.6243093922651937e-05,
337
+ "loss": 0.9959,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.524861878453039,
342
+ "grad_norm": 41.47410583496094,
343
+ "learning_rate": 1.608962553713935e-05,
344
+ "loss": 0.9151,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.662983425414365,
349
+ "grad_norm": 48.764469146728516,
350
+ "learning_rate": 1.5936157151626766e-05,
351
+ "loss": 1.0737,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.801104972375691,
356
+ "grad_norm": 58.2398567199707,
357
+ "learning_rate": 1.5782688766114183e-05,
358
+ "loss": 1.1723,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.939226519337017,
363
+ "grad_norm": 87.3487777709961,
364
+ "learning_rate": 1.5629220380601598e-05,
365
+ "loss": 1.2165,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.5446735395189003,
371
+ "eval_loss": 1.4995858669281006,
372
+ "eval_macro_f1": 0.5244655563767523,
373
+ "eval_runtime": 1.3167,
374
+ "eval_samples_per_second": 442.024,
375
+ "eval_steps_per_second": 28.101,
376
+ "eval_weighted_f1": 0.525747727022952,
377
+ "step": 1086
378
+ },
379
+ {
380
+ "epoch": 6.077348066298343,
381
+ "grad_norm": 43.04379653930664,
382
+ "learning_rate": 1.5475751995089012e-05,
383
+ "loss": 0.8937,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.2154696132596685,
388
+ "grad_norm": 50.99885940551758,
389
+ "learning_rate": 1.532228360957643e-05,
390
+ "loss": 0.4374,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.3535911602209945,
395
+ "grad_norm": 16.83222198486328,
396
+ "learning_rate": 1.5168815224063844e-05,
397
+ "loss": 0.5404,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.49171270718232,
402
+ "grad_norm": 79.08232879638672,
403
+ "learning_rate": 1.5015346838551258e-05,
404
+ "loss": 0.4842,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.629834254143646,
409
+ "grad_norm": 12.174334526062012,
410
+ "learning_rate": 1.4861878453038676e-05,
411
+ "loss": 0.6169,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.767955801104972,
416
+ "grad_norm": 40.94103240966797,
417
+ "learning_rate": 1.470841006752609e-05,
418
+ "loss": 0.7293,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.906077348066298,
423
+ "grad_norm": 71.6591796875,
424
+ "learning_rate": 1.4554941682013506e-05,
425
+ "loss": 0.5296,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 7.0,
430
+ "eval_accuracy": 0.5670103092783505,
431
+ "eval_loss": 1.8478572368621826,
432
+ "eval_macro_f1": 0.5649927173197177,
433
+ "eval_runtime": 1.2776,
434
+ "eval_samples_per_second": 455.551,
435
+ "eval_steps_per_second": 28.961,
436
+ "eval_weighted_f1": 0.563813073865988,
437
+ "step": 1267
438
+ },
439
+ {
440
+ "epoch": 7.044198895027624,
441
+ "grad_norm": 47.853267669677734,
442
+ "learning_rate": 1.4401473296500922e-05,
443
+ "loss": 0.5994,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.18232044198895,
448
+ "grad_norm": 11.483506202697754,
449
+ "learning_rate": 1.4248004910988338e-05,
450
+ "loss": 0.2776,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.320441988950276,
455
+ "grad_norm": 66.74475860595703,
456
+ "learning_rate": 1.4094536525475753e-05,
457
+ "loss": 0.3624,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.458563535911602,
462
+ "grad_norm": 30.32476806640625,
463
+ "learning_rate": 1.394106813996317e-05,
464
+ "loss": 0.3493,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.596685082872928,
469
+ "grad_norm": 117.52507019042969,
470
+ "learning_rate": 1.3787599754450585e-05,
471
+ "loss": 0.377,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.734806629834254,
476
+ "grad_norm": 27.136577606201172,
477
+ "learning_rate": 1.3634131368937999e-05,
478
+ "loss": 0.4173,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.87292817679558,
483
+ "grad_norm": 16.560192108154297,
484
+ "learning_rate": 1.3480662983425417e-05,
485
+ "loss": 0.4672,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 8.0,
490
+ "eval_accuracy": 0.5618556701030928,
491
+ "eval_loss": 2.380108594894409,
492
+ "eval_macro_f1": 0.5536149797719663,
493
+ "eval_runtime": 1.181,
494
+ "eval_samples_per_second": 492.792,
495
+ "eval_steps_per_second": 31.329,
496
+ "eval_weighted_f1": 0.5546006188884759,
497
+ "step": 1448
498
+ },
499
+ {
500
+ "epoch": 8.011049723756907,
501
+ "grad_norm": 11.204202651977539,
502
+ "learning_rate": 1.3327194597912831e-05,
503
+ "loss": 0.2238,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.149171270718233,
508
+ "grad_norm": 42.98948669433594,
509
+ "learning_rate": 1.3173726212400245e-05,
510
+ "loss": 0.1752,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.287292817679559,
515
+ "grad_norm": 78.01802062988281,
516
+ "learning_rate": 1.3020257826887662e-05,
517
+ "loss": 0.1865,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.425414364640885,
522
+ "grad_norm": 66.42144012451172,
523
+ "learning_rate": 1.2866789441375078e-05,
524
+ "loss": 0.3063,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.56353591160221,
529
+ "grad_norm": 59.27225112915039,
530
+ "learning_rate": 1.2713321055862494e-05,
531
+ "loss": 0.1966,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.701657458563536,
536
+ "grad_norm": 10.438326835632324,
537
+ "learning_rate": 1.2559852670349908e-05,
538
+ "loss": 0.2691,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.839779005524862,
543
+ "grad_norm": 9.469968795776367,
544
+ "learning_rate": 1.2406384284837326e-05,
545
+ "loss": 0.2183,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.977900552486188,
550
+ "grad_norm": 99.45462799072266,
551
+ "learning_rate": 1.225291589932474e-05,
552
+ "loss": 0.3943,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 9.0,
557
+ "eval_accuracy": 0.5876288659793815,
558
+ "eval_loss": 2.641221523284912,
559
+ "eval_macro_f1": 0.5891044915695222,
560
+ "eval_runtime": 1.2356,
561
+ "eval_samples_per_second": 471.013,
562
+ "eval_steps_per_second": 29.944,
563
+ "eval_weighted_f1": 0.588638521942443,
564
+ "step": 1629
565
+ }
566
+ ],
567
+ "logging_steps": 25,
568
+ "max_steps": 3620,
569
+ "num_input_tokens_seen": 0,
570
+ "num_train_epochs": 20,
571
+ "save_steps": 500,
572
+ "stateful_callbacks": {
573
+ "EarlyStoppingCallback": {
574
+ "args": {
575
+ "early_stopping_patience": 3,
576
+ "early_stopping_threshold": 0.0
577
+ },
578
+ "attributes": {
579
+ "early_stopping_patience_counter": 0
580
+ }
581
+ },
582
+ "TrainerControl": {
583
+ "args": {
584
+ "should_epoch_stop": false,
585
+ "should_evaluate": false,
586
+ "should_log": false,
587
+ "should_save": true,
588
+ "should_training_stop": false
589
+ },
590
+ "attributes": {}
591
+ }
592
+ },
593
+ "total_flos": 0.0,
594
+ "train_batch_size": 8,
595
+ "trial_name": null,
596
+ "trial_params": null
597
+ }
ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-2172/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-2172/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-2172/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-2172/trainer_state.json ADDED
@@ -0,0 +1,777 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1629,
3
+ "best_metric": 0.5891044915695222,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-02-s43-20260831T012427Z-1a169064/trainer/checkpoint-1629",
5
+ "epoch": 12.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2172,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13812154696132597,
14
+ "grad_norm": 27.335662841796875,
15
+ "learning_rate": 1.3259668508287293e-06,
16
+ "loss": 4.9035,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27624309392265195,
21
+ "grad_norm": 21.327495574951172,
22
+ "learning_rate": 2.707182320441989e-06,
23
+ "loss": 4.541,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4143646408839779,
28
+ "grad_norm": 22.216156005859375,
29
+ "learning_rate": 4.088397790055249e-06,
30
+ "loss": 4.4314,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5524861878453039,
35
+ "grad_norm": 27.660001754760742,
36
+ "learning_rate": 5.469613259668509e-06,
37
+ "loss": 4.4338,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6906077348066298,
42
+ "grad_norm": 25.91851806640625,
43
+ "learning_rate": 6.850828729281769e-06,
44
+ "loss": 4.4312,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8287292817679558,
49
+ "grad_norm": 31.54214859008789,
50
+ "learning_rate": 8.232044198895029e-06,
51
+ "loss": 4.3709,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9668508287292817,
56
+ "grad_norm": 34.3691520690918,
57
+ "learning_rate": 9.613259668508288e-06,
58
+ "loss": 4.3934,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4381443298969072,
64
+ "eval_loss": 1.053492784500122,
65
+ "eval_macro_f1": 0.39091448290736547,
66
+ "eval_runtime": 1.2234,
67
+ "eval_samples_per_second": 475.721,
68
+ "eval_steps_per_second": 30.243,
69
+ "eval_weighted_f1": 0.39632724315466367,
70
+ "step": 181
71
+ },
72
+ {
73
+ "epoch": 1.1049723756906078,
74
+ "grad_norm": 25.648290634155273,
75
+ "learning_rate": 1.0994475138121547e-05,
76
+ "loss": 4.224,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2430939226519337,
81
+ "grad_norm": 19.31011390686035,
82
+ "learning_rate": 1.2375690607734806e-05,
83
+ "loss": 4.18,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3812154696132597,
88
+ "grad_norm": 38.72648620605469,
89
+ "learning_rate": 1.3756906077348067e-05,
90
+ "loss": 4.2667,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.5193370165745856,
95
+ "grad_norm": 26.38353157043457,
96
+ "learning_rate": 1.5138121546961326e-05,
97
+ "loss": 4.2076,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6574585635359116,
102
+ "grad_norm": 57.3071403503418,
103
+ "learning_rate": 1.6519337016574586e-05,
104
+ "loss": 4.3571,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7955801104972375,
109
+ "grad_norm": 18.260107040405273,
110
+ "learning_rate": 1.7900552486187846e-05,
111
+ "loss": 4.2109,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9337016574585635,
116
+ "grad_norm": 14.118206024169922,
117
+ "learning_rate": 1.9281767955801107e-05,
118
+ "loss": 4.3317,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.47594501718213056,
124
+ "eval_loss": 1.0273001194000244,
125
+ "eval_macro_f1": 0.44513648245031207,
126
+ "eval_runtime": 1.2291,
127
+ "eval_samples_per_second": 473.504,
128
+ "eval_steps_per_second": 30.102,
129
+ "eval_weighted_f1": 0.44956769058920576,
130
+ "step": 362
131
+ },
132
+ {
133
+ "epoch": 2.0718232044198897,
134
+ "grad_norm": 18.690710067749023,
135
+ "learning_rate": 1.9926335174953962e-05,
136
+ "loss": 4.0973,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.2099447513812156,
141
+ "grad_norm": 24.73369598388672,
142
+ "learning_rate": 1.9772866789441377e-05,
143
+ "loss": 3.8954,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.3480662983425415,
148
+ "grad_norm": 34.313846588134766,
149
+ "learning_rate": 1.9619398403928794e-05,
150
+ "loss": 3.9883,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.4861878453038675,
155
+ "grad_norm": 21.537168502807617,
156
+ "learning_rate": 1.946593001841621e-05,
157
+ "loss": 3.7673,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.6243093922651934,
162
+ "grad_norm": 21.02507209777832,
163
+ "learning_rate": 1.9312461632903623e-05,
164
+ "loss": 3.5875,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7624309392265194,
169
+ "grad_norm": 30.997478485107422,
170
+ "learning_rate": 1.915899324739104e-05,
171
+ "loss": 3.5076,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.9005524861878453,
176
+ "grad_norm": 26.355712890625,
177
+ "learning_rate": 1.9005524861878455e-05,
178
+ "loss": 3.5718,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.49140893470790376,
184
+ "eval_loss": 1.0402227640151978,
185
+ "eval_macro_f1": 0.4766026445849854,
186
+ "eval_runtime": 1.2221,
187
+ "eval_samples_per_second": 476.225,
188
+ "eval_steps_per_second": 30.276,
189
+ "eval_weighted_f1": 0.47945146096021224,
190
+ "step": 543
191
+ },
192
+ {
193
+ "epoch": 3.0386740331491713,
194
+ "grad_norm": 27.272193908691406,
195
+ "learning_rate": 1.885205647636587e-05,
196
+ "loss": 3.5849,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.1767955801104972,
201
+ "grad_norm": 27.247163772583008,
202
+ "learning_rate": 1.8698588090853287e-05,
203
+ "loss": 2.9377,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.314917127071823,
208
+ "grad_norm": 43.23213195800781,
209
+ "learning_rate": 1.85451197053407e-05,
210
+ "loss": 3.0916,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.453038674033149,
215
+ "grad_norm": 43.39284896850586,
216
+ "learning_rate": 1.8391651319828116e-05,
217
+ "loss": 2.7921,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.591160220994475,
222
+ "grad_norm": 41.5806999206543,
223
+ "learning_rate": 1.823818293431553e-05,
224
+ "loss": 2.8743,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.729281767955801,
229
+ "grad_norm": 37.16095733642578,
230
+ "learning_rate": 1.8084714548802948e-05,
231
+ "loss": 2.8339,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.867403314917127,
236
+ "grad_norm": 74.77664184570312,
237
+ "learning_rate": 1.7931246163290362e-05,
238
+ "loss": 2.8439,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 4.0,
243
+ "eval_accuracy": 0.5584192439862543,
244
+ "eval_loss": 1.016396403312683,
245
+ "eval_macro_f1": 0.5380304714259044,
246
+ "eval_runtime": 1.2131,
247
+ "eval_samples_per_second": 479.762,
248
+ "eval_steps_per_second": 30.5,
249
+ "eval_weighted_f1": 0.5391344861520974,
250
+ "step": 724
251
+ },
252
+ {
253
+ "epoch": 4.005524861878453,
254
+ "grad_norm": 27.540451049804688,
255
+ "learning_rate": 1.7777777777777777e-05,
256
+ "loss": 2.8797,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.143646408839779,
261
+ "grad_norm": 39.97393035888672,
262
+ "learning_rate": 1.7624309392265194e-05,
263
+ "loss": 1.9942,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.281767955801105,
268
+ "grad_norm": 41.98676681518555,
269
+ "learning_rate": 1.747084100675261e-05,
270
+ "loss": 1.9718,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.419889502762431,
275
+ "grad_norm": 28.36094093322754,
276
+ "learning_rate": 1.7317372621240026e-05,
277
+ "loss": 1.8035,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.558011049723757,
282
+ "grad_norm": 39.36656188964844,
283
+ "learning_rate": 1.716390423572744e-05,
284
+ "loss": 1.897,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.696132596685083,
289
+ "grad_norm": 49.739192962646484,
290
+ "learning_rate": 1.701043585021486e-05,
291
+ "loss": 2.1822,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.834254143646409,
296
+ "grad_norm": 29.114208221435547,
297
+ "learning_rate": 1.6856967464702273e-05,
298
+ "loss": 1.947,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.972375690607735,
303
+ "grad_norm": 40.23611831665039,
304
+ "learning_rate": 1.670349907918969e-05,
305
+ "loss": 1.8745,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.5721649484536082,
311
+ "eval_loss": 1.1483445167541504,
312
+ "eval_macro_f1": 0.560929301297984,
313
+ "eval_runtime": 1.2323,
314
+ "eval_samples_per_second": 472.27,
315
+ "eval_steps_per_second": 30.024,
316
+ "eval_weighted_f1": 0.5569620726176342,
317
+ "step": 905
318
+ },
319
+ {
320
+ "epoch": 5.110497237569061,
321
+ "grad_norm": 27.892236709594727,
322
+ "learning_rate": 1.6550030693677105e-05,
323
+ "loss": 1.2467,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.248618784530387,
328
+ "grad_norm": 77.62355041503906,
329
+ "learning_rate": 1.639656230816452e-05,
330
+ "loss": 1.0786,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.386740331491713,
335
+ "grad_norm": 41.20305252075195,
336
+ "learning_rate": 1.6243093922651937e-05,
337
+ "loss": 0.9959,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.524861878453039,
342
+ "grad_norm": 41.47410583496094,
343
+ "learning_rate": 1.608962553713935e-05,
344
+ "loss": 0.9151,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.662983425414365,
349
+ "grad_norm": 48.764469146728516,
350
+ "learning_rate": 1.5936157151626766e-05,
351
+ "loss": 1.0737,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.801104972375691,
356
+ "grad_norm": 58.2398567199707,
357
+ "learning_rate": 1.5782688766114183e-05,
358
+ "loss": 1.1723,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.939226519337017,
363
+ "grad_norm": 87.3487777709961,
364
+ "learning_rate": 1.5629220380601598e-05,
365
+ "loss": 1.2165,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.5446735395189003,
371
+ "eval_loss": 1.4995858669281006,
372
+ "eval_macro_f1": 0.5244655563767523,
373
+ "eval_runtime": 1.3167,
374
+ "eval_samples_per_second": 442.024,
375
+ "eval_steps_per_second": 28.101,
376
+ "eval_weighted_f1": 0.525747727022952,
377
+ "step": 1086
378
+ },
379
+ {
380
+ "epoch": 6.077348066298343,
381
+ "grad_norm": 43.04379653930664,
382
+ "learning_rate": 1.5475751995089012e-05,
383
+ "loss": 0.8937,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.2154696132596685,
388
+ "grad_norm": 50.99885940551758,
389
+ "learning_rate": 1.532228360957643e-05,
390
+ "loss": 0.4374,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.3535911602209945,
395
+ "grad_norm": 16.83222198486328,
396
+ "learning_rate": 1.5168815224063844e-05,
397
+ "loss": 0.5404,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.49171270718232,
402
+ "grad_norm": 79.08232879638672,
403
+ "learning_rate": 1.5015346838551258e-05,
404
+ "loss": 0.4842,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.629834254143646,
409
+ "grad_norm": 12.174334526062012,
410
+ "learning_rate": 1.4861878453038676e-05,
411
+ "loss": 0.6169,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.767955801104972,
416
+ "grad_norm": 40.94103240966797,
417
+ "learning_rate": 1.470841006752609e-05,
418
+ "loss": 0.7293,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.906077348066298,
423
+ "grad_norm": 71.6591796875,
424
+ "learning_rate": 1.4554941682013506e-05,
425
+ "loss": 0.5296,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 7.0,
430
+ "eval_accuracy": 0.5670103092783505,
431
+ "eval_loss": 1.8478572368621826,
432
+ "eval_macro_f1": 0.5649927173197177,
433
+ "eval_runtime": 1.2776,
434
+ "eval_samples_per_second": 455.551,
435
+ "eval_steps_per_second": 28.961,
436
+ "eval_weighted_f1": 0.563813073865988,
437
+ "step": 1267
438
+ },
439
+ {
440
+ "epoch": 7.044198895027624,
441
+ "grad_norm": 47.853267669677734,
442
+ "learning_rate": 1.4401473296500922e-05,
443
+ "loss": 0.5994,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.18232044198895,
448
+ "grad_norm": 11.483506202697754,
449
+ "learning_rate": 1.4248004910988338e-05,
450
+ "loss": 0.2776,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.320441988950276,
455
+ "grad_norm": 66.74475860595703,
456
+ "learning_rate": 1.4094536525475753e-05,
457
+ "loss": 0.3624,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.458563535911602,
462
+ "grad_norm": 30.32476806640625,
463
+ "learning_rate": 1.394106813996317e-05,
464
+ "loss": 0.3493,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.596685082872928,
469
+ "grad_norm": 117.52507019042969,
470
+ "learning_rate": 1.3787599754450585e-05,
471
+ "loss": 0.377,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.734806629834254,
476
+ "grad_norm": 27.136577606201172,
477
+ "learning_rate": 1.3634131368937999e-05,
478
+ "loss": 0.4173,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.87292817679558,
483
+ "grad_norm": 16.560192108154297,
484
+ "learning_rate": 1.3480662983425417e-05,
485
+ "loss": 0.4672,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 8.0,
490
+ "eval_accuracy": 0.5618556701030928,
491
+ "eval_loss": 2.380108594894409,
492
+ "eval_macro_f1": 0.5536149797719663,
493
+ "eval_runtime": 1.181,
494
+ "eval_samples_per_second": 492.792,
495
+ "eval_steps_per_second": 31.329,
496
+ "eval_weighted_f1": 0.5546006188884759,
497
+ "step": 1448
498
+ },
499
+ {
500
+ "epoch": 8.011049723756907,
501
+ "grad_norm": 11.204202651977539,
502
+ "learning_rate": 1.3327194597912831e-05,
503
+ "loss": 0.2238,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.149171270718233,
508
+ "grad_norm": 42.98948669433594,
509
+ "learning_rate": 1.3173726212400245e-05,
510
+ "loss": 0.1752,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.287292817679559,
515
+ "grad_norm": 78.01802062988281,
516
+ "learning_rate": 1.3020257826887662e-05,
517
+ "loss": 0.1865,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.425414364640885,
522
+ "grad_norm": 66.42144012451172,
523
+ "learning_rate": 1.2866789441375078e-05,
524
+ "loss": 0.3063,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.56353591160221,
529
+ "grad_norm": 59.27225112915039,
530
+ "learning_rate": 1.2713321055862494e-05,
531
+ "loss": 0.1966,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.701657458563536,
536
+ "grad_norm": 10.438326835632324,
537
+ "learning_rate": 1.2559852670349908e-05,
538
+ "loss": 0.2691,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.839779005524862,
543
+ "grad_norm": 9.469968795776367,
544
+ "learning_rate": 1.2406384284837326e-05,
545
+ "loss": 0.2183,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.977900552486188,
550
+ "grad_norm": 99.45462799072266,
551
+ "learning_rate": 1.225291589932474e-05,
552
+ "loss": 0.3943,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 9.0,
557
+ "eval_accuracy": 0.5876288659793815,
558
+ "eval_loss": 2.641221523284912,
559
+ "eval_macro_f1": 0.5891044915695222,
560
+ "eval_runtime": 1.2356,
561
+ "eval_samples_per_second": 471.013,
562
+ "eval_steps_per_second": 29.944,
563
+ "eval_weighted_f1": 0.588638521942443,
564
+ "step": 1629
565
+ },
566
+ {
567
+ "epoch": 9.116022099447514,
568
+ "grad_norm": 76.10488891601562,
569
+ "learning_rate": 1.2099447513812154e-05,
570
+ "loss": 0.2742,
571
+ "step": 1650
572
+ },
573
+ {
574
+ "epoch": 9.25414364640884,
575
+ "grad_norm": 12.389909744262695,
576
+ "learning_rate": 1.1945979128299572e-05,
577
+ "loss": 0.1632,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.392265193370166,
582
+ "grad_norm": 2.2620766162872314,
583
+ "learning_rate": 1.1792510742786986e-05,
584
+ "loss": 0.2473,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.530386740331492,
589
+ "grad_norm": 40.60436248779297,
590
+ "learning_rate": 1.1639042357274402e-05,
591
+ "loss": 0.1408,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.668508287292818,
596
+ "grad_norm": 4.49537467956543,
597
+ "learning_rate": 1.1485573971761818e-05,
598
+ "loss": 0.2387,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 9.806629834254144,
603
+ "grad_norm": 3.7760696411132812,
604
+ "learning_rate": 1.1332105586249233e-05,
605
+ "loss": 0.149,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 9.94475138121547,
610
+ "grad_norm": 53.04218292236328,
611
+ "learning_rate": 1.1178637200736649e-05,
612
+ "loss": 0.2103,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 10.0,
617
+ "eval_accuracy": 0.563573883161512,
618
+ "eval_loss": 3.3874754905700684,
619
+ "eval_macro_f1": 0.5615303872243169,
620
+ "eval_runtime": 1.2204,
621
+ "eval_samples_per_second": 476.88,
622
+ "eval_steps_per_second": 30.317,
623
+ "eval_weighted_f1": 0.5626884582319744,
624
+ "step": 1810
625
+ },
626
+ {
627
+ "epoch": 10.082872928176796,
628
+ "grad_norm": 14.091012001037598,
629
+ "learning_rate": 1.1025168815224065e-05,
630
+ "loss": 0.2321,
631
+ "step": 1825
632
+ },
633
+ {
634
+ "epoch": 10.220994475138122,
635
+ "grad_norm": 0.43849682807922363,
636
+ "learning_rate": 1.087170042971148e-05,
637
+ "loss": 0.0478,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.359116022099448,
642
+ "grad_norm": 59.10844421386719,
643
+ "learning_rate": 1.0718232044198895e-05,
644
+ "loss": 0.0807,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.497237569060774,
649
+ "grad_norm": 0.4257087707519531,
650
+ "learning_rate": 1.0564763658686313e-05,
651
+ "loss": 0.034,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.6353591160221,
656
+ "grad_norm": 0.8232848048210144,
657
+ "learning_rate": 1.0411295273173727e-05,
658
+ "loss": 0.1595,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 10.773480662983426,
663
+ "grad_norm": 79.92752075195312,
664
+ "learning_rate": 1.0257826887661141e-05,
665
+ "loss": 0.1427,
666
+ "step": 1950
667
+ },
668
+ {
669
+ "epoch": 10.911602209944752,
670
+ "grad_norm": 6.015427112579346,
671
+ "learning_rate": 1.010435850214856e-05,
672
+ "loss": 0.1513,
673
+ "step": 1975
674
+ },
675
+ {
676
+ "epoch": 11.0,
677
+ "eval_accuracy": 0.5601374570446735,
678
+ "eval_loss": 3.653582811355591,
679
+ "eval_macro_f1": 0.5640452894794149,
680
+ "eval_runtime": 1.3803,
681
+ "eval_samples_per_second": 421.65,
682
+ "eval_steps_per_second": 26.806,
683
+ "eval_weighted_f1": 0.5626940172444153,
684
+ "step": 1991
685
+ },
686
+ {
687
+ "epoch": 11.049723756906078,
688
+ "grad_norm": 0.014614409767091274,
689
+ "learning_rate": 9.950890116635974e-06,
690
+ "loss": 0.0849,
691
+ "step": 2000
692
+ },
693
+ {
694
+ "epoch": 11.187845303867404,
695
+ "grad_norm": 10.184608459472656,
696
+ "learning_rate": 9.79742173112339e-06,
697
+ "loss": 0.0456,
698
+ "step": 2025
699
+ },
700
+ {
701
+ "epoch": 11.32596685082873,
702
+ "grad_norm": 0.01762835681438446,
703
+ "learning_rate": 9.643953345610806e-06,
704
+ "loss": 0.0611,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 11.464088397790055,
709
+ "grad_norm": 20.771657943725586,
710
+ "learning_rate": 9.49048496009822e-06,
711
+ "loss": 0.1612,
712
+ "step": 2075
713
+ },
714
+ {
715
+ "epoch": 11.602209944751381,
716
+ "grad_norm": 0.03344396874308586,
717
+ "learning_rate": 9.337016574585636e-06,
718
+ "loss": 0.1203,
719
+ "step": 2100
720
+ },
721
+ {
722
+ "epoch": 11.740331491712707,
723
+ "grad_norm": 64.49190521240234,
724
+ "learning_rate": 9.183548189073052e-06,
725
+ "loss": 0.1822,
726
+ "step": 2125
727
+ },
728
+ {
729
+ "epoch": 11.878453038674033,
730
+ "grad_norm": 6.624955177307129,
731
+ "learning_rate": 9.030079803560468e-06,
732
+ "loss": 0.0596,
733
+ "step": 2150
734
+ },
735
+ {
736
+ "epoch": 12.0,
737
+ "eval_accuracy": 0.5498281786941581,
738
+ "eval_loss": 3.892807960510254,
739
+ "eval_macro_f1": 0.5523687908836651,
740
+ "eval_runtime": 1.1969,
741
+ "eval_samples_per_second": 486.24,
742
+ "eval_steps_per_second": 30.912,
743
+ "eval_weighted_f1": 0.5519241456772066,
744
+ "step": 2172
745
+ }
746
+ ],
747
+ "logging_steps": 25,
748
+ "max_steps": 3620,
749
+ "num_input_tokens_seen": 0,
750
+ "num_train_epochs": 20,
751
+ "save_steps": 500,
752
+ "stateful_callbacks": {
753
+ "EarlyStoppingCallback": {
754
+ "args": {
755
+ "early_stopping_patience": 3,
756
+ "early_stopping_threshold": 0.0
757
+ },
758
+ "attributes": {
759
+ "early_stopping_patience_counter": 3
760
+ }
761
+ },
762
+ "TrainerControl": {
763
+ "args": {
764
+ "should_epoch_stop": false,
765
+ "should_evaluate": false,
766
+ "should_log": false,
767
+ "should_save": true,
768
+ "should_training_stop": true
769
+ },
770
+ "attributes": {}
771
+ }
772
+ },
773
+ "total_flos": 0.0,
774
+ "train_batch_size": 8,
775
+ "trial_name": null,
776
+ "trial_params": null
777
+ }
ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464/trainer_state.json ADDED
@@ -0,0 +1,537 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1464,
3
+ "best_metric": 0.5684456981786657,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464",
5
+ "epoch": 8.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1464,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13717421124828533,
14
+ "grad_norm": 35.773765563964844,
15
+ "learning_rate": 1.3114754098360657e-06,
16
+ "loss": 4.8983,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27434842249657065,
21
+ "grad_norm": 18.36159896850586,
22
+ "learning_rate": 2.677595628415301e-06,
23
+ "loss": 4.5364,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.411522633744856,
28
+ "grad_norm": 20.749332427978516,
29
+ "learning_rate": 4.043715846994536e-06,
30
+ "loss": 4.4473,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5486968449931413,
35
+ "grad_norm": 34.27757263183594,
36
+ "learning_rate": 5.409836065573772e-06,
37
+ "loss": 4.4199,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6858710562414266,
42
+ "grad_norm": 23.1390323638916,
43
+ "learning_rate": 6.775956284153006e-06,
44
+ "loss": 4.4464,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.823045267489712,
49
+ "grad_norm": 17.340579986572266,
50
+ "learning_rate": 8.14207650273224e-06,
51
+ "loss": 4.3953,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9602194787379973,
56
+ "grad_norm": 23.77947235107422,
57
+ "learning_rate": 9.508196721311476e-06,
58
+ "loss": 4.3967,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.37603993344425957,
64
+ "eval_loss": 1.0887583494186401,
65
+ "eval_macro_f1": 0.3648876427036547,
66
+ "eval_runtime": 1.3004,
67
+ "eval_samples_per_second": 462.179,
68
+ "eval_steps_per_second": 29.223,
69
+ "eval_weighted_f1": 0.3668154403572677,
70
+ "step": 183
71
+ },
72
+ {
73
+ "epoch": 1.093278463648834,
74
+ "grad_norm": 22.9132080078125,
75
+ "learning_rate": 1.0874316939890712e-05,
76
+ "loss": 4.0832,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2304526748971194,
81
+ "grad_norm": 22.48322105407715,
82
+ "learning_rate": 1.2240437158469946e-05,
83
+ "loss": 4.1635,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3676268861454046,
88
+ "grad_norm": 18.87447738647461,
89
+ "learning_rate": 1.3606557377049181e-05,
90
+ "loss": 4.2072,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.50480109739369,
95
+ "grad_norm": 27.639368057250977,
96
+ "learning_rate": 1.4972677595628417e-05,
97
+ "loss": 4.1868,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6419753086419753,
102
+ "grad_norm": 20.872695922851562,
103
+ "learning_rate": 1.633879781420765e-05,
104
+ "loss": 4.2447,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7791495198902605,
109
+ "grad_norm": 22.119314193725586,
110
+ "learning_rate": 1.7704918032786887e-05,
111
+ "loss": 4.1773,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.916323731138546,
116
+ "grad_norm": 27.60935401916504,
117
+ "learning_rate": 1.9071038251366122e-05,
118
+ "loss": 4.2426,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.47920133111480867,
124
+ "eval_loss": 1.0199211835861206,
125
+ "eval_macro_f1": 0.4736183562360454,
126
+ "eval_runtime": 1.2121,
127
+ "eval_samples_per_second": 495.814,
128
+ "eval_steps_per_second": 31.349,
129
+ "eval_weighted_f1": 0.4723706298525388,
130
+ "step": 366
131
+ },
132
+ {
133
+ "epoch": 2.049382716049383,
134
+ "grad_norm": 35.469505310058594,
135
+ "learning_rate": 1.995142683667274e-05,
136
+ "loss": 3.8538,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.186556927297668,
141
+ "grad_norm": 30.557981491088867,
142
+ "learning_rate": 1.9799635701275045e-05,
143
+ "loss": 3.7359,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.3237311385459534,
148
+ "grad_norm": 50.85703659057617,
149
+ "learning_rate": 1.9647844565877354e-05,
150
+ "loss": 3.6142,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.460905349794239,
155
+ "grad_norm": 28.61817169189453,
156
+ "learning_rate": 1.9496053430479663e-05,
157
+ "loss": 3.6408,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.598079561042524,
162
+ "grad_norm": 23.514291763305664,
163
+ "learning_rate": 1.934426229508197e-05,
164
+ "loss": 3.5242,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7352537722908092,
169
+ "grad_norm": 31.297636032104492,
170
+ "learning_rate": 1.9192471159684274e-05,
171
+ "loss": 3.6192,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8724279835390947,
176
+ "grad_norm": 31.54410171508789,
177
+ "learning_rate": 1.9040680024286583e-05,
178
+ "loss": 3.6805,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.5324459234608985,
184
+ "eval_loss": 0.9632903337478638,
185
+ "eval_macro_f1": 0.5342286294532401,
186
+ "eval_runtime": 1.2891,
187
+ "eval_samples_per_second": 466.228,
188
+ "eval_steps_per_second": 29.479,
189
+ "eval_weighted_f1": 0.5341721692518161,
190
+ "step": 549
191
+ },
192
+ {
193
+ "epoch": 3.0054869684499312,
194
+ "grad_norm": 18.252853393554688,
195
+ "learning_rate": 1.888888888888889e-05,
196
+ "loss": 3.5446,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.1426611796982167,
201
+ "grad_norm": 32.142784118652344,
202
+ "learning_rate": 1.8737097753491198e-05,
203
+ "loss": 2.8502,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.279835390946502,
208
+ "grad_norm": 36.734375,
209
+ "learning_rate": 1.8585306618093504e-05,
210
+ "loss": 2.8287,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.4170096021947876,
215
+ "grad_norm": 38.70542526245117,
216
+ "learning_rate": 1.8433515482695813e-05,
217
+ "loss": 2.8043,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5541838134430725,
222
+ "grad_norm": 64.33454132080078,
223
+ "learning_rate": 1.828172434729812e-05,
224
+ "loss": 3.0055,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.691358024691358,
229
+ "grad_norm": 48.271095275878906,
230
+ "learning_rate": 1.8129933211900424e-05,
231
+ "loss": 2.6911,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.8285322359396434,
236
+ "grad_norm": 35.47870635986328,
237
+ "learning_rate": 1.7978142076502733e-05,
238
+ "loss": 2.5437,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.9657064471879284,
243
+ "grad_norm": 35.509010314941406,
244
+ "learning_rate": 1.7826350941105042e-05,
245
+ "loss": 2.756,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.5590682196339434,
251
+ "eval_loss": 1.0417397022247314,
252
+ "eval_macro_f1": 0.5605725768541334,
253
+ "eval_runtime": 1.207,
254
+ "eval_samples_per_second": 497.927,
255
+ "eval_steps_per_second": 31.483,
256
+ "eval_weighted_f1": 0.5613459696101107,
257
+ "step": 732
258
+ },
259
+ {
260
+ "epoch": 4.098765432098766,
261
+ "grad_norm": 33.41615676879883,
262
+ "learning_rate": 1.7674559805707348e-05,
263
+ "loss": 2.2763,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.23593964334705,
268
+ "grad_norm": 20.14259910583496,
269
+ "learning_rate": 1.7522768670309657e-05,
270
+ "loss": 2.1488,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.373113854595336,
275
+ "grad_norm": 50.4369010925293,
276
+ "learning_rate": 1.7370977534911962e-05,
277
+ "loss": 1.8718,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.510288065843621,
282
+ "grad_norm": 52.44806671142578,
283
+ "learning_rate": 1.7219186399514268e-05,
284
+ "loss": 1.7847,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.647462277091907,
289
+ "grad_norm": 75.71940612792969,
290
+ "learning_rate": 1.7067395264116577e-05,
291
+ "loss": 1.692,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.784636488340192,
296
+ "grad_norm": 32.079856872558594,
297
+ "learning_rate": 1.6915604128718883e-05,
298
+ "loss": 1.944,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.921810699588478,
303
+ "grad_norm": 27.68158531188965,
304
+ "learning_rate": 1.6763812993321192e-05,
305
+ "loss": 1.8544,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.562396006655574,
311
+ "eval_loss": 1.261223554611206,
312
+ "eval_macro_f1": 0.5554997453625848,
313
+ "eval_runtime": 1.2461,
314
+ "eval_samples_per_second": 482.317,
315
+ "eval_steps_per_second": 30.496,
316
+ "eval_weighted_f1": 0.5561173758772788,
317
+ "step": 915
318
+ },
319
+ {
320
+ "epoch": 5.054869684499314,
321
+ "grad_norm": 25.147499084472656,
322
+ "learning_rate": 1.66120218579235e-05,
323
+ "loss": 1.5044,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.1920438957476,
328
+ "grad_norm": 23.746217727661133,
329
+ "learning_rate": 1.6460230722525807e-05,
330
+ "loss": 1.1644,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.329218106995885,
335
+ "grad_norm": 51.73383331298828,
336
+ "learning_rate": 1.6308439587128112e-05,
337
+ "loss": 1.2062,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.4663923182441705,
342
+ "grad_norm": 110.5195083618164,
343
+ "learning_rate": 1.615664845173042e-05,
344
+ "loss": 1.1908,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.603566529492455,
349
+ "grad_norm": 65.53054809570312,
350
+ "learning_rate": 1.6004857316332727e-05,
351
+ "loss": 1.3523,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.7407407407407405,
356
+ "grad_norm": 68.0005874633789,
357
+ "learning_rate": 1.5853066180935036e-05,
358
+ "loss": 0.9882,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.877914951989026,
363
+ "grad_norm": 42.40581512451172,
364
+ "learning_rate": 1.570127504553734e-05,
365
+ "loss": 1.4759,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.5640599001663894,
371
+ "eval_loss": 1.4078166484832764,
372
+ "eval_macro_f1": 0.5641121258995321,
373
+ "eval_runtime": 1.3863,
374
+ "eval_samples_per_second": 433.514,
375
+ "eval_steps_per_second": 27.41,
376
+ "eval_weighted_f1": 0.5640349086891602,
377
+ "step": 1098
378
+ },
379
+ {
380
+ "epoch": 6.0109739368998625,
381
+ "grad_norm": 19.456941604614258,
382
+ "learning_rate": 1.5549483910139647e-05,
383
+ "loss": 1.0929,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.148148148148148,
388
+ "grad_norm": 32.80638122558594,
389
+ "learning_rate": 1.5397692774741956e-05,
390
+ "loss": 0.5395,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.285322359396433,
395
+ "grad_norm": 35.48299026489258,
396
+ "learning_rate": 1.5245901639344264e-05,
397
+ "loss": 0.6457,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.422496570644719,
402
+ "grad_norm": 32.192237854003906,
403
+ "learning_rate": 1.5094110503946571e-05,
404
+ "loss": 0.5244,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.559670781893004,
409
+ "grad_norm": 15.71074104309082,
410
+ "learning_rate": 1.4942319368548878e-05,
411
+ "loss": 0.8643,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.69684499314129,
416
+ "grad_norm": 82.7958984375,
417
+ "learning_rate": 1.4790528233151184e-05,
418
+ "loss": 0.8479,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.834019204389575,
423
+ "grad_norm": 36.42034149169922,
424
+ "learning_rate": 1.4638737097753493e-05,
425
+ "loss": 0.5515,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.97119341563786,
430
+ "grad_norm": 30.871334075927734,
431
+ "learning_rate": 1.44869459623558e-05,
432
+ "loss": 0.5979,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 7.0,
437
+ "eval_accuracy": 0.5607321131447587,
438
+ "eval_loss": 1.8727744817733765,
439
+ "eval_macro_f1": 0.5574076732690569,
440
+ "eval_runtime": 1.4157,
441
+ "eval_samples_per_second": 424.522,
442
+ "eval_steps_per_second": 26.842,
443
+ "eval_weighted_f1": 0.5586560321939252,
444
+ "step": 1281
445
+ },
446
+ {
447
+ "epoch": 7.104252400548697,
448
+ "grad_norm": 31.63418960571289,
449
+ "learning_rate": 1.4335154826958106e-05,
450
+ "loss": 0.4713,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.2414266117969825,
455
+ "grad_norm": 48.247108459472656,
456
+ "learning_rate": 1.4183363691560415e-05,
457
+ "loss": 0.3244,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.378600823045267,
462
+ "grad_norm": 33.2430305480957,
463
+ "learning_rate": 1.403157255616272e-05,
464
+ "loss": 0.3822,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.5157750342935525,
469
+ "grad_norm": 47.77069091796875,
470
+ "learning_rate": 1.3879781420765028e-05,
471
+ "loss": 0.3044,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.652949245541838,
476
+ "grad_norm": 7.046760559082031,
477
+ "learning_rate": 1.3727990285367337e-05,
478
+ "loss": 0.3778,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.790123456790123,
483
+ "grad_norm": 15.364222526550293,
484
+ "learning_rate": 1.3576199149969643e-05,
485
+ "loss": 0.3444,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.927297668038409,
490
+ "grad_norm": 72.87935638427734,
491
+ "learning_rate": 1.342440801457195e-05,
492
+ "loss": 0.3421,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 8.0,
497
+ "eval_accuracy": 0.5690515806988353,
498
+ "eval_loss": 2.3249778747558594,
499
+ "eval_macro_f1": 0.5684456981786657,
500
+ "eval_runtime": 1.286,
501
+ "eval_samples_per_second": 467.341,
502
+ "eval_steps_per_second": 29.549,
503
+ "eval_weighted_f1": 0.5683382411770878,
504
+ "step": 1464
505
+ }
506
+ ],
507
+ "logging_steps": 25,
508
+ "max_steps": 3660,
509
+ "num_input_tokens_seen": 0,
510
+ "num_train_epochs": 20,
511
+ "save_steps": 500,
512
+ "stateful_callbacks": {
513
+ "EarlyStoppingCallback": {
514
+ "args": {
515
+ "early_stopping_patience": 3,
516
+ "early_stopping_threshold": 0.0
517
+ },
518
+ "attributes": {
519
+ "early_stopping_patience_counter": 0
520
+ }
521
+ },
522
+ "TrainerControl": {
523
+ "args": {
524
+ "should_epoch_stop": false,
525
+ "should_evaluate": false,
526
+ "should_log": false,
527
+ "should_save": true,
528
+ "should_training_stop": false
529
+ },
530
+ "attributes": {}
531
+ }
532
+ },
533
+ "total_flos": 0.0,
534
+ "train_batch_size": 8,
535
+ "trial_name": null,
536
+ "trial_params": null
537
+ }
ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-2013/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-2013/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-2013/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-2013/trainer_state.json ADDED
@@ -0,0 +1,724 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1464,
3
+ "best_metric": 0.5684456981786657,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-03-s43-20260831T013703Z-1a169064/trainer/checkpoint-1464",
5
+ "epoch": 11.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2013,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13717421124828533,
14
+ "grad_norm": 35.773765563964844,
15
+ "learning_rate": 1.3114754098360657e-06,
16
+ "loss": 4.8983,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27434842249657065,
21
+ "grad_norm": 18.36159896850586,
22
+ "learning_rate": 2.677595628415301e-06,
23
+ "loss": 4.5364,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.411522633744856,
28
+ "grad_norm": 20.749332427978516,
29
+ "learning_rate": 4.043715846994536e-06,
30
+ "loss": 4.4473,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5486968449931413,
35
+ "grad_norm": 34.27757263183594,
36
+ "learning_rate": 5.409836065573772e-06,
37
+ "loss": 4.4199,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6858710562414266,
42
+ "grad_norm": 23.1390323638916,
43
+ "learning_rate": 6.775956284153006e-06,
44
+ "loss": 4.4464,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.823045267489712,
49
+ "grad_norm": 17.340579986572266,
50
+ "learning_rate": 8.14207650273224e-06,
51
+ "loss": 4.3953,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9602194787379973,
56
+ "grad_norm": 23.77947235107422,
57
+ "learning_rate": 9.508196721311476e-06,
58
+ "loss": 4.3967,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.37603993344425957,
64
+ "eval_loss": 1.0887583494186401,
65
+ "eval_macro_f1": 0.3648876427036547,
66
+ "eval_runtime": 1.3004,
67
+ "eval_samples_per_second": 462.179,
68
+ "eval_steps_per_second": 29.223,
69
+ "eval_weighted_f1": 0.3668154403572677,
70
+ "step": 183
71
+ },
72
+ {
73
+ "epoch": 1.093278463648834,
74
+ "grad_norm": 22.9132080078125,
75
+ "learning_rate": 1.0874316939890712e-05,
76
+ "loss": 4.0832,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2304526748971194,
81
+ "grad_norm": 22.48322105407715,
82
+ "learning_rate": 1.2240437158469946e-05,
83
+ "loss": 4.1635,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3676268861454046,
88
+ "grad_norm": 18.87447738647461,
89
+ "learning_rate": 1.3606557377049181e-05,
90
+ "loss": 4.2072,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.50480109739369,
95
+ "grad_norm": 27.639368057250977,
96
+ "learning_rate": 1.4972677595628417e-05,
97
+ "loss": 4.1868,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6419753086419753,
102
+ "grad_norm": 20.872695922851562,
103
+ "learning_rate": 1.633879781420765e-05,
104
+ "loss": 4.2447,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7791495198902605,
109
+ "grad_norm": 22.119314193725586,
110
+ "learning_rate": 1.7704918032786887e-05,
111
+ "loss": 4.1773,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.916323731138546,
116
+ "grad_norm": 27.60935401916504,
117
+ "learning_rate": 1.9071038251366122e-05,
118
+ "loss": 4.2426,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.47920133111480867,
124
+ "eval_loss": 1.0199211835861206,
125
+ "eval_macro_f1": 0.4736183562360454,
126
+ "eval_runtime": 1.2121,
127
+ "eval_samples_per_second": 495.814,
128
+ "eval_steps_per_second": 31.349,
129
+ "eval_weighted_f1": 0.4723706298525388,
130
+ "step": 366
131
+ },
132
+ {
133
+ "epoch": 2.049382716049383,
134
+ "grad_norm": 35.469505310058594,
135
+ "learning_rate": 1.995142683667274e-05,
136
+ "loss": 3.8538,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.186556927297668,
141
+ "grad_norm": 30.557981491088867,
142
+ "learning_rate": 1.9799635701275045e-05,
143
+ "loss": 3.7359,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.3237311385459534,
148
+ "grad_norm": 50.85703659057617,
149
+ "learning_rate": 1.9647844565877354e-05,
150
+ "loss": 3.6142,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.460905349794239,
155
+ "grad_norm": 28.61817169189453,
156
+ "learning_rate": 1.9496053430479663e-05,
157
+ "loss": 3.6408,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.598079561042524,
162
+ "grad_norm": 23.514291763305664,
163
+ "learning_rate": 1.934426229508197e-05,
164
+ "loss": 3.5242,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7352537722908092,
169
+ "grad_norm": 31.297636032104492,
170
+ "learning_rate": 1.9192471159684274e-05,
171
+ "loss": 3.6192,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8724279835390947,
176
+ "grad_norm": 31.54410171508789,
177
+ "learning_rate": 1.9040680024286583e-05,
178
+ "loss": 3.6805,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.5324459234608985,
184
+ "eval_loss": 0.9632903337478638,
185
+ "eval_macro_f1": 0.5342286294532401,
186
+ "eval_runtime": 1.2891,
187
+ "eval_samples_per_second": 466.228,
188
+ "eval_steps_per_second": 29.479,
189
+ "eval_weighted_f1": 0.5341721692518161,
190
+ "step": 549
191
+ },
192
+ {
193
+ "epoch": 3.0054869684499312,
194
+ "grad_norm": 18.252853393554688,
195
+ "learning_rate": 1.888888888888889e-05,
196
+ "loss": 3.5446,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.1426611796982167,
201
+ "grad_norm": 32.142784118652344,
202
+ "learning_rate": 1.8737097753491198e-05,
203
+ "loss": 2.8502,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.279835390946502,
208
+ "grad_norm": 36.734375,
209
+ "learning_rate": 1.8585306618093504e-05,
210
+ "loss": 2.8287,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.4170096021947876,
215
+ "grad_norm": 38.70542526245117,
216
+ "learning_rate": 1.8433515482695813e-05,
217
+ "loss": 2.8043,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5541838134430725,
222
+ "grad_norm": 64.33454132080078,
223
+ "learning_rate": 1.828172434729812e-05,
224
+ "loss": 3.0055,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.691358024691358,
229
+ "grad_norm": 48.271095275878906,
230
+ "learning_rate": 1.8129933211900424e-05,
231
+ "loss": 2.6911,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.8285322359396434,
236
+ "grad_norm": 35.47870635986328,
237
+ "learning_rate": 1.7978142076502733e-05,
238
+ "loss": 2.5437,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.9657064471879284,
243
+ "grad_norm": 35.509010314941406,
244
+ "learning_rate": 1.7826350941105042e-05,
245
+ "loss": 2.756,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.5590682196339434,
251
+ "eval_loss": 1.0417397022247314,
252
+ "eval_macro_f1": 0.5605725768541334,
253
+ "eval_runtime": 1.207,
254
+ "eval_samples_per_second": 497.927,
255
+ "eval_steps_per_second": 31.483,
256
+ "eval_weighted_f1": 0.5613459696101107,
257
+ "step": 732
258
+ },
259
+ {
260
+ "epoch": 4.098765432098766,
261
+ "grad_norm": 33.41615676879883,
262
+ "learning_rate": 1.7674559805707348e-05,
263
+ "loss": 2.2763,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.23593964334705,
268
+ "grad_norm": 20.14259910583496,
269
+ "learning_rate": 1.7522768670309657e-05,
270
+ "loss": 2.1488,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.373113854595336,
275
+ "grad_norm": 50.4369010925293,
276
+ "learning_rate": 1.7370977534911962e-05,
277
+ "loss": 1.8718,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.510288065843621,
282
+ "grad_norm": 52.44806671142578,
283
+ "learning_rate": 1.7219186399514268e-05,
284
+ "loss": 1.7847,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.647462277091907,
289
+ "grad_norm": 75.71940612792969,
290
+ "learning_rate": 1.7067395264116577e-05,
291
+ "loss": 1.692,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.784636488340192,
296
+ "grad_norm": 32.079856872558594,
297
+ "learning_rate": 1.6915604128718883e-05,
298
+ "loss": 1.944,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.921810699588478,
303
+ "grad_norm": 27.68158531188965,
304
+ "learning_rate": 1.6763812993321192e-05,
305
+ "loss": 1.8544,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.562396006655574,
311
+ "eval_loss": 1.261223554611206,
312
+ "eval_macro_f1": 0.5554997453625848,
313
+ "eval_runtime": 1.2461,
314
+ "eval_samples_per_second": 482.317,
315
+ "eval_steps_per_second": 30.496,
316
+ "eval_weighted_f1": 0.5561173758772788,
317
+ "step": 915
318
+ },
319
+ {
320
+ "epoch": 5.054869684499314,
321
+ "grad_norm": 25.147499084472656,
322
+ "learning_rate": 1.66120218579235e-05,
323
+ "loss": 1.5044,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.1920438957476,
328
+ "grad_norm": 23.746217727661133,
329
+ "learning_rate": 1.6460230722525807e-05,
330
+ "loss": 1.1644,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.329218106995885,
335
+ "grad_norm": 51.73383331298828,
336
+ "learning_rate": 1.6308439587128112e-05,
337
+ "loss": 1.2062,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.4663923182441705,
342
+ "grad_norm": 110.5195083618164,
343
+ "learning_rate": 1.615664845173042e-05,
344
+ "loss": 1.1908,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.603566529492455,
349
+ "grad_norm": 65.53054809570312,
350
+ "learning_rate": 1.6004857316332727e-05,
351
+ "loss": 1.3523,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.7407407407407405,
356
+ "grad_norm": 68.0005874633789,
357
+ "learning_rate": 1.5853066180935036e-05,
358
+ "loss": 0.9882,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.877914951989026,
363
+ "grad_norm": 42.40581512451172,
364
+ "learning_rate": 1.570127504553734e-05,
365
+ "loss": 1.4759,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.5640599001663894,
371
+ "eval_loss": 1.4078166484832764,
372
+ "eval_macro_f1": 0.5641121258995321,
373
+ "eval_runtime": 1.3863,
374
+ "eval_samples_per_second": 433.514,
375
+ "eval_steps_per_second": 27.41,
376
+ "eval_weighted_f1": 0.5640349086891602,
377
+ "step": 1098
378
+ },
379
+ {
380
+ "epoch": 6.0109739368998625,
381
+ "grad_norm": 19.456941604614258,
382
+ "learning_rate": 1.5549483910139647e-05,
383
+ "loss": 1.0929,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.148148148148148,
388
+ "grad_norm": 32.80638122558594,
389
+ "learning_rate": 1.5397692774741956e-05,
390
+ "loss": 0.5395,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.285322359396433,
395
+ "grad_norm": 35.48299026489258,
396
+ "learning_rate": 1.5245901639344264e-05,
397
+ "loss": 0.6457,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.422496570644719,
402
+ "grad_norm": 32.192237854003906,
403
+ "learning_rate": 1.5094110503946571e-05,
404
+ "loss": 0.5244,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.559670781893004,
409
+ "grad_norm": 15.71074104309082,
410
+ "learning_rate": 1.4942319368548878e-05,
411
+ "loss": 0.8643,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.69684499314129,
416
+ "grad_norm": 82.7958984375,
417
+ "learning_rate": 1.4790528233151184e-05,
418
+ "loss": 0.8479,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.834019204389575,
423
+ "grad_norm": 36.42034149169922,
424
+ "learning_rate": 1.4638737097753493e-05,
425
+ "loss": 0.5515,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.97119341563786,
430
+ "grad_norm": 30.871334075927734,
431
+ "learning_rate": 1.44869459623558e-05,
432
+ "loss": 0.5979,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 7.0,
437
+ "eval_accuracy": 0.5607321131447587,
438
+ "eval_loss": 1.8727744817733765,
439
+ "eval_macro_f1": 0.5574076732690569,
440
+ "eval_runtime": 1.4157,
441
+ "eval_samples_per_second": 424.522,
442
+ "eval_steps_per_second": 26.842,
443
+ "eval_weighted_f1": 0.5586560321939252,
444
+ "step": 1281
445
+ },
446
+ {
447
+ "epoch": 7.104252400548697,
448
+ "grad_norm": 31.63418960571289,
449
+ "learning_rate": 1.4335154826958106e-05,
450
+ "loss": 0.4713,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.2414266117969825,
455
+ "grad_norm": 48.247108459472656,
456
+ "learning_rate": 1.4183363691560415e-05,
457
+ "loss": 0.3244,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.378600823045267,
462
+ "grad_norm": 33.2430305480957,
463
+ "learning_rate": 1.403157255616272e-05,
464
+ "loss": 0.3822,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.5157750342935525,
469
+ "grad_norm": 47.77069091796875,
470
+ "learning_rate": 1.3879781420765028e-05,
471
+ "loss": 0.3044,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.652949245541838,
476
+ "grad_norm": 7.046760559082031,
477
+ "learning_rate": 1.3727990285367337e-05,
478
+ "loss": 0.3778,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.790123456790123,
483
+ "grad_norm": 15.364222526550293,
484
+ "learning_rate": 1.3576199149969643e-05,
485
+ "loss": 0.3444,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.927297668038409,
490
+ "grad_norm": 72.87935638427734,
491
+ "learning_rate": 1.342440801457195e-05,
492
+ "loss": 0.3421,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 8.0,
497
+ "eval_accuracy": 0.5690515806988353,
498
+ "eval_loss": 2.3249778747558594,
499
+ "eval_macro_f1": 0.5684456981786657,
500
+ "eval_runtime": 1.286,
501
+ "eval_samples_per_second": 467.341,
502
+ "eval_steps_per_second": 29.549,
503
+ "eval_weighted_f1": 0.5683382411770878,
504
+ "step": 1464
505
+ },
506
+ {
507
+ "epoch": 8.060356652949245,
508
+ "grad_norm": 6.102505207061768,
509
+ "learning_rate": 1.3272616879174259e-05,
510
+ "loss": 0.2532,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.197530864197532,
515
+ "grad_norm": 7.357082366943359,
516
+ "learning_rate": 1.3120825743776565e-05,
517
+ "loss": 0.1877,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.334705075445816,
522
+ "grad_norm": 59.2963981628418,
523
+ "learning_rate": 1.2969034608378872e-05,
524
+ "loss": 0.2746,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.4718792866941,
529
+ "grad_norm": 84.53719329833984,
530
+ "learning_rate": 1.2817243472981178e-05,
531
+ "loss": 0.3552,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.609053497942387,
536
+ "grad_norm": 1.5999681949615479,
537
+ "learning_rate": 1.2665452337583487e-05,
538
+ "loss": 0.2961,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.746227709190672,
543
+ "grad_norm": 67.35011291503906,
544
+ "learning_rate": 1.2513661202185794e-05,
545
+ "loss": 0.3804,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.883401920438958,
550
+ "grad_norm": 38.6851692199707,
551
+ "learning_rate": 1.23618700667881e-05,
552
+ "loss": 0.2625,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 9.0,
557
+ "eval_accuracy": 0.5640599001663894,
558
+ "eval_loss": 2.5666680335998535,
559
+ "eval_macro_f1": 0.5644424766673183,
560
+ "eval_runtime": 1.3159,
561
+ "eval_samples_per_second": 456.712,
562
+ "eval_steps_per_second": 28.877,
563
+ "eval_weighted_f1": 0.5646726620762842,
564
+ "step": 1647
565
+ },
566
+ {
567
+ "epoch": 9.016460905349794,
568
+ "grad_norm": 10.008074760437012,
569
+ "learning_rate": 1.2210078931390409e-05,
570
+ "loss": 0.2872,
571
+ "step": 1650
572
+ },
573
+ {
574
+ "epoch": 9.15363511659808,
575
+ "grad_norm": 34.07567596435547,
576
+ "learning_rate": 1.2058287795992716e-05,
577
+ "loss": 0.1522,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.290809327846365,
582
+ "grad_norm": 43.5909309387207,
583
+ "learning_rate": 1.1906496660595022e-05,
584
+ "loss": 0.1825,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.427983539094651,
589
+ "grad_norm": 108.45124816894531,
590
+ "learning_rate": 1.1754705525197329e-05,
591
+ "loss": 0.2192,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.565157750342935,
596
+ "grad_norm": 0.9083254933357239,
597
+ "learning_rate": 1.1602914389799638e-05,
598
+ "loss": 0.3218,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 9.70233196159122,
603
+ "grad_norm": 2.381828784942627,
604
+ "learning_rate": 1.1451123254401944e-05,
605
+ "loss": 0.2772,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 9.839506172839506,
610
+ "grad_norm": 55.93046951293945,
611
+ "learning_rate": 1.1299332119004251e-05,
612
+ "loss": 0.2073,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 9.976680384087791,
617
+ "grad_norm": 1.002413272857666,
618
+ "learning_rate": 1.1147540983606557e-05,
619
+ "loss": 0.1354,
620
+ "step": 1825
621
+ },
622
+ {
623
+ "epoch": 10.0,
624
+ "eval_accuracy": 0.5657237936772047,
625
+ "eval_loss": 3.0506722927093506,
626
+ "eval_macro_f1": 0.5681382504100586,
627
+ "eval_runtime": 1.2479,
628
+ "eval_samples_per_second": 481.598,
629
+ "eval_steps_per_second": 30.45,
630
+ "eval_weighted_f1": 0.5674863299664008,
631
+ "step": 1830
632
+ },
633
+ {
634
+ "epoch": 10.109739368998628,
635
+ "grad_norm": 13.557258605957031,
636
+ "learning_rate": 1.0995749848208866e-05,
637
+ "loss": 0.3142,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.246913580246913,
642
+ "grad_norm": 8.840352058410645,
643
+ "learning_rate": 1.0843958712811173e-05,
644
+ "loss": 0.1526,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.3840877914952,
649
+ "grad_norm": 0.4212711453437805,
650
+ "learning_rate": 1.0692167577413479e-05,
651
+ "loss": 0.0293,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.521262002743484,
656
+ "grad_norm": 2.6239118576049805,
657
+ "learning_rate": 1.0540376442015788e-05,
658
+ "loss": 0.0775,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 10.65843621399177,
663
+ "grad_norm": 57.85154724121094,
664
+ "learning_rate": 1.0388585306618095e-05,
665
+ "loss": 0.1152,
666
+ "step": 1950
667
+ },
668
+ {
669
+ "epoch": 10.795610425240055,
670
+ "grad_norm": 90.2874526977539,
671
+ "learning_rate": 1.02367941712204e-05,
672
+ "loss": 0.1906,
673
+ "step": 1975
674
+ },
675
+ {
676
+ "epoch": 10.932784636488341,
677
+ "grad_norm": 0.3817863464355469,
678
+ "learning_rate": 1.008500303582271e-05,
679
+ "loss": 0.1577,
680
+ "step": 2000
681
+ },
682
+ {
683
+ "epoch": 11.0,
684
+ "eval_accuracy": 0.56738768718802,
685
+ "eval_loss": 3.4996042251586914,
686
+ "eval_macro_f1": 0.5663161611796399,
687
+ "eval_runtime": 1.3561,
688
+ "eval_samples_per_second": 443.175,
689
+ "eval_steps_per_second": 28.021,
690
+ "eval_weighted_f1": 0.5647008847277125,
691
+ "step": 2013
692
+ }
693
+ ],
694
+ "logging_steps": 25,
695
+ "max_steps": 3660,
696
+ "num_input_tokens_seen": 0,
697
+ "num_train_epochs": 20,
698
+ "save_steps": 500,
699
+ "stateful_callbacks": {
700
+ "EarlyStoppingCallback": {
701
+ "args": {
702
+ "early_stopping_patience": 3,
703
+ "early_stopping_threshold": 0.0
704
+ },
705
+ "attributes": {
706
+ "early_stopping_patience_counter": 3
707
+ }
708
+ },
709
+ "TrainerControl": {
710
+ "args": {
711
+ "should_epoch_stop": false,
712
+ "should_evaluate": false,
713
+ "should_log": false,
714
+ "should_save": true,
715
+ "should_training_stop": true
716
+ },
717
+ "attributes": {}
718
+ }
719
+ },
720
+ "total_flos": 0.0,
721
+ "train_batch_size": 8,
722
+ "trial_name": null,
723
+ "trial_params": null
724
+ }
ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288/trainer_state.json ADDED
@@ -0,0 +1,823 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2288,
3
+ "best_metric": 0.5859148961439932,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288",
5
+ "epoch": 13.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2288,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.14245014245014245,
14
+ "grad_norm": 27.69466209411621,
15
+ "learning_rate": 1.3636363636363636e-06,
16
+ "loss": 4.8352,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2849002849002849,
21
+ "grad_norm": 19.542036056518555,
22
+ "learning_rate": 2.784090909090909e-06,
23
+ "loss": 4.4519,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.42735042735042733,
28
+ "grad_norm": 26.295495986938477,
29
+ "learning_rate": 4.204545454545455e-06,
30
+ "loss": 4.442,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5698005698005698,
35
+ "grad_norm": 26.776668548583984,
36
+ "learning_rate": 5.625e-06,
37
+ "loss": 4.5038,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.7122507122507122,
42
+ "grad_norm": 23.936386108398438,
43
+ "learning_rate": 7.045454545454546e-06,
44
+ "loss": 4.362,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8547008547008547,
49
+ "grad_norm": 26.568832397460938,
50
+ "learning_rate": 8.465909090909091e-06,
51
+ "loss": 4.4832,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9971509971509972,
56
+ "grad_norm": 18.616518020629883,
57
+ "learning_rate": 9.886363636363637e-06,
58
+ "loss": 4.3872,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.43686502177068215,
64
+ "eval_loss": 1.0517677068710327,
65
+ "eval_macro_f1": 0.3942313052416193,
66
+ "eval_runtime": 1.6887,
67
+ "eval_samples_per_second": 408.006,
68
+ "eval_steps_per_second": 26.056,
69
+ "eval_weighted_f1": 0.39853194871240416,
70
+ "step": 176
71
+ },
72
+ {
73
+ "epoch": 1.1367521367521367,
74
+ "grad_norm": 25.454792022705078,
75
+ "learning_rate": 1.1306818181818183e-05,
76
+ "loss": 4.1347,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2792022792022792,
81
+ "grad_norm": 24.505388259887695,
82
+ "learning_rate": 1.2727272727272728e-05,
83
+ "loss": 4.2989,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.4216524216524218,
88
+ "grad_norm": 22.538040161132812,
89
+ "learning_rate": 1.4147727272727274e-05,
90
+ "loss": 4.235,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.564102564102564,
95
+ "grad_norm": 18.85667610168457,
96
+ "learning_rate": 1.5568181818181822e-05,
97
+ "loss": 4.2795,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.7065527065527064,
102
+ "grad_norm": 24.444332122802734,
103
+ "learning_rate": 1.6988636363636368e-05,
104
+ "loss": 4.2162,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.849002849002849,
109
+ "grad_norm": 19.221044540405273,
110
+ "learning_rate": 1.840909090909091e-05,
111
+ "loss": 4.186,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9914529914529915,
116
+ "grad_norm": 27.694293975830078,
117
+ "learning_rate": 1.9829545454545456e-05,
118
+ "loss": 4.1084,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.5094339622641509,
124
+ "eval_loss": 0.986926257610321,
125
+ "eval_macro_f1": 0.4962483502863326,
126
+ "eval_runtime": 1.4691,
127
+ "eval_samples_per_second": 468.987,
128
+ "eval_steps_per_second": 29.95,
129
+ "eval_weighted_f1": 0.49832918236349383,
130
+ "step": 352
131
+ },
132
+ {
133
+ "epoch": 2.131054131054131,
134
+ "grad_norm": 27.87932777404785,
135
+ "learning_rate": 1.9861111111111114e-05,
136
+ "loss": 3.7175,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.2735042735042734,
141
+ "grad_norm": 26.27695083618164,
142
+ "learning_rate": 1.9703282828282832e-05,
143
+ "loss": 3.5301,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.415954415954416,
148
+ "grad_norm": 40.19943618774414,
149
+ "learning_rate": 1.9545454545454546e-05,
150
+ "loss": 3.6274,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.5584045584045585,
155
+ "grad_norm": 43.94453811645508,
156
+ "learning_rate": 1.9387626262626263e-05,
157
+ "loss": 3.8303,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.700854700854701,
162
+ "grad_norm": 47.385799407958984,
163
+ "learning_rate": 1.922979797979798e-05,
164
+ "loss": 3.5218,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.8433048433048436,
169
+ "grad_norm": 29.534732818603516,
170
+ "learning_rate": 1.90719696969697e-05,
171
+ "loss": 3.6893,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.9857549857549857,
176
+ "grad_norm": 37.31429672241211,
177
+ "learning_rate": 1.8914141414141416e-05,
178
+ "loss": 3.5932,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.5544267053701016,
184
+ "eval_loss": 0.9797604084014893,
185
+ "eval_macro_f1": 0.5478060457189247,
186
+ "eval_runtime": 1.3112,
187
+ "eval_samples_per_second": 525.49,
188
+ "eval_steps_per_second": 33.558,
189
+ "eval_weighted_f1": 0.5493016960716426,
190
+ "step": 528
191
+ },
192
+ {
193
+ "epoch": 3.1253561253561255,
194
+ "grad_norm": 67.17666625976562,
195
+ "learning_rate": 1.8756313131313133e-05,
196
+ "loss": 2.8348,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.267806267806268,
201
+ "grad_norm": 55.27144241333008,
202
+ "learning_rate": 1.859848484848485e-05,
203
+ "loss": 2.7264,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.41025641025641,
208
+ "grad_norm": 57.96361541748047,
209
+ "learning_rate": 1.844065656565657e-05,
210
+ "loss": 2.9081,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.5527065527065527,
215
+ "grad_norm": 27.539400100708008,
216
+ "learning_rate": 1.8282828282828286e-05,
217
+ "loss": 2.9182,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.695156695156695,
222
+ "grad_norm": 45.49375534057617,
223
+ "learning_rate": 1.8125e-05,
224
+ "loss": 2.7568,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.8376068376068377,
229
+ "grad_norm": 23.619964599609375,
230
+ "learning_rate": 1.7967171717171718e-05,
231
+ "loss": 2.5585,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.98005698005698,
236
+ "grad_norm": 33.051025390625,
237
+ "learning_rate": 1.7809343434343435e-05,
238
+ "loss": 2.848,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 4.0,
243
+ "eval_accuracy": 0.5515239477503628,
244
+ "eval_loss": 1.0126099586486816,
245
+ "eval_macro_f1": 0.5522044976182564,
246
+ "eval_runtime": 1.3267,
247
+ "eval_samples_per_second": 519.346,
248
+ "eval_steps_per_second": 33.166,
249
+ "eval_weighted_f1": 0.5524713353890915,
250
+ "step": 704
251
+ },
252
+ {
253
+ "epoch": 4.119658119658119,
254
+ "grad_norm": 41.599609375,
255
+ "learning_rate": 1.7651515151515153e-05,
256
+ "loss": 2.1393,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.262108262108262,
261
+ "grad_norm": 33.376155853271484,
262
+ "learning_rate": 1.749368686868687e-05,
263
+ "loss": 2.003,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.404558404558404,
268
+ "grad_norm": 62.67402648925781,
269
+ "learning_rate": 1.7335858585858588e-05,
270
+ "loss": 1.6959,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.547008547008547,
275
+ "grad_norm": 55.37060546875,
276
+ "learning_rate": 1.7178030303030305e-05,
277
+ "loss": 1.7717,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.689458689458689,
282
+ "grad_norm": 55.278926849365234,
283
+ "learning_rate": 1.7020202020202023e-05,
284
+ "loss": 1.9422,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.831908831908832,
289
+ "grad_norm": 35.771244049072266,
290
+ "learning_rate": 1.6862373737373737e-05,
291
+ "loss": 1.9357,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.9743589743589745,
296
+ "grad_norm": 43.035099029541016,
297
+ "learning_rate": 1.6704545454545454e-05,
298
+ "loss": 1.7887,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 5.0,
303
+ "eval_accuracy": 0.5529753265602322,
304
+ "eval_loss": 1.307969093322754,
305
+ "eval_macro_f1": 0.5500873943650679,
306
+ "eval_runtime": 1.4687,
307
+ "eval_samples_per_second": 469.117,
308
+ "eval_steps_per_second": 29.958,
309
+ "eval_weighted_f1": 0.5511325286130112,
310
+ "step": 880
311
+ },
312
+ {
313
+ "epoch": 5.113960113960114,
314
+ "grad_norm": 47.52789306640625,
315
+ "learning_rate": 1.6546717171717172e-05,
316
+ "loss": 1.3756,
317
+ "step": 900
318
+ },
319
+ {
320
+ "epoch": 5.256410256410256,
321
+ "grad_norm": 67.07496643066406,
322
+ "learning_rate": 1.638888888888889e-05,
323
+ "loss": 0.9039,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.398860398860399,
328
+ "grad_norm": 31.854036331176758,
329
+ "learning_rate": 1.6231060606060607e-05,
330
+ "loss": 1.1748,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.5413105413105415,
335
+ "grad_norm": 106.62139892578125,
336
+ "learning_rate": 1.6073232323232324e-05,
337
+ "loss": 1.195,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.683760683760684,
342
+ "grad_norm": 73.60779571533203,
343
+ "learning_rate": 1.5915404040404042e-05,
344
+ "loss": 1.3651,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.8262108262108265,
349
+ "grad_norm": 51.60421371459961,
350
+ "learning_rate": 1.575757575757576e-05,
351
+ "loss": 1.2479,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.968660968660968,
356
+ "grad_norm": 65.28878784179688,
357
+ "learning_rate": 1.5599747474747477e-05,
358
+ "loss": 1.0724,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 6.0,
363
+ "eval_accuracy": 0.5747460087082729,
364
+ "eval_loss": 1.5242938995361328,
365
+ "eval_macro_f1": 0.5721091381075082,
366
+ "eval_runtime": 1.3765,
367
+ "eval_samples_per_second": 500.541,
368
+ "eval_steps_per_second": 31.965,
369
+ "eval_weighted_f1": 0.5727603574205964,
370
+ "step": 1056
371
+ },
372
+ {
373
+ "epoch": 6.1082621082621085,
374
+ "grad_norm": 26.339204788208008,
375
+ "learning_rate": 1.544191919191919e-05,
376
+ "loss": 0.7805,
377
+ "step": 1075
378
+ },
379
+ {
380
+ "epoch": 6.250712250712251,
381
+ "grad_norm": 53.746768951416016,
382
+ "learning_rate": 1.528409090909091e-05,
383
+ "loss": 0.5242,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.3931623931623935,
388
+ "grad_norm": 34.56966781616211,
389
+ "learning_rate": 1.5126262626262626e-05,
390
+ "loss": 0.7292,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.535612535612536,
395
+ "grad_norm": 74.35704040527344,
396
+ "learning_rate": 1.4968434343434344e-05,
397
+ "loss": 0.7179,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.678062678062678,
402
+ "grad_norm": 28.260852813720703,
403
+ "learning_rate": 1.4810606060606061e-05,
404
+ "loss": 0.7493,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.82051282051282,
409
+ "grad_norm": 61.798500061035156,
410
+ "learning_rate": 1.4652777777777779e-05,
411
+ "loss": 0.7055,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.962962962962963,
416
+ "grad_norm": 39.78801727294922,
417
+ "learning_rate": 1.4494949494949494e-05,
418
+ "loss": 0.75,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 7.0,
423
+ "eval_accuracy": 0.5674891146589259,
424
+ "eval_loss": 1.7215156555175781,
425
+ "eval_macro_f1": 0.5612321340302727,
426
+ "eval_runtime": 1.3148,
427
+ "eval_samples_per_second": 524.042,
428
+ "eval_steps_per_second": 33.466,
429
+ "eval_weighted_f1": 0.5611592964466283,
430
+ "step": 1232
431
+ },
432
+ {
433
+ "epoch": 7.102564102564102,
434
+ "grad_norm": 23.60300064086914,
435
+ "learning_rate": 1.4337121212121212e-05,
436
+ "loss": 0.7275,
437
+ "step": 1250
438
+ },
439
+ {
440
+ "epoch": 7.245014245014245,
441
+ "grad_norm": 18.518918991088867,
442
+ "learning_rate": 1.4179292929292931e-05,
443
+ "loss": 0.3774,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.387464387464387,
448
+ "grad_norm": 74.9653549194336,
449
+ "learning_rate": 1.4021464646464649e-05,
450
+ "loss": 0.4915,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.52991452991453,
455
+ "grad_norm": 35.73414611816406,
456
+ "learning_rate": 1.3863636363636366e-05,
457
+ "loss": 0.4356,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.672364672364672,
462
+ "grad_norm": 3.858736753463745,
463
+ "learning_rate": 1.3705808080808084e-05,
464
+ "loss": 0.315,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.814814814814815,
469
+ "grad_norm": 50.60848617553711,
470
+ "learning_rate": 1.35479797979798e-05,
471
+ "loss": 0.4276,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.957264957264957,
476
+ "grad_norm": 70.68370819091797,
477
+ "learning_rate": 1.3390151515151517e-05,
478
+ "loss": 0.3087,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 8.0,
483
+ "eval_accuracy": 0.5834542815674891,
484
+ "eval_loss": 2.2690086364746094,
485
+ "eval_macro_f1": 0.575839318316549,
486
+ "eval_runtime": 1.5932,
487
+ "eval_samples_per_second": 432.466,
488
+ "eval_steps_per_second": 27.618,
489
+ "eval_weighted_f1": 0.5768206681992663,
490
+ "step": 1408
491
+ },
492
+ {
493
+ "epoch": 8.096866096866098,
494
+ "grad_norm": 31.564485549926758,
495
+ "learning_rate": 1.3232323232323234e-05,
496
+ "loss": 0.2772,
497
+ "step": 1425
498
+ },
499
+ {
500
+ "epoch": 8.239316239316238,
501
+ "grad_norm": 61.35936737060547,
502
+ "learning_rate": 1.3074494949494952e-05,
503
+ "loss": 0.2677,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.381766381766381,
508
+ "grad_norm": 24.228017807006836,
509
+ "learning_rate": 1.2916666666666668e-05,
510
+ "loss": 0.2506,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.524216524216524,
515
+ "grad_norm": 53.46422576904297,
516
+ "learning_rate": 1.2758838383838385e-05,
517
+ "loss": 0.3032,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.666666666666666,
522
+ "grad_norm": 61.28293228149414,
523
+ "learning_rate": 1.2601010101010103e-05,
524
+ "loss": 0.3046,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.809116809116809,
529
+ "grad_norm": 2.150891065597534,
530
+ "learning_rate": 1.244318181818182e-05,
531
+ "loss": 0.4309,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.951566951566951,
536
+ "grad_norm": 115.88075256347656,
537
+ "learning_rate": 1.2285353535353536e-05,
538
+ "loss": 0.2445,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 9.0,
543
+ "eval_accuracy": 0.5573294629898403,
544
+ "eval_loss": 2.739509344100952,
545
+ "eval_macro_f1": 0.5511234758681754,
546
+ "eval_runtime": 1.4544,
547
+ "eval_samples_per_second": 473.741,
548
+ "eval_steps_per_second": 30.253,
549
+ "eval_weighted_f1": 0.5523768897983876,
550
+ "step": 1584
551
+ },
552
+ {
553
+ "epoch": 9.091168091168091,
554
+ "grad_norm": 87.18209075927734,
555
+ "learning_rate": 1.2127525252525254e-05,
556
+ "loss": 0.1399,
557
+ "step": 1600
558
+ },
559
+ {
560
+ "epoch": 9.233618233618234,
561
+ "grad_norm": 28.503408432006836,
562
+ "learning_rate": 1.1969696969696971e-05,
563
+ "loss": 0.1585,
564
+ "step": 1625
565
+ },
566
+ {
567
+ "epoch": 9.376068376068377,
568
+ "grad_norm": 49.824867248535156,
569
+ "learning_rate": 1.1811868686868689e-05,
570
+ "loss": 0.3722,
571
+ "step": 1650
572
+ },
573
+ {
574
+ "epoch": 9.518518518518519,
575
+ "grad_norm": 1.6686387062072754,
576
+ "learning_rate": 1.1654040404040406e-05,
577
+ "loss": 0.2527,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.660968660968662,
582
+ "grad_norm": 0.9758527278900146,
583
+ "learning_rate": 1.1496212121212122e-05,
584
+ "loss": 0.2327,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.803418803418804,
589
+ "grad_norm": 2.719261646270752,
590
+ "learning_rate": 1.133838383838384e-05,
591
+ "loss": 0.1828,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.945868945868947,
596
+ "grad_norm": 7.621354579925537,
597
+ "learning_rate": 1.1180555555555557e-05,
598
+ "loss": 0.277,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 10.0,
603
+ "eval_accuracy": 0.579100145137881,
604
+ "eval_loss": 2.908055305480957,
605
+ "eval_macro_f1": 0.57907237466202,
606
+ "eval_runtime": 1.4314,
607
+ "eval_samples_per_second": 481.333,
608
+ "eval_steps_per_second": 30.738,
609
+ "eval_weighted_f1": 0.5792181523580426,
610
+ "step": 1760
611
+ },
612
+ {
613
+ "epoch": 10.085470085470085,
614
+ "grad_norm": 13.50964641571045,
615
+ "learning_rate": 1.1022727272727275e-05,
616
+ "loss": 0.0596,
617
+ "step": 1775
618
+ },
619
+ {
620
+ "epoch": 10.227920227920228,
621
+ "grad_norm": 1.35202157497406,
622
+ "learning_rate": 1.086489898989899e-05,
623
+ "loss": 0.0965,
624
+ "step": 1800
625
+ },
626
+ {
627
+ "epoch": 10.37037037037037,
628
+ "grad_norm": 47.22248458862305,
629
+ "learning_rate": 1.0707070707070708e-05,
630
+ "loss": 0.1526,
631
+ "step": 1825
632
+ },
633
+ {
634
+ "epoch": 10.512820512820513,
635
+ "grad_norm": 1.046852469444275,
636
+ "learning_rate": 1.0549242424242425e-05,
637
+ "loss": 0.0543,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.655270655270655,
642
+ "grad_norm": 0.43407419323921204,
643
+ "learning_rate": 1.0391414141414143e-05,
644
+ "loss": 0.1675,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.797720797720798,
649
+ "grad_norm": 56.43595504760742,
650
+ "learning_rate": 1.0233585858585859e-05,
651
+ "loss": 0.2615,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.94017094017094,
656
+ "grad_norm": 94.71170043945312,
657
+ "learning_rate": 1.0075757575757576e-05,
658
+ "loss": 0.2073,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 11.0,
663
+ "eval_accuracy": 0.5544267053701016,
664
+ "eval_loss": 3.494981050491333,
665
+ "eval_macro_f1": 0.5537080472229138,
666
+ "eval_runtime": 1.4298,
667
+ "eval_samples_per_second": 481.892,
668
+ "eval_steps_per_second": 30.774,
669
+ "eval_weighted_f1": 0.5546618094147888,
670
+ "step": 1936
671
+ },
672
+ {
673
+ "epoch": 11.079772079772079,
674
+ "grad_norm": 113.4718017578125,
675
+ "learning_rate": 9.917929292929294e-06,
676
+ "loss": 0.19,
677
+ "step": 1950
678
+ },
679
+ {
680
+ "epoch": 11.222222222222221,
681
+ "grad_norm": 114.95403289794922,
682
+ "learning_rate": 9.760101010101011e-06,
683
+ "loss": 0.075,
684
+ "step": 1975
685
+ },
686
+ {
687
+ "epoch": 11.364672364672364,
688
+ "grad_norm": 0.019300974905490875,
689
+ "learning_rate": 9.602272727272727e-06,
690
+ "loss": 0.2343,
691
+ "step": 2000
692
+ },
693
+ {
694
+ "epoch": 11.507122507122507,
695
+ "grad_norm": 121.42213439941406,
696
+ "learning_rate": 9.444444444444445e-06,
697
+ "loss": 0.1058,
698
+ "step": 2025
699
+ },
700
+ {
701
+ "epoch": 11.649572649572649,
702
+ "grad_norm": 43.11870574951172,
703
+ "learning_rate": 9.286616161616162e-06,
704
+ "loss": 0.2155,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 11.792022792022792,
709
+ "grad_norm": 0.13248799741268158,
710
+ "learning_rate": 9.12878787878788e-06,
711
+ "loss": 0.0918,
712
+ "step": 2075
713
+ },
714
+ {
715
+ "epoch": 11.934472934472934,
716
+ "grad_norm": 0.22579947113990784,
717
+ "learning_rate": 8.970959595959597e-06,
718
+ "loss": 0.1035,
719
+ "step": 2100
720
+ },
721
+ {
722
+ "epoch": 12.0,
723
+ "eval_accuracy": 0.5849056603773585,
724
+ "eval_loss": 3.8249480724334717,
725
+ "eval_macro_f1": 0.5753162185259534,
726
+ "eval_runtime": 1.3868,
727
+ "eval_samples_per_second": 496.829,
728
+ "eval_steps_per_second": 31.728,
729
+ "eval_weighted_f1": 0.576045086114766,
730
+ "step": 2112
731
+ },
732
+ {
733
+ "epoch": 12.074074074074074,
734
+ "grad_norm": 0.14796708524227142,
735
+ "learning_rate": 8.813131313131313e-06,
736
+ "loss": 0.2062,
737
+ "step": 2125
738
+ },
739
+ {
740
+ "epoch": 12.216524216524217,
741
+ "grad_norm": 118.33084869384766,
742
+ "learning_rate": 8.65530303030303e-06,
743
+ "loss": 0.0865,
744
+ "step": 2150
745
+ },
746
+ {
747
+ "epoch": 12.35897435897436,
748
+ "grad_norm": 0.20702308416366577,
749
+ "learning_rate": 8.497474747474748e-06,
750
+ "loss": 0.0105,
751
+ "step": 2175
752
+ },
753
+ {
754
+ "epoch": 12.501424501424502,
755
+ "grad_norm": 0.014355039224028587,
756
+ "learning_rate": 8.339646464646465e-06,
757
+ "loss": 0.0552,
758
+ "step": 2200
759
+ },
760
+ {
761
+ "epoch": 12.643874643874645,
762
+ "grad_norm": 1.1931957006454468,
763
+ "learning_rate": 8.181818181818183e-06,
764
+ "loss": 0.0665,
765
+ "step": 2225
766
+ },
767
+ {
768
+ "epoch": 12.786324786324787,
769
+ "grad_norm": 1.0832961797714233,
770
+ "learning_rate": 8.0239898989899e-06,
771
+ "loss": 0.0268,
772
+ "step": 2250
773
+ },
774
+ {
775
+ "epoch": 12.928774928774928,
776
+ "grad_norm": 9.188243865966797,
777
+ "learning_rate": 7.866161616161618e-06,
778
+ "loss": 0.0332,
779
+ "step": 2275
780
+ },
781
+ {
782
+ "epoch": 13.0,
783
+ "eval_accuracy": 0.5863570391872278,
784
+ "eval_loss": 3.6995561122894287,
785
+ "eval_macro_f1": 0.5859148961439932,
786
+ "eval_runtime": 1.3112,
787
+ "eval_samples_per_second": 525.49,
788
+ "eval_steps_per_second": 33.558,
789
+ "eval_weighted_f1": 0.5858398608606373,
790
+ "step": 2288
791
+ }
792
+ ],
793
+ "logging_steps": 25,
794
+ "max_steps": 3520,
795
+ "num_input_tokens_seen": 0,
796
+ "num_train_epochs": 20,
797
+ "save_steps": 500,
798
+ "stateful_callbacks": {
799
+ "EarlyStoppingCallback": {
800
+ "args": {
801
+ "early_stopping_patience": 3,
802
+ "early_stopping_threshold": 0.0
803
+ },
804
+ "attributes": {
805
+ "early_stopping_patience_counter": 0
806
+ }
807
+ },
808
+ "TrainerControl": {
809
+ "args": {
810
+ "should_epoch_stop": false,
811
+ "should_evaluate": false,
812
+ "should_log": false,
813
+ "should_save": true,
814
+ "should_training_stop": false
815
+ },
816
+ "attributes": {}
817
+ }
818
+ },
819
+ "total_flos": 0.0,
820
+ "train_batch_size": 8,
821
+ "trial_name": null,
822
+ "trial_params": null
823
+ }
ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2816/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2816/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2816/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2816/trainer_state.json ADDED
@@ -0,0 +1,1003 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2288,
3
+ "best_metric": 0.5859148961439932,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-04-s43-20260831T014836Z-1a169064/trainer/checkpoint-2288",
5
+ "epoch": 16.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2816,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.14245014245014245,
14
+ "grad_norm": 27.69466209411621,
15
+ "learning_rate": 1.3636363636363636e-06,
16
+ "loss": 4.8352,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2849002849002849,
21
+ "grad_norm": 19.542036056518555,
22
+ "learning_rate": 2.784090909090909e-06,
23
+ "loss": 4.4519,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.42735042735042733,
28
+ "grad_norm": 26.295495986938477,
29
+ "learning_rate": 4.204545454545455e-06,
30
+ "loss": 4.442,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5698005698005698,
35
+ "grad_norm": 26.776668548583984,
36
+ "learning_rate": 5.625e-06,
37
+ "loss": 4.5038,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.7122507122507122,
42
+ "grad_norm": 23.936386108398438,
43
+ "learning_rate": 7.045454545454546e-06,
44
+ "loss": 4.362,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8547008547008547,
49
+ "grad_norm": 26.568832397460938,
50
+ "learning_rate": 8.465909090909091e-06,
51
+ "loss": 4.4832,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9971509971509972,
56
+ "grad_norm": 18.616518020629883,
57
+ "learning_rate": 9.886363636363637e-06,
58
+ "loss": 4.3872,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.43686502177068215,
64
+ "eval_loss": 1.0517677068710327,
65
+ "eval_macro_f1": 0.3942313052416193,
66
+ "eval_runtime": 1.6887,
67
+ "eval_samples_per_second": 408.006,
68
+ "eval_steps_per_second": 26.056,
69
+ "eval_weighted_f1": 0.39853194871240416,
70
+ "step": 176
71
+ },
72
+ {
73
+ "epoch": 1.1367521367521367,
74
+ "grad_norm": 25.454792022705078,
75
+ "learning_rate": 1.1306818181818183e-05,
76
+ "loss": 4.1347,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2792022792022792,
81
+ "grad_norm": 24.505388259887695,
82
+ "learning_rate": 1.2727272727272728e-05,
83
+ "loss": 4.2989,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.4216524216524218,
88
+ "grad_norm": 22.538040161132812,
89
+ "learning_rate": 1.4147727272727274e-05,
90
+ "loss": 4.235,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.564102564102564,
95
+ "grad_norm": 18.85667610168457,
96
+ "learning_rate": 1.5568181818181822e-05,
97
+ "loss": 4.2795,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.7065527065527064,
102
+ "grad_norm": 24.444332122802734,
103
+ "learning_rate": 1.6988636363636368e-05,
104
+ "loss": 4.2162,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.849002849002849,
109
+ "grad_norm": 19.221044540405273,
110
+ "learning_rate": 1.840909090909091e-05,
111
+ "loss": 4.186,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9914529914529915,
116
+ "grad_norm": 27.694293975830078,
117
+ "learning_rate": 1.9829545454545456e-05,
118
+ "loss": 4.1084,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.5094339622641509,
124
+ "eval_loss": 0.986926257610321,
125
+ "eval_macro_f1": 0.4962483502863326,
126
+ "eval_runtime": 1.4691,
127
+ "eval_samples_per_second": 468.987,
128
+ "eval_steps_per_second": 29.95,
129
+ "eval_weighted_f1": 0.49832918236349383,
130
+ "step": 352
131
+ },
132
+ {
133
+ "epoch": 2.131054131054131,
134
+ "grad_norm": 27.87932777404785,
135
+ "learning_rate": 1.9861111111111114e-05,
136
+ "loss": 3.7175,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.2735042735042734,
141
+ "grad_norm": 26.27695083618164,
142
+ "learning_rate": 1.9703282828282832e-05,
143
+ "loss": 3.5301,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.415954415954416,
148
+ "grad_norm": 40.19943618774414,
149
+ "learning_rate": 1.9545454545454546e-05,
150
+ "loss": 3.6274,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.5584045584045585,
155
+ "grad_norm": 43.94453811645508,
156
+ "learning_rate": 1.9387626262626263e-05,
157
+ "loss": 3.8303,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.700854700854701,
162
+ "grad_norm": 47.385799407958984,
163
+ "learning_rate": 1.922979797979798e-05,
164
+ "loss": 3.5218,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.8433048433048436,
169
+ "grad_norm": 29.534732818603516,
170
+ "learning_rate": 1.90719696969697e-05,
171
+ "loss": 3.6893,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.9857549857549857,
176
+ "grad_norm": 37.31429672241211,
177
+ "learning_rate": 1.8914141414141416e-05,
178
+ "loss": 3.5932,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.5544267053701016,
184
+ "eval_loss": 0.9797604084014893,
185
+ "eval_macro_f1": 0.5478060457189247,
186
+ "eval_runtime": 1.3112,
187
+ "eval_samples_per_second": 525.49,
188
+ "eval_steps_per_second": 33.558,
189
+ "eval_weighted_f1": 0.5493016960716426,
190
+ "step": 528
191
+ },
192
+ {
193
+ "epoch": 3.1253561253561255,
194
+ "grad_norm": 67.17666625976562,
195
+ "learning_rate": 1.8756313131313133e-05,
196
+ "loss": 2.8348,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.267806267806268,
201
+ "grad_norm": 55.27144241333008,
202
+ "learning_rate": 1.859848484848485e-05,
203
+ "loss": 2.7264,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.41025641025641,
208
+ "grad_norm": 57.96361541748047,
209
+ "learning_rate": 1.844065656565657e-05,
210
+ "loss": 2.9081,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.5527065527065527,
215
+ "grad_norm": 27.539400100708008,
216
+ "learning_rate": 1.8282828282828286e-05,
217
+ "loss": 2.9182,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.695156695156695,
222
+ "grad_norm": 45.49375534057617,
223
+ "learning_rate": 1.8125e-05,
224
+ "loss": 2.7568,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.8376068376068377,
229
+ "grad_norm": 23.619964599609375,
230
+ "learning_rate": 1.7967171717171718e-05,
231
+ "loss": 2.5585,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.98005698005698,
236
+ "grad_norm": 33.051025390625,
237
+ "learning_rate": 1.7809343434343435e-05,
238
+ "loss": 2.848,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 4.0,
243
+ "eval_accuracy": 0.5515239477503628,
244
+ "eval_loss": 1.0126099586486816,
245
+ "eval_macro_f1": 0.5522044976182564,
246
+ "eval_runtime": 1.3267,
247
+ "eval_samples_per_second": 519.346,
248
+ "eval_steps_per_second": 33.166,
249
+ "eval_weighted_f1": 0.5524713353890915,
250
+ "step": 704
251
+ },
252
+ {
253
+ "epoch": 4.119658119658119,
254
+ "grad_norm": 41.599609375,
255
+ "learning_rate": 1.7651515151515153e-05,
256
+ "loss": 2.1393,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.262108262108262,
261
+ "grad_norm": 33.376155853271484,
262
+ "learning_rate": 1.749368686868687e-05,
263
+ "loss": 2.003,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.404558404558404,
268
+ "grad_norm": 62.67402648925781,
269
+ "learning_rate": 1.7335858585858588e-05,
270
+ "loss": 1.6959,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.547008547008547,
275
+ "grad_norm": 55.37060546875,
276
+ "learning_rate": 1.7178030303030305e-05,
277
+ "loss": 1.7717,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.689458689458689,
282
+ "grad_norm": 55.278926849365234,
283
+ "learning_rate": 1.7020202020202023e-05,
284
+ "loss": 1.9422,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.831908831908832,
289
+ "grad_norm": 35.771244049072266,
290
+ "learning_rate": 1.6862373737373737e-05,
291
+ "loss": 1.9357,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.9743589743589745,
296
+ "grad_norm": 43.035099029541016,
297
+ "learning_rate": 1.6704545454545454e-05,
298
+ "loss": 1.7887,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 5.0,
303
+ "eval_accuracy": 0.5529753265602322,
304
+ "eval_loss": 1.307969093322754,
305
+ "eval_macro_f1": 0.5500873943650679,
306
+ "eval_runtime": 1.4687,
307
+ "eval_samples_per_second": 469.117,
308
+ "eval_steps_per_second": 29.958,
309
+ "eval_weighted_f1": 0.5511325286130112,
310
+ "step": 880
311
+ },
312
+ {
313
+ "epoch": 5.113960113960114,
314
+ "grad_norm": 47.52789306640625,
315
+ "learning_rate": 1.6546717171717172e-05,
316
+ "loss": 1.3756,
317
+ "step": 900
318
+ },
319
+ {
320
+ "epoch": 5.256410256410256,
321
+ "grad_norm": 67.07496643066406,
322
+ "learning_rate": 1.638888888888889e-05,
323
+ "loss": 0.9039,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.398860398860399,
328
+ "grad_norm": 31.854036331176758,
329
+ "learning_rate": 1.6231060606060607e-05,
330
+ "loss": 1.1748,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.5413105413105415,
335
+ "grad_norm": 106.62139892578125,
336
+ "learning_rate": 1.6073232323232324e-05,
337
+ "loss": 1.195,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.683760683760684,
342
+ "grad_norm": 73.60779571533203,
343
+ "learning_rate": 1.5915404040404042e-05,
344
+ "loss": 1.3651,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.8262108262108265,
349
+ "grad_norm": 51.60421371459961,
350
+ "learning_rate": 1.575757575757576e-05,
351
+ "loss": 1.2479,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.968660968660968,
356
+ "grad_norm": 65.28878784179688,
357
+ "learning_rate": 1.5599747474747477e-05,
358
+ "loss": 1.0724,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 6.0,
363
+ "eval_accuracy": 0.5747460087082729,
364
+ "eval_loss": 1.5242938995361328,
365
+ "eval_macro_f1": 0.5721091381075082,
366
+ "eval_runtime": 1.3765,
367
+ "eval_samples_per_second": 500.541,
368
+ "eval_steps_per_second": 31.965,
369
+ "eval_weighted_f1": 0.5727603574205964,
370
+ "step": 1056
371
+ },
372
+ {
373
+ "epoch": 6.1082621082621085,
374
+ "grad_norm": 26.339204788208008,
375
+ "learning_rate": 1.544191919191919e-05,
376
+ "loss": 0.7805,
377
+ "step": 1075
378
+ },
379
+ {
380
+ "epoch": 6.250712250712251,
381
+ "grad_norm": 53.746768951416016,
382
+ "learning_rate": 1.528409090909091e-05,
383
+ "loss": 0.5242,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.3931623931623935,
388
+ "grad_norm": 34.56966781616211,
389
+ "learning_rate": 1.5126262626262626e-05,
390
+ "loss": 0.7292,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.535612535612536,
395
+ "grad_norm": 74.35704040527344,
396
+ "learning_rate": 1.4968434343434344e-05,
397
+ "loss": 0.7179,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.678062678062678,
402
+ "grad_norm": 28.260852813720703,
403
+ "learning_rate": 1.4810606060606061e-05,
404
+ "loss": 0.7493,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.82051282051282,
409
+ "grad_norm": 61.798500061035156,
410
+ "learning_rate": 1.4652777777777779e-05,
411
+ "loss": 0.7055,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.962962962962963,
416
+ "grad_norm": 39.78801727294922,
417
+ "learning_rate": 1.4494949494949494e-05,
418
+ "loss": 0.75,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 7.0,
423
+ "eval_accuracy": 0.5674891146589259,
424
+ "eval_loss": 1.7215156555175781,
425
+ "eval_macro_f1": 0.5612321340302727,
426
+ "eval_runtime": 1.3148,
427
+ "eval_samples_per_second": 524.042,
428
+ "eval_steps_per_second": 33.466,
429
+ "eval_weighted_f1": 0.5611592964466283,
430
+ "step": 1232
431
+ },
432
+ {
433
+ "epoch": 7.102564102564102,
434
+ "grad_norm": 23.60300064086914,
435
+ "learning_rate": 1.4337121212121212e-05,
436
+ "loss": 0.7275,
437
+ "step": 1250
438
+ },
439
+ {
440
+ "epoch": 7.245014245014245,
441
+ "grad_norm": 18.518918991088867,
442
+ "learning_rate": 1.4179292929292931e-05,
443
+ "loss": 0.3774,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.387464387464387,
448
+ "grad_norm": 74.9653549194336,
449
+ "learning_rate": 1.4021464646464649e-05,
450
+ "loss": 0.4915,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.52991452991453,
455
+ "grad_norm": 35.73414611816406,
456
+ "learning_rate": 1.3863636363636366e-05,
457
+ "loss": 0.4356,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.672364672364672,
462
+ "grad_norm": 3.858736753463745,
463
+ "learning_rate": 1.3705808080808084e-05,
464
+ "loss": 0.315,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.814814814814815,
469
+ "grad_norm": 50.60848617553711,
470
+ "learning_rate": 1.35479797979798e-05,
471
+ "loss": 0.4276,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.957264957264957,
476
+ "grad_norm": 70.68370819091797,
477
+ "learning_rate": 1.3390151515151517e-05,
478
+ "loss": 0.3087,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 8.0,
483
+ "eval_accuracy": 0.5834542815674891,
484
+ "eval_loss": 2.2690086364746094,
485
+ "eval_macro_f1": 0.575839318316549,
486
+ "eval_runtime": 1.5932,
487
+ "eval_samples_per_second": 432.466,
488
+ "eval_steps_per_second": 27.618,
489
+ "eval_weighted_f1": 0.5768206681992663,
490
+ "step": 1408
491
+ },
492
+ {
493
+ "epoch": 8.096866096866098,
494
+ "grad_norm": 31.564485549926758,
495
+ "learning_rate": 1.3232323232323234e-05,
496
+ "loss": 0.2772,
497
+ "step": 1425
498
+ },
499
+ {
500
+ "epoch": 8.239316239316238,
501
+ "grad_norm": 61.35936737060547,
502
+ "learning_rate": 1.3074494949494952e-05,
503
+ "loss": 0.2677,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.381766381766381,
508
+ "grad_norm": 24.228017807006836,
509
+ "learning_rate": 1.2916666666666668e-05,
510
+ "loss": 0.2506,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.524216524216524,
515
+ "grad_norm": 53.46422576904297,
516
+ "learning_rate": 1.2758838383838385e-05,
517
+ "loss": 0.3032,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.666666666666666,
522
+ "grad_norm": 61.28293228149414,
523
+ "learning_rate": 1.2601010101010103e-05,
524
+ "loss": 0.3046,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.809116809116809,
529
+ "grad_norm": 2.150891065597534,
530
+ "learning_rate": 1.244318181818182e-05,
531
+ "loss": 0.4309,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.951566951566951,
536
+ "grad_norm": 115.88075256347656,
537
+ "learning_rate": 1.2285353535353536e-05,
538
+ "loss": 0.2445,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 9.0,
543
+ "eval_accuracy": 0.5573294629898403,
544
+ "eval_loss": 2.739509344100952,
545
+ "eval_macro_f1": 0.5511234758681754,
546
+ "eval_runtime": 1.4544,
547
+ "eval_samples_per_second": 473.741,
548
+ "eval_steps_per_second": 30.253,
549
+ "eval_weighted_f1": 0.5523768897983876,
550
+ "step": 1584
551
+ },
552
+ {
553
+ "epoch": 9.091168091168091,
554
+ "grad_norm": 87.18209075927734,
555
+ "learning_rate": 1.2127525252525254e-05,
556
+ "loss": 0.1399,
557
+ "step": 1600
558
+ },
559
+ {
560
+ "epoch": 9.233618233618234,
561
+ "grad_norm": 28.503408432006836,
562
+ "learning_rate": 1.1969696969696971e-05,
563
+ "loss": 0.1585,
564
+ "step": 1625
565
+ },
566
+ {
567
+ "epoch": 9.376068376068377,
568
+ "grad_norm": 49.824867248535156,
569
+ "learning_rate": 1.1811868686868689e-05,
570
+ "loss": 0.3722,
571
+ "step": 1650
572
+ },
573
+ {
574
+ "epoch": 9.518518518518519,
575
+ "grad_norm": 1.6686387062072754,
576
+ "learning_rate": 1.1654040404040406e-05,
577
+ "loss": 0.2527,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.660968660968662,
582
+ "grad_norm": 0.9758527278900146,
583
+ "learning_rate": 1.1496212121212122e-05,
584
+ "loss": 0.2327,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.803418803418804,
589
+ "grad_norm": 2.719261646270752,
590
+ "learning_rate": 1.133838383838384e-05,
591
+ "loss": 0.1828,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.945868945868947,
596
+ "grad_norm": 7.621354579925537,
597
+ "learning_rate": 1.1180555555555557e-05,
598
+ "loss": 0.277,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 10.0,
603
+ "eval_accuracy": 0.579100145137881,
604
+ "eval_loss": 2.908055305480957,
605
+ "eval_macro_f1": 0.57907237466202,
606
+ "eval_runtime": 1.4314,
607
+ "eval_samples_per_second": 481.333,
608
+ "eval_steps_per_second": 30.738,
609
+ "eval_weighted_f1": 0.5792181523580426,
610
+ "step": 1760
611
+ },
612
+ {
613
+ "epoch": 10.085470085470085,
614
+ "grad_norm": 13.50964641571045,
615
+ "learning_rate": 1.1022727272727275e-05,
616
+ "loss": 0.0596,
617
+ "step": 1775
618
+ },
619
+ {
620
+ "epoch": 10.227920227920228,
621
+ "grad_norm": 1.35202157497406,
622
+ "learning_rate": 1.086489898989899e-05,
623
+ "loss": 0.0965,
624
+ "step": 1800
625
+ },
626
+ {
627
+ "epoch": 10.37037037037037,
628
+ "grad_norm": 47.22248458862305,
629
+ "learning_rate": 1.0707070707070708e-05,
630
+ "loss": 0.1526,
631
+ "step": 1825
632
+ },
633
+ {
634
+ "epoch": 10.512820512820513,
635
+ "grad_norm": 1.046852469444275,
636
+ "learning_rate": 1.0549242424242425e-05,
637
+ "loss": 0.0543,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.655270655270655,
642
+ "grad_norm": 0.43407419323921204,
643
+ "learning_rate": 1.0391414141414143e-05,
644
+ "loss": 0.1675,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.797720797720798,
649
+ "grad_norm": 56.43595504760742,
650
+ "learning_rate": 1.0233585858585859e-05,
651
+ "loss": 0.2615,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.94017094017094,
656
+ "grad_norm": 94.71170043945312,
657
+ "learning_rate": 1.0075757575757576e-05,
658
+ "loss": 0.2073,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 11.0,
663
+ "eval_accuracy": 0.5544267053701016,
664
+ "eval_loss": 3.494981050491333,
665
+ "eval_macro_f1": 0.5537080472229138,
666
+ "eval_runtime": 1.4298,
667
+ "eval_samples_per_second": 481.892,
668
+ "eval_steps_per_second": 30.774,
669
+ "eval_weighted_f1": 0.5546618094147888,
670
+ "step": 1936
671
+ },
672
+ {
673
+ "epoch": 11.079772079772079,
674
+ "grad_norm": 113.4718017578125,
675
+ "learning_rate": 9.917929292929294e-06,
676
+ "loss": 0.19,
677
+ "step": 1950
678
+ },
679
+ {
680
+ "epoch": 11.222222222222221,
681
+ "grad_norm": 114.95403289794922,
682
+ "learning_rate": 9.760101010101011e-06,
683
+ "loss": 0.075,
684
+ "step": 1975
685
+ },
686
+ {
687
+ "epoch": 11.364672364672364,
688
+ "grad_norm": 0.019300974905490875,
689
+ "learning_rate": 9.602272727272727e-06,
690
+ "loss": 0.2343,
691
+ "step": 2000
692
+ },
693
+ {
694
+ "epoch": 11.507122507122507,
695
+ "grad_norm": 121.42213439941406,
696
+ "learning_rate": 9.444444444444445e-06,
697
+ "loss": 0.1058,
698
+ "step": 2025
699
+ },
700
+ {
701
+ "epoch": 11.649572649572649,
702
+ "grad_norm": 43.11870574951172,
703
+ "learning_rate": 9.286616161616162e-06,
704
+ "loss": 0.2155,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 11.792022792022792,
709
+ "grad_norm": 0.13248799741268158,
710
+ "learning_rate": 9.12878787878788e-06,
711
+ "loss": 0.0918,
712
+ "step": 2075
713
+ },
714
+ {
715
+ "epoch": 11.934472934472934,
716
+ "grad_norm": 0.22579947113990784,
717
+ "learning_rate": 8.970959595959597e-06,
718
+ "loss": 0.1035,
719
+ "step": 2100
720
+ },
721
+ {
722
+ "epoch": 12.0,
723
+ "eval_accuracy": 0.5849056603773585,
724
+ "eval_loss": 3.8249480724334717,
725
+ "eval_macro_f1": 0.5753162185259534,
726
+ "eval_runtime": 1.3868,
727
+ "eval_samples_per_second": 496.829,
728
+ "eval_steps_per_second": 31.728,
729
+ "eval_weighted_f1": 0.576045086114766,
730
+ "step": 2112
731
+ },
732
+ {
733
+ "epoch": 12.074074074074074,
734
+ "grad_norm": 0.14796708524227142,
735
+ "learning_rate": 8.813131313131313e-06,
736
+ "loss": 0.2062,
737
+ "step": 2125
738
+ },
739
+ {
740
+ "epoch": 12.216524216524217,
741
+ "grad_norm": 118.33084869384766,
742
+ "learning_rate": 8.65530303030303e-06,
743
+ "loss": 0.0865,
744
+ "step": 2150
745
+ },
746
+ {
747
+ "epoch": 12.35897435897436,
748
+ "grad_norm": 0.20702308416366577,
749
+ "learning_rate": 8.497474747474748e-06,
750
+ "loss": 0.0105,
751
+ "step": 2175
752
+ },
753
+ {
754
+ "epoch": 12.501424501424502,
755
+ "grad_norm": 0.014355039224028587,
756
+ "learning_rate": 8.339646464646465e-06,
757
+ "loss": 0.0552,
758
+ "step": 2200
759
+ },
760
+ {
761
+ "epoch": 12.643874643874645,
762
+ "grad_norm": 1.1931957006454468,
763
+ "learning_rate": 8.181818181818183e-06,
764
+ "loss": 0.0665,
765
+ "step": 2225
766
+ },
767
+ {
768
+ "epoch": 12.786324786324787,
769
+ "grad_norm": 1.0832961797714233,
770
+ "learning_rate": 8.0239898989899e-06,
771
+ "loss": 0.0268,
772
+ "step": 2250
773
+ },
774
+ {
775
+ "epoch": 12.928774928774928,
776
+ "grad_norm": 9.188243865966797,
777
+ "learning_rate": 7.866161616161618e-06,
778
+ "loss": 0.0332,
779
+ "step": 2275
780
+ },
781
+ {
782
+ "epoch": 13.0,
783
+ "eval_accuracy": 0.5863570391872278,
784
+ "eval_loss": 3.6995561122894287,
785
+ "eval_macro_f1": 0.5859148961439932,
786
+ "eval_runtime": 1.3112,
787
+ "eval_samples_per_second": 525.49,
788
+ "eval_steps_per_second": 33.558,
789
+ "eval_weighted_f1": 0.5858398608606373,
790
+ "step": 2288
791
+ },
792
+ {
793
+ "epoch": 13.068376068376068,
794
+ "grad_norm": 0.767853319644928,
795
+ "learning_rate": 7.708333333333334e-06,
796
+ "loss": 0.1381,
797
+ "step": 2300
798
+ },
799
+ {
800
+ "epoch": 13.21082621082621,
801
+ "grad_norm": 0.021792540326714516,
802
+ "learning_rate": 7.550505050505051e-06,
803
+ "loss": 0.1077,
804
+ "step": 2325
805
+ },
806
+ {
807
+ "epoch": 13.353276353276353,
808
+ "grad_norm": 0.7662265300750732,
809
+ "learning_rate": 7.392676767676769e-06,
810
+ "loss": 0.0138,
811
+ "step": 2350
812
+ },
813
+ {
814
+ "epoch": 13.495726495726496,
815
+ "grad_norm": 0.37084367871284485,
816
+ "learning_rate": 7.2348484848484855e-06,
817
+ "loss": 0.0508,
818
+ "step": 2375
819
+ },
820
+ {
821
+ "epoch": 13.638176638176638,
822
+ "grad_norm": 0.67719966173172,
823
+ "learning_rate": 7.077020202020203e-06,
824
+ "loss": 0.0817,
825
+ "step": 2400
826
+ },
827
+ {
828
+ "epoch": 13.78062678062678,
829
+ "grad_norm": 0.7890069484710693,
830
+ "learning_rate": 6.91919191919192e-06,
831
+ "loss": 0.0023,
832
+ "step": 2425
833
+ },
834
+ {
835
+ "epoch": 13.923076923076923,
836
+ "grad_norm": 1.2881344556808472,
837
+ "learning_rate": 6.761363636363637e-06,
838
+ "loss": 0.0383,
839
+ "step": 2450
840
+ },
841
+ {
842
+ "epoch": 14.0,
843
+ "eval_accuracy": 0.5674891146589259,
844
+ "eval_loss": 3.994131326675415,
845
+ "eval_macro_f1": 0.5688467474087349,
846
+ "eval_runtime": 1.4256,
847
+ "eval_samples_per_second": 483.321,
848
+ "eval_steps_per_second": 30.865,
849
+ "eval_weighted_f1": 0.568902789224312,
850
+ "step": 2464
851
+ },
852
+ {
853
+ "epoch": 14.062678062678062,
854
+ "grad_norm": 0.1331017166376114,
855
+ "learning_rate": 6.603535353535354e-06,
856
+ "loss": 0.0028,
857
+ "step": 2475
858
+ },
859
+ {
860
+ "epoch": 14.205128205128204,
861
+ "grad_norm": 41.10041046142578,
862
+ "learning_rate": 6.445707070707071e-06,
863
+ "loss": 0.0186,
864
+ "step": 2500
865
+ },
866
+ {
867
+ "epoch": 14.347578347578347,
868
+ "grad_norm": 0.0061663491651415825,
869
+ "learning_rate": 6.287878787878788e-06,
870
+ "loss": 0.0558,
871
+ "step": 2525
872
+ },
873
+ {
874
+ "epoch": 14.49002849002849,
875
+ "grad_norm": 0.08111625909805298,
876
+ "learning_rate": 6.1300505050505055e-06,
877
+ "loss": 0.0375,
878
+ "step": 2550
879
+ },
880
+ {
881
+ "epoch": 14.632478632478632,
882
+ "grad_norm": 0.11624642461538315,
883
+ "learning_rate": 5.972222222222222e-06,
884
+ "loss": 0.0008,
885
+ "step": 2575
886
+ },
887
+ {
888
+ "epoch": 14.774928774928775,
889
+ "grad_norm": 0.00500089954584837,
890
+ "learning_rate": 5.81439393939394e-06,
891
+ "loss": 0.0991,
892
+ "step": 2600
893
+ },
894
+ {
895
+ "epoch": 14.917378917378917,
896
+ "grad_norm": 0.010520764626562595,
897
+ "learning_rate": 5.656565656565657e-06,
898
+ "loss": 0.0149,
899
+ "step": 2625
900
+ },
901
+ {
902
+ "epoch": 15.0,
903
+ "eval_accuracy": 0.5849056603773585,
904
+ "eval_loss": 4.010282516479492,
905
+ "eval_macro_f1": 0.5825035476431479,
906
+ "eval_runtime": 1.3953,
907
+ "eval_samples_per_second": 493.785,
908
+ "eval_steps_per_second": 31.533,
909
+ "eval_weighted_f1": 0.582525742558303,
910
+ "step": 2640
911
+ },
912
+ {
913
+ "epoch": 15.056980056980057,
914
+ "grad_norm": 1.8812940120697021,
915
+ "learning_rate": 5.498737373737374e-06,
916
+ "loss": 0.0541,
917
+ "step": 2650
918
+ },
919
+ {
920
+ "epoch": 15.1994301994302,
921
+ "grad_norm": 0.015146266669034958,
922
+ "learning_rate": 5.340909090909091e-06,
923
+ "loss": 0.0363,
924
+ "step": 2675
925
+ },
926
+ {
927
+ "epoch": 15.341880341880342,
928
+ "grad_norm": 0.38464805483818054,
929
+ "learning_rate": 5.183080808080808e-06,
930
+ "loss": 0.0335,
931
+ "step": 2700
932
+ },
933
+ {
934
+ "epoch": 15.484330484330485,
935
+ "grad_norm": 4.100954055786133,
936
+ "learning_rate": 5.0252525252525255e-06,
937
+ "loss": 0.0404,
938
+ "step": 2725
939
+ },
940
+ {
941
+ "epoch": 15.626780626780628,
942
+ "grad_norm": 0.6361879706382751,
943
+ "learning_rate": 4.867424242424243e-06,
944
+ "loss": 0.0043,
945
+ "step": 2750
946
+ },
947
+ {
948
+ "epoch": 15.76923076923077,
949
+ "grad_norm": 0.0193466879427433,
950
+ "learning_rate": 4.70959595959596e-06,
951
+ "loss": 0.0207,
952
+ "step": 2775
953
+ },
954
+ {
955
+ "epoch": 15.91168091168091,
956
+ "grad_norm": 3.0020270347595215,
957
+ "learning_rate": 4.551767676767677e-06,
958
+ "loss": 0.0374,
959
+ "step": 2800
960
+ },
961
+ {
962
+ "epoch": 16.0,
963
+ "eval_accuracy": 0.5892597968069666,
964
+ "eval_loss": 4.125938415527344,
965
+ "eval_macro_f1": 0.5842694071294178,
966
+ "eval_runtime": 1.3335,
967
+ "eval_samples_per_second": 516.667,
968
+ "eval_steps_per_second": 32.995,
969
+ "eval_weighted_f1": 0.584945473219162,
970
+ "step": 2816
971
+ }
972
+ ],
973
+ "logging_steps": 25,
974
+ "max_steps": 3520,
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-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620/trainer_state.json ADDED
@@ -0,0 +1,590 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1620,
3
+ "best_metric": 0.5835810641427687,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620",
5
+ "epoch": 9.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1620,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1388888888888889,
14
+ "grad_norm": 29.263378143310547,
15
+ "learning_rate": 1.3333333333333334e-06,
16
+ "loss": 4.8502,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2777777777777778,
21
+ "grad_norm": 20.813480377197266,
22
+ "learning_rate": 2.7222222222222224e-06,
23
+ "loss": 4.4757,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4166666666666667,
28
+ "grad_norm": 36.471431732177734,
29
+ "learning_rate": 4.111111111111111e-06,
30
+ "loss": 4.5168,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5555555555555556,
35
+ "grad_norm": 30.926923751831055,
36
+ "learning_rate": 5.500000000000001e-06,
37
+ "loss": 4.4289,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6944444444444444,
42
+ "grad_norm": 27.58977508544922,
43
+ "learning_rate": 6.88888888888889e-06,
44
+ "loss": 4.4247,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8333333333333334,
49
+ "grad_norm": 34.77938461303711,
50
+ "learning_rate": 8.277777777777778e-06,
51
+ "loss": 4.3307,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9722222222222222,
56
+ "grad_norm": 32.43658447265625,
57
+ "learning_rate": 9.666666666666667e-06,
58
+ "loss": 4.378,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4014925373134328,
64
+ "eval_loss": 1.0966709852218628,
65
+ "eval_macro_f1": 0.34144070307099833,
66
+ "eval_runtime": 1.3826,
67
+ "eval_samples_per_second": 484.596,
68
+ "eval_steps_per_second": 30.378,
69
+ "eval_weighted_f1": 0.3363109803996705,
70
+ "step": 180
71
+ },
72
+ {
73
+ "epoch": 1.1111111111111112,
74
+ "grad_norm": 16.528186798095703,
75
+ "learning_rate": 1.1055555555555557e-05,
76
+ "loss": 4.2033,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.25,
81
+ "grad_norm": 17.700929641723633,
82
+ "learning_rate": 1.2444444444444446e-05,
83
+ "loss": 4.3294,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3888888888888888,
88
+ "grad_norm": 17.803096771240234,
89
+ "learning_rate": 1.3833333333333334e-05,
90
+ "loss": 4.2576,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.5277777777777777,
95
+ "grad_norm": 18.98823356628418,
96
+ "learning_rate": 1.5222222222222223e-05,
97
+ "loss": 4.2086,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6666666666666665,
102
+ "grad_norm": 20.693819046020508,
103
+ "learning_rate": 1.6611111111111113e-05,
104
+ "loss": 4.1248,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.8055555555555556,
109
+ "grad_norm": 28.918561935424805,
110
+ "learning_rate": 1.8e-05,
111
+ "loss": 4.2757,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9444444444444444,
116
+ "grad_norm": 19.435781478881836,
117
+ "learning_rate": 1.938888888888889e-05,
118
+ "loss": 4.1187,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.5477611940298508,
124
+ "eval_loss": 0.9657241106033325,
125
+ "eval_macro_f1": 0.5473055994383008,
126
+ "eval_runtime": 1.3516,
127
+ "eval_samples_per_second": 495.711,
128
+ "eval_steps_per_second": 31.074,
129
+ "eval_weighted_f1": 0.5466602803595011,
130
+ "step": 360
131
+ },
132
+ {
133
+ "epoch": 2.0833333333333335,
134
+ "grad_norm": 23.853336334228516,
135
+ "learning_rate": 1.991358024691358e-05,
136
+ "loss": 3.941,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.2222222222222223,
141
+ "grad_norm": 34.18367385864258,
142
+ "learning_rate": 1.975925925925926e-05,
143
+ "loss": 3.7539,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.361111111111111,
148
+ "grad_norm": 44.504329681396484,
149
+ "learning_rate": 1.960493827160494e-05,
150
+ "loss": 3.6522,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.5,
155
+ "grad_norm": 29.060413360595703,
156
+ "learning_rate": 1.945061728395062e-05,
157
+ "loss": 3.6989,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.638888888888889,
162
+ "grad_norm": 31.384469985961914,
163
+ "learning_rate": 1.92962962962963e-05,
164
+ "loss": 3.4029,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7777777777777777,
169
+ "grad_norm": 48.808494567871094,
170
+ "learning_rate": 1.914197530864198e-05,
171
+ "loss": 3.8388,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.9166666666666665,
176
+ "grad_norm": 19.36366844177246,
177
+ "learning_rate": 1.8987654320987657e-05,
178
+ "loss": 3.6407,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.5835820895522388,
184
+ "eval_loss": 0.900704026222229,
185
+ "eval_macro_f1": 0.5730823063833178,
186
+ "eval_runtime": 1.31,
187
+ "eval_samples_per_second": 511.455,
188
+ "eval_steps_per_second": 32.061,
189
+ "eval_weighted_f1": 0.5735315438383259,
190
+ "step": 540
191
+ },
192
+ {
193
+ "epoch": 3.0555555555555554,
194
+ "grad_norm": 24.26601791381836,
195
+ "learning_rate": 1.8833333333333335e-05,
196
+ "loss": 3.3326,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.1944444444444446,
201
+ "grad_norm": 61.2514533996582,
202
+ "learning_rate": 1.8679012345679013e-05,
203
+ "loss": 2.8269,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.3333333333333335,
208
+ "grad_norm": 47.86339569091797,
209
+ "learning_rate": 1.8524691358024694e-05,
210
+ "loss": 2.7367,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.4722222222222223,
215
+ "grad_norm": 38.8129768371582,
216
+ "learning_rate": 1.837037037037037e-05,
217
+ "loss": 2.7173,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.611111111111111,
222
+ "grad_norm": 45.2987174987793,
223
+ "learning_rate": 1.821604938271605e-05,
224
+ "loss": 2.7368,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.75,
229
+ "grad_norm": 37.727134704589844,
230
+ "learning_rate": 1.806172839506173e-05,
231
+ "loss": 2.7808,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.888888888888889,
236
+ "grad_norm": 47.086238861083984,
237
+ "learning_rate": 1.7907407407407408e-05,
238
+ "loss": 2.5216,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 4.0,
243
+ "eval_accuracy": 0.5507462686567164,
244
+ "eval_loss": 1.0473291873931885,
245
+ "eval_macro_f1": 0.5416214803491114,
246
+ "eval_runtime": 1.2777,
247
+ "eval_samples_per_second": 524.367,
248
+ "eval_steps_per_second": 32.871,
249
+ "eval_weighted_f1": 0.5430559342981475,
250
+ "step": 720
251
+ },
252
+ {
253
+ "epoch": 4.027777777777778,
254
+ "grad_norm": 39.34019088745117,
255
+ "learning_rate": 1.775308641975309e-05,
256
+ "loss": 2.6099,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.166666666666667,
261
+ "grad_norm": 34.96302032470703,
262
+ "learning_rate": 1.7598765432098767e-05,
263
+ "loss": 1.8181,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.305555555555555,
268
+ "grad_norm": 46.8112907409668,
269
+ "learning_rate": 1.7444444444444448e-05,
270
+ "loss": 1.743,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.444444444444445,
275
+ "grad_norm": 49.2983512878418,
276
+ "learning_rate": 1.7290123456790125e-05,
277
+ "loss": 1.6882,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.583333333333333,
282
+ "grad_norm": 34.680213928222656,
283
+ "learning_rate": 1.7135802469135806e-05,
284
+ "loss": 1.8312,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.722222222222222,
289
+ "grad_norm": 50.23997497558594,
290
+ "learning_rate": 1.6981481481481484e-05,
291
+ "loss": 1.9134,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.861111111111111,
296
+ "grad_norm": 32.60227584838867,
297
+ "learning_rate": 1.6827160493827162e-05,
298
+ "loss": 1.7515,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 5.0,
303
+ "grad_norm": 61.78374481201172,
304
+ "learning_rate": 1.667283950617284e-05,
305
+ "loss": 1.7827,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.582089552238806,
311
+ "eval_loss": 1.166916012763977,
312
+ "eval_macro_f1": 0.5777411529460815,
313
+ "eval_runtime": 1.2641,
314
+ "eval_samples_per_second": 530.041,
315
+ "eval_steps_per_second": 33.226,
316
+ "eval_weighted_f1": 0.5779446555827403,
317
+ "step": 900
318
+ },
319
+ {
320
+ "epoch": 5.138888888888889,
321
+ "grad_norm": 128.76226806640625,
322
+ "learning_rate": 1.651851851851852e-05,
323
+ "loss": 1.1251,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.277777777777778,
328
+ "grad_norm": 78.46173095703125,
329
+ "learning_rate": 1.6364197530864198e-05,
330
+ "loss": 1.0684,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.416666666666667,
335
+ "grad_norm": 42.87739944458008,
336
+ "learning_rate": 1.6209876543209876e-05,
337
+ "loss": 0.8983,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.555555555555555,
342
+ "grad_norm": 78.20256042480469,
343
+ "learning_rate": 1.6055555555555557e-05,
344
+ "loss": 0.885,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.694444444444445,
349
+ "grad_norm": 60.07840347290039,
350
+ "learning_rate": 1.5901234567901235e-05,
351
+ "loss": 1.0987,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.833333333333333,
356
+ "grad_norm": 69.35907745361328,
357
+ "learning_rate": 1.5746913580246916e-05,
358
+ "loss": 1.11,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.972222222222222,
363
+ "grad_norm": 90.80661010742188,
364
+ "learning_rate": 1.5592592592592593e-05,
365
+ "loss": 1.1472,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.5686567164179105,
371
+ "eval_loss": 1.5736793279647827,
372
+ "eval_macro_f1": 0.5635663296380078,
373
+ "eval_runtime": 1.2472,
374
+ "eval_samples_per_second": 537.197,
375
+ "eval_steps_per_second": 33.675,
376
+ "eval_weighted_f1": 0.5646891761432958,
377
+ "step": 1080
378
+ },
379
+ {
380
+ "epoch": 6.111111111111111,
381
+ "grad_norm": 34.897483825683594,
382
+ "learning_rate": 1.5438271604938274e-05,
383
+ "loss": 0.6438,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.25,
388
+ "grad_norm": 49.88667678833008,
389
+ "learning_rate": 1.5283950617283952e-05,
390
+ "loss": 0.7807,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.388888888888889,
395
+ "grad_norm": 20.57355308532715,
396
+ "learning_rate": 1.5129629629629632e-05,
397
+ "loss": 0.8069,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.527777777777778,
402
+ "grad_norm": 23.29080581665039,
403
+ "learning_rate": 1.4975308641975311e-05,
404
+ "loss": 0.6768,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.666666666666667,
409
+ "grad_norm": 78.04934692382812,
410
+ "learning_rate": 1.4820987654320989e-05,
411
+ "loss": 0.6018,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.805555555555555,
416
+ "grad_norm": 55.91850280761719,
417
+ "learning_rate": 1.4666666666666666e-05,
418
+ "loss": 0.6857,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.944444444444445,
423
+ "grad_norm": 71.79806518554688,
424
+ "learning_rate": 1.4512345679012347e-05,
425
+ "loss": 0.5738,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 7.0,
430
+ "eval_accuracy": 0.5835820895522388,
431
+ "eval_loss": 1.8012571334838867,
432
+ "eval_macro_f1": 0.5791712054768521,
433
+ "eval_runtime": 1.3618,
434
+ "eval_samples_per_second": 492.013,
435
+ "eval_steps_per_second": 30.843,
436
+ "eval_weighted_f1": 0.5794948646379797,
437
+ "step": 1260
438
+ },
439
+ {
440
+ "epoch": 7.083333333333333,
441
+ "grad_norm": 23.227781295776367,
442
+ "learning_rate": 1.4358024691358025e-05,
443
+ "loss": 0.347,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.222222222222222,
448
+ "grad_norm": 21.165639877319336,
449
+ "learning_rate": 1.4203703703703704e-05,
450
+ "loss": 0.364,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.361111111111111,
455
+ "grad_norm": 33.553565979003906,
456
+ "learning_rate": 1.4049382716049384e-05,
457
+ "loss": 0.3212,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.5,
462
+ "grad_norm": 19.577932357788086,
463
+ "learning_rate": 1.3895061728395063e-05,
464
+ "loss": 0.4314,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.638888888888889,
469
+ "grad_norm": 55.02348709106445,
470
+ "learning_rate": 1.3740740740740741e-05,
471
+ "loss": 0.4183,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.777777777777778,
476
+ "grad_norm": 16.349395751953125,
477
+ "learning_rate": 1.3586419753086422e-05,
478
+ "loss": 0.3183,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.916666666666667,
483
+ "grad_norm": 20.643468856811523,
484
+ "learning_rate": 1.34320987654321e-05,
485
+ "loss": 0.3395,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 8.0,
490
+ "eval_accuracy": 0.5656716417910448,
491
+ "eval_loss": 2.199185609817505,
492
+ "eval_macro_f1": 0.5622175675457486,
493
+ "eval_runtime": 1.4462,
494
+ "eval_samples_per_second": 463.269,
495
+ "eval_steps_per_second": 29.041,
496
+ "eval_weighted_f1": 0.562241399623224,
497
+ "step": 1440
498
+ },
499
+ {
500
+ "epoch": 8.055555555555555,
501
+ "grad_norm": 51.110843658447266,
502
+ "learning_rate": 1.3277777777777779e-05,
503
+ "loss": 0.3314,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.194444444444445,
508
+ "grad_norm": 127.76012420654297,
509
+ "learning_rate": 1.3123456790123458e-05,
510
+ "loss": 0.2597,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.333333333333334,
515
+ "grad_norm": 108.4692153930664,
516
+ "learning_rate": 1.2969135802469138e-05,
517
+ "loss": 0.3554,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.472222222222221,
522
+ "grad_norm": 40.320125579833984,
523
+ "learning_rate": 1.2814814814814815e-05,
524
+ "loss": 0.273,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.61111111111111,
529
+ "grad_norm": 24.809839248657227,
530
+ "learning_rate": 1.2660493827160493e-05,
531
+ "loss": 0.2295,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.75,
536
+ "grad_norm": 95.2496337890625,
537
+ "learning_rate": 1.2506172839506174e-05,
538
+ "loss": 0.3116,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.88888888888889,
543
+ "grad_norm": 99.75901794433594,
544
+ "learning_rate": 1.2351851851851852e-05,
545
+ "loss": 0.2636,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 9.0,
550
+ "eval_accuracy": 0.5865671641791045,
551
+ "eval_loss": 2.3661773204803467,
552
+ "eval_macro_f1": 0.5835810641427687,
553
+ "eval_runtime": 1.3414,
554
+ "eval_samples_per_second": 499.488,
555
+ "eval_steps_per_second": 31.311,
556
+ "eval_weighted_f1": 0.5841164452523472,
557
+ "step": 1620
558
+ }
559
+ ],
560
+ "logging_steps": 25,
561
+ "max_steps": 3600,
562
+ "num_input_tokens_seen": 0,
563
+ "num_train_epochs": 20,
564
+ "save_steps": 500,
565
+ "stateful_callbacks": {
566
+ "EarlyStoppingCallback": {
567
+ "args": {
568
+ "early_stopping_patience": 3,
569
+ "early_stopping_threshold": 0.0
570
+ },
571
+ "attributes": {
572
+ "early_stopping_patience_counter": 0
573
+ }
574
+ },
575
+ "TrainerControl": {
576
+ "args": {
577
+ "should_epoch_stop": false,
578
+ "should_evaluate": false,
579
+ "should_log": false,
580
+ "should_save": true,
581
+ "should_training_stop": false
582
+ },
583
+ "attributes": {}
584
+ }
585
+ },
586
+ "total_flos": 0.0,
587
+ "train_batch_size": 8,
588
+ "trial_name": null,
589
+ "trial_params": null
590
+ }
ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-2160/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-2160/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-2160/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-2160/trainer_state.json ADDED
@@ -0,0 +1,777 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1620,
3
+ "best_metric": 0.5835810641427687,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-05-s43-20260831T020453Z-1a169064/trainer/checkpoint-1620",
5
+ "epoch": 12.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2160,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1388888888888889,
14
+ "grad_norm": 29.263378143310547,
15
+ "learning_rate": 1.3333333333333334e-06,
16
+ "loss": 4.8502,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2777777777777778,
21
+ "grad_norm": 20.813480377197266,
22
+ "learning_rate": 2.7222222222222224e-06,
23
+ "loss": 4.4757,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4166666666666667,
28
+ "grad_norm": 36.471431732177734,
29
+ "learning_rate": 4.111111111111111e-06,
30
+ "loss": 4.5168,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5555555555555556,
35
+ "grad_norm": 30.926923751831055,
36
+ "learning_rate": 5.500000000000001e-06,
37
+ "loss": 4.4289,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6944444444444444,
42
+ "grad_norm": 27.58977508544922,
43
+ "learning_rate": 6.88888888888889e-06,
44
+ "loss": 4.4247,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8333333333333334,
49
+ "grad_norm": 34.77938461303711,
50
+ "learning_rate": 8.277777777777778e-06,
51
+ "loss": 4.3307,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9722222222222222,
56
+ "grad_norm": 32.43658447265625,
57
+ "learning_rate": 9.666666666666667e-06,
58
+ "loss": 4.378,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4014925373134328,
64
+ "eval_loss": 1.0966709852218628,
65
+ "eval_macro_f1": 0.34144070307099833,
66
+ "eval_runtime": 1.3826,
67
+ "eval_samples_per_second": 484.596,
68
+ "eval_steps_per_second": 30.378,
69
+ "eval_weighted_f1": 0.3363109803996705,
70
+ "step": 180
71
+ },
72
+ {
73
+ "epoch": 1.1111111111111112,
74
+ "grad_norm": 16.528186798095703,
75
+ "learning_rate": 1.1055555555555557e-05,
76
+ "loss": 4.2033,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.25,
81
+ "grad_norm": 17.700929641723633,
82
+ "learning_rate": 1.2444444444444446e-05,
83
+ "loss": 4.3294,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3888888888888888,
88
+ "grad_norm": 17.803096771240234,
89
+ "learning_rate": 1.3833333333333334e-05,
90
+ "loss": 4.2576,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.5277777777777777,
95
+ "grad_norm": 18.98823356628418,
96
+ "learning_rate": 1.5222222222222223e-05,
97
+ "loss": 4.2086,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6666666666666665,
102
+ "grad_norm": 20.693819046020508,
103
+ "learning_rate": 1.6611111111111113e-05,
104
+ "loss": 4.1248,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.8055555555555556,
109
+ "grad_norm": 28.918561935424805,
110
+ "learning_rate": 1.8e-05,
111
+ "loss": 4.2757,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9444444444444444,
116
+ "grad_norm": 19.435781478881836,
117
+ "learning_rate": 1.938888888888889e-05,
118
+ "loss": 4.1187,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.5477611940298508,
124
+ "eval_loss": 0.9657241106033325,
125
+ "eval_macro_f1": 0.5473055994383008,
126
+ "eval_runtime": 1.3516,
127
+ "eval_samples_per_second": 495.711,
128
+ "eval_steps_per_second": 31.074,
129
+ "eval_weighted_f1": 0.5466602803595011,
130
+ "step": 360
131
+ },
132
+ {
133
+ "epoch": 2.0833333333333335,
134
+ "grad_norm": 23.853336334228516,
135
+ "learning_rate": 1.991358024691358e-05,
136
+ "loss": 3.941,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.2222222222222223,
141
+ "grad_norm": 34.18367385864258,
142
+ "learning_rate": 1.975925925925926e-05,
143
+ "loss": 3.7539,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.361111111111111,
148
+ "grad_norm": 44.504329681396484,
149
+ "learning_rate": 1.960493827160494e-05,
150
+ "loss": 3.6522,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.5,
155
+ "grad_norm": 29.060413360595703,
156
+ "learning_rate": 1.945061728395062e-05,
157
+ "loss": 3.6989,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.638888888888889,
162
+ "grad_norm": 31.384469985961914,
163
+ "learning_rate": 1.92962962962963e-05,
164
+ "loss": 3.4029,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7777777777777777,
169
+ "grad_norm": 48.808494567871094,
170
+ "learning_rate": 1.914197530864198e-05,
171
+ "loss": 3.8388,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.9166666666666665,
176
+ "grad_norm": 19.36366844177246,
177
+ "learning_rate": 1.8987654320987657e-05,
178
+ "loss": 3.6407,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.5835820895522388,
184
+ "eval_loss": 0.900704026222229,
185
+ "eval_macro_f1": 0.5730823063833178,
186
+ "eval_runtime": 1.31,
187
+ "eval_samples_per_second": 511.455,
188
+ "eval_steps_per_second": 32.061,
189
+ "eval_weighted_f1": 0.5735315438383259,
190
+ "step": 540
191
+ },
192
+ {
193
+ "epoch": 3.0555555555555554,
194
+ "grad_norm": 24.26601791381836,
195
+ "learning_rate": 1.8833333333333335e-05,
196
+ "loss": 3.3326,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.1944444444444446,
201
+ "grad_norm": 61.2514533996582,
202
+ "learning_rate": 1.8679012345679013e-05,
203
+ "loss": 2.8269,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.3333333333333335,
208
+ "grad_norm": 47.86339569091797,
209
+ "learning_rate": 1.8524691358024694e-05,
210
+ "loss": 2.7367,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.4722222222222223,
215
+ "grad_norm": 38.8129768371582,
216
+ "learning_rate": 1.837037037037037e-05,
217
+ "loss": 2.7173,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.611111111111111,
222
+ "grad_norm": 45.2987174987793,
223
+ "learning_rate": 1.821604938271605e-05,
224
+ "loss": 2.7368,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.75,
229
+ "grad_norm": 37.727134704589844,
230
+ "learning_rate": 1.806172839506173e-05,
231
+ "loss": 2.7808,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.888888888888889,
236
+ "grad_norm": 47.086238861083984,
237
+ "learning_rate": 1.7907407407407408e-05,
238
+ "loss": 2.5216,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 4.0,
243
+ "eval_accuracy": 0.5507462686567164,
244
+ "eval_loss": 1.0473291873931885,
245
+ "eval_macro_f1": 0.5416214803491114,
246
+ "eval_runtime": 1.2777,
247
+ "eval_samples_per_second": 524.367,
248
+ "eval_steps_per_second": 32.871,
249
+ "eval_weighted_f1": 0.5430559342981475,
250
+ "step": 720
251
+ },
252
+ {
253
+ "epoch": 4.027777777777778,
254
+ "grad_norm": 39.34019088745117,
255
+ "learning_rate": 1.775308641975309e-05,
256
+ "loss": 2.6099,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.166666666666667,
261
+ "grad_norm": 34.96302032470703,
262
+ "learning_rate": 1.7598765432098767e-05,
263
+ "loss": 1.8181,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.305555555555555,
268
+ "grad_norm": 46.8112907409668,
269
+ "learning_rate": 1.7444444444444448e-05,
270
+ "loss": 1.743,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.444444444444445,
275
+ "grad_norm": 49.2983512878418,
276
+ "learning_rate": 1.7290123456790125e-05,
277
+ "loss": 1.6882,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.583333333333333,
282
+ "grad_norm": 34.680213928222656,
283
+ "learning_rate": 1.7135802469135806e-05,
284
+ "loss": 1.8312,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.722222222222222,
289
+ "grad_norm": 50.23997497558594,
290
+ "learning_rate": 1.6981481481481484e-05,
291
+ "loss": 1.9134,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.861111111111111,
296
+ "grad_norm": 32.60227584838867,
297
+ "learning_rate": 1.6827160493827162e-05,
298
+ "loss": 1.7515,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 5.0,
303
+ "grad_norm": 61.78374481201172,
304
+ "learning_rate": 1.667283950617284e-05,
305
+ "loss": 1.7827,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.582089552238806,
311
+ "eval_loss": 1.166916012763977,
312
+ "eval_macro_f1": 0.5777411529460815,
313
+ "eval_runtime": 1.2641,
314
+ "eval_samples_per_second": 530.041,
315
+ "eval_steps_per_second": 33.226,
316
+ "eval_weighted_f1": 0.5779446555827403,
317
+ "step": 900
318
+ },
319
+ {
320
+ "epoch": 5.138888888888889,
321
+ "grad_norm": 128.76226806640625,
322
+ "learning_rate": 1.651851851851852e-05,
323
+ "loss": 1.1251,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.277777777777778,
328
+ "grad_norm": 78.46173095703125,
329
+ "learning_rate": 1.6364197530864198e-05,
330
+ "loss": 1.0684,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.416666666666667,
335
+ "grad_norm": 42.87739944458008,
336
+ "learning_rate": 1.6209876543209876e-05,
337
+ "loss": 0.8983,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.555555555555555,
342
+ "grad_norm": 78.20256042480469,
343
+ "learning_rate": 1.6055555555555557e-05,
344
+ "loss": 0.885,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.694444444444445,
349
+ "grad_norm": 60.07840347290039,
350
+ "learning_rate": 1.5901234567901235e-05,
351
+ "loss": 1.0987,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.833333333333333,
356
+ "grad_norm": 69.35907745361328,
357
+ "learning_rate": 1.5746913580246916e-05,
358
+ "loss": 1.11,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.972222222222222,
363
+ "grad_norm": 90.80661010742188,
364
+ "learning_rate": 1.5592592592592593e-05,
365
+ "loss": 1.1472,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.5686567164179105,
371
+ "eval_loss": 1.5736793279647827,
372
+ "eval_macro_f1": 0.5635663296380078,
373
+ "eval_runtime": 1.2472,
374
+ "eval_samples_per_second": 537.197,
375
+ "eval_steps_per_second": 33.675,
376
+ "eval_weighted_f1": 0.5646891761432958,
377
+ "step": 1080
378
+ },
379
+ {
380
+ "epoch": 6.111111111111111,
381
+ "grad_norm": 34.897483825683594,
382
+ "learning_rate": 1.5438271604938274e-05,
383
+ "loss": 0.6438,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.25,
388
+ "grad_norm": 49.88667678833008,
389
+ "learning_rate": 1.5283950617283952e-05,
390
+ "loss": 0.7807,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.388888888888889,
395
+ "grad_norm": 20.57355308532715,
396
+ "learning_rate": 1.5129629629629632e-05,
397
+ "loss": 0.8069,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.527777777777778,
402
+ "grad_norm": 23.29080581665039,
403
+ "learning_rate": 1.4975308641975311e-05,
404
+ "loss": 0.6768,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.666666666666667,
409
+ "grad_norm": 78.04934692382812,
410
+ "learning_rate": 1.4820987654320989e-05,
411
+ "loss": 0.6018,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.805555555555555,
416
+ "grad_norm": 55.91850280761719,
417
+ "learning_rate": 1.4666666666666666e-05,
418
+ "loss": 0.6857,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.944444444444445,
423
+ "grad_norm": 71.79806518554688,
424
+ "learning_rate": 1.4512345679012347e-05,
425
+ "loss": 0.5738,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 7.0,
430
+ "eval_accuracy": 0.5835820895522388,
431
+ "eval_loss": 1.8012571334838867,
432
+ "eval_macro_f1": 0.5791712054768521,
433
+ "eval_runtime": 1.3618,
434
+ "eval_samples_per_second": 492.013,
435
+ "eval_steps_per_second": 30.843,
436
+ "eval_weighted_f1": 0.5794948646379797,
437
+ "step": 1260
438
+ },
439
+ {
440
+ "epoch": 7.083333333333333,
441
+ "grad_norm": 23.227781295776367,
442
+ "learning_rate": 1.4358024691358025e-05,
443
+ "loss": 0.347,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.222222222222222,
448
+ "grad_norm": 21.165639877319336,
449
+ "learning_rate": 1.4203703703703704e-05,
450
+ "loss": 0.364,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.361111111111111,
455
+ "grad_norm": 33.553565979003906,
456
+ "learning_rate": 1.4049382716049384e-05,
457
+ "loss": 0.3212,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.5,
462
+ "grad_norm": 19.577932357788086,
463
+ "learning_rate": 1.3895061728395063e-05,
464
+ "loss": 0.4314,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.638888888888889,
469
+ "grad_norm": 55.02348709106445,
470
+ "learning_rate": 1.3740740740740741e-05,
471
+ "loss": 0.4183,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.777777777777778,
476
+ "grad_norm": 16.349395751953125,
477
+ "learning_rate": 1.3586419753086422e-05,
478
+ "loss": 0.3183,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.916666666666667,
483
+ "grad_norm": 20.643468856811523,
484
+ "learning_rate": 1.34320987654321e-05,
485
+ "loss": 0.3395,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 8.0,
490
+ "eval_accuracy": 0.5656716417910448,
491
+ "eval_loss": 2.199185609817505,
492
+ "eval_macro_f1": 0.5622175675457486,
493
+ "eval_runtime": 1.4462,
494
+ "eval_samples_per_second": 463.269,
495
+ "eval_steps_per_second": 29.041,
496
+ "eval_weighted_f1": 0.562241399623224,
497
+ "step": 1440
498
+ },
499
+ {
500
+ "epoch": 8.055555555555555,
501
+ "grad_norm": 51.110843658447266,
502
+ "learning_rate": 1.3277777777777779e-05,
503
+ "loss": 0.3314,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.194444444444445,
508
+ "grad_norm": 127.76012420654297,
509
+ "learning_rate": 1.3123456790123458e-05,
510
+ "loss": 0.2597,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.333333333333334,
515
+ "grad_norm": 108.4692153930664,
516
+ "learning_rate": 1.2969135802469138e-05,
517
+ "loss": 0.3554,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.472222222222221,
522
+ "grad_norm": 40.320125579833984,
523
+ "learning_rate": 1.2814814814814815e-05,
524
+ "loss": 0.273,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.61111111111111,
529
+ "grad_norm": 24.809839248657227,
530
+ "learning_rate": 1.2660493827160493e-05,
531
+ "loss": 0.2295,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.75,
536
+ "grad_norm": 95.2496337890625,
537
+ "learning_rate": 1.2506172839506174e-05,
538
+ "loss": 0.3116,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.88888888888889,
543
+ "grad_norm": 99.75901794433594,
544
+ "learning_rate": 1.2351851851851852e-05,
545
+ "loss": 0.2636,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 9.0,
550
+ "eval_accuracy": 0.5865671641791045,
551
+ "eval_loss": 2.3661773204803467,
552
+ "eval_macro_f1": 0.5835810641427687,
553
+ "eval_runtime": 1.3414,
554
+ "eval_samples_per_second": 499.488,
555
+ "eval_steps_per_second": 31.311,
556
+ "eval_weighted_f1": 0.5841164452523472,
557
+ "step": 1620
558
+ },
559
+ {
560
+ "epoch": 9.027777777777779,
561
+ "grad_norm": 8.73697280883789,
562
+ "learning_rate": 1.2197530864197531e-05,
563
+ "loss": 0.2855,
564
+ "step": 1625
565
+ },
566
+ {
567
+ "epoch": 9.166666666666666,
568
+ "grad_norm": 146.2117156982422,
569
+ "learning_rate": 1.204320987654321e-05,
570
+ "loss": 0.1948,
571
+ "step": 1650
572
+ },
573
+ {
574
+ "epoch": 9.305555555555555,
575
+ "grad_norm": 0.29749780893325806,
576
+ "learning_rate": 1.188888888888889e-05,
577
+ "loss": 0.1473,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.444444444444445,
582
+ "grad_norm": 167.3782501220703,
583
+ "learning_rate": 1.1734567901234568e-05,
584
+ "loss": 0.2275,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.583333333333334,
589
+ "grad_norm": 1.9692506790161133,
590
+ "learning_rate": 1.1580246913580249e-05,
591
+ "loss": 0.1645,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.722222222222221,
596
+ "grad_norm": 40.45298385620117,
597
+ "learning_rate": 1.1425925925925927e-05,
598
+ "loss": 0.155,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 9.86111111111111,
603
+ "grad_norm": 28.839811325073242,
604
+ "learning_rate": 1.1271604938271606e-05,
605
+ "loss": 0.2302,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 10.0,
610
+ "grad_norm": 97.46849822998047,
611
+ "learning_rate": 1.1117283950617285e-05,
612
+ "loss": 0.2435,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 10.0,
617
+ "eval_accuracy": 0.5761194029850746,
618
+ "eval_loss": 2.9205551147460938,
619
+ "eval_macro_f1": 0.5780603163067671,
620
+ "eval_runtime": 1.4485,
621
+ "eval_samples_per_second": 462.552,
622
+ "eval_steps_per_second": 28.996,
623
+ "eval_weighted_f1": 0.5785550502673846,
624
+ "step": 1800
625
+ },
626
+ {
627
+ "epoch": 10.13888888888889,
628
+ "grad_norm": 99.0831069946289,
629
+ "learning_rate": 1.0962962962962965e-05,
630
+ "loss": 0.2061,
631
+ "step": 1825
632
+ },
633
+ {
634
+ "epoch": 10.277777777777779,
635
+ "grad_norm": 67.60861206054688,
636
+ "learning_rate": 1.0808641975308642e-05,
637
+ "loss": 0.1757,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.416666666666666,
642
+ "grad_norm": 3.9673125743865967,
643
+ "learning_rate": 1.065432098765432e-05,
644
+ "loss": 0.08,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.555555555555555,
649
+ "grad_norm": 48.27328872680664,
650
+ "learning_rate": 1.0500000000000001e-05,
651
+ "loss": 0.1772,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.694444444444445,
656
+ "grad_norm": 73.0173110961914,
657
+ "learning_rate": 1.0345679012345679e-05,
658
+ "loss": 0.0884,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 10.833333333333334,
663
+ "grad_norm": 8.935346603393555,
664
+ "learning_rate": 1.0191358024691358e-05,
665
+ "loss": 0.1199,
666
+ "step": 1950
667
+ },
668
+ {
669
+ "epoch": 10.972222222222221,
670
+ "grad_norm": 165.04708862304688,
671
+ "learning_rate": 1.003703703703704e-05,
672
+ "loss": 0.3493,
673
+ "step": 1975
674
+ },
675
+ {
676
+ "epoch": 11.0,
677
+ "eval_accuracy": 0.5686567164179105,
678
+ "eval_loss": 3.40043044090271,
679
+ "eval_macro_f1": 0.5651251502962451,
680
+ "eval_runtime": 1.373,
681
+ "eval_samples_per_second": 487.969,
682
+ "eval_steps_per_second": 30.589,
683
+ "eval_weighted_f1": 0.5654850254794346,
684
+ "step": 1980
685
+ },
686
+ {
687
+ "epoch": 11.11111111111111,
688
+ "grad_norm": 54.66333770751953,
689
+ "learning_rate": 9.882716049382717e-06,
690
+ "loss": 0.1321,
691
+ "step": 2000
692
+ },
693
+ {
694
+ "epoch": 11.25,
695
+ "grad_norm": 0.24743914604187012,
696
+ "learning_rate": 9.728395061728396e-06,
697
+ "loss": 0.0864,
698
+ "step": 2025
699
+ },
700
+ {
701
+ "epoch": 11.38888888888889,
702
+ "grad_norm": 18.838035583496094,
703
+ "learning_rate": 9.574074074074074e-06,
704
+ "loss": 0.1644,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 11.527777777777779,
709
+ "grad_norm": 5.2751641273498535,
710
+ "learning_rate": 9.419753086419753e-06,
711
+ "loss": 0.0884,
712
+ "step": 2075
713
+ },
714
+ {
715
+ "epoch": 11.666666666666666,
716
+ "grad_norm": 0.5860305428504944,
717
+ "learning_rate": 9.265432098765433e-06,
718
+ "loss": 0.0985,
719
+ "step": 2100
720
+ },
721
+ {
722
+ "epoch": 11.805555555555555,
723
+ "grad_norm": 94.5871810913086,
724
+ "learning_rate": 9.111111111111112e-06,
725
+ "loss": 0.1289,
726
+ "step": 2125
727
+ },
728
+ {
729
+ "epoch": 11.944444444444445,
730
+ "grad_norm": 3.28253436088562,
731
+ "learning_rate": 8.956790123456792e-06,
732
+ "loss": 0.2197,
733
+ "step": 2150
734
+ },
735
+ {
736
+ "epoch": 12.0,
737
+ "eval_accuracy": 0.5611940298507463,
738
+ "eval_loss": 3.476285219192505,
739
+ "eval_macro_f1": 0.5630112622986339,
740
+ "eval_runtime": 1.3334,
741
+ "eval_samples_per_second": 502.487,
742
+ "eval_steps_per_second": 31.499,
743
+ "eval_weighted_f1": 0.5627683853010238,
744
+ "step": 2160
745
+ }
746
+ ],
747
+ "logging_steps": 25,
748
+ "max_steps": 3600,
749
+ "num_input_tokens_seen": 0,
750
+ "num_train_epochs": 20,
751
+ "save_steps": 500,
752
+ "stateful_callbacks": {
753
+ "EarlyStoppingCallback": {
754
+ "args": {
755
+ "early_stopping_patience": 3,
756
+ "early_stopping_threshold": 0.0
757
+ },
758
+ "attributes": {
759
+ "early_stopping_patience_counter": 3
760
+ }
761
+ },
762
+ "TrainerControl": {
763
+ "args": {
764
+ "should_epoch_stop": false,
765
+ "should_evaluate": false,
766
+ "should_log": false,
767
+ "should_save": true,
768
+ "should_training_stop": true
769
+ },
770
+ "attributes": {}
771
+ }
772
+ },
773
+ "total_flos": 0.0,
774
+ "train_batch_size": 8,
775
+ "trial_name": null,
776
+ "trial_params": null
777
+ }
ood-vifactcheck-visobert-source-source-05-s44-20260831T033516Z-59e46264/trainer/checkpoint-1080/trainer_state.json ADDED
@@ -0,0 +1,410 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1080,
3
+ "best_metric": 0.6042655440726321,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-05-s44-20260831T033516Z-59e46264/trainer/checkpoint-1080",
5
+ "epoch": 6.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1080,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.1388888888888889,
14
+ "grad_norm": 29.567710876464844,
15
+ "learning_rate": 1.3333333333333334e-06,
16
+ "loss": 4.6955,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2777777777777778,
21
+ "grad_norm": 25.813806533813477,
22
+ "learning_rate": 2.7222222222222224e-06,
23
+ "loss": 4.5598,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4166666666666667,
28
+ "grad_norm": 19.19781494140625,
29
+ "learning_rate": 4.111111111111111e-06,
30
+ "loss": 4.5132,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5555555555555556,
35
+ "grad_norm": 22.473264694213867,
36
+ "learning_rate": 5.500000000000001e-06,
37
+ "loss": 4.4759,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6944444444444444,
42
+ "grad_norm": 20.011192321777344,
43
+ "learning_rate": 6.88888888888889e-06,
44
+ "loss": 4.4339,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8333333333333334,
49
+ "grad_norm": 19.52066993713379,
50
+ "learning_rate": 8.277777777777778e-06,
51
+ "loss": 4.4117,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9722222222222222,
56
+ "grad_norm": 27.693166732788086,
57
+ "learning_rate": 9.666666666666667e-06,
58
+ "loss": 4.3486,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4701492537313433,
64
+ "eval_loss": 1.0288771390914917,
65
+ "eval_macro_f1": 0.4700577514479254,
66
+ "eval_runtime": 1.366,
67
+ "eval_samples_per_second": 490.491,
68
+ "eval_steps_per_second": 30.747,
69
+ "eval_weighted_f1": 0.47051479338607033,
70
+ "step": 180
71
+ },
72
+ {
73
+ "epoch": 1.1111111111111112,
74
+ "grad_norm": 25.464757919311523,
75
+ "learning_rate": 1.1055555555555557e-05,
76
+ "loss": 4.1894,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.25,
81
+ "grad_norm": 27.773420333862305,
82
+ "learning_rate": 1.2444444444444446e-05,
83
+ "loss": 4.274,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3888888888888888,
88
+ "grad_norm": 15.416708946228027,
89
+ "learning_rate": 1.3833333333333334e-05,
90
+ "loss": 4.3267,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.5277777777777777,
95
+ "grad_norm": 21.944576263427734,
96
+ "learning_rate": 1.5222222222222223e-05,
97
+ "loss": 4.2147,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6666666666666665,
102
+ "grad_norm": 36.10844802856445,
103
+ "learning_rate": 1.6611111111111113e-05,
104
+ "loss": 4.2238,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.8055555555555556,
109
+ "grad_norm": 27.377803802490234,
110
+ "learning_rate": 1.8e-05,
111
+ "loss": 4.3148,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.9444444444444444,
116
+ "grad_norm": 37.73032760620117,
117
+ "learning_rate": 1.938888888888889e-05,
118
+ "loss": 4.2655,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.46119402985074626,
124
+ "eval_loss": 1.0248003005981445,
125
+ "eval_macro_f1": 0.38143792493571055,
126
+ "eval_runtime": 1.3128,
127
+ "eval_samples_per_second": 510.377,
128
+ "eval_steps_per_second": 31.994,
129
+ "eval_weighted_f1": 0.38331875889622785,
130
+ "step": 360
131
+ },
132
+ {
133
+ "epoch": 2.0833333333333335,
134
+ "grad_norm": 26.852705001831055,
135
+ "learning_rate": 1.991358024691358e-05,
136
+ "loss": 4.0213,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.2222222222222223,
141
+ "grad_norm": 29.27025032043457,
142
+ "learning_rate": 1.975925925925926e-05,
143
+ "loss": 3.8909,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.361111111111111,
148
+ "grad_norm": 28.35458755493164,
149
+ "learning_rate": 1.960493827160494e-05,
150
+ "loss": 3.7029,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.5,
155
+ "grad_norm": 43.68309020996094,
156
+ "learning_rate": 1.945061728395062e-05,
157
+ "loss": 3.7663,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.638888888888889,
162
+ "grad_norm": 55.959068298339844,
163
+ "learning_rate": 1.92962962962963e-05,
164
+ "loss": 3.6489,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7777777777777777,
169
+ "grad_norm": 21.751813888549805,
170
+ "learning_rate": 1.914197530864198e-05,
171
+ "loss": 3.7585,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.9166666666666665,
176
+ "grad_norm": 24.25555992126465,
177
+ "learning_rate": 1.8987654320987657e-05,
178
+ "loss": 3.4667,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.5597014925373134,
184
+ "eval_loss": 0.9681290984153748,
185
+ "eval_macro_f1": 0.5515693772436397,
186
+ "eval_runtime": 1.3381,
187
+ "eval_samples_per_second": 500.702,
188
+ "eval_steps_per_second": 31.387,
189
+ "eval_weighted_f1": 0.5524811982696164,
190
+ "step": 540
191
+ },
192
+ {
193
+ "epoch": 3.0555555555555554,
194
+ "grad_norm": 26.375534057617188,
195
+ "learning_rate": 1.8833333333333335e-05,
196
+ "loss": 3.305,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.1944444444444446,
201
+ "grad_norm": 38.70694351196289,
202
+ "learning_rate": 1.8679012345679013e-05,
203
+ "loss": 2.7149,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.3333333333333335,
208
+ "grad_norm": 34.07916259765625,
209
+ "learning_rate": 1.8524691358024694e-05,
210
+ "loss": 2.8397,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.4722222222222223,
215
+ "grad_norm": 70.33634185791016,
216
+ "learning_rate": 1.837037037037037e-05,
217
+ "loss": 2.8228,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.611111111111111,
222
+ "grad_norm": 53.93187713623047,
223
+ "learning_rate": 1.821604938271605e-05,
224
+ "loss": 2.9202,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.75,
229
+ "grad_norm": 34.200069427490234,
230
+ "learning_rate": 1.806172839506173e-05,
231
+ "loss": 2.9978,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.888888888888889,
236
+ "grad_norm": 44.9887809753418,
237
+ "learning_rate": 1.7907407407407408e-05,
238
+ "loss": 2.8542,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 4.0,
243
+ "eval_accuracy": 0.5492537313432836,
244
+ "eval_loss": 1.037038803100586,
245
+ "eval_macro_f1": 0.5293592289460588,
246
+ "eval_runtime": 1.5042,
247
+ "eval_samples_per_second": 445.433,
248
+ "eval_steps_per_second": 27.923,
249
+ "eval_weighted_f1": 0.5291349526787942,
250
+ "step": 720
251
+ },
252
+ {
253
+ "epoch": 4.027777777777778,
254
+ "grad_norm": 48.614471435546875,
255
+ "learning_rate": 1.775308641975309e-05,
256
+ "loss": 2.5572,
257
+ "step": 725
258
+ },
259
+ {
260
+ "epoch": 4.166666666666667,
261
+ "grad_norm": 40.57429122924805,
262
+ "learning_rate": 1.7598765432098767e-05,
263
+ "loss": 2.0482,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.305555555555555,
268
+ "grad_norm": 54.28365707397461,
269
+ "learning_rate": 1.7444444444444448e-05,
270
+ "loss": 1.8784,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.444444444444445,
275
+ "grad_norm": 30.25566864013672,
276
+ "learning_rate": 1.7290123456790125e-05,
277
+ "loss": 1.8117,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.583333333333333,
282
+ "grad_norm": 51.76329040527344,
283
+ "learning_rate": 1.7135802469135806e-05,
284
+ "loss": 1.7417,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.722222222222222,
289
+ "grad_norm": 43.56016159057617,
290
+ "learning_rate": 1.6981481481481484e-05,
291
+ "loss": 1.9949,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.861111111111111,
296
+ "grad_norm": 46.071571350097656,
297
+ "learning_rate": 1.6827160493827162e-05,
298
+ "loss": 1.8901,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 5.0,
303
+ "grad_norm": 35.79936599731445,
304
+ "learning_rate": 1.667283950617284e-05,
305
+ "loss": 1.9183,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.5432835820895522,
311
+ "eval_loss": 1.3129158020019531,
312
+ "eval_macro_f1": 0.5358274256128672,
313
+ "eval_runtime": 1.3428,
314
+ "eval_samples_per_second": 498.952,
315
+ "eval_steps_per_second": 31.278,
316
+ "eval_weighted_f1": 0.5374324647845226,
317
+ "step": 900
318
+ },
319
+ {
320
+ "epoch": 5.138888888888889,
321
+ "grad_norm": 29.806840896606445,
322
+ "learning_rate": 1.651851851851852e-05,
323
+ "loss": 1.15,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.277777777777778,
328
+ "grad_norm": 62.33220672607422,
329
+ "learning_rate": 1.6364197530864198e-05,
330
+ "loss": 1.0909,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.416666666666667,
335
+ "grad_norm": 47.7425422668457,
336
+ "learning_rate": 1.6209876543209876e-05,
337
+ "loss": 1.0777,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.555555555555555,
342
+ "grad_norm": 47.028865814208984,
343
+ "learning_rate": 1.6055555555555557e-05,
344
+ "loss": 1.0042,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.694444444444445,
349
+ "grad_norm": 63.89030456542969,
350
+ "learning_rate": 1.5901234567901235e-05,
351
+ "loss": 1.2174,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.833333333333333,
356
+ "grad_norm": 126.78224182128906,
357
+ "learning_rate": 1.5746913580246916e-05,
358
+ "loss": 1.2743,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.972222222222222,
363
+ "grad_norm": 52.99421310424805,
364
+ "learning_rate": 1.5592592592592593e-05,
365
+ "loss": 1.2873,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.6044776119402985,
371
+ "eval_loss": 1.2865700721740723,
372
+ "eval_macro_f1": 0.6042655440726321,
373
+ "eval_runtime": 1.3332,
374
+ "eval_samples_per_second": 502.534,
375
+ "eval_steps_per_second": 31.502,
376
+ "eval_weighted_f1": 0.603304163070053,
377
+ "step": 1080
378
+ }
379
+ ],
380
+ "logging_steps": 25,
381
+ "max_steps": 3600,
382
+ "num_input_tokens_seen": 0,
383
+ "num_train_epochs": 20,
384
+ "save_steps": 500,
385
+ "stateful_callbacks": {
386
+ "EarlyStoppingCallback": {
387
+ "args": {
388
+ "early_stopping_patience": 3,
389
+ "early_stopping_threshold": 0.0
390
+ },
391
+ "attributes": {
392
+ "early_stopping_patience_counter": 0
393
+ }
394
+ },
395
+ "TrainerControl": {
396
+ "args": {
397
+ "should_epoch_stop": false,
398
+ "should_evaluate": false,
399
+ "should_log": false,
400
+ "should_save": true,
401
+ "should_training_stop": false
402
+ },
403
+ "attributes": {}
404
+ }
405
+ },
406
+ "total_flos": 0.0,
407
+ "train_batch_size": 8,
408
+ "trial_name": null,
409
+ "trial_params": null
410
+ }
ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035/trainer_state.json ADDED
@@ -0,0 +1,731 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 2035,
3
+ "best_metric": 0.6008961707884494,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-06-s43-20260831T021725Z-1a169064/trainer/checkpoint-2035",
5
+ "epoch": 11.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2035,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13550135501355012,
14
+ "grad_norm": 23.644420623779297,
15
+ "learning_rate": 1.2972972972972974e-06,
16
+ "loss": 4.8413,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27100271002710025,
21
+ "grad_norm": 25.800350189208984,
22
+ "learning_rate": 2.648648648648649e-06,
23
+ "loss": 4.4473,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4065040650406504,
28
+ "grad_norm": 19.654708862304688,
29
+ "learning_rate": 4.000000000000001e-06,
30
+ "loss": 4.4754,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5420054200542005,
35
+ "grad_norm": 27.55068588256836,
36
+ "learning_rate": 5.351351351351351e-06,
37
+ "loss": 4.3814,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6775067750677507,
42
+ "grad_norm": 18.95235824584961,
43
+ "learning_rate": 6.702702702702704e-06,
44
+ "loss": 4.4142,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8130081300813008,
49
+ "grad_norm": 32.0383415222168,
50
+ "learning_rate": 8.054054054054055e-06,
51
+ "loss": 4.3835,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.948509485094851,
56
+ "grad_norm": 27.09031867980957,
57
+ "learning_rate": 9.405405405405407e-06,
58
+ "loss": 4.3894,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4051490514905149,
64
+ "eval_loss": 1.0821291208267212,
65
+ "eval_macro_f1": 0.3925480841139686,
66
+ "eval_runtime": 1.4265,
67
+ "eval_samples_per_second": 517.364,
68
+ "eval_steps_per_second": 32.949,
69
+ "eval_weighted_f1": 0.38991704668114896,
70
+ "step": 185
71
+ },
72
+ {
73
+ "epoch": 1.08130081300813,
74
+ "grad_norm": 37.34614944458008,
75
+ "learning_rate": 1.0756756756756757e-05,
76
+ "loss": 4.1795,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2168021680216803,
81
+ "grad_norm": 22.745908737182617,
82
+ "learning_rate": 1.210810810810811e-05,
83
+ "loss": 4.2444,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3523035230352303,
88
+ "grad_norm": 20.919368743896484,
89
+ "learning_rate": 1.3459459459459461e-05,
90
+ "loss": 4.2387,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4878048780487805,
95
+ "grad_norm": 52.95161437988281,
96
+ "learning_rate": 1.4810810810810811e-05,
97
+ "loss": 4.2627,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6233062330623307,
102
+ "grad_norm": 27.904052734375,
103
+ "learning_rate": 1.6162162162162163e-05,
104
+ "loss": 4.0913,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.758807588075881,
109
+ "grad_norm": 29.80078125,
110
+ "learning_rate": 1.7513513513513517e-05,
111
+ "loss": 4.1792,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.8943089430894309,
116
+ "grad_norm": 22.177852630615234,
117
+ "learning_rate": 1.8864864864864866e-05,
118
+ "loss": 4.2457,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.4959349593495935,
124
+ "eval_loss": 1.0042661428451538,
125
+ "eval_macro_f1": 0.49460544237685083,
126
+ "eval_runtime": 1.7122,
127
+ "eval_samples_per_second": 431.03,
128
+ "eval_steps_per_second": 27.45,
129
+ "eval_weighted_f1": 0.4934151033840644,
130
+ "step": 370
131
+ },
132
+ {
133
+ "epoch": 2.02710027100271,
134
+ "grad_norm": 20.84690284729004,
135
+ "learning_rate": 1.9975975975975978e-05,
136
+ "loss": 3.9331,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.16260162601626,
141
+ "grad_norm": 36.183631896972656,
142
+ "learning_rate": 1.9825825825825828e-05,
143
+ "loss": 3.7031,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2981029810298104,
148
+ "grad_norm": 42.469383239746094,
149
+ "learning_rate": 1.967567567567568e-05,
150
+ "loss": 3.5926,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.4336043360433606,
155
+ "grad_norm": 58.60617446899414,
156
+ "learning_rate": 1.9525525525525525e-05,
157
+ "loss": 3.6548,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.569105691056911,
162
+ "grad_norm": 35.66259765625,
163
+ "learning_rate": 1.9375375375375376e-05,
164
+ "loss": 3.7605,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7046070460704605,
169
+ "grad_norm": 34.129878997802734,
170
+ "learning_rate": 1.9225225225225226e-05,
171
+ "loss": 3.6567,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8401084010840107,
176
+ "grad_norm": 28.44754409790039,
177
+ "learning_rate": 1.9075075075075076e-05,
178
+ "loss": 3.5784,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.975609756097561,
183
+ "grad_norm": 21.952425003051758,
184
+ "learning_rate": 1.8924924924924926e-05,
185
+ "loss": 3.6628,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.5271002710027101,
191
+ "eval_loss": 0.9627034068107605,
192
+ "eval_macro_f1": 0.5204525414299935,
193
+ "eval_runtime": 1.477,
194
+ "eval_samples_per_second": 499.675,
195
+ "eval_steps_per_second": 31.822,
196
+ "eval_weighted_f1": 0.5175579336276049,
197
+ "step": 555
198
+ },
199
+ {
200
+ "epoch": 3.1084010840108403,
201
+ "grad_norm": 28.471473693847656,
202
+ "learning_rate": 1.8774774774774776e-05,
203
+ "loss": 2.8768,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2439024390243905,
208
+ "grad_norm": 29.604429244995117,
209
+ "learning_rate": 1.8624624624624627e-05,
210
+ "loss": 2.8584,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.3794037940379402,
215
+ "grad_norm": 43.4071044921875,
216
+ "learning_rate": 1.8474474474474477e-05,
217
+ "loss": 2.5758,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5149051490514904,
222
+ "grad_norm": 47.71430587768555,
223
+ "learning_rate": 1.8324324324324324e-05,
224
+ "loss": 2.8767,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6504065040650406,
229
+ "grad_norm": 37.3945198059082,
230
+ "learning_rate": 1.8174174174174174e-05,
231
+ "loss": 2.6165,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.785907859078591,
236
+ "grad_norm": 27.95343589782715,
237
+ "learning_rate": 1.8024024024024024e-05,
238
+ "loss": 2.7598,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.921409214092141,
243
+ "grad_norm": 41.22407913208008,
244
+ "learning_rate": 1.7873873873873874e-05,
245
+ "loss": 2.8942,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.5772357723577236,
251
+ "eval_loss": 1.0029407739639282,
252
+ "eval_macro_f1": 0.5682174009734623,
253
+ "eval_runtime": 1.4061,
254
+ "eval_samples_per_second": 524.86,
255
+ "eval_steps_per_second": 33.426,
256
+ "eval_weighted_f1": 0.5694122373702232,
257
+ "step": 740
258
+ },
259
+ {
260
+ "epoch": 4.05420054200542,
261
+ "grad_norm": 26.70307731628418,
262
+ "learning_rate": 1.7723723723723724e-05,
263
+ "loss": 2.57,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.1897018970189706,
268
+ "grad_norm": 41.035099029541016,
269
+ "learning_rate": 1.7573573573573575e-05,
270
+ "loss": 1.7102,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.32520325203252,
275
+ "grad_norm": 47.82585525512695,
276
+ "learning_rate": 1.7423423423423425e-05,
277
+ "loss": 1.869,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.46070460704607,
282
+ "grad_norm": 53.93787384033203,
283
+ "learning_rate": 1.7273273273273275e-05,
284
+ "loss": 1.8035,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.596205962059621,
289
+ "grad_norm": 27.55320930480957,
290
+ "learning_rate": 1.7123123123123125e-05,
291
+ "loss": 1.7709,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.7317073170731705,
296
+ "grad_norm": 65.54625701904297,
297
+ "learning_rate": 1.6972972972972975e-05,
298
+ "loss": 1.7673,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.867208672086721,
303
+ "grad_norm": 66.76729583740234,
304
+ "learning_rate": 1.6822822822822826e-05,
305
+ "loss": 1.7912,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "grad_norm": 109.89456939697266,
311
+ "learning_rate": 1.6672672672672673e-05,
312
+ "loss": 1.9093,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 5.0,
317
+ "eval_accuracy": 0.5636856368563685,
318
+ "eval_loss": 1.1756083965301514,
319
+ "eval_macro_f1": 0.5673510528846676,
320
+ "eval_runtime": 1.3549,
321
+ "eval_samples_per_second": 544.689,
322
+ "eval_steps_per_second": 34.689,
323
+ "eval_weighted_f1": 0.5668583332524822,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.13550135501355,
328
+ "grad_norm": 66.03388214111328,
329
+ "learning_rate": 1.6522522522522523e-05,
330
+ "loss": 0.992,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.2710027100271,
335
+ "grad_norm": 67.87275695800781,
336
+ "learning_rate": 1.6372372372372373e-05,
337
+ "loss": 0.951,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.40650406504065,
342
+ "grad_norm": 116.9877700805664,
343
+ "learning_rate": 1.6222222222222223e-05,
344
+ "loss": 1.1113,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.542005420054201,
349
+ "grad_norm": 69.57601165771484,
350
+ "learning_rate": 1.6072072072072073e-05,
351
+ "loss": 1.305,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.677506775067751,
356
+ "grad_norm": 57.67194747924805,
357
+ "learning_rate": 1.5921921921921924e-05,
358
+ "loss": 1.5146,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.8130081300813,
363
+ "grad_norm": 26.43688201904297,
364
+ "learning_rate": 1.5771771771771774e-05,
365
+ "loss": 1.0235,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.948509485094851,
370
+ "grad_norm": 42.5482063293457,
371
+ "learning_rate": 1.5621621621621624e-05,
372
+ "loss": 1.0116,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 6.0,
377
+ "eval_accuracy": 0.55420054200542,
378
+ "eval_loss": 1.4322667121887207,
379
+ "eval_macro_f1": 0.5567958398020552,
380
+ "eval_runtime": 1.4799,
381
+ "eval_samples_per_second": 498.668,
382
+ "eval_steps_per_second": 31.758,
383
+ "eval_weighted_f1": 0.5555285916479377,
384
+ "step": 1110
385
+ },
386
+ {
387
+ "epoch": 6.08130081300813,
388
+ "grad_norm": 24.943260192871094,
389
+ "learning_rate": 1.5471471471471474e-05,
390
+ "loss": 0.6804,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.2168021680216805,
395
+ "grad_norm": 57.205467224121094,
396
+ "learning_rate": 1.5321321321321324e-05,
397
+ "loss": 0.6917,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.35230352303523,
402
+ "grad_norm": 60.98979568481445,
403
+ "learning_rate": 1.5171171171171173e-05,
404
+ "loss": 0.6462,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.487804878048781,
409
+ "grad_norm": 42.11968994140625,
410
+ "learning_rate": 1.5021021021021023e-05,
411
+ "loss": 0.729,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.623306233062331,
416
+ "grad_norm": 31.105592727661133,
417
+ "learning_rate": 1.4870870870870873e-05,
418
+ "loss": 0.5225,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.7588075880758804,
423
+ "grad_norm": 46.184783935546875,
424
+ "learning_rate": 1.4720720720720722e-05,
425
+ "loss": 0.6345,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.894308943089431,
430
+ "grad_norm": 104.41244506835938,
431
+ "learning_rate": 1.4570570570570572e-05,
432
+ "loss": 0.5265,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 7.0,
437
+ "eval_accuracy": 0.5650406504065041,
438
+ "eval_loss": 1.9319794178009033,
439
+ "eval_macro_f1": 0.5684701440187893,
440
+ "eval_runtime": 1.525,
441
+ "eval_samples_per_second": 483.942,
442
+ "eval_steps_per_second": 30.82,
443
+ "eval_weighted_f1": 0.5683702527633752,
444
+ "step": 1295
445
+ },
446
+ {
447
+ "epoch": 7.02710027100271,
448
+ "grad_norm": 3.5306365489959717,
449
+ "learning_rate": 1.4420420420420422e-05,
450
+ "loss": 0.58,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.16260162601626,
455
+ "grad_norm": 28.436826705932617,
456
+ "learning_rate": 1.4270270270270272e-05,
457
+ "loss": 0.3672,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.29810298102981,
462
+ "grad_norm": 41.26507568359375,
463
+ "learning_rate": 1.4120120120120123e-05,
464
+ "loss": 0.3083,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.43360433604336,
469
+ "grad_norm": 43.98733901977539,
470
+ "learning_rate": 1.3969969969969971e-05,
471
+ "loss": 0.4188,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.569105691056911,
476
+ "grad_norm": 80.8287124633789,
477
+ "learning_rate": 1.3819819819819821e-05,
478
+ "loss": 0.3759,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.7046070460704605,
483
+ "grad_norm": 36.36510467529297,
484
+ "learning_rate": 1.3669669669669672e-05,
485
+ "loss": 0.4157,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.840108401084011,
490
+ "grad_norm": 55.13428497314453,
491
+ "learning_rate": 1.3519519519519522e-05,
492
+ "loss": 0.2661,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 7.975609756097561,
497
+ "grad_norm": 49.47197723388672,
498
+ "learning_rate": 1.3369369369369369e-05,
499
+ "loss": 0.4562,
500
+ "step": 1475
501
+ },
502
+ {
503
+ "epoch": 8.0,
504
+ "eval_accuracy": 0.5691056910569106,
505
+ "eval_loss": 2.113414764404297,
506
+ "eval_macro_f1": 0.5698669130203521,
507
+ "eval_runtime": 1.5495,
508
+ "eval_samples_per_second": 476.273,
509
+ "eval_steps_per_second": 30.332,
510
+ "eval_weighted_f1": 0.5693848500221795,
511
+ "step": 1480
512
+ },
513
+ {
514
+ "epoch": 8.10840108401084,
515
+ "grad_norm": 76.07450103759766,
516
+ "learning_rate": 1.3219219219219219e-05,
517
+ "loss": 0.3138,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.24390243902439,
522
+ "grad_norm": 50.969268798828125,
523
+ "learning_rate": 1.3069069069069069e-05,
524
+ "loss": 0.2577,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.379403794037941,
529
+ "grad_norm": 92.61463165283203,
530
+ "learning_rate": 1.291891891891892e-05,
531
+ "loss": 0.1471,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.51490514905149,
536
+ "grad_norm": 38.85557556152344,
537
+ "learning_rate": 1.276876876876877e-05,
538
+ "loss": 0.3627,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.65040650406504,
543
+ "grad_norm": 2.0656871795654297,
544
+ "learning_rate": 1.2618618618618618e-05,
545
+ "loss": 0.4193,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.785907859078591,
550
+ "grad_norm": 23.408620834350586,
551
+ "learning_rate": 1.2468468468468468e-05,
552
+ "loss": 0.2464,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 8.92140921409214,
557
+ "grad_norm": 141.25840759277344,
558
+ "learning_rate": 1.2318318318318318e-05,
559
+ "loss": 0.4415,
560
+ "step": 1650
561
+ },
562
+ {
563
+ "epoch": 9.0,
564
+ "eval_accuracy": 0.5934959349593496,
565
+ "eval_loss": 2.5610709190368652,
566
+ "eval_macro_f1": 0.5907776037613098,
567
+ "eval_runtime": 1.6025,
568
+ "eval_samples_per_second": 460.52,
569
+ "eval_steps_per_second": 29.328,
570
+ "eval_weighted_f1": 0.5913238827137199,
571
+ "step": 1665
572
+ },
573
+ {
574
+ "epoch": 9.05420054200542,
575
+ "grad_norm": 0.8425450921058655,
576
+ "learning_rate": 1.2168168168168169e-05,
577
+ "loss": 0.2424,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.18970189701897,
582
+ "grad_norm": 5.634618759155273,
583
+ "learning_rate": 1.2018018018018019e-05,
584
+ "loss": 0.3127,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.32520325203252,
589
+ "grad_norm": 16.855060577392578,
590
+ "learning_rate": 1.1867867867867869e-05,
591
+ "loss": 0.1878,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.46070460704607,
596
+ "grad_norm": 91.97650146484375,
597
+ "learning_rate": 1.1717717717717718e-05,
598
+ "loss": 0.2588,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 9.59620596205962,
603
+ "grad_norm": 30.13783073425293,
604
+ "learning_rate": 1.1567567567567568e-05,
605
+ "loss": 0.2151,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 9.731707317073171,
610
+ "grad_norm": 40.8541259765625,
611
+ "learning_rate": 1.1417417417417418e-05,
612
+ "loss": 0.2247,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 9.86720867208672,
617
+ "grad_norm": 0.22134290635585785,
618
+ "learning_rate": 1.1267267267267268e-05,
619
+ "loss": 0.1759,
620
+ "step": 1825
621
+ },
622
+ {
623
+ "epoch": 10.0,
624
+ "grad_norm": 39.33653259277344,
625
+ "learning_rate": 1.1117117117117118e-05,
626
+ "loss": 0.1725,
627
+ "step": 1850
628
+ },
629
+ {
630
+ "epoch": 10.0,
631
+ "eval_accuracy": 0.5745257452574526,
632
+ "eval_loss": 3.2890443801879883,
633
+ "eval_macro_f1": 0.5663151864992984,
634
+ "eval_runtime": 1.4142,
635
+ "eval_samples_per_second": 521.855,
636
+ "eval_steps_per_second": 33.235,
637
+ "eval_weighted_f1": 0.5672364398217392,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.13550135501355,
642
+ "grad_norm": 0.9082435965538025,
643
+ "learning_rate": 1.0966966966966967e-05,
644
+ "loss": 0.1883,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.2710027100271,
649
+ "grad_norm": 4.879505157470703,
650
+ "learning_rate": 1.0816816816816817e-05,
651
+ "loss": 0.1127,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.40650406504065,
656
+ "grad_norm": 0.7202223539352417,
657
+ "learning_rate": 1.0666666666666667e-05,
658
+ "loss": 0.2144,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 10.5420054200542,
663
+ "grad_norm": 10.110024452209473,
664
+ "learning_rate": 1.0516516516516518e-05,
665
+ "loss": 0.2009,
666
+ "step": 1950
667
+ },
668
+ {
669
+ "epoch": 10.677506775067751,
670
+ "grad_norm": 142.56756591796875,
671
+ "learning_rate": 1.0366366366366368e-05,
672
+ "loss": 0.0916,
673
+ "step": 1975
674
+ },
675
+ {
676
+ "epoch": 10.8130081300813,
677
+ "grad_norm": 18.083961486816406,
678
+ "learning_rate": 1.0216216216216216e-05,
679
+ "loss": 0.0476,
680
+ "step": 2000
681
+ },
682
+ {
683
+ "epoch": 10.948509485094851,
684
+ "grad_norm": 92.09305572509766,
685
+ "learning_rate": 1.0066066066066066e-05,
686
+ "loss": 0.0713,
687
+ "step": 2025
688
+ },
689
+ {
690
+ "epoch": 11.0,
691
+ "eval_accuracy": 0.6002710027100271,
692
+ "eval_loss": 3.4880893230438232,
693
+ "eval_macro_f1": 0.6008961707884494,
694
+ "eval_runtime": 1.4136,
695
+ "eval_samples_per_second": 522.079,
696
+ "eval_steps_per_second": 33.249,
697
+ "eval_weighted_f1": 0.6011004064149246,
698
+ "step": 2035
699
+ }
700
+ ],
701
+ "logging_steps": 25,
702
+ "max_steps": 3700,
703
+ "num_input_tokens_seen": 0,
704
+ "num_train_epochs": 20,
705
+ "save_steps": 500,
706
+ "stateful_callbacks": {
707
+ "EarlyStoppingCallback": {
708
+ "args": {
709
+ "early_stopping_patience": 3,
710
+ "early_stopping_threshold": 0.0
711
+ },
712
+ "attributes": {
713
+ "early_stopping_patience_counter": 0
714
+ }
715
+ },
716
+ "TrainerControl": {
717
+ "args": {
718
+ "should_epoch_stop": false,
719
+ "should_evaluate": false,
720
+ "should_log": false,
721
+ "should_save": true,
722
+ "should_training_stop": false
723
+ },
724
+ "attributes": {}
725
+ }
726
+ },
727
+ "total_flos": 0.0,
728
+ "train_batch_size": 8,
729
+ "trial_name": null,
730
+ "trial_params": null
731
+ }
ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110/trainer_state.json ADDED
@@ -0,0 +1,417 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1110,
3
+ "best_metric": 0.570043158695038,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110",
5
+ "epoch": 6.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1110,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13513513513513514,
14
+ "grad_norm": 47.08201599121094,
15
+ "learning_rate": 1.2972972972972974e-06,
16
+ "loss": 4.6644,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2702702702702703,
21
+ "grad_norm": 26.193357467651367,
22
+ "learning_rate": 2.648648648648649e-06,
23
+ "loss": 4.5906,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.40540540540540543,
28
+ "grad_norm": 21.08445930480957,
29
+ "learning_rate": 4.000000000000001e-06,
30
+ "loss": 4.4594,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5405405405405406,
35
+ "grad_norm": 22.938949584960938,
36
+ "learning_rate": 5.351351351351351e-06,
37
+ "loss": 4.4581,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6756756756756757,
42
+ "grad_norm": 28.988496780395508,
43
+ "learning_rate": 6.702702702702704e-06,
44
+ "loss": 4.4553,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8108108108108109,
49
+ "grad_norm": 18.4388484954834,
50
+ "learning_rate": 8.054054054054055e-06,
51
+ "loss": 4.3135,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9459459459459459,
56
+ "grad_norm": 26.072526931762695,
57
+ "learning_rate": 9.405405405405407e-06,
58
+ "loss": 4.3428,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4201954397394137,
64
+ "eval_loss": 1.0680015087127686,
65
+ "eval_macro_f1": 0.41017693935210947,
66
+ "eval_runtime": 1.2819,
67
+ "eval_samples_per_second": 478.966,
68
+ "eval_steps_per_second": 30.423,
69
+ "eval_weighted_f1": 0.4108528036060057,
70
+ "step": 185
71
+ },
72
+ {
73
+ "epoch": 1.0810810810810811,
74
+ "grad_norm": 26.582788467407227,
75
+ "learning_rate": 1.0756756756756757e-05,
76
+ "loss": 4.2848,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2162162162162162,
81
+ "grad_norm": 29.265682220458984,
82
+ "learning_rate": 1.210810810810811e-05,
83
+ "loss": 4.2679,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3513513513513513,
88
+ "grad_norm": 26.10413360595703,
89
+ "learning_rate": 1.3459459459459461e-05,
90
+ "loss": 4.2162,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4864864864864864,
95
+ "grad_norm": 27.424680709838867,
96
+ "learning_rate": 1.4810810810810811e-05,
97
+ "loss": 4.3063,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6216216216216215,
102
+ "grad_norm": 29.469947814941406,
103
+ "learning_rate": 1.6162162162162163e-05,
104
+ "loss": 4.3689,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7567567567567568,
109
+ "grad_norm": 25.6850528717041,
110
+ "learning_rate": 1.7513513513513517e-05,
111
+ "loss": 4.3296,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.8918918918918919,
116
+ "grad_norm": 22.17389488220215,
117
+ "learning_rate": 1.8864864864864866e-05,
118
+ "loss": 4.185,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.49185667752442996,
124
+ "eval_loss": 1.0277191400527954,
125
+ "eval_macro_f1": 0.4787119341593644,
126
+ "eval_runtime": 1.2405,
127
+ "eval_samples_per_second": 494.972,
128
+ "eval_steps_per_second": 31.44,
129
+ "eval_weighted_f1": 0.4786092620393604,
130
+ "step": 370
131
+ },
132
+ {
133
+ "epoch": 2.027027027027027,
134
+ "grad_norm": 31.49361801147461,
135
+ "learning_rate": 1.9975975975975978e-05,
136
+ "loss": 4.1091,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.1621621621621623,
141
+ "grad_norm": 35.481075286865234,
142
+ "learning_rate": 1.9825825825825828e-05,
143
+ "loss": 3.779,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2972972972972974,
148
+ "grad_norm": 37.560726165771484,
149
+ "learning_rate": 1.967567567567568e-05,
150
+ "loss": 3.9835,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.4324324324324325,
155
+ "grad_norm": 61.58547592163086,
156
+ "learning_rate": 1.9525525525525525e-05,
157
+ "loss": 3.5805,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.5675675675675675,
162
+ "grad_norm": 21.816303253173828,
163
+ "learning_rate": 1.9375375375375376e-05,
164
+ "loss": 3.6515,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7027027027027026,
169
+ "grad_norm": 48.2593879699707,
170
+ "learning_rate": 1.9225225225225226e-05,
171
+ "loss": 3.6591,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8378378378378377,
176
+ "grad_norm": 96.28617095947266,
177
+ "learning_rate": 1.9075075075075076e-05,
178
+ "loss": 3.6854,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.972972972972973,
183
+ "grad_norm": 19.8309326171875,
184
+ "learning_rate": 1.8924924924924926e-05,
185
+ "loss": 3.8213,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.5439739413680782,
191
+ "eval_loss": 0.9436128735542297,
192
+ "eval_macro_f1": 0.5409081559355756,
193
+ "eval_runtime": 1.2687,
194
+ "eval_samples_per_second": 483.973,
195
+ "eval_steps_per_second": 30.741,
196
+ "eval_weighted_f1": 0.5406742189122544,
197
+ "step": 555
198
+ },
199
+ {
200
+ "epoch": 3.108108108108108,
201
+ "grad_norm": 56.405128479003906,
202
+ "learning_rate": 1.8774774774774776e-05,
203
+ "loss": 2.9787,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2432432432432434,
208
+ "grad_norm": 33.513641357421875,
209
+ "learning_rate": 1.8624624624624627e-05,
210
+ "loss": 2.7818,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.3783783783783785,
215
+ "grad_norm": 61.72041320800781,
216
+ "learning_rate": 1.8474474474474477e-05,
217
+ "loss": 2.8639,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5135135135135136,
222
+ "grad_norm": 56.814849853515625,
223
+ "learning_rate": 1.8324324324324324e-05,
224
+ "loss": 3.0699,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6486486486486487,
229
+ "grad_norm": 46.38714599609375,
230
+ "learning_rate": 1.8174174174174174e-05,
231
+ "loss": 2.8414,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.7837837837837838,
236
+ "grad_norm": 39.972930908203125,
237
+ "learning_rate": 1.8024024024024024e-05,
238
+ "loss": 2.8263,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.918918918918919,
243
+ "grad_norm": 32.175453186035156,
244
+ "learning_rate": 1.7873873873873874e-05,
245
+ "loss": 2.9646,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.5504885993485342,
251
+ "eval_loss": 0.9520296454429626,
252
+ "eval_macro_f1": 0.5299214434728453,
253
+ "eval_runtime": 1.4242,
254
+ "eval_samples_per_second": 431.108,
255
+ "eval_steps_per_second": 27.383,
256
+ "eval_weighted_f1": 0.5299665268918519,
257
+ "step": 740
258
+ },
259
+ {
260
+ "epoch": 4.054054054054054,
261
+ "grad_norm": 18.94879150390625,
262
+ "learning_rate": 1.7723723723723724e-05,
263
+ "loss": 2.8281,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.1891891891891895,
268
+ "grad_norm": 41.29767608642578,
269
+ "learning_rate": 1.7573573573573575e-05,
270
+ "loss": 2.1991,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.324324324324325,
275
+ "grad_norm": 42.534515380859375,
276
+ "learning_rate": 1.7423423423423425e-05,
277
+ "loss": 2.0389,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.45945945945946,
282
+ "grad_norm": 41.03356170654297,
283
+ "learning_rate": 1.7273273273273275e-05,
284
+ "loss": 1.8074,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.594594594594595,
289
+ "grad_norm": 97.1319580078125,
290
+ "learning_rate": 1.7123123123123125e-05,
291
+ "loss": 2.1413,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.72972972972973,
296
+ "grad_norm": 23.09299087524414,
297
+ "learning_rate": 1.6972972972972975e-05,
298
+ "loss": 2.4603,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.864864864864865,
303
+ "grad_norm": 75.58317565917969,
304
+ "learning_rate": 1.6822822822822826e-05,
305
+ "loss": 2.0532,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "grad_norm": 56.788883209228516,
311
+ "learning_rate": 1.6672672672672673e-05,
312
+ "loss": 2.0503,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 5.0,
317
+ "eval_accuracy": 0.5618892508143323,
318
+ "eval_loss": 1.1100575923919678,
319
+ "eval_macro_f1": 0.5473636502134681,
320
+ "eval_runtime": 1.2916,
321
+ "eval_samples_per_second": 475.371,
322
+ "eval_steps_per_second": 30.195,
323
+ "eval_weighted_f1": 0.547097041671809,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.135135135135135,
328
+ "grad_norm": 33.32196044921875,
329
+ "learning_rate": 1.6522522522522523e-05,
330
+ "loss": 1.1314,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.27027027027027,
335
+ "grad_norm": 41.198726654052734,
336
+ "learning_rate": 1.6372372372372373e-05,
337
+ "loss": 1.2725,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.405405405405405,
342
+ "grad_norm": 102.79741668701172,
343
+ "learning_rate": 1.6222222222222223e-05,
344
+ "loss": 1.3139,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.54054054054054,
349
+ "grad_norm": 48.75640869140625,
350
+ "learning_rate": 1.6072072072072073e-05,
351
+ "loss": 1.397,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.675675675675675,
356
+ "grad_norm": 67.42860412597656,
357
+ "learning_rate": 1.5921921921921924e-05,
358
+ "loss": 1.4075,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.8108108108108105,
363
+ "grad_norm": 44.444580078125,
364
+ "learning_rate": 1.5771771771771774e-05,
365
+ "loss": 1.281,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.945945945945946,
370
+ "grad_norm": 34.64950942993164,
371
+ "learning_rate": 1.5621621621621624e-05,
372
+ "loss": 1.3574,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 6.0,
377
+ "eval_accuracy": 0.5700325732899023,
378
+ "eval_loss": 1.2948631048202515,
379
+ "eval_macro_f1": 0.570043158695038,
380
+ "eval_runtime": 1.2172,
381
+ "eval_samples_per_second": 504.428,
382
+ "eval_steps_per_second": 32.04,
383
+ "eval_weighted_f1": 0.5697932543331644,
384
+ "step": 1110
385
+ }
386
+ ],
387
+ "logging_steps": 25,
388
+ "max_steps": 3700,
389
+ "num_input_tokens_seen": 0,
390
+ "num_train_epochs": 20,
391
+ "save_steps": 500,
392
+ "stateful_callbacks": {
393
+ "EarlyStoppingCallback": {
394
+ "args": {
395
+ "early_stopping_patience": 3,
396
+ "early_stopping_threshold": 0.0
397
+ },
398
+ "attributes": {
399
+ "early_stopping_patience_counter": 0
400
+ }
401
+ },
402
+ "TrainerControl": {
403
+ "args": {
404
+ "should_epoch_stop": false,
405
+ "should_evaluate": false,
406
+ "should_log": false,
407
+ "should_save": true,
408
+ "should_training_stop": false
409
+ },
410
+ "attributes": {}
411
+ }
412
+ },
413
+ "total_flos": 0.0,
414
+ "train_batch_size": 8,
415
+ "trial_name": null,
416
+ "trial_params": null
417
+ }
ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1665/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "cls_token": "<s>",
4
+ "eos_token": "</s>",
5
+ "mask_token": {
6
+ "content": "<mask>",
7
+ "lstrip": true,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false
11
+ },
12
+ "pad_token": "<pad>",
13
+ "sep_token": "</s>",
14
+ "unk_token": "<unk>"
15
+ }
ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1665/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1665/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "15001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
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": 512,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1665/trainer_state.json ADDED
@@ -0,0 +1,604 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1110,
3
+ "best_metric": 0.570043158695038,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-07-s44-20260831T035525Z-59e46264/trainer/checkpoint-1110",
5
+ "epoch": 9.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1665,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13513513513513514,
14
+ "grad_norm": 47.08201599121094,
15
+ "learning_rate": 1.2972972972972974e-06,
16
+ "loss": 4.6644,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.2702702702702703,
21
+ "grad_norm": 26.193357467651367,
22
+ "learning_rate": 2.648648648648649e-06,
23
+ "loss": 4.5906,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.40540540540540543,
28
+ "grad_norm": 21.08445930480957,
29
+ "learning_rate": 4.000000000000001e-06,
30
+ "loss": 4.4594,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5405405405405406,
35
+ "grad_norm": 22.938949584960938,
36
+ "learning_rate": 5.351351351351351e-06,
37
+ "loss": 4.4581,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6756756756756757,
42
+ "grad_norm": 28.988496780395508,
43
+ "learning_rate": 6.702702702702704e-06,
44
+ "loss": 4.4553,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8108108108108109,
49
+ "grad_norm": 18.4388484954834,
50
+ "learning_rate": 8.054054054054055e-06,
51
+ "loss": 4.3135,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9459459459459459,
56
+ "grad_norm": 26.072526931762695,
57
+ "learning_rate": 9.405405405405407e-06,
58
+ "loss": 4.3428,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4201954397394137,
64
+ "eval_loss": 1.0680015087127686,
65
+ "eval_macro_f1": 0.41017693935210947,
66
+ "eval_runtime": 1.2819,
67
+ "eval_samples_per_second": 478.966,
68
+ "eval_steps_per_second": 30.423,
69
+ "eval_weighted_f1": 0.4108528036060057,
70
+ "step": 185
71
+ },
72
+ {
73
+ "epoch": 1.0810810810810811,
74
+ "grad_norm": 26.582788467407227,
75
+ "learning_rate": 1.0756756756756757e-05,
76
+ "loss": 4.2848,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2162162162162162,
81
+ "grad_norm": 29.265682220458984,
82
+ "learning_rate": 1.210810810810811e-05,
83
+ "loss": 4.2679,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3513513513513513,
88
+ "grad_norm": 26.10413360595703,
89
+ "learning_rate": 1.3459459459459461e-05,
90
+ "loss": 4.2162,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4864864864864864,
95
+ "grad_norm": 27.424680709838867,
96
+ "learning_rate": 1.4810810810810811e-05,
97
+ "loss": 4.3063,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6216216216216215,
102
+ "grad_norm": 29.469947814941406,
103
+ "learning_rate": 1.6162162162162163e-05,
104
+ "loss": 4.3689,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7567567567567568,
109
+ "grad_norm": 25.6850528717041,
110
+ "learning_rate": 1.7513513513513517e-05,
111
+ "loss": 4.3296,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.8918918918918919,
116
+ "grad_norm": 22.17389488220215,
117
+ "learning_rate": 1.8864864864864866e-05,
118
+ "loss": 4.185,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.49185667752442996,
124
+ "eval_loss": 1.0277191400527954,
125
+ "eval_macro_f1": 0.4787119341593644,
126
+ "eval_runtime": 1.2405,
127
+ "eval_samples_per_second": 494.972,
128
+ "eval_steps_per_second": 31.44,
129
+ "eval_weighted_f1": 0.4786092620393604,
130
+ "step": 370
131
+ },
132
+ {
133
+ "epoch": 2.027027027027027,
134
+ "grad_norm": 31.49361801147461,
135
+ "learning_rate": 1.9975975975975978e-05,
136
+ "loss": 4.1091,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.1621621621621623,
141
+ "grad_norm": 35.481075286865234,
142
+ "learning_rate": 1.9825825825825828e-05,
143
+ "loss": 3.779,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2972972972972974,
148
+ "grad_norm": 37.560726165771484,
149
+ "learning_rate": 1.967567567567568e-05,
150
+ "loss": 3.9835,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.4324324324324325,
155
+ "grad_norm": 61.58547592163086,
156
+ "learning_rate": 1.9525525525525525e-05,
157
+ "loss": 3.5805,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.5675675675675675,
162
+ "grad_norm": 21.816303253173828,
163
+ "learning_rate": 1.9375375375375376e-05,
164
+ "loss": 3.6515,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7027027027027026,
169
+ "grad_norm": 48.2593879699707,
170
+ "learning_rate": 1.9225225225225226e-05,
171
+ "loss": 3.6591,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8378378378378377,
176
+ "grad_norm": 96.28617095947266,
177
+ "learning_rate": 1.9075075075075076e-05,
178
+ "loss": 3.6854,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.972972972972973,
183
+ "grad_norm": 19.8309326171875,
184
+ "learning_rate": 1.8924924924924926e-05,
185
+ "loss": 3.8213,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.5439739413680782,
191
+ "eval_loss": 0.9436128735542297,
192
+ "eval_macro_f1": 0.5409081559355756,
193
+ "eval_runtime": 1.2687,
194
+ "eval_samples_per_second": 483.973,
195
+ "eval_steps_per_second": 30.741,
196
+ "eval_weighted_f1": 0.5406742189122544,
197
+ "step": 555
198
+ },
199
+ {
200
+ "epoch": 3.108108108108108,
201
+ "grad_norm": 56.405128479003906,
202
+ "learning_rate": 1.8774774774774776e-05,
203
+ "loss": 2.9787,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2432432432432434,
208
+ "grad_norm": 33.513641357421875,
209
+ "learning_rate": 1.8624624624624627e-05,
210
+ "loss": 2.7818,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.3783783783783785,
215
+ "grad_norm": 61.72041320800781,
216
+ "learning_rate": 1.8474474474474477e-05,
217
+ "loss": 2.8639,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5135135135135136,
222
+ "grad_norm": 56.814849853515625,
223
+ "learning_rate": 1.8324324324324324e-05,
224
+ "loss": 3.0699,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6486486486486487,
229
+ "grad_norm": 46.38714599609375,
230
+ "learning_rate": 1.8174174174174174e-05,
231
+ "loss": 2.8414,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.7837837837837838,
236
+ "grad_norm": 39.972930908203125,
237
+ "learning_rate": 1.8024024024024024e-05,
238
+ "loss": 2.8263,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.918918918918919,
243
+ "grad_norm": 32.175453186035156,
244
+ "learning_rate": 1.7873873873873874e-05,
245
+ "loss": 2.9646,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.5504885993485342,
251
+ "eval_loss": 0.9520296454429626,
252
+ "eval_macro_f1": 0.5299214434728453,
253
+ "eval_runtime": 1.4242,
254
+ "eval_samples_per_second": 431.108,
255
+ "eval_steps_per_second": 27.383,
256
+ "eval_weighted_f1": 0.5299665268918519,
257
+ "step": 740
258
+ },
259
+ {
260
+ "epoch": 4.054054054054054,
261
+ "grad_norm": 18.94879150390625,
262
+ "learning_rate": 1.7723723723723724e-05,
263
+ "loss": 2.8281,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.1891891891891895,
268
+ "grad_norm": 41.29767608642578,
269
+ "learning_rate": 1.7573573573573575e-05,
270
+ "loss": 2.1991,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.324324324324325,
275
+ "grad_norm": 42.534515380859375,
276
+ "learning_rate": 1.7423423423423425e-05,
277
+ "loss": 2.0389,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.45945945945946,
282
+ "grad_norm": 41.03356170654297,
283
+ "learning_rate": 1.7273273273273275e-05,
284
+ "loss": 1.8074,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.594594594594595,
289
+ "grad_norm": 97.1319580078125,
290
+ "learning_rate": 1.7123123123123125e-05,
291
+ "loss": 2.1413,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.72972972972973,
296
+ "grad_norm": 23.09299087524414,
297
+ "learning_rate": 1.6972972972972975e-05,
298
+ "loss": 2.4603,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.864864864864865,
303
+ "grad_norm": 75.58317565917969,
304
+ "learning_rate": 1.6822822822822826e-05,
305
+ "loss": 2.0532,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "grad_norm": 56.788883209228516,
311
+ "learning_rate": 1.6672672672672673e-05,
312
+ "loss": 2.0503,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 5.0,
317
+ "eval_accuracy": 0.5618892508143323,
318
+ "eval_loss": 1.1100575923919678,
319
+ "eval_macro_f1": 0.5473636502134681,
320
+ "eval_runtime": 1.2916,
321
+ "eval_samples_per_second": 475.371,
322
+ "eval_steps_per_second": 30.195,
323
+ "eval_weighted_f1": 0.547097041671809,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.135135135135135,
328
+ "grad_norm": 33.32196044921875,
329
+ "learning_rate": 1.6522522522522523e-05,
330
+ "loss": 1.1314,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.27027027027027,
335
+ "grad_norm": 41.198726654052734,
336
+ "learning_rate": 1.6372372372372373e-05,
337
+ "loss": 1.2725,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.405405405405405,
342
+ "grad_norm": 102.79741668701172,
343
+ "learning_rate": 1.6222222222222223e-05,
344
+ "loss": 1.3139,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.54054054054054,
349
+ "grad_norm": 48.75640869140625,
350
+ "learning_rate": 1.6072072072072073e-05,
351
+ "loss": 1.397,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.675675675675675,
356
+ "grad_norm": 67.42860412597656,
357
+ "learning_rate": 1.5921921921921924e-05,
358
+ "loss": 1.4075,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.8108108108108105,
363
+ "grad_norm": 44.444580078125,
364
+ "learning_rate": 1.5771771771771774e-05,
365
+ "loss": 1.281,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.945945945945946,
370
+ "grad_norm": 34.64950942993164,
371
+ "learning_rate": 1.5621621621621624e-05,
372
+ "loss": 1.3574,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 6.0,
377
+ "eval_accuracy": 0.5700325732899023,
378
+ "eval_loss": 1.2948631048202515,
379
+ "eval_macro_f1": 0.570043158695038,
380
+ "eval_runtime": 1.2172,
381
+ "eval_samples_per_second": 504.428,
382
+ "eval_steps_per_second": 32.04,
383
+ "eval_weighted_f1": 0.5697932543331644,
384
+ "step": 1110
385
+ },
386
+ {
387
+ "epoch": 6.081081081081081,
388
+ "grad_norm": 38.24656677246094,
389
+ "learning_rate": 1.5471471471471474e-05,
390
+ "loss": 0.8201,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.216216216216216,
395
+ "grad_norm": 33.79013442993164,
396
+ "learning_rate": 1.5321321321321324e-05,
397
+ "loss": 0.6768,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.351351351351352,
402
+ "grad_norm": 69.95608520507812,
403
+ "learning_rate": 1.5171171171171173e-05,
404
+ "loss": 0.7047,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.486486486486487,
409
+ "grad_norm": 38.98823547363281,
410
+ "learning_rate": 1.5021021021021023e-05,
411
+ "loss": 0.7469,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.621621621621622,
416
+ "grad_norm": 23.87979507446289,
417
+ "learning_rate": 1.4870870870870873e-05,
418
+ "loss": 0.6613,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.756756756756757,
423
+ "grad_norm": 50.64320373535156,
424
+ "learning_rate": 1.4720720720720722e-05,
425
+ "loss": 0.638,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.891891891891892,
430
+ "grad_norm": 51.634010314941406,
431
+ "learning_rate": 1.4570570570570572e-05,
432
+ "loss": 0.7247,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 7.0,
437
+ "eval_accuracy": 0.5667752442996743,
438
+ "eval_loss": 1.616267204284668,
439
+ "eval_macro_f1": 0.5674840047145074,
440
+ "eval_runtime": 1.21,
441
+ "eval_samples_per_second": 507.423,
442
+ "eval_steps_per_second": 32.23,
443
+ "eval_weighted_f1": 0.5671119942672486,
444
+ "step": 1295
445
+ },
446
+ {
447
+ "epoch": 7.027027027027027,
448
+ "grad_norm": 19.774234771728516,
449
+ "learning_rate": 1.4420420420420422e-05,
450
+ "loss": 0.6041,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.162162162162162,
455
+ "grad_norm": 6.243575096130371,
456
+ "learning_rate": 1.4270270270270272e-05,
457
+ "loss": 0.4571,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.297297297297297,
462
+ "grad_norm": 25.344165802001953,
463
+ "learning_rate": 1.4120120120120123e-05,
464
+ "loss": 0.3678,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.4324324324324325,
469
+ "grad_norm": 109.5257568359375,
470
+ "learning_rate": 1.3969969969969971e-05,
471
+ "loss": 0.5097,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.5675675675675675,
476
+ "grad_norm": 17.59688377380371,
477
+ "learning_rate": 1.3819819819819821e-05,
478
+ "loss": 0.4298,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.702702702702703,
483
+ "grad_norm": 42.131534576416016,
484
+ "learning_rate": 1.3669669669669672e-05,
485
+ "loss": 0.4018,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.837837837837838,
490
+ "grad_norm": 58.51371765136719,
491
+ "learning_rate": 1.3519519519519522e-05,
492
+ "loss": 0.4852,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 7.972972972972973,
497
+ "grad_norm": 106.68798828125,
498
+ "learning_rate": 1.3369369369369369e-05,
499
+ "loss": 0.6589,
500
+ "step": 1475
501
+ },
502
+ {
503
+ "epoch": 8.0,
504
+ "eval_accuracy": 0.5195439739413681,
505
+ "eval_loss": 2.4034078121185303,
506
+ "eval_macro_f1": 0.5147580130973256,
507
+ "eval_runtime": 1.2499,
508
+ "eval_samples_per_second": 491.228,
509
+ "eval_steps_per_second": 31.202,
510
+ "eval_weighted_f1": 0.5150605441257019,
511
+ "step": 1480
512
+ },
513
+ {
514
+ "epoch": 8.108108108108109,
515
+ "grad_norm": 20.660078048706055,
516
+ "learning_rate": 1.3219219219219219e-05,
517
+ "loss": 0.3209,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.243243243243244,
522
+ "grad_norm": 33.259765625,
523
+ "learning_rate": 1.3069069069069069e-05,
524
+ "loss": 0.2288,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.378378378378379,
529
+ "grad_norm": 33.1270637512207,
530
+ "learning_rate": 1.291891891891892e-05,
531
+ "loss": 0.3475,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.513513513513514,
536
+ "grad_norm": 33.48015594482422,
537
+ "learning_rate": 1.276876876876877e-05,
538
+ "loss": 0.3586,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.64864864864865,
543
+ "grad_norm": 13.339167594909668,
544
+ "learning_rate": 1.2618618618618618e-05,
545
+ "loss": 0.2511,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.783783783783784,
550
+ "grad_norm": 86.10514831542969,
551
+ "learning_rate": 1.2468468468468468e-05,
552
+ "loss": 0.1936,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 8.91891891891892,
557
+ "grad_norm": 25.579668045043945,
558
+ "learning_rate": 1.2318318318318318e-05,
559
+ "loss": 0.2825,
560
+ "step": 1650
561
+ },
562
+ {
563
+ "epoch": 9.0,
564
+ "eval_accuracy": 0.5667752442996743,
565
+ "eval_loss": 2.8016889095306396,
566
+ "eval_macro_f1": 0.5687973086335752,
567
+ "eval_runtime": 1.2254,
568
+ "eval_samples_per_second": 501.046,
569
+ "eval_steps_per_second": 31.825,
570
+ "eval_weighted_f1": 0.5686154755485238,
571
+ "step": 1665
572
+ }
573
+ ],
574
+ "logging_steps": 25,
575
+ "max_steps": 3700,
576
+ "num_input_tokens_seen": 0,
577
+ "num_train_epochs": 20,
578
+ "save_steps": 500,
579
+ "stateful_callbacks": {
580
+ "EarlyStoppingCallback": {
581
+ "args": {
582
+ "early_stopping_patience": 3,
583
+ "early_stopping_threshold": 0.0
584
+ },
585
+ "attributes": {
586
+ "early_stopping_patience_counter": 3
587
+ }
588
+ },
589
+ "TrainerControl": {
590
+ "args": {
591
+ "should_epoch_stop": false,
592
+ "should_evaluate": false,
593
+ "should_log": false,
594
+ "should_save": true,
595
+ "should_training_stop": true
596
+ },
597
+ "attributes": {}
598
+ }
599
+ },
600
+ "total_flos": 0.0,
601
+ "train_batch_size": 8,
602
+ "trial_name": null,
603
+ "trial_params": null
604
+ }