PhaaNe commited on
Commit
6554329
·
verified ·
1 Parent(s): 1a1acf8

Add files using upload-large-folder tool

Browse files
Files changed (20) hide show
  1. ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638/special_tokens_map.json +15 -0
  2. ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638/tokenizer.json +0 -0
  3. ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638/tokenizer_config.json +56 -0
  4. ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638/trainer_state.json +597 -0
  5. ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-2184/special_tokens_map.json +15 -0
  6. ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-2184/tokenizer.json +0 -0
  7. ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-2184/tokenizer_config.json +56 -0
  8. ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-2184/trainer_state.json +784 -0
  9. ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-855/special_tokens_map.json +15 -0
  10. ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-855/tokenizer.json +0 -0
  11. ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-855/tokenizer_config.json +56 -0
  12. ood-vifactcheck-visobert-source-source-05-s44-20260831T033516Z-59e46264/trainer/checkpoint-1620/special_tokens_map.json +15 -0
  13. ood-vifactcheck-visobert-source-source-05-s44-20260831T033516Z-59e46264/trainer/checkpoint-1620/tokenizer.json +0 -0
  14. ood-vifactcheck-visobert-source-source-05-s44-20260831T033516Z-59e46264/trainer/checkpoint-1620/tokenizer_config.json +56 -0
  15. ood-vifactcheck-visobert-source-source-05-s44-20260831T033516Z-59e46264/trainer/checkpoint-1620/trainer_state.json +590 -0
  16. ood-vifactcheck-visobert-source-source-06-s42-20260831T003231Z-f21e2f7e/trainer/checkpoint-1480/tokenizer.json +0 -0
  17. ood-vifactcheck-visobert-source-source-06-s42-20260831T003231Z-f21e2f7e/trainer/checkpoint-1480/trainer_state.json +544 -0
  18. ood-vifactcheck-visobert-source-source-06-s44-20260831T034444Z-59e46264/trainer/checkpoint-1295/special_tokens_map.json +15 -0
  19. ood-vifactcheck-visobert-source-source-06-s44-20260831T034444Z-59e46264/trainer/checkpoint-1295/tokenizer_config.json +56 -0
  20. ood-vifactcheck-visobert-source-source-06-s44-20260831T034444Z-59e46264/trainer/checkpoint-1295/trainer_state.json +477 -0
ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638/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-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638/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-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638/trainer_state.json ADDED
@@ -0,0 +1,597 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1638,
3
+ "best_metric": 0.5896711509781644,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638",
5
+ "epoch": 9.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1638,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13793103448275862,
14
+ "grad_norm": 26.575096130371094,
15
+ "learning_rate": 1.3186813186813187e-06,
16
+ "loss": 4.9635,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27586206896551724,
21
+ "grad_norm": 35.562255859375,
22
+ "learning_rate": 2.6923076923076923e-06,
23
+ "loss": 4.473,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.41379310344827586,
28
+ "grad_norm": 19.921688079833984,
29
+ "learning_rate": 4.065934065934066e-06,
30
+ "loss": 4.481,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5517241379310345,
35
+ "grad_norm": 18.724918365478516,
36
+ "learning_rate": 5.43956043956044e-06,
37
+ "loss": 4.4112,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6896551724137931,
42
+ "grad_norm": 22.945472717285156,
43
+ "learning_rate": 6.813186813186814e-06,
44
+ "loss": 4.3912,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8275862068965517,
49
+ "grad_norm": 27.29220199584961,
50
+ "learning_rate": 8.186813186813188e-06,
51
+ "loss": 4.3679,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9655172413793104,
56
+ "grad_norm": 19.696191787719727,
57
+ "learning_rate": 9.560439560439562e-06,
58
+ "loss": 4.351,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.45773524720893144,
64
+ "eval_loss": 1.050980567932129,
65
+ "eval_macro_f1": 0.45385903483259726,
66
+ "eval_runtime": 1.1797,
67
+ "eval_samples_per_second": 531.489,
68
+ "eval_steps_per_second": 33.907,
69
+ "eval_weighted_f1": 0.4554201845332243,
70
+ "step": 182
71
+ },
72
+ {
73
+ "epoch": 1.0993103448275863,
74
+ "grad_norm": 32.07797622680664,
75
+ "learning_rate": 1.0934065934065933e-05,
76
+ "loss": 4.1139,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.237241379310345,
81
+ "grad_norm": 15.515581130981445,
82
+ "learning_rate": 1.230769230769231e-05,
83
+ "loss": 4.2158,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3751724137931034,
88
+ "grad_norm": 22.117534637451172,
89
+ "learning_rate": 1.3681318681318683e-05,
90
+ "loss": 4.2599,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.513103448275862,
95
+ "grad_norm": 17.774995803833008,
96
+ "learning_rate": 1.5054945054945056e-05,
97
+ "loss": 4.2231,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6510344827586207,
102
+ "grad_norm": 27.439960479736328,
103
+ "learning_rate": 1.642857142857143e-05,
104
+ "loss": 4.2672,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7889655172413792,
109
+ "grad_norm": 22.73525047302246,
110
+ "learning_rate": 1.78021978021978e-05,
111
+ "loss": 4.2553,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.926896551724138,
116
+ "grad_norm": 23.057146072387695,
117
+ "learning_rate": 1.9175824175824178e-05,
118
+ "loss": 4.2416,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.4864433811802233,
124
+ "eval_loss": 1.0021166801452637,
125
+ "eval_macro_f1": 0.488904678955645,
126
+ "eval_runtime": 1.2773,
127
+ "eval_samples_per_second": 490.888,
128
+ "eval_steps_per_second": 31.317,
129
+ "eval_weighted_f1": 0.4881585306493283,
130
+ "step": 364
131
+ },
132
+ {
133
+ "epoch": 2.0606896551724136,
134
+ "grad_norm": 49.71544647216797,
135
+ "learning_rate": 1.993894993894994e-05,
136
+ "loss": 3.7333,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.1986206896551725,
141
+ "grad_norm": 28.5469913482666,
142
+ "learning_rate": 1.9786324786324788e-05,
143
+ "loss": 3.9059,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.336551724137931,
148
+ "grad_norm": 37.46480178833008,
149
+ "learning_rate": 1.9633699633699634e-05,
150
+ "loss": 3.4297,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.47448275862069,
155
+ "grad_norm": 41.697845458984375,
156
+ "learning_rate": 1.948107448107448e-05,
157
+ "loss": 3.5016,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.6124137931034483,
162
+ "grad_norm": 35.8870735168457,
163
+ "learning_rate": 1.932844932844933e-05,
164
+ "loss": 3.682,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.750344827586207,
169
+ "grad_norm": 34.63261413574219,
170
+ "learning_rate": 1.9175824175824178e-05,
171
+ "loss": 3.685,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8882758620689657,
176
+ "grad_norm": 32.974422454833984,
177
+ "learning_rate": 1.9023199023199025e-05,
178
+ "loss": 3.7675,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.507177033492823,
184
+ "eval_loss": 0.9895545244216919,
185
+ "eval_macro_f1": 0.4967913731077999,
186
+ "eval_runtime": 1.2454,
187
+ "eval_samples_per_second": 503.436,
188
+ "eval_steps_per_second": 32.117,
189
+ "eval_weighted_f1": 0.49262741180781616,
190
+ "step": 546
191
+ },
192
+ {
193
+ "epoch": 3.0220689655172412,
194
+ "grad_norm": 30.783283233642578,
195
+ "learning_rate": 1.887057387057387e-05,
196
+ "loss": 3.4839,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.16,
201
+ "grad_norm": 79.42498779296875,
202
+ "learning_rate": 1.8717948717948718e-05,
203
+ "loss": 2.86,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2979310344827586,
208
+ "grad_norm": 41.3476448059082,
209
+ "learning_rate": 1.8565323565323568e-05,
210
+ "loss": 2.9271,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.435862068965517,
215
+ "grad_norm": 31.817678451538086,
216
+ "learning_rate": 1.8412698412698415e-05,
217
+ "loss": 3.078,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.573793103448276,
222
+ "grad_norm": 30.916423797607422,
223
+ "learning_rate": 1.826007326007326e-05,
224
+ "loss": 2.6603,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.7117241379310344,
229
+ "grad_norm": 28.779693603515625,
230
+ "learning_rate": 1.8107448107448108e-05,
231
+ "loss": 2.8651,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.849655172413793,
236
+ "grad_norm": 40.502933502197266,
237
+ "learning_rate": 1.7954822954822958e-05,
238
+ "loss": 2.7305,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.9875862068965517,
243
+ "grad_norm": 40.181915283203125,
244
+ "learning_rate": 1.78021978021978e-05,
245
+ "loss": 2.6274,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.5518341307814992,
251
+ "eval_loss": 1.062201976776123,
252
+ "eval_macro_f1": 0.5417668087507094,
253
+ "eval_runtime": 1.2279,
254
+ "eval_samples_per_second": 510.611,
255
+ "eval_steps_per_second": 32.575,
256
+ "eval_weighted_f1": 0.5429427985347581,
257
+ "step": 728
258
+ },
259
+ {
260
+ "epoch": 4.121379310344827,
261
+ "grad_norm": 53.15599822998047,
262
+ "learning_rate": 1.764957264957265e-05,
263
+ "loss": 2.0303,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.259310344827586,
268
+ "grad_norm": 43.258941650390625,
269
+ "learning_rate": 1.74969474969475e-05,
270
+ "loss": 1.8058,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.397241379310345,
275
+ "grad_norm": 69.51874542236328,
276
+ "learning_rate": 1.7344322344322345e-05,
277
+ "loss": 1.7058,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.535172413793103,
282
+ "grad_norm": 45.85198211669922,
283
+ "learning_rate": 1.7191697191697195e-05,
284
+ "loss": 2.1799,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.673103448275862,
289
+ "grad_norm": 57.6066780090332,
290
+ "learning_rate": 1.703907203907204e-05,
291
+ "loss": 1.7652,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.811034482758621,
296
+ "grad_norm": 38.48773956298828,
297
+ "learning_rate": 1.688644688644689e-05,
298
+ "loss": 1.9997,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.94896551724138,
303
+ "grad_norm": 38.666595458984375,
304
+ "learning_rate": 1.6733821733821735e-05,
305
+ "loss": 1.9021,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.5374800637958532,
311
+ "eval_loss": 1.168395757675171,
312
+ "eval_macro_f1": 0.5388034546709805,
313
+ "eval_runtime": 1.2689,
314
+ "eval_samples_per_second": 494.123,
315
+ "eval_steps_per_second": 31.523,
316
+ "eval_weighted_f1": 0.5393512866960177,
317
+ "step": 910
318
+ },
319
+ {
320
+ "epoch": 5.082758620689655,
321
+ "grad_norm": 41.31229782104492,
322
+ "learning_rate": 1.6581196581196585e-05,
323
+ "loss": 1.385,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.220689655172414,
328
+ "grad_norm": 101.49195861816406,
329
+ "learning_rate": 1.642857142857143e-05,
330
+ "loss": 1.0578,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.358620689655172,
335
+ "grad_norm": 43.01548767089844,
336
+ "learning_rate": 1.627594627594628e-05,
337
+ "loss": 0.9764,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.496551724137931,
342
+ "grad_norm": 58.39358139038086,
343
+ "learning_rate": 1.6123321123321125e-05,
344
+ "loss": 1.1107,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.63448275862069,
349
+ "grad_norm": 87.73316192626953,
350
+ "learning_rate": 1.5970695970695972e-05,
351
+ "loss": 1.1031,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.772413793103448,
356
+ "grad_norm": 60.960079193115234,
357
+ "learning_rate": 1.581807081807082e-05,
358
+ "loss": 1.0898,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.910344827586207,
363
+ "grad_norm": 44.55670166015625,
364
+ "learning_rate": 1.5665445665445665e-05,
365
+ "loss": 1.0198,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.5534290271132376,
371
+ "eval_loss": 1.4985970258712769,
372
+ "eval_macro_f1": 0.554246483931539,
373
+ "eval_runtime": 1.2881,
374
+ "eval_samples_per_second": 486.77,
375
+ "eval_steps_per_second": 31.054,
376
+ "eval_weighted_f1": 0.5546838726601748,
377
+ "step": 1092
378
+ },
379
+ {
380
+ "epoch": 6.0441379310344825,
381
+ "grad_norm": 23.108503341674805,
382
+ "learning_rate": 1.5512820512820516e-05,
383
+ "loss": 0.9103,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.182068965517241,
388
+ "grad_norm": 32.50311279296875,
389
+ "learning_rate": 1.5360195360195362e-05,
390
+ "loss": 0.6263,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.32,
395
+ "grad_norm": 20.798681259155273,
396
+ "learning_rate": 1.5207570207570207e-05,
397
+ "loss": 0.4834,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.457931034482758,
402
+ "grad_norm": 39.38999557495117,
403
+ "learning_rate": 1.5054945054945056e-05,
404
+ "loss": 0.6147,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.595862068965517,
409
+ "grad_norm": 124.87044525146484,
410
+ "learning_rate": 1.4902319902319904e-05,
411
+ "loss": 0.6238,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.733793103448276,
416
+ "grad_norm": 74.66791534423828,
417
+ "learning_rate": 1.4749694749694752e-05,
418
+ "loss": 0.653,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.871724137931034,
423
+ "grad_norm": 41.76373291015625,
424
+ "learning_rate": 1.4597069597069597e-05,
425
+ "loss": 0.7231,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 7.0,
430
+ "eval_accuracy": 0.5502392344497608,
431
+ "eval_loss": 1.7950698137283325,
432
+ "eval_macro_f1": 0.5524237990246955,
433
+ "eval_runtime": 1.2696,
434
+ "eval_samples_per_second": 493.869,
435
+ "eval_steps_per_second": 31.507,
436
+ "eval_weighted_f1": 0.5523545923272882,
437
+ "step": 1274
438
+ },
439
+ {
440
+ "epoch": 7.0055172413793105,
441
+ "grad_norm": 45.92796325683594,
442
+ "learning_rate": 1.4444444444444446e-05,
443
+ "loss": 0.5914,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.1434482758620685,
448
+ "grad_norm": 80.1987533569336,
449
+ "learning_rate": 1.4291819291819293e-05,
450
+ "loss": 0.3786,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.281379310344827,
455
+ "grad_norm": 31.46441078186035,
456
+ "learning_rate": 1.4139194139194141e-05,
457
+ "loss": 0.4155,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.419310344827586,
462
+ "grad_norm": 58.27080535888672,
463
+ "learning_rate": 1.3986568986568988e-05,
464
+ "loss": 0.3563,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.557241379310344,
469
+ "grad_norm": 27.80811309814453,
470
+ "learning_rate": 1.3833943833943834e-05,
471
+ "loss": 0.2784,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.695172413793103,
476
+ "grad_norm": 123.95394134521484,
477
+ "learning_rate": 1.3681318681318683e-05,
478
+ "loss": 0.4474,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.833103448275862,
483
+ "grad_norm": 56.654483795166016,
484
+ "learning_rate": 1.3528693528693531e-05,
485
+ "loss": 0.5541,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.971034482758621,
490
+ "grad_norm": 38.215049743652344,
491
+ "learning_rate": 1.3376068376068376e-05,
492
+ "loss": 0.322,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 8.0,
497
+ "eval_accuracy": 0.569377990430622,
498
+ "eval_loss": 2.231452226638794,
499
+ "eval_macro_f1": 0.5675325770229606,
500
+ "eval_runtime": 1.2795,
501
+ "eval_samples_per_second": 490.046,
502
+ "eval_steps_per_second": 31.263,
503
+ "eval_weighted_f1": 0.5678424630824139,
504
+ "step": 1456
505
+ },
506
+ {
507
+ "epoch": 8.104827586206897,
508
+ "grad_norm": 9.044121742248535,
509
+ "learning_rate": 1.3223443223443224e-05,
510
+ "loss": 0.1792,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.242758620689655,
515
+ "grad_norm": 75.3625259399414,
516
+ "learning_rate": 1.3070818070818073e-05,
517
+ "loss": 0.1228,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.380689655172414,
522
+ "grad_norm": 63.682559967041016,
523
+ "learning_rate": 1.291819291819292e-05,
524
+ "loss": 0.3982,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.518620689655172,
529
+ "grad_norm": 83.37100219726562,
530
+ "learning_rate": 1.2765567765567766e-05,
531
+ "loss": 0.2978,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.65655172413793,
536
+ "grad_norm": 79.4727554321289,
537
+ "learning_rate": 1.2612942612942613e-05,
538
+ "loss": 0.4788,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.79448275862069,
543
+ "grad_norm": 37.42170715332031,
544
+ "learning_rate": 1.2460317460317461e-05,
545
+ "loss": 0.2772,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.932413793103448,
550
+ "grad_norm": 16.293792724609375,
551
+ "learning_rate": 1.230769230769231e-05,
552
+ "loss": 0.3163,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 9.0,
557
+ "eval_accuracy": 0.5885167464114832,
558
+ "eval_loss": 2.641322374343872,
559
+ "eval_macro_f1": 0.5896711509781644,
560
+ "eval_runtime": 1.4931,
561
+ "eval_samples_per_second": 419.94,
562
+ "eval_steps_per_second": 26.79,
563
+ "eval_weighted_f1": 0.5891970686195149,
564
+ "step": 1638
565
+ }
566
+ ],
567
+ "logging_steps": 25,
568
+ "max_steps": 3640,
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-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-2184/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-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-2184/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-2184/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-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-2184/trainer_state.json ADDED
@@ -0,0 +1,784 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1638,
3
+ "best_metric": 0.5896711509781644,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-00-s43-20260831T010354Z-1a169064/trainer/checkpoint-1638",
5
+ "epoch": 12.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2184,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.13793103448275862,
14
+ "grad_norm": 26.575096130371094,
15
+ "learning_rate": 1.3186813186813187e-06,
16
+ "loss": 4.9635,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27586206896551724,
21
+ "grad_norm": 35.562255859375,
22
+ "learning_rate": 2.6923076923076923e-06,
23
+ "loss": 4.473,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.41379310344827586,
28
+ "grad_norm": 19.921688079833984,
29
+ "learning_rate": 4.065934065934066e-06,
30
+ "loss": 4.481,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5517241379310345,
35
+ "grad_norm": 18.724918365478516,
36
+ "learning_rate": 5.43956043956044e-06,
37
+ "loss": 4.4112,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6896551724137931,
42
+ "grad_norm": 22.945472717285156,
43
+ "learning_rate": 6.813186813186814e-06,
44
+ "loss": 4.3912,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8275862068965517,
49
+ "grad_norm": 27.29220199584961,
50
+ "learning_rate": 8.186813186813188e-06,
51
+ "loss": 4.3679,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.9655172413793104,
56
+ "grad_norm": 19.696191787719727,
57
+ "learning_rate": 9.560439560439562e-06,
58
+ "loss": 4.351,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.45773524720893144,
64
+ "eval_loss": 1.050980567932129,
65
+ "eval_macro_f1": 0.45385903483259726,
66
+ "eval_runtime": 1.1797,
67
+ "eval_samples_per_second": 531.489,
68
+ "eval_steps_per_second": 33.907,
69
+ "eval_weighted_f1": 0.4554201845332243,
70
+ "step": 182
71
+ },
72
+ {
73
+ "epoch": 1.0993103448275863,
74
+ "grad_norm": 32.07797622680664,
75
+ "learning_rate": 1.0934065934065933e-05,
76
+ "loss": 4.1139,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.237241379310345,
81
+ "grad_norm": 15.515581130981445,
82
+ "learning_rate": 1.230769230769231e-05,
83
+ "loss": 4.2158,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3751724137931034,
88
+ "grad_norm": 22.117534637451172,
89
+ "learning_rate": 1.3681318681318683e-05,
90
+ "loss": 4.2599,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.513103448275862,
95
+ "grad_norm": 17.774995803833008,
96
+ "learning_rate": 1.5054945054945056e-05,
97
+ "loss": 4.2231,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6510344827586207,
102
+ "grad_norm": 27.439960479736328,
103
+ "learning_rate": 1.642857142857143e-05,
104
+ "loss": 4.2672,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.7889655172413792,
109
+ "grad_norm": 22.73525047302246,
110
+ "learning_rate": 1.78021978021978e-05,
111
+ "loss": 4.2553,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.926896551724138,
116
+ "grad_norm": 23.057146072387695,
117
+ "learning_rate": 1.9175824175824178e-05,
118
+ "loss": 4.2416,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.4864433811802233,
124
+ "eval_loss": 1.0021166801452637,
125
+ "eval_macro_f1": 0.488904678955645,
126
+ "eval_runtime": 1.2773,
127
+ "eval_samples_per_second": 490.888,
128
+ "eval_steps_per_second": 31.317,
129
+ "eval_weighted_f1": 0.4881585306493283,
130
+ "step": 364
131
+ },
132
+ {
133
+ "epoch": 2.0606896551724136,
134
+ "grad_norm": 49.71544647216797,
135
+ "learning_rate": 1.993894993894994e-05,
136
+ "loss": 3.7333,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.1986206896551725,
141
+ "grad_norm": 28.5469913482666,
142
+ "learning_rate": 1.9786324786324788e-05,
143
+ "loss": 3.9059,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.336551724137931,
148
+ "grad_norm": 37.46480178833008,
149
+ "learning_rate": 1.9633699633699634e-05,
150
+ "loss": 3.4297,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.47448275862069,
155
+ "grad_norm": 41.697845458984375,
156
+ "learning_rate": 1.948107448107448e-05,
157
+ "loss": 3.5016,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.6124137931034483,
162
+ "grad_norm": 35.8870735168457,
163
+ "learning_rate": 1.932844932844933e-05,
164
+ "loss": 3.682,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.750344827586207,
169
+ "grad_norm": 34.63261413574219,
170
+ "learning_rate": 1.9175824175824178e-05,
171
+ "loss": 3.685,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8882758620689657,
176
+ "grad_norm": 32.974422454833984,
177
+ "learning_rate": 1.9023199023199025e-05,
178
+ "loss": 3.7675,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 3.0,
183
+ "eval_accuracy": 0.507177033492823,
184
+ "eval_loss": 0.9895545244216919,
185
+ "eval_macro_f1": 0.4967913731077999,
186
+ "eval_runtime": 1.2454,
187
+ "eval_samples_per_second": 503.436,
188
+ "eval_steps_per_second": 32.117,
189
+ "eval_weighted_f1": 0.49262741180781616,
190
+ "step": 546
191
+ },
192
+ {
193
+ "epoch": 3.0220689655172412,
194
+ "grad_norm": 30.783283233642578,
195
+ "learning_rate": 1.887057387057387e-05,
196
+ "loss": 3.4839,
197
+ "step": 550
198
+ },
199
+ {
200
+ "epoch": 3.16,
201
+ "grad_norm": 79.42498779296875,
202
+ "learning_rate": 1.8717948717948718e-05,
203
+ "loss": 2.86,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2979310344827586,
208
+ "grad_norm": 41.3476448059082,
209
+ "learning_rate": 1.8565323565323568e-05,
210
+ "loss": 2.9271,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.435862068965517,
215
+ "grad_norm": 31.817678451538086,
216
+ "learning_rate": 1.8412698412698415e-05,
217
+ "loss": 3.078,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.573793103448276,
222
+ "grad_norm": 30.916423797607422,
223
+ "learning_rate": 1.826007326007326e-05,
224
+ "loss": 2.6603,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.7117241379310344,
229
+ "grad_norm": 28.779693603515625,
230
+ "learning_rate": 1.8107448107448108e-05,
231
+ "loss": 2.8651,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.849655172413793,
236
+ "grad_norm": 40.502933502197266,
237
+ "learning_rate": 1.7954822954822958e-05,
238
+ "loss": 2.7305,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.9875862068965517,
243
+ "grad_norm": 40.181915283203125,
244
+ "learning_rate": 1.78021978021978e-05,
245
+ "loss": 2.6274,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.5518341307814992,
251
+ "eval_loss": 1.062201976776123,
252
+ "eval_macro_f1": 0.5417668087507094,
253
+ "eval_runtime": 1.2279,
254
+ "eval_samples_per_second": 510.611,
255
+ "eval_steps_per_second": 32.575,
256
+ "eval_weighted_f1": 0.5429427985347581,
257
+ "step": 728
258
+ },
259
+ {
260
+ "epoch": 4.121379310344827,
261
+ "grad_norm": 53.15599822998047,
262
+ "learning_rate": 1.764957264957265e-05,
263
+ "loss": 2.0303,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.259310344827586,
268
+ "grad_norm": 43.258941650390625,
269
+ "learning_rate": 1.74969474969475e-05,
270
+ "loss": 1.8058,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.397241379310345,
275
+ "grad_norm": 69.51874542236328,
276
+ "learning_rate": 1.7344322344322345e-05,
277
+ "loss": 1.7058,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.535172413793103,
282
+ "grad_norm": 45.85198211669922,
283
+ "learning_rate": 1.7191697191697195e-05,
284
+ "loss": 2.1799,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.673103448275862,
289
+ "grad_norm": 57.6066780090332,
290
+ "learning_rate": 1.703907203907204e-05,
291
+ "loss": 1.7652,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.811034482758621,
296
+ "grad_norm": 38.48773956298828,
297
+ "learning_rate": 1.688644688644689e-05,
298
+ "loss": 1.9997,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.94896551724138,
303
+ "grad_norm": 38.666595458984375,
304
+ "learning_rate": 1.6733821733821735e-05,
305
+ "loss": 1.9021,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "eval_accuracy": 0.5374800637958532,
311
+ "eval_loss": 1.168395757675171,
312
+ "eval_macro_f1": 0.5388034546709805,
313
+ "eval_runtime": 1.2689,
314
+ "eval_samples_per_second": 494.123,
315
+ "eval_steps_per_second": 31.523,
316
+ "eval_weighted_f1": 0.5393512866960177,
317
+ "step": 910
318
+ },
319
+ {
320
+ "epoch": 5.082758620689655,
321
+ "grad_norm": 41.31229782104492,
322
+ "learning_rate": 1.6581196581196585e-05,
323
+ "loss": 1.385,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.220689655172414,
328
+ "grad_norm": 101.49195861816406,
329
+ "learning_rate": 1.642857142857143e-05,
330
+ "loss": 1.0578,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.358620689655172,
335
+ "grad_norm": 43.01548767089844,
336
+ "learning_rate": 1.627594627594628e-05,
337
+ "loss": 0.9764,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.496551724137931,
342
+ "grad_norm": 58.39358139038086,
343
+ "learning_rate": 1.6123321123321125e-05,
344
+ "loss": 1.1107,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.63448275862069,
349
+ "grad_norm": 87.73316192626953,
350
+ "learning_rate": 1.5970695970695972e-05,
351
+ "loss": 1.1031,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.772413793103448,
356
+ "grad_norm": 60.960079193115234,
357
+ "learning_rate": 1.581807081807082e-05,
358
+ "loss": 1.0898,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.910344827586207,
363
+ "grad_norm": 44.55670166015625,
364
+ "learning_rate": 1.5665445665445665e-05,
365
+ "loss": 1.0198,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 6.0,
370
+ "eval_accuracy": 0.5534290271132376,
371
+ "eval_loss": 1.4985970258712769,
372
+ "eval_macro_f1": 0.554246483931539,
373
+ "eval_runtime": 1.2881,
374
+ "eval_samples_per_second": 486.77,
375
+ "eval_steps_per_second": 31.054,
376
+ "eval_weighted_f1": 0.5546838726601748,
377
+ "step": 1092
378
+ },
379
+ {
380
+ "epoch": 6.0441379310344825,
381
+ "grad_norm": 23.108503341674805,
382
+ "learning_rate": 1.5512820512820516e-05,
383
+ "loss": 0.9103,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.182068965517241,
388
+ "grad_norm": 32.50311279296875,
389
+ "learning_rate": 1.5360195360195362e-05,
390
+ "loss": 0.6263,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.32,
395
+ "grad_norm": 20.798681259155273,
396
+ "learning_rate": 1.5207570207570207e-05,
397
+ "loss": 0.4834,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.457931034482758,
402
+ "grad_norm": 39.38999557495117,
403
+ "learning_rate": 1.5054945054945056e-05,
404
+ "loss": 0.6147,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.595862068965517,
409
+ "grad_norm": 124.87044525146484,
410
+ "learning_rate": 1.4902319902319904e-05,
411
+ "loss": 0.6238,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.733793103448276,
416
+ "grad_norm": 74.66791534423828,
417
+ "learning_rate": 1.4749694749694752e-05,
418
+ "loss": 0.653,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.871724137931034,
423
+ "grad_norm": 41.76373291015625,
424
+ "learning_rate": 1.4597069597069597e-05,
425
+ "loss": 0.7231,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 7.0,
430
+ "eval_accuracy": 0.5502392344497608,
431
+ "eval_loss": 1.7950698137283325,
432
+ "eval_macro_f1": 0.5524237990246955,
433
+ "eval_runtime": 1.2696,
434
+ "eval_samples_per_second": 493.869,
435
+ "eval_steps_per_second": 31.507,
436
+ "eval_weighted_f1": 0.5523545923272882,
437
+ "step": 1274
438
+ },
439
+ {
440
+ "epoch": 7.0055172413793105,
441
+ "grad_norm": 45.92796325683594,
442
+ "learning_rate": 1.4444444444444446e-05,
443
+ "loss": 0.5914,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.1434482758620685,
448
+ "grad_norm": 80.1987533569336,
449
+ "learning_rate": 1.4291819291819293e-05,
450
+ "loss": 0.3786,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.281379310344827,
455
+ "grad_norm": 31.46441078186035,
456
+ "learning_rate": 1.4139194139194141e-05,
457
+ "loss": 0.4155,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.419310344827586,
462
+ "grad_norm": 58.27080535888672,
463
+ "learning_rate": 1.3986568986568988e-05,
464
+ "loss": 0.3563,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.557241379310344,
469
+ "grad_norm": 27.80811309814453,
470
+ "learning_rate": 1.3833943833943834e-05,
471
+ "loss": 0.2784,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.695172413793103,
476
+ "grad_norm": 123.95394134521484,
477
+ "learning_rate": 1.3681318681318683e-05,
478
+ "loss": 0.4474,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.833103448275862,
483
+ "grad_norm": 56.654483795166016,
484
+ "learning_rate": 1.3528693528693531e-05,
485
+ "loss": 0.5541,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.971034482758621,
490
+ "grad_norm": 38.215049743652344,
491
+ "learning_rate": 1.3376068376068376e-05,
492
+ "loss": 0.322,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 8.0,
497
+ "eval_accuracy": 0.569377990430622,
498
+ "eval_loss": 2.231452226638794,
499
+ "eval_macro_f1": 0.5675325770229606,
500
+ "eval_runtime": 1.2795,
501
+ "eval_samples_per_second": 490.046,
502
+ "eval_steps_per_second": 31.263,
503
+ "eval_weighted_f1": 0.5678424630824139,
504
+ "step": 1456
505
+ },
506
+ {
507
+ "epoch": 8.104827586206897,
508
+ "grad_norm": 9.044121742248535,
509
+ "learning_rate": 1.3223443223443224e-05,
510
+ "loss": 0.1792,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.242758620689655,
515
+ "grad_norm": 75.3625259399414,
516
+ "learning_rate": 1.3070818070818073e-05,
517
+ "loss": 0.1228,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.380689655172414,
522
+ "grad_norm": 63.682559967041016,
523
+ "learning_rate": 1.291819291819292e-05,
524
+ "loss": 0.3982,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.518620689655172,
529
+ "grad_norm": 83.37100219726562,
530
+ "learning_rate": 1.2765567765567766e-05,
531
+ "loss": 0.2978,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.65655172413793,
536
+ "grad_norm": 79.4727554321289,
537
+ "learning_rate": 1.2612942612942613e-05,
538
+ "loss": 0.4788,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.79448275862069,
543
+ "grad_norm": 37.42170715332031,
544
+ "learning_rate": 1.2460317460317461e-05,
545
+ "loss": 0.2772,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 8.932413793103448,
550
+ "grad_norm": 16.293792724609375,
551
+ "learning_rate": 1.230769230769231e-05,
552
+ "loss": 0.3163,
553
+ "step": 1625
554
+ },
555
+ {
556
+ "epoch": 9.0,
557
+ "eval_accuracy": 0.5885167464114832,
558
+ "eval_loss": 2.641322374343872,
559
+ "eval_macro_f1": 0.5896711509781644,
560
+ "eval_runtime": 1.4931,
561
+ "eval_samples_per_second": 419.94,
562
+ "eval_steps_per_second": 26.79,
563
+ "eval_weighted_f1": 0.5891970686195149,
564
+ "step": 1638
565
+ },
566
+ {
567
+ "epoch": 9.066206896551725,
568
+ "grad_norm": 0.9830600023269653,
569
+ "learning_rate": 1.2155067155067155e-05,
570
+ "loss": 0.1527,
571
+ "step": 1650
572
+ },
573
+ {
574
+ "epoch": 9.204137931034483,
575
+ "grad_norm": 55.66547775268555,
576
+ "learning_rate": 1.2002442002442003e-05,
577
+ "loss": 0.1769,
578
+ "step": 1675
579
+ },
580
+ {
581
+ "epoch": 9.34206896551724,
582
+ "grad_norm": 25.83209228515625,
583
+ "learning_rate": 1.1849816849816852e-05,
584
+ "loss": 0.1895,
585
+ "step": 1700
586
+ },
587
+ {
588
+ "epoch": 9.48,
589
+ "grad_norm": 24.788331985473633,
590
+ "learning_rate": 1.1697191697191698e-05,
591
+ "loss": 0.2111,
592
+ "step": 1725
593
+ },
594
+ {
595
+ "epoch": 9.617931034482758,
596
+ "grad_norm": 115.94087982177734,
597
+ "learning_rate": 1.1544566544566545e-05,
598
+ "loss": 0.2086,
599
+ "step": 1750
600
+ },
601
+ {
602
+ "epoch": 9.755862068965516,
603
+ "grad_norm": 10.962858200073242,
604
+ "learning_rate": 1.1391941391941392e-05,
605
+ "loss": 0.2576,
606
+ "step": 1775
607
+ },
608
+ {
609
+ "epoch": 9.893793103448276,
610
+ "grad_norm": 11.381913185119629,
611
+ "learning_rate": 1.123931623931624e-05,
612
+ "loss": 0.2265,
613
+ "step": 1800
614
+ },
615
+ {
616
+ "epoch": 10.0,
617
+ "eval_accuracy": 0.5582137161084529,
618
+ "eval_loss": 3.2752583026885986,
619
+ "eval_macro_f1": 0.5575698420396434,
620
+ "eval_runtime": 1.3019,
621
+ "eval_samples_per_second": 481.618,
622
+ "eval_steps_per_second": 30.725,
623
+ "eval_weighted_f1": 0.5580266980369962,
624
+ "step": 1820
625
+ },
626
+ {
627
+ "epoch": 10.027586206896551,
628
+ "grad_norm": 13.521539688110352,
629
+ "learning_rate": 1.1086691086691088e-05,
630
+ "loss": 0.1742,
631
+ "step": 1825
632
+ },
633
+ {
634
+ "epoch": 10.16551724137931,
635
+ "grad_norm": 53.98459243774414,
636
+ "learning_rate": 1.0934065934065933e-05,
637
+ "loss": 0.1475,
638
+ "step": 1850
639
+ },
640
+ {
641
+ "epoch": 10.303448275862069,
642
+ "grad_norm": 24.670190811157227,
643
+ "learning_rate": 1.0781440781440782e-05,
644
+ "loss": 0.0892,
645
+ "step": 1875
646
+ },
647
+ {
648
+ "epoch": 10.441379310344828,
649
+ "grad_norm": 0.26385369896888733,
650
+ "learning_rate": 1.062881562881563e-05,
651
+ "loss": 0.183,
652
+ "step": 1900
653
+ },
654
+ {
655
+ "epoch": 10.579310344827586,
656
+ "grad_norm": 0.3430776596069336,
657
+ "learning_rate": 1.0476190476190477e-05,
658
+ "loss": 0.1811,
659
+ "step": 1925
660
+ },
661
+ {
662
+ "epoch": 10.717241379310344,
663
+ "grad_norm": 53.72909164428711,
664
+ "learning_rate": 1.0323565323565325e-05,
665
+ "loss": 0.2266,
666
+ "step": 1950
667
+ },
668
+ {
669
+ "epoch": 10.855172413793104,
670
+ "grad_norm": 12.592849731445312,
671
+ "learning_rate": 1.0170940170940172e-05,
672
+ "loss": 0.2719,
673
+ "step": 1975
674
+ },
675
+ {
676
+ "epoch": 10.993103448275862,
677
+ "grad_norm": 89.65878295898438,
678
+ "learning_rate": 1.0018315018315019e-05,
679
+ "loss": 0.108,
680
+ "step": 2000
681
+ },
682
+ {
683
+ "epoch": 11.0,
684
+ "eval_accuracy": 0.5486443381180224,
685
+ "eval_loss": 3.7517478466033936,
686
+ "eval_macro_f1": 0.5517201292729753,
687
+ "eval_runtime": 1.4669,
688
+ "eval_samples_per_second": 427.419,
689
+ "eval_steps_per_second": 27.268,
690
+ "eval_weighted_f1": 0.5512357550130131,
691
+ "step": 2002
692
+ },
693
+ {
694
+ "epoch": 11.126896551724139,
695
+ "grad_norm": 19.571704864501953,
696
+ "learning_rate": 9.865689865689867e-06,
697
+ "loss": 0.0642,
698
+ "step": 2025
699
+ },
700
+ {
701
+ "epoch": 11.264827586206897,
702
+ "grad_norm": 0.30397388339042664,
703
+ "learning_rate": 9.713064713064714e-06,
704
+ "loss": 0.062,
705
+ "step": 2050
706
+ },
707
+ {
708
+ "epoch": 11.402758620689655,
709
+ "grad_norm": 86.4576187133789,
710
+ "learning_rate": 9.560439560439562e-06,
711
+ "loss": 0.0723,
712
+ "step": 2075
713
+ },
714
+ {
715
+ "epoch": 11.540689655172415,
716
+ "grad_norm": 120.63780975341797,
717
+ "learning_rate": 9.407814407814409e-06,
718
+ "loss": 0.0874,
719
+ "step": 2100
720
+ },
721
+ {
722
+ "epoch": 11.678620689655173,
723
+ "grad_norm": 2.5405189990997314,
724
+ "learning_rate": 9.255189255189256e-06,
725
+ "loss": 0.1203,
726
+ "step": 2125
727
+ },
728
+ {
729
+ "epoch": 11.81655172413793,
730
+ "grad_norm": 102.72106170654297,
731
+ "learning_rate": 9.102564102564104e-06,
732
+ "loss": 0.1588,
733
+ "step": 2150
734
+ },
735
+ {
736
+ "epoch": 11.95448275862069,
737
+ "grad_norm": 0.8757811784744263,
738
+ "learning_rate": 8.94993894993895e-06,
739
+ "loss": 0.1741,
740
+ "step": 2175
741
+ },
742
+ {
743
+ "epoch": 12.0,
744
+ "eval_accuracy": 0.5486443381180224,
745
+ "eval_loss": 3.927983522415161,
746
+ "eval_macro_f1": 0.5513051600402706,
747
+ "eval_runtime": 1.4525,
748
+ "eval_samples_per_second": 431.664,
749
+ "eval_steps_per_second": 27.538,
750
+ "eval_weighted_f1": 0.5508568592546262,
751
+ "step": 2184
752
+ }
753
+ ],
754
+ "logging_steps": 25,
755
+ "max_steps": 3640,
756
+ "num_input_tokens_seen": 0,
757
+ "num_train_epochs": 20,
758
+ "save_steps": 500,
759
+ "stateful_callbacks": {
760
+ "EarlyStoppingCallback": {
761
+ "args": {
762
+ "early_stopping_patience": 3,
763
+ "early_stopping_threshold": 0.0
764
+ },
765
+ "attributes": {
766
+ "early_stopping_patience_counter": 3
767
+ }
768
+ },
769
+ "TrainerControl": {
770
+ "args": {
771
+ "should_epoch_stop": false,
772
+ "should_evaluate": false,
773
+ "should_log": false,
774
+ "should_save": true,
775
+ "should_training_stop": true
776
+ },
777
+ "attributes": {}
778
+ }
779
+ },
780
+ "total_flos": 0.0,
781
+ "train_batch_size": 8,
782
+ "trial_name": null,
783
+ "trial_params": null
784
+ }
ood-vifactcheck-visobert-source-source-01-s43-20260831T011628Z-1a169064/trainer/checkpoint-855/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-855/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-855/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-s44-20260831T033516Z-59e46264/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-s44-20260831T033516Z-59e46264/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-s44-20260831T033516Z-59e46264/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-s44-20260831T033516Z-59e46264/trainer/checkpoint-1620/trainer_state.json ADDED
@@ -0,0 +1,590 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 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": 29.567710876464844,
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
+ "epoch": 6.111111111111111,
381
+ "grad_norm": 71.3809585571289,
382
+ "learning_rate": 1.5438271604938274e-05,
383
+ "loss": 0.5578,
384
+ "step": 1100
385
+ },
386
+ {
387
+ "epoch": 6.25,
388
+ "grad_norm": 58.372955322265625,
389
+ "learning_rate": 1.5283950617283952e-05,
390
+ "loss": 0.6027,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.388888888888889,
395
+ "grad_norm": 59.98765182495117,
396
+ "learning_rate": 1.5129629629629632e-05,
397
+ "loss": 0.5687,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.527777777777778,
402
+ "grad_norm": 34.29204177856445,
403
+ "learning_rate": 1.4975308641975311e-05,
404
+ "loss": 0.4372,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.666666666666667,
409
+ "grad_norm": 62.86279296875,
410
+ "learning_rate": 1.4820987654320989e-05,
411
+ "loss": 0.8426,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.805555555555555,
416
+ "grad_norm": 80.83411407470703,
417
+ "learning_rate": 1.4666666666666666e-05,
418
+ "loss": 0.6073,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.944444444444445,
423
+ "grad_norm": 51.71305847167969,
424
+ "learning_rate": 1.4512345679012347e-05,
425
+ "loss": 0.7848,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 7.0,
430
+ "eval_accuracy": 0.5552238805970149,
431
+ "eval_loss": 1.8124223947525024,
432
+ "eval_macro_f1": 0.5581148995698336,
433
+ "eval_runtime": 1.3307,
434
+ "eval_samples_per_second": 503.493,
435
+ "eval_steps_per_second": 31.562,
436
+ "eval_weighted_f1": 0.5584028996283782,
437
+ "step": 1260
438
+ },
439
+ {
440
+ "epoch": 7.083333333333333,
441
+ "grad_norm": 31.017576217651367,
442
+ "learning_rate": 1.4358024691358025e-05,
443
+ "loss": 0.5165,
444
+ "step": 1275
445
+ },
446
+ {
447
+ "epoch": 7.222222222222222,
448
+ "grad_norm": 43.640228271484375,
449
+ "learning_rate": 1.4203703703703704e-05,
450
+ "loss": 0.3103,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.361111111111111,
455
+ "grad_norm": 43.77082824707031,
456
+ "learning_rate": 1.4049382716049384e-05,
457
+ "loss": 0.4042,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.5,
462
+ "grad_norm": 19.751436233520508,
463
+ "learning_rate": 1.3895061728395063e-05,
464
+ "loss": 0.3762,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.638888888888889,
469
+ "grad_norm": 25.015094757080078,
470
+ "learning_rate": 1.3740740740740741e-05,
471
+ "loss": 0.5138,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.777777777777778,
476
+ "grad_norm": 32.85478210449219,
477
+ "learning_rate": 1.3586419753086422e-05,
478
+ "loss": 0.378,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.916666666666667,
483
+ "grad_norm": 53.80678939819336,
484
+ "learning_rate": 1.34320987654321e-05,
485
+ "loss": 0.3303,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 8.0,
490
+ "eval_accuracy": 0.5865671641791045,
491
+ "eval_loss": 2.1597299575805664,
492
+ "eval_macro_f1": 0.5904484160136129,
493
+ "eval_runtime": 1.3041,
494
+ "eval_samples_per_second": 513.75,
495
+ "eval_steps_per_second": 32.205,
496
+ "eval_weighted_f1": 0.5903436051015648,
497
+ "step": 1440
498
+ },
499
+ {
500
+ "epoch": 8.055555555555555,
501
+ "grad_norm": 51.572452545166016,
502
+ "learning_rate": 1.3277777777777779e-05,
503
+ "loss": 0.3088,
504
+ "step": 1450
505
+ },
506
+ {
507
+ "epoch": 8.194444444444445,
508
+ "grad_norm": 4.196484088897705,
509
+ "learning_rate": 1.3123456790123458e-05,
510
+ "loss": 0.2177,
511
+ "step": 1475
512
+ },
513
+ {
514
+ "epoch": 8.333333333333334,
515
+ "grad_norm": 49.75428009033203,
516
+ "learning_rate": 1.2969135802469138e-05,
517
+ "loss": 0.2965,
518
+ "step": 1500
519
+ },
520
+ {
521
+ "epoch": 8.472222222222221,
522
+ "grad_norm": 40.72304916381836,
523
+ "learning_rate": 1.2814814814814815e-05,
524
+ "loss": 0.2597,
525
+ "step": 1525
526
+ },
527
+ {
528
+ "epoch": 8.61111111111111,
529
+ "grad_norm": 7.4955949783325195,
530
+ "learning_rate": 1.2660493827160493e-05,
531
+ "loss": 0.1463,
532
+ "step": 1550
533
+ },
534
+ {
535
+ "epoch": 8.75,
536
+ "grad_norm": 52.93016815185547,
537
+ "learning_rate": 1.2506172839506174e-05,
538
+ "loss": 0.2276,
539
+ "step": 1575
540
+ },
541
+ {
542
+ "epoch": 8.88888888888889,
543
+ "grad_norm": 107.95427703857422,
544
+ "learning_rate": 1.2351851851851852e-05,
545
+ "loss": 0.3302,
546
+ "step": 1600
547
+ },
548
+ {
549
+ "epoch": 9.0,
550
+ "eval_accuracy": 0.5626865671641791,
551
+ "eval_loss": 2.6830008029937744,
552
+ "eval_macro_f1": 0.5651507918342434,
553
+ "eval_runtime": 1.5159,
554
+ "eval_samples_per_second": 441.986,
555
+ "eval_steps_per_second": 27.707,
556
+ "eval_weighted_f1": 0.5652171331160957,
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": 3
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": true
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-06-s42-20260831T003231Z-f21e2f7e/trainer/checkpoint-1480/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
ood-vifactcheck-visobert-source-source-06-s42-20260831T003231Z-f21e2f7e/trainer/checkpoint-1480/trainer_state.json ADDED
@@ -0,0 +1,544 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1480,
3
+ "best_metric": 0.58596362354947,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-06-s42-20260831T003231Z-f21e2f7e/trainer/checkpoint-1480",
5
+ "epoch": 8.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1480,
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": 20.357709884643555,
15
+ "learning_rate": 1.2972972972972974e-06,
16
+ "loss": 4.6716,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27100271002710025,
21
+ "grad_norm": 25.524555206298828,
22
+ "learning_rate": 2.648648648648649e-06,
23
+ "loss": 4.4931,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4065040650406504,
28
+ "grad_norm": 30.60616111755371,
29
+ "learning_rate": 4.000000000000001e-06,
30
+ "loss": 4.451,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5420054200542005,
35
+ "grad_norm": 28.622480392456055,
36
+ "learning_rate": 5.351351351351351e-06,
37
+ "loss": 4.4259,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6775067750677507,
42
+ "grad_norm": 27.390562057495117,
43
+ "learning_rate": 6.702702702702704e-06,
44
+ "loss": 4.4619,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8130081300813008,
49
+ "grad_norm": 28.70516586303711,
50
+ "learning_rate": 8.054054054054055e-06,
51
+ "loss": 4.3448,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.948509485094851,
56
+ "grad_norm": 20.170974731445312,
57
+ "learning_rate": 9.405405405405407e-06,
58
+ "loss": 4.4515,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.4159891598915989,
64
+ "eval_loss": 1.0661125183105469,
65
+ "eval_macro_f1": 0.3927685716277139,
66
+ "eval_runtime": 1.5313,
67
+ "eval_samples_per_second": 481.955,
68
+ "eval_steps_per_second": 30.694,
69
+ "eval_weighted_f1": 0.3954841313234317,
70
+ "step": 185
71
+ },
72
+ {
73
+ "epoch": 1.08130081300813,
74
+ "grad_norm": 19.588459014892578,
75
+ "learning_rate": 1.0756756756756757e-05,
76
+ "loss": 4.194,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2168021680216803,
81
+ "grad_norm": 22.15048599243164,
82
+ "learning_rate": 1.210810810810811e-05,
83
+ "loss": 4.3269,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3523035230352303,
88
+ "grad_norm": 20.58547592163086,
89
+ "learning_rate": 1.3459459459459461e-05,
90
+ "loss": 4.1278,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4878048780487805,
95
+ "grad_norm": 47.9154052734375,
96
+ "learning_rate": 1.4810810810810811e-05,
97
+ "loss": 4.2197,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6233062330623307,
102
+ "grad_norm": 26.602832794189453,
103
+ "learning_rate": 1.6162162162162163e-05,
104
+ "loss": 4.1971,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.758807588075881,
109
+ "grad_norm": 16.841196060180664,
110
+ "learning_rate": 1.7513513513513517e-05,
111
+ "loss": 4.1891,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.8943089430894309,
116
+ "grad_norm": 20.704299926757812,
117
+ "learning_rate": 1.8864864864864866e-05,
118
+ "loss": 4.3157,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.39295392953929537,
124
+ "eval_loss": 1.1406210660934448,
125
+ "eval_macro_f1": 0.33649005900799467,
126
+ "eval_runtime": 1.5201,
127
+ "eval_samples_per_second": 485.486,
128
+ "eval_steps_per_second": 30.918,
129
+ "eval_weighted_f1": 0.3303605175364009,
130
+ "step": 370
131
+ },
132
+ {
133
+ "epoch": 2.02710027100271,
134
+ "grad_norm": 23.664655685424805,
135
+ "learning_rate": 1.9975975975975978e-05,
136
+ "loss": 3.9594,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.16260162601626,
141
+ "grad_norm": 31.630964279174805,
142
+ "learning_rate": 1.9825825825825828e-05,
143
+ "loss": 3.7926,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2981029810298104,
148
+ "grad_norm": 46.86356735229492,
149
+ "learning_rate": 1.967567567567568e-05,
150
+ "loss": 3.797,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.4336043360433606,
155
+ "grad_norm": 35.555259704589844,
156
+ "learning_rate": 1.9525525525525525e-05,
157
+ "loss": 3.6979,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.569105691056911,
162
+ "grad_norm": 21.405046463012695,
163
+ "learning_rate": 1.9375375375375376e-05,
164
+ "loss": 3.5943,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7046070460704605,
169
+ "grad_norm": 44.80155944824219,
170
+ "learning_rate": 1.9225225225225226e-05,
171
+ "loss": 3.4968,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8401084010840107,
176
+ "grad_norm": 51.13383483886719,
177
+ "learning_rate": 1.9075075075075076e-05,
178
+ "loss": 3.7023,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.975609756097561,
183
+ "grad_norm": 22.87540626525879,
184
+ "learning_rate": 1.8924924924924926e-05,
185
+ "loss": 3.6388,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.5365853658536586,
191
+ "eval_loss": 0.9346652030944824,
192
+ "eval_macro_f1": 0.5358602523969763,
193
+ "eval_runtime": 1.4809,
194
+ "eval_samples_per_second": 498.353,
195
+ "eval_steps_per_second": 31.738,
196
+ "eval_weighted_f1": 0.5349633509782014,
197
+ "step": 555
198
+ },
199
+ {
200
+ "epoch": 3.1084010840108403,
201
+ "grad_norm": 46.90874481201172,
202
+ "learning_rate": 1.8774774774774776e-05,
203
+ "loss": 2.9077,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2439024390243905,
208
+ "grad_norm": 63.010780334472656,
209
+ "learning_rate": 1.8624624624624627e-05,
210
+ "loss": 2.7618,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.3794037940379402,
215
+ "grad_norm": 28.250125885009766,
216
+ "learning_rate": 1.8474474474474477e-05,
217
+ "loss": 2.7634,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5149051490514904,
222
+ "grad_norm": 43.522178649902344,
223
+ "learning_rate": 1.8324324324324324e-05,
224
+ "loss": 2.965,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6504065040650406,
229
+ "grad_norm": 44.616695404052734,
230
+ "learning_rate": 1.8174174174174174e-05,
231
+ "loss": 2.7232,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.785907859078591,
236
+ "grad_norm": 64.59855651855469,
237
+ "learning_rate": 1.8024024024024024e-05,
238
+ "loss": 2.7811,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.921409214092141,
243
+ "grad_norm": 48.08187484741211,
244
+ "learning_rate": 1.7873873873873874e-05,
245
+ "loss": 2.7656,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.5718157181571816,
251
+ "eval_loss": 0.9289685487747192,
252
+ "eval_macro_f1": 0.5694014592940978,
253
+ "eval_runtime": 1.6006,
254
+ "eval_samples_per_second": 461.087,
255
+ "eval_steps_per_second": 29.365,
256
+ "eval_weighted_f1": 0.5699293802086568,
257
+ "step": 740
258
+ },
259
+ {
260
+ "epoch": 4.05420054200542,
261
+ "grad_norm": 37.47281265258789,
262
+ "learning_rate": 1.7723723723723724e-05,
263
+ "loss": 2.4828,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.1897018970189706,
268
+ "grad_norm": 50.268280029296875,
269
+ "learning_rate": 1.7573573573573575e-05,
270
+ "loss": 1.8789,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.32520325203252,
275
+ "grad_norm": 49.67667007446289,
276
+ "learning_rate": 1.7423423423423425e-05,
277
+ "loss": 2.0543,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.46070460704607,
282
+ "grad_norm": 43.392860412597656,
283
+ "learning_rate": 1.7273273273273275e-05,
284
+ "loss": 1.9123,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.596205962059621,
289
+ "grad_norm": 69.00615692138672,
290
+ "learning_rate": 1.7123123123123125e-05,
291
+ "loss": 1.8992,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.7317073170731705,
296
+ "grad_norm": 38.3200569152832,
297
+ "learning_rate": 1.6972972972972975e-05,
298
+ "loss": 1.9871,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.867208672086721,
303
+ "grad_norm": 62.512229919433594,
304
+ "learning_rate": 1.6822822822822826e-05,
305
+ "loss": 1.9823,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "grad_norm": 22.450998306274414,
311
+ "learning_rate": 1.6672672672672673e-05,
312
+ "loss": 2.102,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 5.0,
317
+ "eval_accuracy": 0.5772357723577236,
318
+ "eval_loss": 1.02829909324646,
319
+ "eval_macro_f1": 0.5762566137566137,
320
+ "eval_runtime": 1.5422,
321
+ "eval_samples_per_second": 478.545,
322
+ "eval_steps_per_second": 30.476,
323
+ "eval_weighted_f1": 0.5769282057899944,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.13550135501355,
328
+ "grad_norm": 50.8524169921875,
329
+ "learning_rate": 1.6522522522522523e-05,
330
+ "loss": 1.2148,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.2710027100271,
335
+ "grad_norm": 29.982929229736328,
336
+ "learning_rate": 1.6372372372372373e-05,
337
+ "loss": 1.1953,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.40650406504065,
342
+ "grad_norm": 44.76668930053711,
343
+ "learning_rate": 1.6222222222222223e-05,
344
+ "loss": 1.0117,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.542005420054201,
349
+ "grad_norm": 70.34624481201172,
350
+ "learning_rate": 1.6072072072072073e-05,
351
+ "loss": 1.2226,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.677506775067751,
356
+ "grad_norm": 90.42357635498047,
357
+ "learning_rate": 1.5921921921921924e-05,
358
+ "loss": 1.0973,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.8130081300813,
363
+ "grad_norm": 79.31895446777344,
364
+ "learning_rate": 1.5771771771771774e-05,
365
+ "loss": 1.2688,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.948509485094851,
370
+ "grad_norm": 70.91158294677734,
371
+ "learning_rate": 1.5621621621621624e-05,
372
+ "loss": 1.1066,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 6.0,
377
+ "eval_accuracy": 0.5704607046070461,
378
+ "eval_loss": 1.3678768873214722,
379
+ "eval_macro_f1": 0.5727491315228105,
380
+ "eval_runtime": 1.5767,
381
+ "eval_samples_per_second": 468.062,
382
+ "eval_steps_per_second": 29.809,
383
+ "eval_weighted_f1": 0.5731608927877845,
384
+ "step": 1110
385
+ },
386
+ {
387
+ "epoch": 6.08130081300813,
388
+ "grad_norm": 25.119922637939453,
389
+ "learning_rate": 1.5471471471471474e-05,
390
+ "loss": 0.9272,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.2168021680216805,
395
+ "grad_norm": 79.80097961425781,
396
+ "learning_rate": 1.5321321321321324e-05,
397
+ "loss": 0.4529,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.35230352303523,
402
+ "grad_norm": 58.90565872192383,
403
+ "learning_rate": 1.5171171171171173e-05,
404
+ "loss": 0.624,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.487804878048781,
409
+ "grad_norm": 47.078086853027344,
410
+ "learning_rate": 1.5021021021021023e-05,
411
+ "loss": 0.8501,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.623306233062331,
416
+ "grad_norm": 77.01166534423828,
417
+ "learning_rate": 1.4870870870870873e-05,
418
+ "loss": 0.5812,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.7588075880758804,
423
+ "grad_norm": 43.48472213745117,
424
+ "learning_rate": 1.4720720720720722e-05,
425
+ "loss": 0.8517,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.894308943089431,
430
+ "grad_norm": 21.808334350585938,
431
+ "learning_rate": 1.4570570570570572e-05,
432
+ "loss": 0.5582,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 7.0,
437
+ "eval_accuracy": 0.5623306233062331,
438
+ "eval_loss": 1.7512412071228027,
439
+ "eval_macro_f1": 0.5648117316305371,
440
+ "eval_runtime": 1.559,
441
+ "eval_samples_per_second": 473.381,
442
+ "eval_steps_per_second": 30.148,
443
+ "eval_weighted_f1": 0.5644118241533768,
444
+ "step": 1295
445
+ },
446
+ {
447
+ "epoch": 7.02710027100271,
448
+ "grad_norm": 15.119675636291504,
449
+ "learning_rate": 1.4420420420420422e-05,
450
+ "loss": 0.6254,
451
+ "step": 1300
452
+ },
453
+ {
454
+ "epoch": 7.16260162601626,
455
+ "grad_norm": 31.58051300048828,
456
+ "learning_rate": 1.4270270270270272e-05,
457
+ "loss": 0.3593,
458
+ "step": 1325
459
+ },
460
+ {
461
+ "epoch": 7.29810298102981,
462
+ "grad_norm": 47.139869689941406,
463
+ "learning_rate": 1.4120120120120123e-05,
464
+ "loss": 0.3728,
465
+ "step": 1350
466
+ },
467
+ {
468
+ "epoch": 7.43360433604336,
469
+ "grad_norm": 42.259700775146484,
470
+ "learning_rate": 1.3969969969969971e-05,
471
+ "loss": 0.3558,
472
+ "step": 1375
473
+ },
474
+ {
475
+ "epoch": 7.569105691056911,
476
+ "grad_norm": 43.38686752319336,
477
+ "learning_rate": 1.3819819819819821e-05,
478
+ "loss": 0.401,
479
+ "step": 1400
480
+ },
481
+ {
482
+ "epoch": 7.7046070460704605,
483
+ "grad_norm": 9.104645729064941,
484
+ "learning_rate": 1.3669669669669672e-05,
485
+ "loss": 0.3715,
486
+ "step": 1425
487
+ },
488
+ {
489
+ "epoch": 7.840108401084011,
490
+ "grad_norm": 24.254037857055664,
491
+ "learning_rate": 1.3519519519519522e-05,
492
+ "loss": 0.3486,
493
+ "step": 1450
494
+ },
495
+ {
496
+ "epoch": 7.975609756097561,
497
+ "grad_norm": 62.172889709472656,
498
+ "learning_rate": 1.3369369369369369e-05,
499
+ "loss": 0.331,
500
+ "step": 1475
501
+ },
502
+ {
503
+ "epoch": 8.0,
504
+ "eval_accuracy": 0.5880758807588076,
505
+ "eval_loss": 2.284409999847412,
506
+ "eval_macro_f1": 0.58596362354947,
507
+ "eval_runtime": 1.5115,
508
+ "eval_samples_per_second": 488.244,
509
+ "eval_steps_per_second": 31.094,
510
+ "eval_weighted_f1": 0.5861948528770883,
511
+ "step": 1480
512
+ }
513
+ ],
514
+ "logging_steps": 25,
515
+ "max_steps": 3700,
516
+ "num_input_tokens_seen": 0,
517
+ "num_train_epochs": 20,
518
+ "save_steps": 500,
519
+ "stateful_callbacks": {
520
+ "EarlyStoppingCallback": {
521
+ "args": {
522
+ "early_stopping_patience": 3,
523
+ "early_stopping_threshold": 0.0
524
+ },
525
+ "attributes": {
526
+ "early_stopping_patience_counter": 0
527
+ }
528
+ },
529
+ "TrainerControl": {
530
+ "args": {
531
+ "should_epoch_stop": false,
532
+ "should_evaluate": false,
533
+ "should_log": false,
534
+ "should_save": true,
535
+ "should_training_stop": false
536
+ },
537
+ "attributes": {}
538
+ }
539
+ },
540
+ "total_flos": 0.0,
541
+ "train_batch_size": 8,
542
+ "trial_name": null,
543
+ "trial_params": null
544
+ }
ood-vifactcheck-visobert-source-source-06-s44-20260831T034444Z-59e46264/trainer/checkpoint-1295/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-s44-20260831T034444Z-59e46264/trainer/checkpoint-1295/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-s44-20260831T034444Z-59e46264/trainer/checkpoint-1295/trainer_state.json ADDED
@@ -0,0 +1,477 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1295,
3
+ "best_metric": 0.5823950641665662,
4
+ "best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-visobert-source-source-06-s44-20260831T034444Z-59e46264/trainer/checkpoint-1295",
5
+ "epoch": 7.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1295,
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": 43.989139556884766,
15
+ "learning_rate": 1.2972972972972974e-06,
16
+ "loss": 4.6491,
17
+ "step": 25
18
+ },
19
+ {
20
+ "epoch": 0.27100271002710025,
21
+ "grad_norm": 19.655298233032227,
22
+ "learning_rate": 2.648648648648649e-06,
23
+ "loss": 4.5705,
24
+ "step": 50
25
+ },
26
+ {
27
+ "epoch": 0.4065040650406504,
28
+ "grad_norm": 30.5579891204834,
29
+ "learning_rate": 4.000000000000001e-06,
30
+ "loss": 4.4999,
31
+ "step": 75
32
+ },
33
+ {
34
+ "epoch": 0.5420054200542005,
35
+ "grad_norm": 20.41074562072754,
36
+ "learning_rate": 5.351351351351351e-06,
37
+ "loss": 4.4274,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.6775067750677507,
42
+ "grad_norm": 30.664974212646484,
43
+ "learning_rate": 6.702702702702704e-06,
44
+ "loss": 4.4315,
45
+ "step": 125
46
+ },
47
+ {
48
+ "epoch": 0.8130081300813008,
49
+ "grad_norm": 34.81926345825195,
50
+ "learning_rate": 8.054054054054055e-06,
51
+ "loss": 4.4835,
52
+ "step": 150
53
+ },
54
+ {
55
+ "epoch": 0.948509485094851,
56
+ "grad_norm": 22.79718017578125,
57
+ "learning_rate": 9.405405405405407e-06,
58
+ "loss": 4.3666,
59
+ "step": 175
60
+ },
61
+ {
62
+ "epoch": 1.0,
63
+ "eval_accuracy": 0.42276422764227645,
64
+ "eval_loss": 1.0699764490127563,
65
+ "eval_macro_f1": 0.4241342152866007,
66
+ "eval_runtime": 1.5197,
67
+ "eval_samples_per_second": 485.63,
68
+ "eval_steps_per_second": 30.928,
69
+ "eval_weighted_f1": 0.4250737362705159,
70
+ "step": 185
71
+ },
72
+ {
73
+ "epoch": 1.08130081300813,
74
+ "grad_norm": 29.870878219604492,
75
+ "learning_rate": 1.0756756756756757e-05,
76
+ "loss": 4.1584,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 1.2168021680216803,
81
+ "grad_norm": 27.648677825927734,
82
+ "learning_rate": 1.210810810810811e-05,
83
+ "loss": 4.2634,
84
+ "step": 225
85
+ },
86
+ {
87
+ "epoch": 1.3523035230352303,
88
+ "grad_norm": 33.50834274291992,
89
+ "learning_rate": 1.3459459459459461e-05,
90
+ "loss": 4.2228,
91
+ "step": 250
92
+ },
93
+ {
94
+ "epoch": 1.4878048780487805,
95
+ "grad_norm": 25.84048080444336,
96
+ "learning_rate": 1.4810810810810811e-05,
97
+ "loss": 4.2876,
98
+ "step": 275
99
+ },
100
+ {
101
+ "epoch": 1.6233062330623307,
102
+ "grad_norm": 46.06636428833008,
103
+ "learning_rate": 1.6162162162162163e-05,
104
+ "loss": 4.2973,
105
+ "step": 300
106
+ },
107
+ {
108
+ "epoch": 1.758807588075881,
109
+ "grad_norm": 44.92756652832031,
110
+ "learning_rate": 1.7513513513513517e-05,
111
+ "loss": 4.3512,
112
+ "step": 325
113
+ },
114
+ {
115
+ "epoch": 1.8943089430894309,
116
+ "grad_norm": 23.320087432861328,
117
+ "learning_rate": 1.8864864864864866e-05,
118
+ "loss": 4.2766,
119
+ "step": 350
120
+ },
121
+ {
122
+ "epoch": 2.0,
123
+ "eval_accuracy": 0.4349593495934959,
124
+ "eval_loss": 1.0430904626846313,
125
+ "eval_macro_f1": 0.42940890242703683,
126
+ "eval_runtime": 1.4271,
127
+ "eval_samples_per_second": 517.141,
128
+ "eval_steps_per_second": 32.934,
129
+ "eval_weighted_f1": 0.42798523454751686,
130
+ "step": 370
131
+ },
132
+ {
133
+ "epoch": 2.02710027100271,
134
+ "grad_norm": 14.736526489257812,
135
+ "learning_rate": 1.9975975975975978e-05,
136
+ "loss": 4.0595,
137
+ "step": 375
138
+ },
139
+ {
140
+ "epoch": 2.16260162601626,
141
+ "grad_norm": 30.984207153320312,
142
+ "learning_rate": 1.9825825825825828e-05,
143
+ "loss": 3.7606,
144
+ "step": 400
145
+ },
146
+ {
147
+ "epoch": 2.2981029810298104,
148
+ "grad_norm": 31.585784912109375,
149
+ "learning_rate": 1.967567567567568e-05,
150
+ "loss": 3.7101,
151
+ "step": 425
152
+ },
153
+ {
154
+ "epoch": 2.4336043360433606,
155
+ "grad_norm": 23.017831802368164,
156
+ "learning_rate": 1.9525525525525525e-05,
157
+ "loss": 3.7168,
158
+ "step": 450
159
+ },
160
+ {
161
+ "epoch": 2.569105691056911,
162
+ "grad_norm": 29.353456497192383,
163
+ "learning_rate": 1.9375375375375376e-05,
164
+ "loss": 3.615,
165
+ "step": 475
166
+ },
167
+ {
168
+ "epoch": 2.7046070460704605,
169
+ "grad_norm": 26.89971351623535,
170
+ "learning_rate": 1.9225225225225226e-05,
171
+ "loss": 3.8652,
172
+ "step": 500
173
+ },
174
+ {
175
+ "epoch": 2.8401084010840107,
176
+ "grad_norm": 45.5249137878418,
177
+ "learning_rate": 1.9075075075075076e-05,
178
+ "loss": 3.7074,
179
+ "step": 525
180
+ },
181
+ {
182
+ "epoch": 2.975609756097561,
183
+ "grad_norm": 44.861595153808594,
184
+ "learning_rate": 1.8924924924924926e-05,
185
+ "loss": 3.7543,
186
+ "step": 550
187
+ },
188
+ {
189
+ "epoch": 3.0,
190
+ "eval_accuracy": 0.532520325203252,
191
+ "eval_loss": 0.9882494807243347,
192
+ "eval_macro_f1": 0.5273770535452779,
193
+ "eval_runtime": 1.5519,
194
+ "eval_samples_per_second": 475.56,
195
+ "eval_steps_per_second": 30.286,
196
+ "eval_weighted_f1": 0.528912302622507,
197
+ "step": 555
198
+ },
199
+ {
200
+ "epoch": 3.1084010840108403,
201
+ "grad_norm": 61.48655700683594,
202
+ "learning_rate": 1.8774774774774776e-05,
203
+ "loss": 3.0042,
204
+ "step": 575
205
+ },
206
+ {
207
+ "epoch": 3.2439024390243905,
208
+ "grad_norm": 83.6844253540039,
209
+ "learning_rate": 1.8624624624624627e-05,
210
+ "loss": 2.9475,
211
+ "step": 600
212
+ },
213
+ {
214
+ "epoch": 3.3794037940379402,
215
+ "grad_norm": 43.365089416503906,
216
+ "learning_rate": 1.8474474474474477e-05,
217
+ "loss": 2.9404,
218
+ "step": 625
219
+ },
220
+ {
221
+ "epoch": 3.5149051490514904,
222
+ "grad_norm": 53.747764587402344,
223
+ "learning_rate": 1.8324324324324324e-05,
224
+ "loss": 2.9198,
225
+ "step": 650
226
+ },
227
+ {
228
+ "epoch": 3.6504065040650406,
229
+ "grad_norm": 85.7112808227539,
230
+ "learning_rate": 1.8174174174174174e-05,
231
+ "loss": 2.7972,
232
+ "step": 675
233
+ },
234
+ {
235
+ "epoch": 3.785907859078591,
236
+ "grad_norm": 35.51165008544922,
237
+ "learning_rate": 1.8024024024024024e-05,
238
+ "loss": 2.8307,
239
+ "step": 700
240
+ },
241
+ {
242
+ "epoch": 3.921409214092141,
243
+ "grad_norm": 35.78587341308594,
244
+ "learning_rate": 1.7873873873873874e-05,
245
+ "loss": 2.7418,
246
+ "step": 725
247
+ },
248
+ {
249
+ "epoch": 4.0,
250
+ "eval_accuracy": 0.556910569105691,
251
+ "eval_loss": 1.0178123712539673,
252
+ "eval_macro_f1": 0.5594083643154767,
253
+ "eval_runtime": 1.5145,
254
+ "eval_samples_per_second": 487.277,
255
+ "eval_steps_per_second": 31.033,
256
+ "eval_weighted_f1": 0.5583768105792863,
257
+ "step": 740
258
+ },
259
+ {
260
+ "epoch": 4.05420054200542,
261
+ "grad_norm": 41.14622497558594,
262
+ "learning_rate": 1.7723723723723724e-05,
263
+ "loss": 2.5614,
264
+ "step": 750
265
+ },
266
+ {
267
+ "epoch": 4.1897018970189706,
268
+ "grad_norm": 35.917537689208984,
269
+ "learning_rate": 1.7573573573573575e-05,
270
+ "loss": 1.9798,
271
+ "step": 775
272
+ },
273
+ {
274
+ "epoch": 4.32520325203252,
275
+ "grad_norm": 35.059120178222656,
276
+ "learning_rate": 1.7423423423423425e-05,
277
+ "loss": 2.0427,
278
+ "step": 800
279
+ },
280
+ {
281
+ "epoch": 4.46070460704607,
282
+ "grad_norm": 58.0841064453125,
283
+ "learning_rate": 1.7273273273273275e-05,
284
+ "loss": 2.0326,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 4.596205962059621,
289
+ "grad_norm": 46.54280090332031,
290
+ "learning_rate": 1.7123123123123125e-05,
291
+ "loss": 2.0397,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 4.7317073170731705,
296
+ "grad_norm": 89.54529571533203,
297
+ "learning_rate": 1.6972972972972975e-05,
298
+ "loss": 2.2818,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 4.867208672086721,
303
+ "grad_norm": 55.32621383666992,
304
+ "learning_rate": 1.6822822822822826e-05,
305
+ "loss": 1.8006,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.0,
310
+ "grad_norm": 52.9863395690918,
311
+ "learning_rate": 1.6672672672672673e-05,
312
+ "loss": 1.9312,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 5.0,
317
+ "eval_accuracy": 0.5636856368563685,
318
+ "eval_loss": 1.101500153541565,
319
+ "eval_macro_f1": 0.5577557883244757,
320
+ "eval_runtime": 1.5043,
321
+ "eval_samples_per_second": 490.586,
322
+ "eval_steps_per_second": 31.243,
323
+ "eval_weighted_f1": 0.5585066121177835,
324
+ "step": 925
325
+ },
326
+ {
327
+ "epoch": 5.13550135501355,
328
+ "grad_norm": 49.05887222290039,
329
+ "learning_rate": 1.6522522522522523e-05,
330
+ "loss": 1.1363,
331
+ "step": 950
332
+ },
333
+ {
334
+ "epoch": 5.2710027100271,
335
+ "grad_norm": 54.03490447998047,
336
+ "learning_rate": 1.6372372372372373e-05,
337
+ "loss": 1.188,
338
+ "step": 975
339
+ },
340
+ {
341
+ "epoch": 5.40650406504065,
342
+ "grad_norm": 55.72603225708008,
343
+ "learning_rate": 1.6222222222222223e-05,
344
+ "loss": 1.3471,
345
+ "step": 1000
346
+ },
347
+ {
348
+ "epoch": 5.542005420054201,
349
+ "grad_norm": 59.89711380004883,
350
+ "learning_rate": 1.6072072072072073e-05,
351
+ "loss": 1.2589,
352
+ "step": 1025
353
+ },
354
+ {
355
+ "epoch": 5.677506775067751,
356
+ "grad_norm": 80.13398742675781,
357
+ "learning_rate": 1.5921921921921924e-05,
358
+ "loss": 1.2965,
359
+ "step": 1050
360
+ },
361
+ {
362
+ "epoch": 5.8130081300813,
363
+ "grad_norm": 63.47343444824219,
364
+ "learning_rate": 1.5771771771771774e-05,
365
+ "loss": 1.2344,
366
+ "step": 1075
367
+ },
368
+ {
369
+ "epoch": 5.948509485094851,
370
+ "grad_norm": 47.332698822021484,
371
+ "learning_rate": 1.5621621621621624e-05,
372
+ "loss": 1.1108,
373
+ "step": 1100
374
+ },
375
+ {
376
+ "epoch": 6.0,
377
+ "eval_accuracy": 0.5745257452574526,
378
+ "eval_loss": 1.4827275276184082,
379
+ "eval_macro_f1": 0.5773063185140571,
380
+ "eval_runtime": 1.4462,
381
+ "eval_samples_per_second": 510.299,
382
+ "eval_steps_per_second": 32.499,
383
+ "eval_weighted_f1": 0.57609308706819,
384
+ "step": 1110
385
+ },
386
+ {
387
+ "epoch": 6.08130081300813,
388
+ "grad_norm": 88.07003021240234,
389
+ "learning_rate": 1.5471471471471474e-05,
390
+ "loss": 0.8647,
391
+ "step": 1125
392
+ },
393
+ {
394
+ "epoch": 6.2168021680216805,
395
+ "grad_norm": 43.04215621948242,
396
+ "learning_rate": 1.5321321321321324e-05,
397
+ "loss": 0.5985,
398
+ "step": 1150
399
+ },
400
+ {
401
+ "epoch": 6.35230352303523,
402
+ "grad_norm": 64.628173828125,
403
+ "learning_rate": 1.5171171171171173e-05,
404
+ "loss": 0.6384,
405
+ "step": 1175
406
+ },
407
+ {
408
+ "epoch": 6.487804878048781,
409
+ "grad_norm": 73.82109832763672,
410
+ "learning_rate": 1.5021021021021023e-05,
411
+ "loss": 0.8202,
412
+ "step": 1200
413
+ },
414
+ {
415
+ "epoch": 6.623306233062331,
416
+ "grad_norm": 61.590763092041016,
417
+ "learning_rate": 1.4870870870870873e-05,
418
+ "loss": 0.7373,
419
+ "step": 1225
420
+ },
421
+ {
422
+ "epoch": 6.7588075880758804,
423
+ "grad_norm": 33.31949996948242,
424
+ "learning_rate": 1.4720720720720722e-05,
425
+ "loss": 0.6751,
426
+ "step": 1250
427
+ },
428
+ {
429
+ "epoch": 6.894308943089431,
430
+ "grad_norm": 33.728031158447266,
431
+ "learning_rate": 1.4570570570570572e-05,
432
+ "loss": 0.6632,
433
+ "step": 1275
434
+ },
435
+ {
436
+ "epoch": 7.0,
437
+ "eval_accuracy": 0.5826558265582655,
438
+ "eval_loss": 1.7230784893035889,
439
+ "eval_macro_f1": 0.5823950641665662,
440
+ "eval_runtime": 1.6641,
441
+ "eval_samples_per_second": 443.489,
442
+ "eval_steps_per_second": 28.244,
443
+ "eval_weighted_f1": 0.5822306689981528,
444
+ "step": 1295
445
+ }
446
+ ],
447
+ "logging_steps": 25,
448
+ "max_steps": 3700,
449
+ "num_input_tokens_seen": 0,
450
+ "num_train_epochs": 20,
451
+ "save_steps": 500,
452
+ "stateful_callbacks": {
453
+ "EarlyStoppingCallback": {
454
+ "args": {
455
+ "early_stopping_patience": 3,
456
+ "early_stopping_threshold": 0.0
457
+ },
458
+ "attributes": {
459
+ "early_stopping_patience_counter": 0
460
+ }
461
+ },
462
+ "TrainerControl": {
463
+ "args": {
464
+ "should_epoch_stop": false,
465
+ "should_evaluate": false,
466
+ "should_log": false,
467
+ "should_save": true,
468
+ "should_training_stop": false
469
+ },
470
+ "attributes": {}
471
+ }
472
+ },
473
+ "total_flos": 0.0,
474
+ "train_batch_size": 8,
475
+ "trial_name": null,
476
+ "trial_params": null
477
+ }