Theoreticallyhugo commited on
Commit
a7e8532
1 Parent(s): 3aa2af1

Training in progress, epoch 25, checkpoint

Browse files
checkpoint-2025/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1b73f56b1847d35e71764267aa752fab14522e8c144c15b31942410c51fe4ab8
3
  size 592318676
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:694ee1119f848759c1fb22ac30f8fc42c8019c9db5092790265894486ba64a26
3
  size 592318676
checkpoint-2025/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1f72969164ec760bf1a013784733075b1bc959441bf68ab754ddf4514861fc46
3
  size 1014645498
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62633f4227cc36a01dd7f2306de6caaea4ff62be41ed51f2611ccb4681e33e68
3
  size 1014645498
checkpoint-2025/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3f9321bc11ea6bca18575e8269cfc17c24700dfe4e6bb3c56fb55d02fcde9a55
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8d867bba4be3f9f8d77fd27fedf6b41a732d2f6a832fe7f80aba4e5fcdeb034
3
  size 14244
checkpoint-2025/trainer_state.json CHANGED
@@ -11,1003 +11,1003 @@
11
  {
12
  "epoch": 1.0,
13
  "eval_B": {
14
- "f1-score": 0.8266764382557715,
15
- "precision": 0.7866108786610879,
16
- "recall": 0.871042471042471,
17
- "support": 1295.0
18
  },
19
  "eval_I": {
20
- "f1-score": 0.9410658617447671,
21
- "precision": 0.9344077040239768,
22
- "recall": 0.9478195863443808,
23
- "support": 20065.0
24
  },
25
  "eval_O": {
26
- "f1-score": 0.8562443302086484,
27
- "precision": 0.8789421405512788,
28
- "recall": 0.8346893055064262,
29
- "support": 8481.0
30
  },
31
- "eval_accuracy": 0.912335377500754,
32
- "eval_loss": 0.2450982630252838,
33
  "eval_macro avg": {
34
- "f1-score": 0.8746622100697291,
35
- "precision": 0.8666535744121145,
36
- "recall": 0.8845171209644261,
37
- "support": 29841.0
38
- },
39
- "eval_runtime": 4.6269,
40
- "eval_samples_per_second": 17.29,
41
- "eval_steps_per_second": 2.161,
42
  "eval_weighted avg": {
43
- "f1-score": 0.9119949287205363,
44
- "precision": 0.9122301519092053,
45
- "recall": 0.912335377500754,
46
- "support": 29841.0
47
  },
48
  "step": 81
49
  },
50
  {
51
  "epoch": 2.0,
52
  "eval_B": {
53
- "f1-score": 0.8833395452851286,
54
- "precision": 0.8537463976945245,
55
- "recall": 0.915057915057915,
56
- "support": 1295.0
57
  },
58
  "eval_I": {
59
- "f1-score": 0.9549518227873249,
60
- "precision": 0.9516903430183636,
61
- "recall": 0.958235733864939,
62
- "support": 20065.0
63
  },
64
  "eval_O": {
65
- "f1-score": 0.88924750463212,
66
- "precision": 0.9016969696969697,
67
- "recall": 0.877137130055418,
68
- "support": 8481.0
69
  },
70
- "eval_accuracy": 0.9333132267685399,
71
- "eval_loss": 0.20076116919517517,
72
  "eval_macro avg": {
73
- "f1-score": 0.9091796242348579,
74
- "precision": 0.9023779034699526,
75
- "recall": 0.916810259659424,
76
- "support": 29841.0
77
- },
78
- "eval_runtime": 4.6524,
79
- "eval_samples_per_second": 17.195,
80
- "eval_steps_per_second": 2.149,
81
  "eval_weighted avg": {
82
- "f1-score": 0.9331705077630417,
83
- "precision": 0.9332314707173981,
84
- "recall": 0.9333132267685399,
85
- "support": 29841.0
86
  },
87
  "step": 162
88
  },
89
  {
90
  "epoch": 3.0,
91
  "eval_B": {
92
- "f1-score": 0.8927335640138409,
93
- "precision": 0.8889739663093414,
94
- "recall": 0.8965250965250965,
95
- "support": 1295.0
96
  },
97
  "eval_I": {
98
- "f1-score": 0.9590235396687009,
99
- "precision": 0.9586653386454184,
100
- "recall": 0.9593820084724645,
101
- "support": 20065.0
102
  },
103
  "eval_O": {
104
- "f1-score": 0.9006849315068494,
105
- "precision": 0.9020697811945594,
106
- "recall": 0.8993043273198915,
107
- "support": 8481.0
108
  },
109
- "eval_accuracy": 0.9395797727958178,
110
- "eval_loss": 0.19863653182983398,
111
  "eval_macro avg": {
112
- "f1-score": 0.9174806783964637,
113
- "precision": 0.9165696953831063,
114
- "recall": 0.9184038107724842,
115
- "support": 29841.0
116
- },
117
- "eval_runtime": 4.655,
118
- "eval_samples_per_second": 17.186,
119
- "eval_steps_per_second": 2.148,
120
  "eval_weighted avg": {
121
- "f1-score": 0.9395665759512081,
122
- "precision": 0.9395561516236713,
123
- "recall": 0.9395797727958178,
124
- "support": 29841.0
125
  },
126
  "step": 243
127
  },
128
  {
129
  "epoch": 4.0,
130
  "eval_B": {
131
- "f1-score": 0.8999616711383672,
132
- "precision": 0.893455098934551,
133
- "recall": 0.9065637065637066,
134
- "support": 1295.0
135
  },
136
  "eval_I": {
137
- "f1-score": 0.9612657113512985,
138
- "precision": 0.9436362763454799,
139
- "recall": 0.9795664091701969,
140
- "support": 20065.0
141
  },
142
  "eval_O": {
143
- "f1-score": 0.8996847765622102,
144
- "precision": 0.9454403741231489,
145
- "recall": 0.8581535196321188,
146
- "support": 8481.0
147
  },
148
- "eval_accuracy": 0.9418920277470594,
149
- "eval_loss": 0.19743724167346954,
150
  "eval_macro avg": {
151
- "f1-score": 0.920304053017292,
152
- "precision": 0.9275105831343932,
153
- "recall": 0.9147612117886741,
154
- "support": 29841.0
155
- },
156
- "eval_runtime": 4.6615,
157
- "eval_samples_per_second": 17.162,
158
- "eval_steps_per_second": 2.145,
159
  "eval_weighted avg": {
160
- "f1-score": 0.9411036309913238,
161
- "precision": 0.9419713163409645,
162
- "recall": 0.9418920277470594,
163
- "support": 29841.0
164
  },
165
  "step": 324
166
  },
167
  {
168
  "epoch": 5.0,
169
  "eval_B": {
170
- "f1-score": 0.8953846153846153,
171
- "precision": 0.8919540229885058,
172
- "recall": 0.8988416988416988,
173
- "support": 1295.0
174
  },
175
  "eval_I": {
176
- "f1-score": 0.9602426713742261,
177
- "precision": 0.9581224570804803,
178
- "recall": 0.9623722900573137,
179
- "support": 20065.0
180
  },
181
  "eval_O": {
182
- "f1-score": 0.9022119433078337,
183
- "precision": 0.9075399665950847,
184
- "recall": 0.8969461148449476,
185
- "support": 8481.0
186
  },
187
- "eval_accuracy": 0.9410207432726785,
188
- "eval_loss": 0.2340461015701294,
189
  "eval_macro avg": {
190
- "f1-score": 0.9192797433555584,
191
- "precision": 0.9192054822213569,
192
- "recall": 0.9193867012479867,
193
- "support": 29841.0
194
- },
195
- "eval_runtime": 4.6719,
196
- "eval_samples_per_second": 17.124,
197
- "eval_steps_per_second": 2.14,
198
  "eval_weighted avg": {
199
- "f1-score": 0.9409353496612266,
200
- "precision": 0.9408751053176122,
201
- "recall": 0.9410207432726785,
202
- "support": 29841.0
203
  },
204
  "step": 405
205
  },
206
  {
207
  "epoch": 6.0,
208
  "eval_B": {
209
- "f1-score": 0.8929831438651511,
210
- "precision": 0.9068471337579618,
211
- "recall": 0.8795366795366796,
212
- "support": 1295.0
213
  },
214
  "eval_I": {
215
- "f1-score": 0.9522293798155866,
216
- "precision": 0.9655225409836066,
217
- "recall": 0.9392972838275604,
218
- "support": 20065.0
219
  },
220
  "eval_O": {
221
- "f1-score": 0.8885216003647556,
222
- "precision": 0.8599007170435742,
223
- "recall": 0.919113312109421,
224
- "support": 8481.0
225
  },
226
- "eval_accuracy": 0.930967460875976,
227
- "eval_loss": 0.21774999797344208,
228
  "eval_macro avg": {
229
- "f1-score": 0.9112447080151643,
230
- "precision": 0.9107567972617142,
231
- "recall": 0.9126490918245537,
232
- "support": 29841.0
233
- },
234
- "eval_runtime": 4.647,
235
- "eval_samples_per_second": 17.215,
236
- "eval_steps_per_second": 2.152,
237
  "eval_weighted avg": {
238
- "f1-score": 0.9315521386682287,
239
- "precision": 0.9329578366776978,
240
- "recall": 0.930967460875976,
241
- "support": 29841.0
242
  },
243
  "step": 486
244
  },
245
  {
246
  "epoch": 6.17,
247
- "grad_norm": 9.310327529907227,
248
  "learning_rate": 1.7530864197530865e-05,
249
- "loss": 0.1627,
250
  "step": 500
251
  },
252
  {
253
  "epoch": 7.0,
254
  "eval_B": {
255
- "f1-score": 0.8936490041337843,
256
- "precision": 0.8704245973645681,
257
- "recall": 0.9181467181467181,
258
- "support": 1295.0
259
  },
260
  "eval_I": {
261
- "f1-score": 0.9599271850630982,
262
- "precision": 0.9477800446905664,
263
- "recall": 0.9723897333665587,
264
- "support": 20065.0
265
  },
266
  "eval_O": {
267
- "f1-score": 0.8979841172877214,
268
- "precision": 0.9316770186335404,
269
- "recall": 0.8666430845419172,
270
- "support": 8481.0
271
  },
272
- "eval_accuracy": 0.9399819040916859,
273
- "eval_loss": 0.27376389503479004,
274
  "eval_macro avg": {
275
- "f1-score": 0.9171867688282013,
276
- "precision": 0.9166272202295582,
277
- "recall": 0.9190598453517312,
278
- "support": 29841.0
279
- },
280
- "eval_runtime": 4.6671,
281
- "eval_samples_per_second": 17.141,
282
- "eval_steps_per_second": 2.143,
283
  "eval_weighted avg": {
284
- "f1-score": 0.9394463230910989,
285
- "precision": 0.9398464945991887,
286
- "recall": 0.9399819040916859,
287
- "support": 29841.0
288
  },
289
  "step": 567
290
  },
291
  {
292
  "epoch": 8.0,
293
  "eval_B": {
294
- "f1-score": 0.8948545861297539,
295
- "precision": 0.8651766402307137,
296
- "recall": 0.9266409266409267,
297
- "support": 1295.0
298
  },
299
  "eval_I": {
300
- "f1-score": 0.9590139880661254,
301
- "precision": 0.9414702069429106,
302
- "recall": 0.9772240219287316,
303
- "support": 20065.0
304
  },
305
  "eval_O": {
306
- "f1-score": 0.8927241122423641,
307
- "precision": 0.9427035531663825,
308
- "recall": 0.8477773847423653,
309
- "support": 8481.0
310
  },
311
- "eval_accuracy": 0.9382393351429241,
312
- "eval_loss": 0.3103167414665222,
313
  "eval_macro avg": {
314
- "f1-score": 0.9155308954794145,
315
- "precision": 0.916450133446669,
316
- "recall": 0.9172141111040079,
317
- "support": 29841.0
318
- },
319
- "eval_runtime": 4.6339,
320
- "eval_samples_per_second": 17.264,
321
- "eval_steps_per_second": 2.158,
322
  "eval_weighted avg": {
323
- "f1-score": 0.9373896838414373,
324
- "precision": 0.9385098450391195,
325
- "recall": 0.9382393351429241,
326
- "support": 29841.0
327
  },
328
  "step": 648
329
  },
330
  {
331
  "epoch": 9.0,
332
  "eval_B": {
333
- "f1-score": 0.9011235955056179,
334
- "precision": 0.8749090909090909,
335
- "recall": 0.9289575289575289,
336
- "support": 1295.0
337
  },
338
  "eval_I": {
339
- "f1-score": 0.9615119937082186,
340
- "precision": 0.9485040973670175,
341
- "recall": 0.9748816346872664,
342
- "support": 20065.0
343
  },
344
  "eval_O": {
345
- "f1-score": 0.9013722126929674,
346
- "precision": 0.9380339155935229,
347
- "recall": 0.8674684589081476,
348
- "support": 8481.0
349
  },
350
- "eval_accuracy": 0.9423611809255722,
351
- "eval_loss": 0.32883673906326294,
352
  "eval_macro avg": {
353
- "f1-score": 0.9213359339689347,
354
- "precision": 0.9204823679565438,
355
- "recall": 0.9237692075176476,
356
- "support": 29841.0
357
- },
358
- "eval_runtime": 4.6714,
359
- "eval_samples_per_second": 17.126,
360
- "eval_steps_per_second": 2.141,
361
  "eval_weighted avg": {
362
- "f1-score": 0.9417992341337168,
363
- "precision": 0.9423346276781992,
364
- "recall": 0.9423611809255722,
365
- "support": 29841.0
366
  },
367
  "step": 729
368
  },
369
  {
370
  "epoch": 10.0,
371
  "eval_B": {
372
- "f1-score": 0.8935361216730039,
373
- "precision": 0.8801498127340824,
374
- "recall": 0.9073359073359073,
375
- "support": 1295.0
376
  },
377
  "eval_I": {
378
- "f1-score": 0.9632332861646931,
379
- "precision": 0.9596359319351009,
380
- "recall": 0.9668577124345876,
381
- "support": 20065.0
382
  },
383
  "eval_O": {
384
- "f1-score": 0.9095462405342556,
385
- "precision": 0.9200241254523522,
386
- "recall": 0.8993043273198915,
387
- "support": 8481.0
388
  },
389
- "eval_accuracy": 0.9450755671726819,
390
- "eval_loss": 0.3041447699069977,
391
  "eval_macro avg": {
392
- "f1-score": 0.9221052161239841,
393
- "precision": 0.9199366233738452,
394
- "recall": 0.9244993156967954,
395
- "support": 29841.0
396
- },
397
- "eval_runtime": 4.6532,
398
- "eval_samples_per_second": 17.193,
399
- "eval_steps_per_second": 2.149,
400
  "eval_weighted avg": {
401
- "f1-score": 0.9449504651463465,
402
- "precision": 0.9449285744355027,
403
- "recall": 0.9450755671726819,
404
- "support": 29841.0
405
  },
406
  "step": 810
407
  },
408
  {
409
  "epoch": 11.0,
410
  "eval_B": {
411
- "f1-score": 0.8924122310305775,
412
- "precision": 0.8729689807976366,
413
- "recall": 0.9127413127413128,
414
- "support": 1295.0
415
  },
416
  "eval_I": {
417
- "f1-score": 0.9588723786993913,
418
- "precision": 0.9483330083837005,
419
- "recall": 0.9696486419137802,
420
- "support": 20065.0
421
  },
422
  "eval_O": {
423
- "f1-score": 0.8956965718453683,
424
- "precision": 0.9243507715468574,
425
- "recall": 0.8687654757693668,
426
- "support": 8481.0
427
  },
428
- "eval_accuracy": 0.9385074226735028,
429
- "eval_loss": 0.3531416654586792,
430
  "eval_macro avg": {
431
- "f1-score": 0.9156603938584458,
432
- "precision": 0.9152175869093981,
433
- "recall": 0.9170518101414866,
434
- "support": 29841.0
435
- },
436
- "eval_runtime": 4.6508,
437
- "eval_samples_per_second": 17.201,
438
- "eval_steps_per_second": 2.15,
439
  "eval_weighted avg": {
440
- "f1-score": 0.938033267772811,
441
- "precision": 0.9382465579853486,
442
- "recall": 0.9385074226735028,
443
- "support": 29841.0
444
  },
445
  "step": 891
446
  },
447
  {
448
  "epoch": 12.0,
449
  "eval_B": {
450
- "f1-score": 0.8970199924556771,
451
- "precision": 0.8768436578171092,
452
- "recall": 0.9181467181467181,
453
- "support": 1295.0
454
  },
455
  "eval_I": {
456
- "f1-score": 0.9595004897159647,
457
- "precision": 0.9431046931407943,
458
- "recall": 0.9764764515325193,
459
- "support": 20065.0
460
  },
461
  "eval_O": {
462
- "f1-score": 0.8953122104873078,
463
- "precision": 0.9400778210116731,
464
- "recall": 0.8546162009197029,
465
- "support": 8481.0
466
  },
467
- "eval_accuracy": 0.9393116852652391,
468
- "eval_loss": 0.4237186312675476,
469
  "eval_macro avg": {
470
- "f1-score": 0.9172775642196499,
471
- "precision": 0.9200087239898588,
472
- "recall": 0.9164131235329801,
473
- "support": 29841.0
474
- },
475
- "eval_runtime": 4.6388,
476
- "eval_samples_per_second": 17.246,
477
- "eval_steps_per_second": 2.156,
478
  "eval_weighted avg": {
479
- "f1-score": 0.9385463313402296,
480
- "precision": 0.9393689288141548,
481
- "recall": 0.9393116852652391,
482
- "support": 29841.0
483
  },
484
  "step": 972
485
  },
486
  {
487
  "epoch": 12.35,
488
- "grad_norm": 0.5223535895347595,
489
  "learning_rate": 1.506172839506173e-05,
490
- "loss": 0.0196,
491
  "step": 1000
492
  },
493
  {
494
  "epoch": 13.0,
495
  "eval_B": {
496
- "f1-score": 0.8891411275066211,
497
- "precision": 0.8716617210682492,
498
- "recall": 0.9073359073359073,
499
- "support": 1295.0
500
  },
501
  "eval_I": {
502
- "f1-score": 0.9610692322951795,
503
- "precision": 0.9563736860287223,
504
- "recall": 0.9658111138798904,
505
- "support": 20065.0
506
  },
507
  "eval_O": {
508
- "f1-score": 0.9043145233678416,
509
- "precision": 0.918104495747266,
510
- "recall": 0.8909326730338404,
511
- "support": 8481.0
512
  },
513
- "eval_accuracy": 0.9419925605710264,
514
- "eval_loss": 0.4308999180793762,
515
  "eval_macro avg": {
516
- "f1-score": 0.9181749610565474,
517
- "precision": 0.9153799676147458,
518
- "recall": 0.9213598980832126,
519
- "support": 29841.0
520
- },
521
- "eval_runtime": 4.6555,
522
- "eval_samples_per_second": 17.184,
523
- "eval_steps_per_second": 2.148,
524
  "eval_weighted avg": {
525
- "f1-score": 0.9418177466843106,
526
- "precision": 0.9418211242043584,
527
- "recall": 0.9419925605710264,
528
- "support": 29841.0
529
  },
530
  "step": 1053
531
  },
532
  {
533
  "epoch": 14.0,
534
  "eval_B": {
535
- "f1-score": 0.9025601834161253,
536
- "precision": 0.8933434190620272,
537
- "recall": 0.911969111969112,
538
- "support": 1295.0
539
  },
540
  "eval_I": {
541
- "f1-score": 0.960806215404473,
542
- "precision": 0.9570764513895758,
543
- "recall": 0.9645651632195366,
544
- "support": 20065.0
545
  },
546
  "eval_O": {
547
- "f1-score": 0.9030873763261413,
548
- "precision": 0.9131011208870676,
549
- "recall": 0.8932908855087843,
550
- "support": 8481.0
551
  },
552
- "eval_accuracy": 0.9420260715123487,
553
- "eval_loss": 0.4198700487613678,
554
  "eval_macro avg": {
555
- "f1-score": 0.9221512583822467,
556
- "precision": 0.9211736637795568,
557
- "recall": 0.9232750535658111,
558
- "support": 29841.0
559
- },
560
- "eval_runtime": 4.6834,
561
- "eval_samples_per_second": 17.082,
562
- "eval_steps_per_second": 2.135,
563
  "eval_weighted avg": {
564
- "f1-score": 0.9418744743217934,
565
- "precision": 0.9418125843993292,
566
- "recall": 0.9420260715123487,
567
- "support": 29841.0
568
  },
569
  "step": 1134
570
  },
571
  {
572
  "epoch": 15.0,
573
  "eval_B": {
574
- "f1-score": 0.8935681470137825,
575
- "precision": 0.8861047835990888,
576
- "recall": 0.9011583011583012,
577
- "support": 1295.0
578
  },
579
  "eval_I": {
580
- "f1-score": 0.9633398694432723,
581
- "precision": 0.9595530063291139,
582
- "recall": 0.9671567405930726,
583
- "support": 20065.0
584
  },
585
  "eval_O": {
586
- "f1-score": 0.9100768726535964,
587
- "precision": 0.92,
588
- "recall": 0.9003655229336163,
589
- "support": 8481.0
590
  },
591
- "eval_accuracy": 0.9453101437619382,
592
- "eval_loss": 0.39179423451423645,
593
  "eval_macro avg": {
594
- "f1-score": 0.9223282963702171,
595
- "precision": 0.9218859299760677,
596
- "recall": 0.9228935215616634,
597
- "support": 29841.0
598
- },
599
- "eval_runtime": 4.6645,
600
- "eval_samples_per_second": 17.151,
601
- "eval_steps_per_second": 2.144,
602
  "eval_weighted avg": {
603
- "f1-score": 0.945174330208011,
604
- "precision": 0.9451243847979119,
605
- "recall": 0.9453101437619382,
606
- "support": 29841.0
607
  },
608
  "step": 1215
609
  },
610
  {
611
  "epoch": 16.0,
612
  "eval_B": {
613
- "f1-score": 0.9022614028363357,
614
- "precision": 0.895738203957382,
615
- "recall": 0.9088803088803089,
616
- "support": 1295.0
617
  },
618
  "eval_I": {
619
- "f1-score": 0.9632045864531594,
620
- "precision": 0.9552494853445741,
621
- "recall": 0.9712932967854473,
622
- "support": 20065.0
623
  },
624
  "eval_O": {
625
- "f1-score": 0.9083463808262073,
626
- "precision": 0.9282461538461538,
627
- "recall": 0.8892819243013795,
628
- "support": 8481.0
629
  },
630
- "eval_accuracy": 0.9452766328206159,
631
- "eval_loss": 0.44052162766456604,
632
  "eval_macro avg": {
633
- "f1-score": 0.9246041233719008,
634
- "precision": 0.92641128104937,
635
- "recall": 0.9231518433223785,
636
- "support": 29841.0
637
- },
638
- "eval_runtime": 4.6376,
639
- "eval_samples_per_second": 17.25,
640
- "eval_steps_per_second": 2.156,
641
  "eval_weighted avg": {
642
- "f1-score": 0.9449688080038458,
643
- "precision": 0.944992377210312,
644
- "recall": 0.9452766328206159,
645
- "support": 29841.0
646
  },
647
  "step": 1296
648
  },
649
  {
650
  "epoch": 17.0,
651
  "eval_B": {
652
- "f1-score": 0.8923649906890131,
653
- "precision": 0.8618705035971223,
654
- "recall": 0.9250965250965251,
655
- "support": 1295.0
656
  },
657
  "eval_I": {
658
- "f1-score": 0.9599605522682445,
659
- "precision": 0.9498902171261283,
660
- "recall": 0.97024669823075,
661
- "support": 20065.0
662
  },
663
  "eval_O": {
664
- "f1-score": 0.9001642635517431,
665
- "precision": 0.9298642533936652,
666
- "recall": 0.8723027944817828,
667
- "support": 8481.0
668
  },
669
- "eval_accuracy": 0.9404510572701987,
670
- "eval_loss": 0.4657233655452728,
671
  "eval_macro avg": {
672
- "f1-score": 0.9174966021696669,
673
- "precision": 0.9138749913723053,
674
- "recall": 0.9225486726030193,
675
- "support": 29841.0
676
- },
677
- "eval_runtime": 4.6396,
678
- "eval_samples_per_second": 17.243,
679
- "eval_steps_per_second": 2.155,
680
  "eval_weighted avg": {
681
- "f1-score": 0.940032648483192,
682
- "precision": 0.9403789498282803,
683
- "recall": 0.9404510572701987,
684
- "support": 29841.0
685
  },
686
  "step": 1377
687
  },
688
  {
689
  "epoch": 18.0,
690
  "eval_B": {
691
- "f1-score": 0.9014302280633939,
692
- "precision": 0.9024767801857585,
693
- "recall": 0.9003861003861003,
694
- "support": 1295.0
695
  },
696
  "eval_I": {
697
- "f1-score": 0.9638021281301453,
698
- "precision": 0.9506496024820632,
699
- "recall": 0.9773236979815599,
700
- "support": 20065.0
701
  },
702
  "eval_O": {
703
- "f1-score": 0.9083038653822705,
704
- "precision": 0.9404115641964399,
705
- "recall": 0.87831623629289,
706
- "support": 8481.0
707
  },
708
- "eval_accuracy": 0.9458463188230958,
709
- "eval_loss": 0.4731404185295105,
710
  "eval_macro avg": {
711
- "f1-score": 0.9245120738586032,
712
- "precision": 0.9311793156214204,
713
- "recall": 0.9186753448868501,
714
- "support": 29841.0
715
- },
716
- "eval_runtime": 4.6405,
717
- "eval_samples_per_second": 17.24,
718
- "eval_steps_per_second": 2.155,
719
  "eval_weighted avg": {
720
- "f1-score": 0.9453224398840687,
721
- "precision": 0.9456493475450943,
722
- "recall": 0.9458463188230958,
723
- "support": 29841.0
724
  },
725
  "step": 1458
726
  },
727
  {
728
  "epoch": 18.52,
729
- "grad_norm": 1.548362135887146,
730
  "learning_rate": 1.2592592592592593e-05,
731
- "loss": 0.0038,
732
  "step": 1500
733
  },
734
  {
735
  "epoch": 19.0,
736
  "eval_B": {
737
- "f1-score": 0.8891402714932127,
738
- "precision": 0.868828297715549,
739
- "recall": 0.9104247104247104,
740
- "support": 1295.0
741
  },
742
  "eval_I": {
743
- "f1-score": 0.9603099696984749,
744
- "precision": 0.9571718572065158,
745
- "recall": 0.9634687266384251,
746
- "support": 20065.0
747
  },
748
  "eval_O": {
749
- "f1-score": 0.9027910305343512,
750
- "precision": 0.913358271992277,
751
- "recall": 0.8924655111425539,
752
- "support": 8481.0
753
  },
754
- "eval_accuracy": 0.9409872323313562,
755
- "eval_loss": 0.49517783522605896,
756
  "eval_macro avg": {
757
- "f1-score": 0.9174137572420129,
758
- "precision": 0.9131194756381139,
759
- "recall": 0.9221196494018965,
760
- "support": 29841.0
761
- },
762
- "eval_runtime": 4.6377,
763
- "eval_samples_per_second": 17.25,
764
- "eval_steps_per_second": 2.156,
765
  "eval_weighted avg": {
766
- "f1-score": 0.9408741973642117,
767
- "precision": 0.9408859443435835,
768
- "recall": 0.9409872323313562,
769
- "support": 29841.0
770
  },
771
  "step": 1539
772
  },
773
  {
774
  "epoch": 20.0,
775
  "eval_B": {
776
- "f1-score": 0.899884925201381,
777
- "precision": 0.8940548780487805,
778
- "recall": 0.9057915057915058,
779
- "support": 1295.0
780
  },
781
  "eval_I": {
782
- "f1-score": 0.9634953846153845,
783
- "precision": 0.951896887159533,
784
- "recall": 0.9753800149514079,
785
- "support": 20065.0
786
  },
787
  "eval_O": {
788
- "f1-score": 0.9073556231003039,
789
- "precision": 0.9365039528171665,
790
- "recall": 0.8799669850253508,
791
- "support": 8481.0
792
  },
793
- "eval_accuracy": 0.9452431218792936,
794
- "eval_loss": 0.45907479524612427,
795
  "eval_macro avg": {
796
- "f1-score": 0.9235786443056897,
797
- "precision": 0.9274852393418267,
798
- "recall": 0.9203795019227549,
799
- "support": 29841.0
800
- },
801
- "eval_runtime": 4.6356,
802
- "eval_samples_per_second": 17.258,
803
- "eval_steps_per_second": 2.157,
804
  "eval_weighted avg": {
805
- "f1-score": 0.9447796290324438,
806
- "precision": 0.9450119678218422,
807
- "recall": 0.9452431218792936,
808
- "support": 29841.0
809
  },
810
  "step": 1620
811
  },
812
  {
813
  "epoch": 21.0,
814
  "eval_B": {
815
- "f1-score": 0.8944337811900192,
816
- "precision": 0.8893129770992366,
817
- "recall": 0.8996138996138996,
818
- "support": 1295.0
819
  },
820
  "eval_I": {
821
- "f1-score": 0.9624703322784811,
822
- "precision": 0.9549624687239366,
823
- "recall": 0.9700971841515076,
824
- "support": 20065.0
825
  },
826
  "eval_O": {
827
- "f1-score": 0.9066089361958025,
828
- "precision": 0.9251350024545901,
829
- "recall": 0.8888102818063908,
830
- "support": 8481.0
831
  },
832
- "eval_accuracy": 0.9439361951677223,
833
- "eval_loss": 0.4614206850528717,
834
  "eval_macro avg": {
835
- "f1-score": 0.9211710165547675,
836
- "precision": 0.9231368160925877,
837
- "recall": 0.9195071218572659,
838
- "support": 29841.0
839
- },
840
- "eval_runtime": 4.6636,
841
- "eval_samples_per_second": 17.154,
842
- "eval_steps_per_second": 2.144,
843
  "eval_weighted avg": {
844
- "f1-score": 0.9436416122678664,
845
- "precision": 0.9436363458364894,
846
- "recall": 0.9439361951677223,
847
- "support": 29841.0
848
  },
849
  "step": 1701
850
  },
851
  {
852
  "epoch": 22.0,
853
  "eval_B": {
854
- "f1-score": 0.9011072928598702,
855
- "precision": 0.8912386706948641,
856
- "recall": 0.9111969111969112,
857
- "support": 1295.0
858
  },
859
  "eval_I": {
860
- "f1-score": 0.9635141147970535,
861
- "precision": 0.9558585511795576,
862
- "recall": 0.9712932967854473,
863
- "support": 20065.0
864
  },
865
  "eval_O": {
866
- "f1-score": 0.9095068938527304,
867
- "precision": 0.9292568897637795,
868
- "recall": 0.8905789411625987,
869
- "support": 8481.0
870
  },
871
- "eval_accuracy": 0.9457457859991287,
872
- "eval_loss": 0.4724768102169037,
873
  "eval_macro avg": {
874
- "f1-score": 0.9247094338365515,
875
- "precision": 0.9254513705460671,
876
- "recall": 0.9243563830483191,
877
- "support": 29841.0
878
- },
879
- "eval_runtime": 4.6333,
880
- "eval_samples_per_second": 17.266,
881
- "eval_steps_per_second": 2.158,
882
  "eval_weighted avg": {
883
- "f1-score": 0.9454566745223489,
884
- "precision": 0.9454939040264833,
885
- "recall": 0.9457457859991287,
886
- "support": 29841.0
887
  },
888
  "step": 1782
889
  },
890
  {
891
  "epoch": 23.0,
892
  "eval_B": {
893
- "f1-score": 0.8956916099773242,
894
- "precision": 0.8771280532938565,
895
- "recall": 0.915057915057915,
896
- "support": 1295.0
897
  },
898
  "eval_I": {
899
- "f1-score": 0.9624542808091362,
900
- "precision": 0.9609957269204015,
901
- "recall": 0.9639172688761525,
902
- "support": 20065.0
903
  },
904
  "eval_O": {
905
- "f1-score": 0.9084001187295934,
906
- "precision": 0.914753706360593,
907
- "recall": 0.9021341822898243,
908
- "support": 8481.0
909
  },
910
- "eval_accuracy": 0.9442377936396233,
911
- "eval_loss": 0.47468113899230957,
912
  "eval_macro avg": {
913
- "f1-score": 0.9221820031720179,
914
- "precision": 0.9176258288582838,
915
- "recall": 0.927036455407964,
916
- "support": 29841.0
917
- },
918
- "eval_runtime": 4.6476,
919
- "eval_samples_per_second": 17.213,
920
- "eval_steps_per_second": 2.152,
921
  "eval_weighted avg": {
922
- "f1-score": 0.9441944702356367,
923
- "precision": 0.9442138759866489,
924
- "recall": 0.9442377936396233,
925
- "support": 29841.0
926
  },
927
  "step": 1863
928
  },
929
  {
930
  "epoch": 24.0,
931
  "eval_B": {
932
- "f1-score": 0.9008390541571321,
933
- "precision": 0.8899773926149209,
934
- "recall": 0.911969111969112,
935
- "support": 1295.0
936
  },
937
  "eval_I": {
938
- "f1-score": 0.9639726399486381,
939
- "precision": 0.9553151918559123,
940
- "recall": 0.9727884375778719,
941
- "support": 20065.0
942
  },
943
  "eval_O": {
944
- "f1-score": 0.9101008271448409,
945
- "precision": 0.9325661964860184,
946
- "recall": 0.8886923711826435,
947
- "support": 8481.0
948
  },
949
- "eval_accuracy": 0.9462484501189639,
950
- "eval_loss": 0.4712923467159271,
951
  "eval_macro avg": {
952
- "f1-score": 0.9249708404168704,
953
- "precision": 0.9259529269856172,
954
- "recall": 0.9244833069098758,
955
- "support": 29841.0
956
- },
957
- "eval_runtime": 4.6509,
958
- "eval_samples_per_second": 17.201,
959
- "eval_steps_per_second": 2.15,
960
  "eval_weighted avg": {
961
- "f1-score": 0.9459221443892064,
962
- "precision": 0.9460143413566277,
963
- "recall": 0.9462484501189639,
964
- "support": 29841.0
965
  },
966
  "step": 1944
967
  },
968
  {
969
  "epoch": 24.69,
970
- "grad_norm": 0.04072243347764015,
971
  "learning_rate": 1.0123456790123458e-05,
972
- "loss": 0.0014,
973
  "step": 2000
974
  },
975
  {
976
  "epoch": 25.0,
977
  "eval_B": {
978
- "f1-score": 0.8943768996960487,
979
- "precision": 0.8803290949887809,
980
- "recall": 0.9088803088803089,
981
- "support": 1295.0
982
  },
983
  "eval_I": {
984
- "f1-score": 0.9629187495336865,
985
- "precision": 0.9610305798252582,
986
- "recall": 0.9648143533516073,
987
- "support": 20065.0
988
  },
989
  "eval_O": {
990
- "f1-score": 0.9089721512974288,
991
- "precision": 0.9155502392344498,
992
- "recall": 0.9024879141610659,
993
- "support": 8481.0
994
  },
995
- "eval_accuracy": 0.9446734358768137,
996
- "eval_loss": 0.48565673828125,
997
  "eval_macro avg": {
998
- "f1-score": 0.922089266842388,
999
- "precision": 0.9189699713494962,
1000
- "recall": 0.925394192130994,
1001
- "support": 29841.0
1002
- },
1003
- "eval_runtime": 4.6314,
1004
- "eval_samples_per_second": 17.274,
1005
- "eval_steps_per_second": 2.159,
1006
  "eval_weighted avg": {
1007
- "f1-score": 0.9446122988389564,
1008
- "precision": 0.944602605179171,
1009
- "recall": 0.9446734358768137,
1010
- "support": 29841.0
1011
  },
1012
  "step": 2025
1013
  }
@@ -1017,7 +1017,7 @@
1017
  "num_input_tokens_seen": 0,
1018
  "num_train_epochs": 50,
1019
  "save_steps": 500,
1020
- "total_flos": 3594666318390000.0,
1021
  "train_batch_size": 4,
1022
  "trial_name": null,
1023
  "trial_params": null
 
11
  {
12
  "epoch": 1.0,
13
  "eval_B": {
14
+ "f1-score": 0.7905866302864938,
15
+ "precision": 0.7768096514745308,
16
+ "recall": 0.8048611111111111,
17
+ "support": 1440.0
18
  },
19
  "eval_I": {
20
+ "f1-score": 0.9005331227169513,
21
+ "precision": 0.9637593111099372,
22
+ "recall": 0.8450919534905267,
23
+ "support": 21587.0
24
  },
25
  "eval_O": {
26
+ "f1-score": 0.8288892663043479,
27
+ "precision": 0.7463108800367001,
28
+ "recall": 0.9320156593144275,
29
+ "support": 10473.0
30
  },
31
+ "eval_accuracy": 0.8705373134328358,
32
+ "eval_loss": 0.32930415868759155,
33
  "eval_macro avg": {
34
+ "f1-score": 0.8400030064359308,
35
+ "precision": 0.8289599475403894,
36
+ "recall": 0.8606562413053552,
37
+ "support": 33500.0
38
+ },
39
+ "eval_runtime": 4.7349,
40
+ "eval_samples_per_second": 17.107,
41
+ "eval_steps_per_second": 2.323,
42
  "eval_weighted avg": {
43
+ "f1-score": 0.8734092702599646,
44
+ "precision": 0.8877430445874,
45
+ "recall": 0.8705373134328358,
46
+ "support": 33500.0
47
  },
48
  "step": 81
49
  },
50
  {
51
  "epoch": 2.0,
52
  "eval_B": {
53
+ "f1-score": 0.8656218739579861,
54
+ "precision": 0.8325849903784477,
55
+ "recall": 0.9013888888888889,
56
+ "support": 1440.0
57
  },
58
  "eval_I": {
59
+ "f1-score": 0.9547168948727912,
60
+ "precision": 0.9423736462093862,
61
+ "recall": 0.9673877796822161,
62
+ "support": 21587.0
63
  },
64
  "eval_O": {
65
+ "f1-score": 0.9013528191962081,
66
+ "precision": 0.9332379102341274,
67
+ "recall": 0.8715745249689678,
68
+ "support": 10473.0
69
  },
70
+ "eval_accuracy": 0.9345970149253732,
71
+ "eval_loss": 0.2162298709154129,
72
  "eval_macro avg": {
73
+ "f1-score": 0.9072305293423284,
74
+ "precision": 0.902732182273987,
75
+ "recall": 0.9134503978466909,
76
+ "support": 33500.0
77
+ },
78
+ "eval_runtime": 4.753,
79
+ "eval_samples_per_second": 17.042,
80
+ "eval_steps_per_second": 2.314,
81
  "eval_weighted avg": {
82
+ "f1-score": 0.9342040950316517,
83
+ "precision": 0.9347982961417612,
84
+ "recall": 0.9345970149253732,
85
+ "support": 33500.0
86
  },
87
  "step": 162
88
  },
89
  {
90
  "epoch": 3.0,
91
  "eval_B": {
92
+ "f1-score": 0.8603587759409076,
93
+ "precision": 0.8717034925160371,
94
+ "recall": 0.8493055555555555,
95
+ "support": 1440.0
96
  },
97
  "eval_I": {
98
+ "f1-score": 0.9434195218082959,
99
+ "precision": 0.9677057963955188,
100
+ "recall": 0.9203224162690509,
101
+ "support": 21587.0
102
  },
103
  "eval_O": {
104
+ "f1-score": 0.8879310344827586,
105
+ "precision": 0.8459410391631365,
106
+ "recall": 0.9343072663038289,
107
+ "support": 10473.0
108
  },
109
+ "eval_accuracy": 0.9216417910447762,
110
+ "eval_loss": 0.24082203209400177,
111
  "eval_macro avg": {
112
+ "f1-score": 0.8972364440773206,
113
+ "precision": 0.8951167760248975,
114
+ "recall": 0.9013117460428117,
115
+ "support": 33500.0
116
+ },
117
+ "eval_runtime": 4.7554,
118
+ "eval_samples_per_second": 17.033,
119
+ "eval_steps_per_second": 2.313,
120
  "eval_weighted avg": {
121
+ "f1-score": 0.9225019575751797,
122
+ "precision": 0.9255121957960801,
123
+ "recall": 0.9216417910447762,
124
+ "support": 33500.0
125
  },
126
  "step": 243
127
  },
128
  {
129
  "epoch": 4.0,
130
  "eval_B": {
131
+ "f1-score": 0.8804020100502512,
132
+ "precision": 0.8504854368932039,
133
+ "recall": 0.9125,
134
+ "support": 1440.0
135
  },
136
  "eval_I": {
137
+ "f1-score": 0.9589184992350025,
138
+ "precision": 0.945595388218339,
139
+ "recall": 0.9726224116366332,
140
+ "support": 21587.0
141
  },
142
  "eval_O": {
143
+ "f1-score": 0.9091178797468354,
144
+ "precision": 0.9427750999897446,
145
+ "recall": 0.877780960565263,
146
+ "support": 10473.0
147
  },
148
+ "eval_accuracy": 0.9403880597014925,
149
+ "eval_loss": 0.22083505988121033,
150
  "eval_macro avg": {
151
+ "f1-score": 0.9161461296773631,
152
+ "precision": 0.9129519750337626,
153
+ "recall": 0.9209677907339654,
154
+ "support": 33500.0
155
+ },
156
+ "eval_runtime": 4.7449,
157
+ "eval_samples_per_second": 17.071,
158
+ "eval_steps_per_second": 2.318,
159
  "eval_weighted avg": {
160
+ "f1-score": 0.9399744505088647,
161
+ "precision": 0.9406253819936745,
162
+ "recall": 0.9403880597014925,
163
+ "support": 33500.0
164
  },
165
  "step": 324
166
  },
167
  {
168
  "epoch": 5.0,
169
  "eval_B": {
170
+ "f1-score": 0.8879781420765027,
171
+ "precision": 0.8736559139784946,
172
+ "recall": 0.9027777777777778,
173
+ "support": 1440.0
174
  },
175
  "eval_I": {
176
+ "f1-score": 0.9613201259451634,
177
+ "precision": 0.9539317642765919,
178
+ "recall": 0.9688238291564367,
179
+ "support": 21587.0
180
  },
181
  "eval_O": {
182
+ "f1-score": 0.916006030835076,
183
+ "precision": 0.9334853291038858,
184
+ "recall": 0.899169292466342,
185
+ "support": 10473.0
186
  },
187
+ "eval_accuracy": 0.9442089552238806,
188
+ "eval_loss": 0.2257838249206543,
189
  "eval_macro avg": {
190
+ "f1-score": 0.921768099618914,
191
+ "precision": 0.9203576691196576,
192
+ "recall": 0.9235902998001855,
193
+ "support": 33500.0
194
+ },
195
+ "eval_runtime": 4.7845,
196
+ "eval_samples_per_second": 16.93,
197
+ "eval_steps_per_second": 2.299,
198
  "eval_weighted avg": {
199
+ "f1-score": 0.9440011117702735,
200
+ "precision": 0.9440889959185916,
201
+ "recall": 0.9442089552238806,
202
+ "support": 33500.0
203
  },
204
  "step": 405
205
  },
206
  {
207
  "epoch": 6.0,
208
  "eval_B": {
209
+ "f1-score": 0.8873579056148813,
210
+ "precision": 0.8803827751196173,
211
+ "recall": 0.8944444444444445,
212
+ "support": 1440.0
213
  },
214
  "eval_I": {
215
+ "f1-score": 0.9622060984570168,
216
+ "precision": 0.9539267015706806,
217
+ "recall": 0.9706304720433594,
218
+ "support": 21587.0
219
  },
220
  "eval_O": {
221
+ "f1-score": 0.9182769530299342,
222
+ "precision": 0.9365567911040509,
223
+ "recall": 0.9006970304592762,
224
+ "support": 10473.0
225
  },
226
+ "eval_accuracy": 0.9454925373134329,
227
+ "eval_loss": 0.2242102175951004,
228
  "eval_macro avg": {
229
+ "f1-score": 0.9226136523672773,
230
+ "precision": 0.923622089264783,
231
+ "recall": 0.9219239823156933,
232
+ "support": 33500.0
233
+ },
234
+ "eval_runtime": 4.7715,
235
+ "eval_samples_per_second": 16.976,
236
+ "eval_steps_per_second": 2.305,
237
  "eval_weighted avg": {
238
+ "f1-score": 0.945255312255509,
239
+ "precision": 0.9453351097376493,
240
+ "recall": 0.9454925373134329,
241
+ "support": 33500.0
242
  },
243
  "step": 486
244
  },
245
  {
246
  "epoch": 6.17,
247
+ "grad_norm": 1.8705081939697266,
248
  "learning_rate": 1.7530864197530865e-05,
249
+ "loss": 0.1601,
250
  "step": 500
251
  },
252
  {
253
  "epoch": 7.0,
254
  "eval_B": {
255
+ "f1-score": 0.8876749743939911,
256
+ "precision": 0.8730691739422431,
257
+ "recall": 0.9027777777777778,
258
+ "support": 1440.0
259
  },
260
  "eval_I": {
261
+ "f1-score": 0.9611415113943341,
262
+ "precision": 0.9581108451482232,
263
+ "recall": 0.9641914114976606,
264
+ "support": 21587.0
265
  },
266
  "eval_O": {
267
+ "f1-score": 0.9164739884393064,
268
+ "precision": 0.9247594050743657,
269
+ "recall": 0.9083357204239473,
270
+ "support": 10473.0
271
  },
272
+ "eval_accuracy": 0.944089552238806,
273
+ "eval_loss": 0.22517083585262299,
274
  "eval_macro avg": {
275
+ "f1-score": 0.9217634914092105,
276
+ "precision": 0.9186464747216107,
277
+ "recall": 0.9251016365664619,
278
+ "support": 33500.0
279
+ },
280
+ "eval_runtime": 4.7926,
281
+ "eval_samples_per_second": 16.901,
282
+ "eval_steps_per_second": 2.295,
283
  "eval_weighted avg": {
284
+ "f1-score": 0.9440192791200506,
285
+ "precision": 0.9440287663891153,
286
+ "recall": 0.944089552238806,
287
+ "support": 33500.0
288
  },
289
  "step": 567
290
  },
291
  {
292
  "epoch": 8.0,
293
  "eval_B": {
294
+ "f1-score": 0.894540522210919,
295
+ "precision": 0.8740888005301524,
296
+ "recall": 0.9159722222222222,
297
+ "support": 1440.0
298
  },
299
  "eval_I": {
300
+ "f1-score": 0.9638018385639943,
301
+ "precision": 0.9540255968049378,
302
+ "recall": 0.9737805160513272,
303
+ "support": 21587.0
304
  },
305
  "eval_O": {
306
+ "f1-score": 0.9203132648066569,
307
+ "precision": 0.9441598875163202,
308
+ "recall": 0.8976415544734078,
309
+ "support": 10473.0
310
  },
311
+ "eval_accuracy": 0.9474925373134329,
312
+ "eval_loss": 0.24869142472743988,
313
  "eval_macro avg": {
314
+ "f1-score": 0.9262185418605234,
315
+ "precision": 0.9240914282838034,
316
+ "recall": 0.9291314309156524,
317
+ "support": 33500.0
318
+ },
319
+ "eval_runtime": 4.7743,
320
+ "eval_samples_per_second": 16.966,
321
+ "eval_steps_per_second": 2.304,
322
  "eval_weighted avg": {
323
+ "f1-score": 0.947228939205516,
324
+ "precision": 0.9475052218791054,
325
+ "recall": 0.9474925373134329,
326
+ "support": 33500.0
327
  },
328
  "step": 648
329
  },
330
  {
331
  "epoch": 9.0,
332
  "eval_B": {
333
+ "f1-score": 0.8851236868858015,
334
+ "precision": 0.8643282594308405,
335
+ "recall": 0.9069444444444444,
336
+ "support": 1440.0
337
  },
338
  "eval_I": {
339
+ "f1-score": 0.9617001723147618,
340
+ "precision": 0.9540067462849849,
341
+ "recall": 0.9695186918052532,
342
+ "support": 21587.0
343
  },
344
  "eval_O": {
345
+ "f1-score": 0.9165854609237966,
346
+ "precision": 0.9358272808675754,
347
+ "recall": 0.8981189725961998,
348
+ "support": 10473.0
349
  },
350
+ "eval_accuracy": 0.9445074626865672,
351
+ "eval_loss": 0.3045330345630646,
352
  "eval_macro avg": {
353
+ "f1-score": 0.9211364400414532,
354
+ "precision": 0.9180540955278002,
355
+ "recall": 0.9248607029486324,
356
+ "support": 33500.0
357
+ },
358
+ "eval_runtime": 4.7599,
359
+ "eval_samples_per_second": 17.017,
360
+ "eval_steps_per_second": 2.311,
361
  "eval_weighted avg": {
362
+ "f1-score": 0.9443044555560967,
363
+ "precision": 0.9444685205421044,
364
+ "recall": 0.9445074626865672,
365
+ "support": 33500.0
366
  },
367
  "step": 729
368
  },
369
  {
370
  "epoch": 10.0,
371
  "eval_B": {
372
+ "f1-score": 0.8815068493150685,
373
+ "precision": 0.8695945945945946,
374
+ "recall": 0.89375,
375
+ "support": 1440.0
376
  },
377
  "eval_I": {
378
+ "f1-score": 0.956297959947595,
379
+ "precision": 0.9660159757999717,
380
+ "recall": 0.9467735211006625,
381
+ "support": 21587.0
382
  },
383
  "eval_O": {
384
+ "f1-score": 0.9101049868766404,
385
+ "precision": 0.8937678357728068,
386
+ "recall": 0.9270505108373914,
387
+ "support": 10473.0
388
  },
389
+ "eval_accuracy": 0.9383283582089552,
390
+ "eval_loss": 0.32627996802330017,
391
  "eval_macro avg": {
392
+ "f1-score": 0.9159699320464346,
393
+ "precision": 0.909792802055791,
394
+ "recall": 0.9225246773126846,
395
+ "support": 33500.0
396
+ },
397
+ "eval_runtime": 4.7534,
398
+ "eval_samples_per_second": 17.04,
399
+ "eval_steps_per_second": 2.314,
400
  "eval_weighted avg": {
401
+ "f1-score": 0.9386418940884026,
402
+ "precision": 0.9392845859659347,
403
+ "recall": 0.9383283582089552,
404
+ "support": 33500.0
405
  },
406
  "step": 810
407
  },
408
  {
409
  "epoch": 11.0,
410
  "eval_B": {
411
+ "f1-score": 0.8901136755080951,
412
+ "precision": 0.8831168831168831,
413
+ "recall": 0.8972222222222223,
414
+ "support": 1440.0
415
  },
416
  "eval_I": {
417
+ "f1-score": 0.9620073001078947,
418
+ "precision": 0.9535359970874671,
419
+ "recall": 0.9706304720433594,
420
+ "support": 21587.0
421
  },
422
  "eval_O": {
423
+ "f1-score": 0.917413322945072,
424
+ "precision": 0.9361025539103647,
425
+ "recall": 0.8994557433400172,
426
+ "support": 10473.0
427
  },
428
+ "eval_accuracy": 0.9452238805970149,
429
+ "eval_loss": 0.3123624622821808,
430
  "eval_macro avg": {
431
+ "f1-score": 0.9231780995203539,
432
+ "precision": 0.9242518113715716,
433
+ "recall": 0.922436145868533,
434
+ "support": 33500.0
435
+ },
436
+ "eval_runtime": 4.7539,
437
+ "eval_samples_per_second": 17.039,
438
+ "eval_steps_per_second": 2.314,
439
  "eval_weighted avg": {
440
+ "f1-score": 0.9449756719810304,
441
+ "precision": 0.9450588635199616,
442
+ "recall": 0.9452238805970149,
443
+ "support": 33500.0
444
  },
445
  "step": 891
446
  },
447
  {
448
  "epoch": 12.0,
449
  "eval_B": {
450
+ "f1-score": 0.8800819952169457,
451
+ "precision": 0.8661735036987223,
452
+ "recall": 0.8944444444444445,
453
+ "support": 1440.0
454
  },
455
  "eval_I": {
456
+ "f1-score": 0.9595463311598861,
457
+ "precision": 0.9512039692293687,
458
+ "recall": 0.9680363181544448,
459
+ "support": 21587.0
460
  },
461
  "eval_O": {
462
+ "f1-score": 0.9115367743822197,
463
+ "precision": 0.9310035842293907,
464
+ "recall": 0.8928673732454884,
465
+ "support": 10473.0
466
  },
467
+ "eval_accuracy": 0.9413731343283582,
468
+ "eval_loss": 0.3427491784095764,
469
  "eval_macro avg": {
470
+ "f1-score": 0.9170550335863505,
471
+ "precision": 0.9161270190524938,
472
+ "recall": 0.9184493786147926,
473
+ "support": 33500.0
474
+ },
475
+ "eval_runtime": 4.7455,
476
+ "eval_samples_per_second": 17.069,
477
+ "eval_steps_per_second": 2.318,
478
  "eval_weighted avg": {
479
+ "f1-score": 0.9411214734915179,
480
+ "precision": 0.9412337452750731,
481
+ "recall": 0.9413731343283582,
482
+ "support": 33500.0
483
  },
484
  "step": 972
485
  },
486
  {
487
  "epoch": 12.35,
488
+ "grad_norm": 0.47551417350769043,
489
  "learning_rate": 1.506172839506173e-05,
490
+ "loss": 0.0222,
491
  "step": 1000
492
  },
493
  {
494
  "epoch": 13.0,
495
  "eval_B": {
496
+ "f1-score": 0.8896434634974534,
497
+ "precision": 0.8704318936877077,
498
+ "recall": 0.9097222222222222,
499
+ "support": 1440.0
500
  },
501
  "eval_I": {
502
+ "f1-score": 0.9620136667970457,
503
+ "precision": 0.9556591698665204,
504
+ "recall": 0.9684532357437347,
505
+ "support": 21587.0
506
  },
507
  "eval_O": {
508
+ "f1-score": 0.9183177933177933,
509
+ "precision": 0.9343808676746714,
510
+ "recall": 0.9027976701995608,
511
+ "support": 10473.0
512
  },
513
+ "eval_accuracy": 0.9454029850746268,
514
+ "eval_loss": 0.3816670775413513,
515
  "eval_macro avg": {
516
+ "f1-score": 0.9233249745374308,
517
+ "precision": 0.9201573104096331,
518
+ "recall": 0.9269910427218392,
519
+ "support": 33500.0
520
+ },
521
+ "eval_runtime": 4.7247,
522
+ "eval_samples_per_second": 17.144,
523
+ "eval_steps_per_second": 2.328,
524
  "eval_weighted avg": {
525
+ "f1-score": 0.945242324238848,
526
+ "precision": 0.9453435001186778,
527
+ "recall": 0.9454029850746268,
528
+ "support": 33500.0
529
  },
530
  "step": 1053
531
  },
532
  {
533
  "epoch": 14.0,
534
  "eval_B": {
535
+ "f1-score": 0.8865272540281112,
536
+ "precision": 0.8754231550440081,
537
+ "recall": 0.8979166666666667,
538
+ "support": 1440.0
539
  },
540
  "eval_I": {
541
+ "f1-score": 0.9599206715247671,
542
+ "precision": 0.9557331129172981,
543
+ "recall": 0.9641450873210728,
544
+ "support": 21587.0
545
  },
546
  "eval_O": {
547
+ "f1-score": 0.9141367826632559,
548
+ "precision": 0.9242631270739801,
549
+ "recall": 0.9042299245679366,
550
+ "support": 10473.0
551
  },
552
+ "eval_accuracy": 0.9425671641791045,
553
+ "eval_loss": 0.3827010989189148,
554
  "eval_macro avg": {
555
+ "f1-score": 0.9201949027387114,
556
+ "precision": 0.9184731316784287,
557
+ "recall": 0.9220972261852255,
558
+ "support": 33500.0
559
+ },
560
+ "eval_runtime": 4.7319,
561
+ "eval_samples_per_second": 17.118,
562
+ "eval_steps_per_second": 2.325,
563
  "eval_weighted avg": {
564
+ "f1-score": 0.9424525763235196,
565
+ "precision": 0.9424426203479068,
566
+ "recall": 0.9425671641791045,
567
+ "support": 33500.0
568
  },
569
  "step": 1134
570
  },
571
  {
572
  "epoch": 15.0,
573
  "eval_B": {
574
+ "f1-score": 0.8871919472405415,
575
+ "precision": 0.8868841082581541,
576
+ "recall": 0.8875,
577
+ "support": 1440.0
578
  },
579
  "eval_I": {
580
+ "f1-score": 0.9616856471022177,
581
+ "precision": 0.9571841585975862,
582
+ "recall": 0.9662296752675221,
583
+ "support": 21587.0
584
  },
585
  "eval_O": {
586
+ "f1-score": 0.9181813798756087,
587
+ "precision": 0.9273470977795092,
588
+ "recall": 0.9091950730449728,
589
+ "support": 10473.0
590
  },
591
+ "eval_accuracy": 0.9450149253731344,
592
+ "eval_loss": 0.39016807079315186,
593
  "eval_macro avg": {
594
+ "f1-score": 0.9223529914061226,
595
+ "precision": 0.9238051215450831,
596
+ "recall": 0.920974916104165,
597
+ "support": 33500.0
598
+ },
599
+ "eval_runtime": 4.7376,
600
+ "eval_samples_per_second": 17.097,
601
+ "eval_steps_per_second": 2.322,
602
  "eval_weighted avg": {
603
+ "f1-score": 0.944882927148036,
604
+ "precision": 0.9448344388830637,
605
+ "recall": 0.9450149253731344,
606
+ "support": 33500.0
607
  },
608
  "step": 1215
609
  },
610
  {
611
  "epoch": 16.0,
612
  "eval_B": {
613
+ "f1-score": 0.8897338403041825,
614
+ "precision": 0.8857536132140399,
615
+ "recall": 0.89375,
616
+ "support": 1440.0
617
  },
618
  "eval_I": {
619
+ "f1-score": 0.9608010520729992,
620
+ "precision": 0.9570693142121713,
621
+ "recall": 0.9645620049103627,
622
+ "support": 21587.0
623
  },
624
  "eval_O": {
625
+ "f1-score": 0.9159121556540166,
626
+ "precision": 0.9240112719852298,
627
+ "recall": 0.9079537859257137,
628
+ "support": 10473.0
629
  },
630
+ "eval_accuracy": 0.943820895522388,
631
+ "eval_loss": 0.40087994933128357,
632
  "eval_macro avg": {
633
+ "f1-score": 0.9221490160103993,
634
+ "precision": 0.9222780664704805,
635
+ "recall": 0.9220885969453588,
636
+ "support": 33500.0
637
+ },
638
+ "eval_runtime": 4.8145,
639
+ "eval_samples_per_second": 16.824,
640
+ "eval_steps_per_second": 2.285,
641
  "eval_weighted avg": {
642
+ "f1-score": 0.9437127476806678,
643
+ "precision": 0.9436689713560499,
644
+ "recall": 0.943820895522388,
645
+ "support": 33500.0
646
  },
647
  "step": 1296
648
  },
649
  {
650
  "epoch": 17.0,
651
  "eval_B": {
652
+ "f1-score": 0.8904347826086957,
653
+ "precision": 0.89198606271777,
654
+ "recall": 0.8888888888888888,
655
+ "support": 1440.0
656
  },
657
  "eval_I": {
658
+ "f1-score": 0.9615171394252979,
659
+ "precision": 0.9589034784611841,
660
+ "recall": 0.9641450873210728,
661
+ "support": 21587.0
662
  },
663
  "eval_O": {
664
+ "f1-score": 0.9182546920750733,
665
+ "precision": 0.9232625482625483,
666
+ "recall": 0.9133008689009835,
667
+ "support": 10473.0
668
  },
669
+ "eval_accuracy": 0.9450149253731344,
670
+ "eval_loss": 0.42787015438079834,
671
  "eval_macro avg": {
672
+ "f1-score": 0.9234022047030224,
673
+ "precision": 0.9247173631471673,
674
+ "recall": 0.9221116150369818,
675
+ "support": 33500.0
676
+ },
677
+ "eval_runtime": 4.7453,
678
+ "eval_samples_per_second": 17.07,
679
+ "eval_steps_per_second": 2.318,
680
  "eval_weighted avg": {
681
+ "f1-score": 0.9449366556964975,
682
+ "precision": 0.9448847160539952,
683
+ "recall": 0.9450149253731344,
684
+ "support": 33500.0
685
  },
686
  "step": 1377
687
  },
688
  {
689
  "epoch": 18.0,
690
  "eval_B": {
691
+ "f1-score": 0.8895027624309393,
692
+ "precision": 0.8846153846153846,
693
+ "recall": 0.8944444444444445,
694
+ "support": 1440.0
695
  },
696
  "eval_I": {
697
+ "f1-score": 0.9607206210979795,
698
+ "precision": 0.9555912007332722,
699
+ "recall": 0.9659054060314078,
700
+ "support": 21587.0
701
  },
702
  "eval_O": {
703
+ "f1-score": 0.9159781610861478,
704
+ "precision": 0.9271322378716745,
705
+ "recall": 0.905089277188962,
706
+ "support": 10473.0
707
  },
708
+ "eval_accuracy": 0.943820895522388,
709
+ "eval_loss": 0.422620564699173,
710
  "eval_macro avg": {
711
+ "f1-score": 0.9220671815383555,
712
+ "precision": 0.9224462744067772,
713
+ "recall": 0.9218130425549381,
714
+ "support": 33500.0
715
+ },
716
+ "eval_runtime": 4.7703,
717
+ "eval_samples_per_second": 16.98,
718
+ "eval_steps_per_second": 2.306,
719
  "eval_weighted avg": {
720
+ "f1-score": 0.9436716210924735,
721
+ "precision": 0.9436432636210552,
722
+ "recall": 0.943820895522388,
723
+ "support": 33500.0
724
  },
725
  "step": 1458
726
  },
727
  {
728
  "epoch": 18.52,
729
+ "grad_norm": 0.1027584820985794,
730
  "learning_rate": 1.2592592592592593e-05,
731
+ "loss": 0.0054,
732
  "step": 1500
733
  },
734
  {
735
  "epoch": 19.0,
736
  "eval_B": {
737
+ "f1-score": 0.8924247665167762,
738
+ "precision": 0.8890420399724328,
739
+ "recall": 0.8958333333333334,
740
+ "support": 1440.0
741
  },
742
  "eval_I": {
743
+ "f1-score": 0.9599648245857633,
744
+ "precision": 0.959121387283237,
745
+ "recall": 0.9608097466067541,
746
+ "support": 21587.0
747
  },
748
  "eval_O": {
749
+ "f1-score": 0.9158252380724505,
750
+ "precision": 0.9179777436684574,
751
+ "recall": 0.913682803399217,
752
+ "support": 10473.0
753
  },
754
+ "eval_accuracy": 0.9432835820895522,
755
+ "eval_loss": 0.4375981390476227,
756
  "eval_macro avg": {
757
+ "f1-score": 0.9227382763916633,
758
+ "precision": 0.9220470569747091,
759
+ "recall": 0.9234419611131015,
760
+ "support": 33500.0
761
+ },
762
+ "eval_runtime": 4.7487,
763
+ "eval_samples_per_second": 17.057,
764
+ "eval_steps_per_second": 2.316,
765
  "eval_weighted avg": {
766
+ "f1-score": 0.9432623895656658,
767
+ "precision": 0.9432464129636805,
768
+ "recall": 0.9432835820895522,
769
+ "support": 33500.0
770
  },
771
  "step": 1539
772
  },
773
  {
774
  "epoch": 20.0,
775
  "eval_B": {
776
+ "f1-score": 0.8907273354015857,
777
+ "precision": 0.8843258042436687,
778
+ "recall": 0.8972222222222223,
779
+ "support": 1440.0
780
  },
781
  "eval_I": {
782
+ "f1-score": 0.9620682542229149,
783
+ "precision": 0.9520958083832335,
784
+ "recall": 0.972251818223931,
785
+ "support": 21587.0
786
  },
787
  "eval_O": {
788
+ "f1-score": 0.9177252296267345,
789
+ "precision": 0.9396698349174587,
790
+ "recall": 0.8967822018523823,
791
+ "support": 10473.0
792
  },
793
+ "eval_accuracy": 0.9454328358208955,
794
+ "eval_loss": 0.44444534182548523,
795
  "eval_macro avg": {
796
+ "f1-score": 0.9235069397504118,
797
+ "precision": 0.9253638158481202,
798
+ "recall": 0.922085414099512,
799
+ "support": 33500.0
800
+ },
801
+ "eval_runtime": 4.7631,
802
+ "eval_samples_per_second": 17.006,
803
+ "eval_steps_per_second": 2.309,
804
  "eval_weighted avg": {
805
+ "f1-score": 0.9451388387095264,
806
+ "precision": 0.9452980165603072,
807
+ "recall": 0.9454328358208955,
808
+ "support": 33500.0
809
  },
810
  "step": 1620
811
  },
812
  {
813
  "epoch": 21.0,
814
  "eval_B": {
815
+ "f1-score": 0.890790825059911,
816
+ "precision": 0.8784604996623903,
817
+ "recall": 0.9034722222222222,
818
+ "support": 1440.0
819
  },
820
  "eval_I": {
821
+ "f1-score": 0.9615375783600082,
822
+ "precision": 0.9533284764593388,
823
+ "recall": 0.9698892852179553,
824
+ "support": 21587.0
825
  },
826
  "eval_O": {
827
+ "f1-score": 0.9162201656113005,
828
+ "precision": 0.9351695336581486,
829
+ "recall": 0.8980234889716414,
830
+ "support": 10473.0
831
  },
832
+ "eval_accuracy": 0.9445671641791045,
833
+ "eval_loss": 0.4363422691822052,
834
  "eval_macro avg": {
835
+ "f1-score": 0.9228495230104065,
836
+ "precision": 0.9223195032599593,
837
+ "recall": 0.9237949988039397,
838
+ "support": 33500.0
839
+ },
840
+ "eval_runtime": 4.7902,
841
+ "eval_samples_per_second": 16.91,
842
+ "eval_steps_per_second": 2.296,
843
  "eval_weighted avg": {
844
+ "f1-score": 0.9443290831818184,
845
+ "precision": 0.9444332975177129,
846
+ "recall": 0.9445671641791045,
847
+ "support": 33500.0
848
  },
849
  "step": 1701
850
  },
851
  {
852
  "epoch": 22.0,
853
  "eval_B": {
854
+ "f1-score": 0.8907216494845361,
855
+ "precision": 0.8816326530612245,
856
+ "recall": 0.9,
857
+ "support": 1440.0
858
  },
859
  "eval_I": {
860
+ "f1-score": 0.961758728985775,
861
+ "precision": 0.9588801399825022,
862
+ "recall": 0.9646546532635383,
863
+ "support": 21587.0
864
  },
865
  "eval_O": {
866
+ "f1-score": 0.9186952756663139,
867
+ "precision": 0.9258217783380199,
868
+ "recall": 0.9116776472834909,
869
+ "support": 10473.0
870
  },
871
+ "eval_accuracy": 0.9453134328358209,
872
+ "eval_loss": 0.3998025953769684,
873
  "eval_macro avg": {
874
+ "f1-score": 0.9237252180455416,
875
+ "precision": 0.9221115237939155,
876
+ "recall": 0.925444100182343,
877
+ "support": 33500.0
878
+ },
879
+ "eval_runtime": 4.7313,
880
+ "eval_samples_per_second": 17.12,
881
+ "eval_steps_per_second": 2.325,
882
  "eval_weighted avg": {
883
+ "f1-score": 0.9452424023858794,
884
+ "precision": 0.9452247190072991,
885
+ "recall": 0.9453134328358209,
886
+ "support": 33500.0
887
  },
888
  "step": 1782
889
  },
890
  {
891
  "epoch": 23.0,
892
  "eval_B": {
893
+ "f1-score": 0.8938053097345132,
894
+ "precision": 0.8765020026702269,
895
+ "recall": 0.9118055555555555,
896
+ "support": 1440.0
897
  },
898
  "eval_I": {
899
+ "f1-score": 0.9620159689850926,
900
+ "precision": 0.9584770313146641,
901
+ "recall": 0.9655811367952934,
902
+ "support": 21587.0
903
  },
904
  "eval_O": {
905
+ "f1-score": 0.918660748745658,
906
+ "precision": 0.9284251584592882,
907
+ "recall": 0.9090995894204144,
908
+ "support": 10473.0
909
  },
910
+ "eval_accuracy": 0.9456119402985075,
911
+ "eval_loss": 0.4420088231563568,
912
  "eval_macro avg": {
913
+ "f1-score": 0.9248273424884212,
914
+ "precision": 0.9211347308147264,
915
+ "recall": 0.9288287605904211,
916
+ "support": 33500.0
917
+ },
918
+ "eval_runtime": 4.7375,
919
+ "eval_samples_per_second": 17.098,
920
+ "eval_steps_per_second": 2.322,
921
  "eval_weighted avg": {
922
+ "f1-score": 0.9455299220929005,
923
+ "precision": 0.9455583057725045,
924
+ "recall": 0.9456119402985075,
925
+ "support": 33500.0
926
  },
927
  "step": 1863
928
  },
929
  {
930
  "epoch": 24.0,
931
  "eval_B": {
932
+ "f1-score": 0.89482699554642,
933
+ "precision": 0.8830290736984449,
934
+ "recall": 0.9069444444444444,
935
+ "support": 1440.0
936
  },
937
  "eval_I": {
938
+ "f1-score": 0.962135833103575,
939
+ "precision": 0.9544605005242285,
940
+ "recall": 0.969935609394543,
941
+ "support": 21587.0
942
  },
943
  "eval_O": {
944
+ "f1-score": 0.9185192391885975,
945
+ "precision": 0.9362356207854026,
946
+ "recall": 0.9014608994557434,
947
+ "support": 10473.0
948
  },
949
+ "eval_accuracy": 0.945820895522388,
950
+ "eval_loss": 0.4822453558444977,
951
  "eval_macro avg": {
952
+ "f1-score": 0.9251606892795309,
953
+ "precision": 0.924575065002692,
954
+ "recall": 0.9261136510982436,
955
+ "support": 33500.0
956
+ },
957
+ "eval_runtime": 4.7345,
958
+ "eval_samples_per_second": 17.108,
959
+ "eval_steps_per_second": 2.323,
960
  "eval_weighted avg": {
961
+ "f1-score": 0.9456068386512209,
962
+ "precision": 0.945692428281427,
963
+ "recall": 0.945820895522388,
964
+ "support": 33500.0
965
  },
966
  "step": 1944
967
  },
968
  {
969
  "epoch": 24.69,
970
+ "grad_norm": 0.10262901335954666,
971
  "learning_rate": 1.0123456790123458e-05,
972
+ "loss": 0.0037,
973
  "step": 2000
974
  },
975
  {
976
  "epoch": 25.0,
977
  "eval_B": {
978
+ "f1-score": 0.8864342423196411,
979
+ "precision": 0.8812628689087165,
980
+ "recall": 0.8916666666666667,
981
+ "support": 1440.0
982
  },
983
  "eval_I": {
984
+ "f1-score": 0.9595285234126801,
985
+ "precision": 0.9537320946409775,
986
+ "recall": 0.9653958400889424,
987
+ "support": 21587.0
988
  },
989
  "eval_O": {
990
+ "f1-score": 0.9129445923058311,
991
+ "precision": 0.9255298273155416,
992
+ "recall": 0.9006970304592762,
993
+ "support": 10473.0
994
  },
995
+ "eval_accuracy": 0.942,
996
+ "eval_loss": 0.46240004897117615,
997
  "eval_macro avg": {
998
+ "f1-score": 0.9196357860127174,
999
+ "precision": 0.9201749302884119,
1000
+ "recall": 0.9192531790716284,
1001
+ "support": 33500.0
1002
+ },
1003
+ "eval_runtime": 4.7408,
1004
+ "eval_samples_per_second": 17.086,
1005
+ "eval_steps_per_second": 2.32,
1006
  "eval_weighted avg": {
1007
+ "f1-score": 0.941823171912501,
1008
+ "precision": 0.9418002131259403,
1009
+ "recall": 0.942,
1010
+ "support": 33500.0
1011
  },
1012
  "step": 2025
1013
  }
 
1017
  "num_input_tokens_seen": 0,
1018
  "num_train_epochs": 50,
1019
  "save_steps": 500,
1020
+ "total_flos": 3583502758395000.0,
1021
  "train_batch_size": 4,
1022
  "trial_name": null,
1023
  "trial_params": null
checkpoint-2025/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d229d7bc9127d222c880edfd3fa9418a9ccede656fe27f7f0aa04990c9605be4
3
  size 4856
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8ca363922308e3892528dd031742ffedc372734b464a042d09444af9700901e
3
  size 4856