Theoreticallyhugo commited on
Commit
911648d
1 Parent(s): 79429ca

Training in progress, epoch 14, checkpoint

Browse files
checkpoint-1134/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c62609734eb7f961b4813a38da037fe34e81df227912551ab858b74d8c722563
3
  size 592318676
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81be16fcef6861639ab1d7dc2a0b57bb0395142c0eb4bc342266adcf0fd537bc
3
  size 592318676
checkpoint-1134/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1917018ff9d6a615d0ad43348d16a1d59f130a47586a238a6b0e0673cd472575
3
  size 1014645498
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7d52305fd80970a572bd126e844e6d8687f1d7669713a1cb571941f2611b7e9
3
  size 1014645498
checkpoint-1134/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:14e677a8e4cf0651cd6287bdf9c4606c6b0d01f2faa8f50f92dc07071e2f16e0
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:597b4af46402de550809c9388e42f2a2ec66bffe613bfb15b48cafe52a9de8a6
3
  size 1064
checkpoint-1134/trainer_state.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
- "best_metric": null,
3
- "best_model_checkpoint": null,
4
  "epoch": 14.0,
5
  "eval_steps": 500,
6
  "global_step": 1134,
@@ -11,568 +11,568 @@
11
  {
12
  "epoch": 1.0,
13
  "eval_B": {
14
- "f1-score": 0.8331822302810515,
15
- "precision": 0.7901977644024075,
16
- "recall": 0.8811121764141898,
17
- "support": 1043.0
18
  },
19
  "eval_I": {
20
- "f1-score": 0.9471294753175105,
21
- "precision": 0.9443616777446711,
22
- "recall": 0.9499135446685879,
23
- "support": 17350.0
24
  },
25
  "eval_O": {
26
- "f1-score": 0.8992868897421833,
27
- "precision": 0.9103731674811195,
28
- "recall": 0.8884673748103187,
29
- "support": 9226.0
30
  },
31
- "eval_accuracy": 0.9267895289474637,
32
- "eval_loss": 0.20478281378746033,
33
  "eval_macro avg": {
34
- "f1-score": 0.8931995317802484,
35
- "precision": 0.8816442032093993,
36
- "recall": 0.9064976986310321,
37
- "support": 27619.0
38
- },
39
- "eval_runtime": 4.6128,
40
- "eval_samples_per_second": 17.343,
41
- "eval_steps_per_second": 2.168,
42
  "eval_weighted avg": {
43
- "f1-score": 0.9268447919078652,
44
- "precision": 0.9271861479533134,
45
- "recall": 0.9267895289474637,
46
- "support": 27619.0
47
  },
48
  "step": 81
49
  },
50
  {
51
  "epoch": 2.0,
52
  "eval_B": {
53
- "f1-score": 0.8682547399124939,
54
- "precision": 0.8806706114398422,
55
- "recall": 0.8561840843720039,
56
- "support": 1043.0
57
  },
58
  "eval_I": {
59
- "f1-score": 0.9590983771942466,
60
- "precision": 0.9592919333448654,
61
- "recall": 0.9589048991354466,
62
- "support": 17350.0
63
  },
64
  "eval_O": {
65
- "f1-score": 0.9226525313717006,
66
- "precision": 0.9208594256100194,
67
- "recall": 0.9244526338608281,
68
- "support": 9226.0
69
  },
70
- "eval_accuracy": 0.9435171439950758,
71
- "eval_loss": 0.16574698686599731,
72
  "eval_macro avg": {
73
- "f1-score": 0.9166685494928136,
74
- "precision": 0.9202739901315756,
75
- "recall": 0.9131805391227594,
76
- "support": 27619.0
77
- },
78
- "eval_runtime": 4.6238,
79
- "eval_samples_per_second": 17.302,
80
- "eval_steps_per_second": 2.163,
81
  "eval_weighted avg": {
82
- "f1-score": 0.9434932036816766,
83
- "precision": 0.9434846863370583,
84
- "recall": 0.9435171439950758,
85
- "support": 27619.0
86
  },
87
  "step": 162
88
  },
89
  {
90
  "epoch": 3.0,
91
  "eval_B": {
92
- "f1-score": 0.8791725434884815,
93
- "precision": 0.8625461254612546,
94
- "recall": 0.8964525407478428,
95
- "support": 1043.0
96
  },
97
  "eval_I": {
98
- "f1-score": 0.9588372159825352,
99
- "precision": 0.9557349825345015,
100
- "recall": 0.9619596541786744,
101
- "support": 17350.0
102
  },
103
  "eval_O": {
104
- "f1-score": 0.9204284621270084,
105
- "precision": 0.9282407407407407,
106
- "recall": 0.9127465857359636,
107
- "support": 9226.0
108
  },
109
- "eval_accuracy": 0.9430464535283681,
110
- "eval_loss": 0.15394224226474762,
111
  "eval_macro avg": {
112
- "f1-score": 0.9194794071993417,
113
- "precision": 0.9155072829121655,
114
- "recall": 0.9237195935541602,
115
- "support": 27619.0
116
- },
117
- "eval_runtime": 4.6233,
118
- "eval_samples_per_second": 17.303,
119
- "eval_steps_per_second": 2.163,
120
  "eval_weighted avg": {
121
- "f1-score": 0.9429985029052194,
122
- "precision": 0.9430314866542512,
123
- "recall": 0.9430464535283681,
124
- "support": 27619.0
125
  },
126
  "step": 243
127
  },
128
  {
129
  "epoch": 4.0,
130
  "eval_B": {
131
- "f1-score": 0.8800746616892207,
132
- "precision": 0.8572727272727273,
133
- "recall": 0.9041227229146692,
134
- "support": 1043.0
135
  },
136
  "eval_I": {
137
- "f1-score": 0.9537844036697248,
138
- "precision": 0.9488876212207644,
139
- "recall": 0.9587319884726225,
140
- "support": 17350.0
141
  },
142
  "eval_O": {
143
- "f1-score": 0.9100192149327477,
144
- "precision": 0.9220157970853265,
145
- "recall": 0.8983308042488619,
146
- "support": 9226.0
147
  },
148
- "eval_accuracy": 0.9364929939534379,
149
- "eval_loss": 0.19801080226898193,
150
  "eval_macro avg": {
151
- "f1-score": 0.9146260934305644,
152
- "precision": 0.9093920485262728,
153
- "recall": 0.9203951718787179,
154
- "support": 27619.0
155
- },
156
- "eval_runtime": 4.6132,
157
- "eval_samples_per_second": 17.342,
158
- "eval_steps_per_second": 2.168,
159
  "eval_weighted avg": {
160
- "f1-score": 0.9363812792925563,
161
- "precision": 0.9364514800186444,
162
- "recall": 0.9364929939534379,
163
- "support": 27619.0
164
  },
165
  "step": 324
166
  },
167
  {
168
  "epoch": 5.0,
169
  "eval_B": {
170
- "f1-score": 0.883877159309021,
171
- "precision": 0.8847262247838616,
172
- "recall": 0.8830297219558965,
173
- "support": 1043.0
174
  },
175
  "eval_I": {
176
- "f1-score": 0.9591500482927107,
177
- "precision": 0.9456643513331839,
178
- "recall": 0.9730259365994236,
179
- "support": 17350.0
180
  },
181
  "eval_O": {
182
- "f1-score": 0.9191176470588236,
183
- "precision": 0.9454503781801513,
184
- "recall": 0.8942120095382614,
185
- "support": 9226.0
186
  },
187
- "eval_accuracy": 0.9432999022412107,
188
- "eval_loss": 0.20074255764484406,
189
  "eval_macro avg": {
190
- "f1-score": 0.9207149515535183,
191
- "precision": 0.9252803180990656,
192
- "recall": 0.9167558893645271,
193
- "support": 27619.0
194
- },
195
- "eval_runtime": 4.6092,
196
- "eval_samples_per_second": 17.356,
197
- "eval_steps_per_second": 2.17,
198
  "eval_weighted avg": {
199
- "f1-score": 0.9429348139614955,
200
- "precision": 0.9432916158141273,
201
- "recall": 0.9432999022412107,
202
- "support": 27619.0
203
  },
204
  "step": 405
205
  },
206
  {
207
  "epoch": 6.0,
208
  "eval_B": {
209
- "f1-score": 0.8897338403041826,
210
- "precision": 0.882186616399623,
211
- "recall": 0.8974113135186961,
212
- "support": 1043.0
213
  },
214
  "eval_I": {
215
- "f1-score": 0.9624160683661308,
216
- "precision": 0.9707969272268809,
217
- "recall": 0.954178674351585,
218
- "support": 17350.0
219
  },
220
  "eval_O": {
221
- "f1-score": 0.9297955261331482,
222
- "precision": 0.9161493950552341,
223
- "recall": 0.9438543247344461,
224
- "support": 9226.0
225
  },
226
- "eval_accuracy": 0.9485861182519281,
227
- "eval_loss": 0.16565027832984924,
228
  "eval_macro avg": {
229
- "f1-score": 0.9273151449344872,
230
- "precision": 0.9230443128939126,
231
- "recall": 0.9318147708682424,
232
- "support": 27619.0
233
- },
234
- "eval_runtime": 4.6219,
235
- "eval_samples_per_second": 17.309,
236
- "eval_steps_per_second": 2.164,
237
  "eval_weighted avg": {
238
- "f1-score": 0.9487745648174827,
239
- "precision": 0.9491959030765336,
240
- "recall": 0.9485861182519281,
241
- "support": 27619.0
242
  },
243
  "step": 486
244
  },
245
  {
246
  "epoch": 6.17,
247
- "grad_norm": 4.767655849456787,
248
- "learning_rate": 1.7530864197530865e-05,
249
- "loss": 0.1585,
250
  "step": 500
251
  },
252
  {
253
  "epoch": 7.0,
254
  "eval_B": {
255
- "f1-score": 0.8802588996763754,
256
- "precision": 0.85,
257
- "recall": 0.912751677852349,
258
- "support": 1043.0
259
  },
260
  "eval_I": {
261
- "f1-score": 0.958423523458803,
262
- "precision": 0.9354661691269275,
263
- "recall": 0.9825360230547551,
264
- "support": 17350.0
265
  },
266
  "eval_O": {
267
- "f1-score": 0.9139526911210146,
268
- "precision": 0.9664088931851136,
269
- "recall": 0.8668978972469109,
270
- "support": 9226.0
271
  },
272
- "eval_accuracy": 0.9412723125384699,
273
- "eval_loss": 0.2591131329536438,
274
  "eval_macro avg": {
275
- "f1-score": 0.9175450380853977,
276
- "precision": 0.917291687437347,
277
- "recall": 0.9207285327180049,
278
- "support": 27619.0
279
- },
280
- "eval_runtime": 4.617,
281
- "eval_samples_per_second": 17.327,
282
- "eval_steps_per_second": 2.166,
283
  "eval_weighted avg": {
284
- "f1-score": 0.9406164485555297,
285
- "precision": 0.9425749115781907,
286
- "recall": 0.9412723125384699,
287
- "support": 27619.0
288
  },
289
  "step": 567
290
  },
291
  {
292
  "epoch": 8.0,
293
  "eval_B": {
294
- "f1-score": 0.8687910028116214,
295
- "precision": 0.84967919340055,
296
- "recall": 0.8887823585810163,
297
- "support": 1043.0
298
  },
299
  "eval_I": {
300
- "f1-score": 0.9595010203788336,
301
- "precision": 0.956997878562009,
302
- "recall": 0.9620172910662824,
303
- "support": 17350.0
304
  },
305
  "eval_O": {
306
- "f1-score": 0.9212035166275324,
307
- "precision": 0.9282491471332673,
308
- "recall": 0.9142640364188164,
309
- "support": 9226.0
310
  },
311
- "eval_accuracy": 0.9432999022412107,
312
- "eval_loss": 0.2693409323692322,
313
  "eval_macro avg": {
314
- "f1-score": 0.9164985132726624,
315
- "precision": 0.9116420730319422,
316
- "recall": 0.9216878953553717,
317
- "support": 27619.0
318
- },
319
- "eval_runtime": 4.6153,
320
- "eval_samples_per_second": 17.333,
321
- "eval_steps_per_second": 2.167,
322
  "eval_weighted avg": {
323
- "f1-score": 0.9432823550422136,
324
- "precision": 0.9433417293609165,
325
- "recall": 0.9432999022412107,
326
- "support": 27619.0
327
  },
328
  "step": 648
329
  },
330
  {
331
  "epoch": 9.0,
332
  "eval_B": {
333
- "f1-score": 0.8645343367826904,
334
- "precision": 0.8485687903970452,
335
- "recall": 0.8811121764141898,
336
- "support": 1043.0
337
  },
338
  "eval_I": {
339
- "f1-score": 0.9567003348057282,
340
- "precision": 0.934535260814599,
341
- "recall": 0.9799423631123919,
342
- "support": 17350.0
343
  },
344
  "eval_O": {
345
- "f1-score": 0.9104673003585861,
346
- "precision": 0.9586479683567062,
347
- "recall": 0.8668978972469109,
348
- "support": 9226.0
349
  },
350
- "eval_accuracy": 0.9384481697382236,
351
- "eval_loss": 0.3087053894996643,
352
  "eval_macro avg": {
353
- "f1-score": 0.910567323982335,
354
- "precision": 0.9139173398561168,
355
- "recall": 0.9093174789244975,
356
- "support": 27619.0
357
- },
358
- "eval_runtime": 4.6156,
359
- "eval_samples_per_second": 17.333,
360
- "eval_steps_per_second": 2.167,
361
  "eval_weighted avg": {
362
- "f1-score": 0.9377758584761231,
363
- "precision": 0.9393435743356523,
364
- "recall": 0.9384481697382236,
365
- "support": 27619.0
366
  },
367
  "step": 729
368
  },
369
  {
370
  "epoch": 10.0,
371
  "eval_B": {
372
- "f1-score": 0.8803126526624329,
373
- "precision": 0.897410358565737,
374
- "recall": 0.8638542665388304,
375
- "support": 1043.0
376
  },
377
  "eval_I": {
378
- "f1-score": 0.9590983771942466,
379
- "precision": 0.9592919333448654,
380
- "recall": 0.9589048991354466,
381
- "support": 17350.0
382
  },
383
  "eval_O": {
384
- "f1-score": 0.9220456265542221,
385
- "precision": 0.9197584124245038,
386
- "recall": 0.9243442445263386,
387
- "support": 9226.0
388
  },
389
- "eval_accuracy": 0.9437705927079184,
390
- "eval_loss": 0.2796257436275482,
391
  "eval_macro avg": {
392
- "f1-score": 0.9204855521369671,
393
- "precision": 0.9254869014450354,
394
- "recall": 0.9157011367335386,
395
- "support": 27619.0
396
- },
397
- "eval_runtime": 4.641,
398
- "eval_samples_per_second": 17.238,
399
- "eval_steps_per_second": 2.155,
400
  "eval_weighted avg": {
401
- "f1-score": 0.9437458232244597,
402
- "precision": 0.9437490553802075,
403
- "recall": 0.9437705927079184,
404
- "support": 27619.0
405
  },
406
  "step": 810
407
  },
408
  {
409
  "epoch": 11.0,
410
  "eval_B": {
411
- "f1-score": 0.8830917874396135,
412
- "precision": 0.8899707887049659,
413
- "recall": 0.8763183125599233,
414
- "support": 1043.0
415
  },
416
  "eval_I": {
417
- "f1-score": 0.9574634882843123,
418
- "precision": 0.9647182727751448,
419
- "recall": 0.9503170028818444,
420
- "support": 17350.0
421
  },
422
  "eval_O": {
423
- "f1-score": 0.9200619426496501,
424
- "precision": 0.9067466582465004,
425
- "recall": 0.9337741166269239,
426
- "support": 9226.0
427
  },
428
- "eval_accuracy": 0.9419964517180202,
429
- "eval_loss": 0.3198166787624359,
430
  "eval_macro avg": {
431
- "f1-score": 0.9202057394578587,
432
- "precision": 0.9204785732422037,
433
- "recall": 0.9201364773562305,
434
- "support": 27619.0
435
- },
436
- "eval_runtime": 4.6345,
437
- "eval_samples_per_second": 17.262,
438
- "eval_steps_per_second": 2.158,
439
  "eval_weighted avg": {
440
- "f1-score": 0.942161111514465,
441
- "precision": 0.9425303680165918,
442
- "recall": 0.9419964517180202,
443
- "support": 27619.0
444
  },
445
  "step": 891
446
  },
447
  {
448
  "epoch": 12.0,
449
  "eval_B": {
450
- "f1-score": 0.871306005719733,
451
- "precision": 0.8663507109004739,
452
- "recall": 0.8763183125599233,
453
- "support": 1043.0
454
  },
455
  "eval_I": {
456
- "f1-score": 0.9576855173984852,
457
- "precision": 0.9464205312922107,
458
- "recall": 0.9692219020172911,
459
- "support": 17350.0
460
  },
461
  "eval_O": {
462
- "f1-score": 0.9162135168127844,
463
- "precision": 0.9386084583901774,
464
- "recall": 0.8948623455451984,
465
- "support": 9226.0
466
  },
467
- "eval_accuracy": 0.9408740359897172,
468
- "eval_loss": 0.35264310240745544,
469
  "eval_macro avg": {
470
- "f1-score": 0.9150683466436677,
471
- "precision": 0.917126566860954,
472
- "recall": 0.9134675200408043,
473
- "support": 27619.0
474
- },
475
- "eval_runtime": 4.6326,
476
- "eval_samples_per_second": 17.269,
477
- "eval_steps_per_second": 2.159,
478
  "eval_weighted avg": {
479
- "f1-score": 0.9405699625961891,
480
- "precision": 0.9407871989028143,
481
- "recall": 0.9408740359897172,
482
- "support": 27619.0
483
  },
484
  "step": 972
485
  },
486
  {
487
  "epoch": 12.35,
488
- "grad_norm": 0.3602512776851654,
489
- "learning_rate": 1.506172839506173e-05,
490
- "loss": 0.021,
491
  "step": 1000
492
  },
493
  {
494
  "epoch": 13.0,
495
  "eval_B": {
496
- "f1-score": 0.877092300334768,
497
- "precision": 0.875,
498
- "recall": 0.8791946308724832,
499
- "support": 1043.0
500
  },
501
  "eval_I": {
502
- "f1-score": 0.9597242793665263,
503
- "precision": 0.9486991778353419,
504
- "recall": 0.9710086455331413,
505
- "support": 17350.0
506
  },
507
  "eval_O": {
508
- "f1-score": 0.9203392649259936,
509
- "precision": 0.9419040054464994,
510
- "recall": 0.8997398655972252,
511
- "support": 9226.0
512
  },
513
- "eval_accuracy": 0.943734385748941,
514
- "eval_loss": 0.3594338297843933,
515
  "eval_macro avg": {
516
- "f1-score": 0.9190519482090961,
517
- "precision": 0.9218677277606138,
518
- "recall": 0.9166477140009498,
519
- "support": 27619.0
520
- },
521
- "eval_runtime": 4.6042,
522
- "eval_samples_per_second": 17.376,
523
- "eval_steps_per_second": 2.172,
524
  "eval_weighted avg": {
525
- "f1-score": 0.943447393984779,
526
- "precision": 0.9436461164304495,
527
- "recall": 0.943734385748941,
528
- "support": 27619.0
529
  },
530
  "step": 1053
531
  },
532
  {
533
  "epoch": 14.0,
534
  "eval_B": {
535
- "f1-score": 0.8754052802223252,
536
- "precision": 0.8467741935483871,
537
- "recall": 0.9060402684563759,
538
- "support": 1043.0
539
  },
540
  "eval_I": {
541
- "f1-score": 0.9574973153224439,
542
- "precision": 0.9392880904856953,
543
- "recall": 0.9764265129682997,
544
- "support": 17350.0
545
  },
546
  "eval_O": {
547
- "f1-score": 0.9130164471825015,
548
- "precision": 0.9539388213062477,
549
- "recall": 0.8754606546715803,
550
- "support": 9226.0
551
  },
552
- "eval_accuracy": 0.9400412759332344,
553
- "eval_loss": 0.3840242028236389,
554
  "eval_macro avg": {
555
- "f1-score": 0.9153063475757568,
556
- "precision": 0.9133337017801101,
557
- "recall": 0.9193091453654186,
558
- "support": 27619.0
559
- },
560
- "eval_runtime": 4.6182,
561
- "eval_samples_per_second": 17.323,
562
- "eval_steps_per_second": 2.165,
563
  "eval_weighted avg": {
564
- "f1-score": 0.9395385738014427,
565
- "precision": 0.9406884180878825,
566
- "recall": 0.9400412759332344,
567
- "support": 27619.0
568
  },
569
  "step": 1134
570
  }
571
  ],
572
  "logging_steps": 500,
573
- "max_steps": 4050,
574
  "num_input_tokens_seen": 0,
575
- "num_train_epochs": 50,
576
  "save_steps": 500,
577
  "total_flos": 2013013138298400.0,
578
  "train_batch_size": 4,
 
1
  {
2
+ "best_metric": 0.2252955436706543,
3
+ "best_model_checkpoint": "longformer-spans/checkpoint-162",
4
  "epoch": 14.0,
5
  "eval_steps": 500,
6
  "global_step": 1134,
 
11
  {
12
  "epoch": 1.0,
13
  "eval_B": {
14
+ "f1-score": 0.816,
15
+ "precision": 0.7461594732991953,
16
+ "recall": 0.9002647837599294,
17
+ "support": 1133.0
18
  },
19
  "eval_I": {
20
+ "f1-score": 0.9321376763813793,
21
+ "precision": 0.9024103768767235,
22
+ "recall": 0.9638902525500463,
23
+ "support": 18333.0
24
  },
25
  "eval_O": {
26
+ "f1-score": 0.8560784528570645,
27
+ "precision": 0.931782945736434,
28
+ "recall": 0.7917511147142278,
29
+ "support": 9868.0
30
  },
31
+ "eval_accuracy": 0.9035249198881844,
32
+ "eval_loss": 0.2619660496711731,
33
  "eval_macro avg": {
34
+ "f1-score": 0.8680720430794812,
35
+ "precision": 0.860117598637451,
36
+ "recall": 0.8853020503414012,
37
+ "support": 29334.0
38
+ },
39
+ "eval_runtime": 5.5032,
40
+ "eval_samples_per_second": 14.537,
41
+ "eval_steps_per_second": 1.817,
42
  "eval_weighted avg": {
43
+ "f1-score": 0.9020655278480035,
44
+ "precision": 0.9062562975065145,
45
+ "recall": 0.9035249198881844,
46
+ "support": 29334.0
47
  },
48
  "step": 81
49
  },
50
  {
51
  "epoch": 2.0,
52
  "eval_B": {
53
+ "f1-score": 0.8512256973795435,
54
+ "precision": 0.8167072181670721,
55
+ "recall": 0.8887908208296558,
56
+ "support": 1133.0
57
  },
58
  "eval_I": {
59
+ "f1-score": 0.9416781438711729,
60
+ "precision": 0.9152551099212274,
61
+ "recall": 0.9696721758577429,
62
+ "support": 18333.0
63
  },
64
  "eval_O": {
65
+ "f1-score": 0.8778173190984578,
66
+ "precision": 0.9380041484212952,
67
+ "recall": 0.8248885285772193,
68
+ "support": 9868.0
69
  },
70
+ "eval_accuracy": 0.9178427763005387,
71
+ "eval_loss": 0.2252955436706543,
72
  "eval_macro avg": {
73
+ "f1-score": 0.8902403867830581,
74
+ "precision": 0.8899888255031981,
75
+ "recall": 0.8944505084215394,
76
+ "support": 29334.0
77
+ },
78
+ "eval_runtime": 5.4972,
79
+ "eval_samples_per_second": 14.553,
80
+ "eval_steps_per_second": 1.819,
81
  "eval_weighted avg": {
82
+ "f1-score": 0.9167016237671239,
83
+ "precision": 0.9191015935430046,
84
+ "recall": 0.9178427763005387,
85
+ "support": 29334.0
86
  },
87
  "step": 162
88
  },
89
  {
90
  "epoch": 3.0,
91
  "eval_B": {
92
+ "f1-score": 0.8531120331950207,
93
+ "precision": 0.8050117462803446,
94
+ "recall": 0.9073256840247131,
95
+ "support": 1133.0
96
  },
97
  "eval_I": {
98
+ "f1-score": 0.9469915571230095,
99
+ "precision": 0.9280963603037444,
100
+ "recall": 0.9666721213112965,
101
+ "support": 18333.0
102
  },
103
  "eval_O": {
104
+ "f1-score": 0.8903876792352629,
105
+ "precision": 0.9353938852934612,
106
+ "recall": 0.8495135792460479,
107
+ "support": 9868.0
108
  },
109
+ "eval_accuracy": 0.9249676143724006,
110
+ "eval_loss": 0.22786012291908264,
111
  "eval_macro avg": {
112
+ "f1-score": 0.8968304231844311,
113
+ "precision": 0.8895006639591835,
114
+ "recall": 0.9078371281940192,
115
+ "support": 29334.0
116
+ },
117
+ "eval_runtime": 5.5036,
118
+ "eval_samples_per_second": 14.536,
119
+ "eval_steps_per_second": 1.817,
120
  "eval_weighted avg": {
121
+ "f1-score": 0.9243239165827936,
122
+ "precision": 0.9257972230878861,
123
+ "recall": 0.9249676143724006,
124
+ "support": 29334.0
125
  },
126
  "step": 243
127
  },
128
  {
129
  "epoch": 4.0,
130
  "eval_B": {
131
+ "f1-score": 0.8567807351077312,
132
+ "precision": 0.8217179902755267,
133
+ "recall": 0.8949691085613416,
134
+ "support": 1133.0
135
  },
136
  "eval_I": {
137
+ "f1-score": 0.9472597903427299,
138
+ "precision": 0.9432635621180161,
139
+ "recall": 0.9512900234549719,
140
+ "support": 18333.0
141
  },
142
  "eval_O": {
143
+ "f1-score": 0.8979927100980543,
144
+ "precision": 0.9099989595255437,
145
+ "recall": 0.8862991487636805,
146
+ "support": 9868.0
147
  },
148
+ "eval_accuracy": 0.9272516533715143,
149
+ "eval_loss": 0.23897655308246613,
150
  "eval_macro avg": {
151
+ "f1-score": 0.9006777451828384,
152
+ "precision": 0.8916601706396955,
153
+ "recall": 0.910852760259998,
154
+ "support": 29334.0
155
+ },
156
+ "eval_runtime": 5.4904,
157
+ "eval_samples_per_second": 14.571,
158
+ "eval_steps_per_second": 1.821,
159
  "eval_weighted avg": {
160
+ "f1-score": 0.9271915992526735,
161
+ "precision": 0.9273787107073643,
162
+ "recall": 0.9272516533715143,
163
+ "support": 29334.0
164
  },
165
  "step": 324
166
  },
167
  {
168
  "epoch": 5.0,
169
  "eval_B": {
170
+ "f1-score": 0.8624407072013798,
171
+ "precision": 0.8431703204047217,
172
+ "recall": 0.8826125330979699,
173
+ "support": 1133.0
174
  },
175
  "eval_I": {
176
+ "f1-score": 0.948190701170407,
177
+ "precision": 0.9335059992600032,
178
+ "recall": 0.9633447880870561,
179
+ "support": 18333.0
180
  },
181
  "eval_O": {
182
+ "f1-score": 0.8955333298423835,
183
+ "precision": 0.9265359193845487,
184
+ "recall": 0.8665383056343737,
185
+ "support": 9868.0
186
  },
187
+ "eval_accuracy": 0.9276607349832958,
188
+ "eval_loss": 0.25390708446502686,
189
  "eval_macro avg": {
190
+ "f1-score": 0.9020549127380568,
191
+ "precision": 0.9010707463497579,
192
+ "recall": 0.9041652089397999,
193
+ "support": 29334.0
194
+ },
195
+ "eval_runtime": 5.4919,
196
+ "eval_samples_per_second": 14.567,
197
+ "eval_steps_per_second": 1.821,
198
  "eval_weighted avg": {
199
+ "f1-score": 0.9271646670996412,
200
+ "precision": 0.9276721180179627,
201
+ "recall": 0.9276607349832958,
202
+ "support": 29334.0
203
  },
204
  "step": 405
205
  },
206
  {
207
  "epoch": 6.0,
208
  "eval_B": {
209
+ "f1-score": 0.8601036269430052,
210
+ "precision": 0.841927303465765,
211
+ "recall": 0.8790820829655781,
212
+ "support": 1133.0
213
  },
214
  "eval_I": {
215
+ "f1-score": 0.9474285714285714,
216
+ "precision": 0.9452679589509693,
217
+ "recall": 0.9495990836197021,
218
+ "support": 18333.0
219
  },
220
  "eval_O": {
221
+ "f1-score": 0.8983777165595348,
222
+ "precision": 0.9045613314156564,
223
+ "recall": 0.8922780705310093,
224
+ "support": 9868.0
225
  },
226
+ "eval_accuracy": 0.9275925547146656,
227
+ "eval_loss": 0.29299432039260864,
228
  "eval_macro avg": {
229
+ "f1-score": 0.9019699716437038,
230
+ "precision": 0.8972521979441302,
231
+ "recall": 0.9069864123720964,
232
+ "support": 29334.0
233
+ },
234
+ "eval_runtime": 5.4971,
235
+ "eval_samples_per_second": 14.553,
236
+ "eval_steps_per_second": 1.819,
237
  "eval_weighted avg": {
238
+ "f1-score": 0.9275549436263691,
239
+ "precision": 0.9275827485063247,
240
+ "recall": 0.9275925547146656,
241
+ "support": 29334.0
242
  },
243
  "step": 486
244
  },
245
  {
246
  "epoch": 6.17,
247
+ "grad_norm": 2.853372573852539,
248
+ "learning_rate": 1.3827160493827162e-05,
249
+ "loss": 0.1621,
250
  "step": 500
251
  },
252
  {
253
  "epoch": 7.0,
254
  "eval_B": {
255
+ "f1-score": 0.8665526090675792,
256
+ "precision": 0.8406639004149378,
257
+ "recall": 0.8940864960282436,
258
+ "support": 1133.0
259
  },
260
  "eval_I": {
261
+ "f1-score": 0.9492722371967655,
262
+ "precision": 0.9382959450098577,
263
+ "recall": 0.9605083728795069,
264
+ "support": 18333.0
265
  },
266
  "eval_O": {
267
+ "f1-score": 0.8984919396775871,
268
+ "precision": 0.9227729117709891,
269
+ "recall": 0.8754560194568302,
270
+ "support": 9868.0
271
  },
272
+ "eval_accuracy": 0.9293311515647371,
273
+ "eval_loss": 0.3148973286151886,
274
  "eval_macro avg": {
275
+ "f1-score": 0.9047722619806439,
276
+ "precision": 0.9005775857319281,
277
+ "recall": 0.9100169627881934,
278
+ "support": 29334.0
279
+ },
280
+ "eval_runtime": 5.4821,
281
+ "eval_samples_per_second": 14.593,
282
+ "eval_steps_per_second": 1.824,
283
  "eval_weighted avg": {
284
+ "f1-score": 0.9289946986889036,
285
+ "precision": 0.9293030221719495,
286
+ "recall": 0.9293311515647371,
287
+ "support": 29334.0
288
  },
289
  "step": 567
290
  },
291
  {
292
  "epoch": 8.0,
293
  "eval_B": {
294
+ "f1-score": 0.8624629707998307,
295
+ "precision": 0.8284552845528456,
296
+ "recall": 0.8993821712268314,
297
+ "support": 1133.0
298
  },
299
  "eval_I": {
300
+ "f1-score": 0.9467755410030451,
301
+ "precision": 0.9356556940449557,
302
+ "recall": 0.9581628756886489,
303
+ "support": 18333.0
304
  },
305
  "eval_O": {
306
+ "f1-score": 0.8934263985831857,
307
+ "precision": 0.9191854233654877,
308
+ "recall": 0.8690717470612079,
309
+ "support": 9868.0
310
  },
311
+ "eval_accuracy": 0.9259221381332242,
312
+ "eval_loss": 0.34766319394111633,
313
  "eval_macro avg": {
314
+ "f1-score": 0.9008883034620205,
315
+ "precision": 0.8944321339877629,
316
+ "recall": 0.9088722646588961,
317
+ "support": 29334.0
318
+ },
319
+ "eval_runtime": 5.5202,
320
+ "eval_samples_per_second": 14.492,
321
+ "eval_steps_per_second": 1.812,
322
  "eval_weighted avg": {
323
+ "f1-score": 0.9255723133682386,
324
+ "precision": 0.9259745494680297,
325
+ "recall": 0.9259221381332242,
326
+ "support": 29334.0
327
  },
328
  "step": 648
329
  },
330
  {
331
  "epoch": 9.0,
332
  "eval_B": {
333
+ "f1-score": 0.8654012079378774,
334
+ "precision": 0.8464135021097047,
335
+ "recall": 0.8852603706972639,
336
+ "support": 1133.0
337
  },
338
  "eval_I": {
339
+ "f1-score": 0.9474813007694164,
340
+ "precision": 0.9316216786166175,
341
+ "recall": 0.9638902525500463,
342
+ "support": 18333.0
343
  },
344
  "eval_O": {
345
+ "f1-score": 0.8933802299333298,
346
+ "precision": 0.9268053588933667,
347
+ "recall": 0.8622821240372922,
348
+ "support": 9868.0
349
  },
350
+ "eval_accuracy": 0.9266721210881571,
351
+ "eval_loss": 0.3807723820209503,
352
  "eval_macro avg": {
353
+ "f1-score": 0.9020875795468745,
354
+ "precision": 0.9016135132065629,
355
+ "recall": 0.9038109157615342,
356
+ "support": 29334.0
357
+ },
358
+ "eval_runtime": 5.532,
359
+ "eval_samples_per_second": 14.461,
360
+ "eval_steps_per_second": 1.808,
361
  "eval_weighted avg": {
362
+ "f1-score": 0.9261113508073029,
363
+ "precision": 0.9267103706800465,
364
+ "recall": 0.9266721210881571,
365
+ "support": 29334.0
366
  },
367
  "step": 729
368
  },
369
  {
370
  "epoch": 10.0,
371
  "eval_B": {
372
+ "f1-score": 0.8593548387096774,
373
+ "precision": 0.8380872483221476,
374
+ "recall": 0.881729920564872,
375
+ "support": 1133.0
376
  },
377
  "eval_I": {
378
+ "f1-score": 0.9447995351539802,
379
+ "precision": 0.9158687080751703,
380
+ "recall": 0.9756177385043364,
381
+ "support": 18333.0
382
  },
383
  "eval_O": {
384
+ "f1-score": 0.8826362209837131,
385
+ "precision": 0.9469406710786021,
386
+ "recall": 0.8265099310903932,
387
+ "support": 9868.0
388
  },
389
+ "eval_accuracy": 0.9218313220154087,
390
+ "eval_loss": 0.46634167432785034,
391
  "eval_macro avg": {
392
+ "f1-score": 0.8955968649491236,
393
+ "precision": 0.9002988758253068,
394
+ "recall": 0.8946191967198672,
395
+ "support": 29334.0
396
+ },
397
+ "eval_runtime": 5.5106,
398
+ "eval_samples_per_second": 14.517,
399
+ "eval_steps_per_second": 1.815,
400
  "eval_weighted avg": {
401
+ "f1-score": 0.9205874800198834,
402
+ "precision": 0.9233171207368492,
403
+ "recall": 0.9218313220154087,
404
+ "support": 29334.0
405
  },
406
  "step": 810
407
  },
408
  {
409
  "epoch": 11.0,
410
  "eval_B": {
411
+ "f1-score": 0.8585028126352229,
412
+ "precision": 0.8421052631578947,
413
+ "recall": 0.8755516328331863,
414
+ "support": 1133.0
415
  },
416
  "eval_I": {
417
+ "f1-score": 0.9467715680954965,
418
+ "precision": 0.941814648890808,
419
+ "recall": 0.9517809414716631,
420
+ "support": 18333.0
421
  },
422
  "eval_O": {
423
+ "f1-score": 0.8955223880597014,
424
+ "precision": 0.906636203136359,
425
+ "recall": 0.8846777462505067,
426
+ "support": 9868.0
427
  },
428
+ "eval_accuracy": 0.9262630394763756,
429
+ "eval_loss": 0.39979249238967896,
430
  "eval_macro avg": {
431
+ "f1-score": 0.9002655895968069,
432
+ "precision": 0.8968520383950206,
433
+ "recall": 0.9040034401851186,
434
+ "support": 29334.0
435
+ },
436
+ "eval_runtime": 5.5098,
437
+ "eval_samples_per_second": 14.52,
438
+ "eval_steps_per_second": 1.815,
439
  "eval_weighted avg": {
440
+ "f1-score": 0.9261219666592889,
441
+ "precision": 0.9261293813943774,
442
+ "recall": 0.9262630394763756,
443
+ "support": 29334.0
444
  },
445
  "step": 891
446
  },
447
  {
448
  "epoch": 12.0,
449
  "eval_B": {
450
+ "f1-score": 0.8661758336942399,
451
+ "precision": 0.8503401360544217,
452
+ "recall": 0.8826125330979699,
453
+ "support": 1133.0
454
  },
455
  "eval_I": {
456
+ "f1-score": 0.9483851819874267,
457
+ "precision": 0.9383342231713828,
458
+ "recall": 0.9586537937053401,
459
+ "support": 18333.0
460
  },
461
  "eval_O": {
462
+ "f1-score": 0.8972844112769486,
463
+ "precision": 0.9182223165040305,
464
+ "recall": 0.8772800972841508,
465
+ "support": 9868.0
466
  },
467
+ "eval_accuracy": 0.9283425376695984,
468
+ "eval_loss": 0.4524174630641937,
469
  "eval_macro avg": {
470
+ "f1-score": 0.9039484756528716,
471
+ "precision": 0.9022988919099451,
472
+ "recall": 0.906182141362487,
473
+ "support": 29334.0
474
+ },
475
+ "eval_runtime": 5.5194,
476
+ "eval_samples_per_second": 14.494,
477
+ "eval_steps_per_second": 1.812,
478
  "eval_weighted avg": {
479
+ "f1-score": 0.9280195449455239,
480
+ "precision": 0.9281698543264606,
481
+ "recall": 0.9283425376695984,
482
+ "support": 29334.0
483
  },
484
  "step": 972
485
  },
486
  {
487
  "epoch": 12.35,
488
+ "grad_norm": 1.163489818572998,
489
+ "learning_rate": 7.654320987654322e-06,
490
+ "loss": 0.0212,
491
  "step": 1000
492
  },
493
  {
494
  "epoch": 13.0,
495
  "eval_B": {
496
+ "f1-score": 0.8624407072013798,
497
+ "precision": 0.8431703204047217,
498
+ "recall": 0.8826125330979699,
499
+ "support": 1133.0
500
  },
501
  "eval_I": {
502
+ "f1-score": 0.94720379658092,
503
+ "precision": 0.9365968111768783,
504
+ "recall": 0.9580537827960508,
505
+ "support": 18333.0
506
  },
507
  "eval_O": {
508
+ "f1-score": 0.8942532315838654,
509
+ "precision": 0.9167642362959021,
510
+ "recall": 0.8728212403729225,
511
+ "support": 9868.0
512
  },
513
+ "eval_accuracy": 0.9264675802822663,
514
+ "eval_loss": 0.45370540022850037,
515
  "eval_macro avg": {
516
+ "f1-score": 0.9012992451220551,
517
+ "precision": 0.8988437892925006,
518
+ "recall": 0.9044958520889811,
519
+ "support": 29334.0
520
+ },
521
+ "eval_runtime": 5.5225,
522
+ "eval_samples_per_second": 14.486,
523
+ "eval_steps_per_second": 1.811,
524
  "eval_weighted avg": {
525
+ "f1-score": 0.9261172500595471,
526
+ "precision": 0.926316588126141,
527
+ "recall": 0.9264675802822663,
528
+ "support": 29334.0
529
  },
530
  "step": 1053
531
  },
532
  {
533
  "epoch": 14.0,
534
  "eval_B": {
535
+ "f1-score": 0.8689595124074879,
536
+ "precision": 0.8573883161512027,
537
+ "recall": 0.880847308031774,
538
+ "support": 1133.0
539
  },
540
  "eval_I": {
541
+ "f1-score": 0.9480582004921365,
542
+ "precision": 0.9300970873786408,
543
+ "recall": 0.9667266677575956,
544
+ "support": 18333.0
545
  },
546
  "eval_O": {
547
+ "f1-score": 0.8934309645472265,
548
+ "precision": 0.9303346132748217,
549
+ "recall": 0.8593433319821646,
550
+ "support": 9868.0
551
  },
552
+ "eval_accuracy": 0.9272857435058294,
553
+ "eval_loss": 0.49018412828445435,
554
  "eval_macro avg": {
555
+ "f1-score": 0.9034828924822836,
556
+ "precision": 0.9059400056015551,
557
+ "recall": 0.9023057692571781,
558
+ "support": 29334.0
559
+ },
560
+ "eval_runtime": 5.5116,
561
+ "eval_samples_per_second": 14.515,
562
+ "eval_steps_per_second": 1.814,
563
  "eval_weighted avg": {
564
+ "f1-score": 0.9266264019680934,
565
+ "precision": 0.9273686789700647,
566
+ "recall": 0.9272857435058294,
567
+ "support": 29334.0
568
  },
569
  "step": 1134
570
  }
571
  ],
572
  "logging_steps": 500,
573
+ "max_steps": 1620,
574
  "num_input_tokens_seen": 0,
575
+ "num_train_epochs": 20,
576
  "save_steps": 500,
577
  "total_flos": 2013013138298400.0,
578
  "train_batch_size": 4,
checkpoint-1134/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:821ca433db9ddb761291ab61fc06d651ad248cd9d560cdb25f7f905e5aaff14e
3
  size 4856
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1094850123635c13193a80398a7bb1e351dd871416d3ce265431b259bf603290
3
  size 4856