DashLuuu commited on
Commit
c0d747c
·
verified ·
1 Parent(s): d5f34ef

Delete trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +0 -2434
trainer_state.json DELETED
@@ -1,2434 +0,0 @@
1
- {
2
- "best_global_step": null,
3
- "best_metric": null,
4
- "best_model_checkpoint": null,
5
- "epoch": 1.1588377723970944,
6
- "eval_steps": 2000,
7
- "global_step": 300,
8
- "is_hyper_param_search": false,
9
- "is_local_process_zero": true,
10
- "is_world_process_zero": true,
11
- "log_history": [
12
- {
13
- "epoch": 0.00387409200968523,
14
- "grad_norm": 1.598986029624939,
15
- "learning_rate": 2.0000000000000002e-07,
16
- "loss": 0.40103477239608765,
17
- "step": 1,
18
- "token_acc": 0.8705013179702646
19
- },
20
- {
21
- "epoch": 0.00774818401937046,
22
- "grad_norm": 1.988427758216858,
23
- "learning_rate": 4.0000000000000003e-07,
24
- "loss": 0.4314175248146057,
25
- "step": 2,
26
- "token_acc": 0.8610088406262493
27
- },
28
- {
29
- "epoch": 0.01162227602905569,
30
- "grad_norm": 1.6525965929031372,
31
- "learning_rate": 6.000000000000001e-07,
32
- "loss": 0.41751521825790405,
33
- "step": 3,
34
- "token_acc": 0.8659394954574845
35
- },
36
- {
37
- "epoch": 0.01549636803874092,
38
- "grad_norm": 1.3594496250152588,
39
- "learning_rate": 8.000000000000001e-07,
40
- "loss": 0.39516761898994446,
41
- "step": 4,
42
- "token_acc": 0.8712739341656057
43
- },
44
- {
45
- "epoch": 0.01937046004842615,
46
- "grad_norm": 1.4459697008132935,
47
- "learning_rate": 1.0000000000000002e-06,
48
- "loss": 0.41443824768066406,
49
- "step": 5,
50
- "token_acc": 0.8673064711013153
51
- },
52
- {
53
- "epoch": 0.02324455205811138,
54
- "grad_norm": 1.165871024131775,
55
- "learning_rate": 1.2000000000000002e-06,
56
- "loss": 0.3951181471347809,
57
- "step": 6,
58
- "token_acc": 0.8717731277799119
59
- },
60
- {
61
- "epoch": 0.02711864406779661,
62
- "grad_norm": 1.150416374206543,
63
- "learning_rate": 1.4000000000000001e-06,
64
- "loss": 0.40562719106674194,
65
- "step": 7,
66
- "token_acc": 0.8683618627898853
67
- },
68
- {
69
- "epoch": 0.03099273607748184,
70
- "grad_norm": 0.7621377110481262,
71
- "learning_rate": 1.6000000000000001e-06,
72
- "loss": 0.4054454565048218,
73
- "step": 8,
74
- "token_acc": 0.8672108063124587
75
- },
76
- {
77
- "epoch": 0.03486682808716707,
78
- "grad_norm": 0.588590681552887,
79
- "learning_rate": 1.8000000000000001e-06,
80
- "loss": 0.383542001247406,
81
- "step": 9,
82
- "token_acc": 0.8732824386699718
83
- },
84
- {
85
- "epoch": 0.0387409200968523,
86
- "grad_norm": 0.5067570805549622,
87
- "learning_rate": 2.0000000000000003e-06,
88
- "loss": 0.3769374489784241,
89
- "step": 10,
90
- "token_acc": 0.8749500487669789
91
- },
92
- {
93
- "epoch": 0.04261501210653753,
94
- "grad_norm": 0.6109248995780945,
95
- "learning_rate": 2.2e-06,
96
- "loss": 0.3687226176261902,
97
- "step": 11,
98
- "token_acc": 0.8778429629931872
99
- },
100
- {
101
- "epoch": 0.04648910411622276,
102
- "grad_norm": 0.6168301701545715,
103
- "learning_rate": 2.4000000000000003e-06,
104
- "loss": 0.3631238639354706,
105
- "step": 12,
106
- "token_acc": 0.8792909317747671
107
- },
108
- {
109
- "epoch": 0.05036319612590799,
110
- "grad_norm": 0.5205990076065063,
111
- "learning_rate": 2.6e-06,
112
- "loss": 0.37530872225761414,
113
- "step": 13,
114
- "token_acc": 0.8747995859550826
115
- },
116
- {
117
- "epoch": 0.05423728813559322,
118
- "grad_norm": 0.4970836639404297,
119
- "learning_rate": 2.8000000000000003e-06,
120
- "loss": 0.33857205510139465,
121
- "step": 14,
122
- "token_acc": 0.8863650931395268
123
- },
124
- {
125
- "epoch": 0.05811138014527845,
126
- "grad_norm": 0.4103075861930847,
127
- "learning_rate": 3e-06,
128
- "loss": 0.38399845361709595,
129
- "step": 15,
130
- "token_acc": 0.8722473100295478
131
- },
132
- {
133
- "epoch": 0.06198547215496368,
134
- "grad_norm": 0.505113959312439,
135
- "learning_rate": 3.2000000000000003e-06,
136
- "loss": 0.37927311658859253,
137
- "step": 16,
138
- "token_acc": 0.8732506907722828
139
- },
140
- {
141
- "epoch": 0.06585956416464891,
142
- "grad_norm": 0.4578634202480316,
143
- "learning_rate": 3.4000000000000005e-06,
144
- "loss": 0.388744592666626,
145
- "step": 17,
146
- "token_acc": 0.8707925977418891
147
- },
148
- {
149
- "epoch": 0.06973365617433414,
150
- "grad_norm": 0.40881460905075073,
151
- "learning_rate": 3.6000000000000003e-06,
152
- "loss": 0.37862884998321533,
153
- "step": 18,
154
- "token_acc": 0.8738148420049672
155
- },
156
- {
157
- "epoch": 0.07360774818401937,
158
- "grad_norm": 0.3267415165901184,
159
- "learning_rate": 3.8000000000000005e-06,
160
- "loss": 0.3523765206336975,
161
- "step": 19,
162
- "token_acc": 0.8821479488850912
163
- },
164
- {
165
- "epoch": 0.0774818401937046,
166
- "grad_norm": 0.3520510196685791,
167
- "learning_rate": 4.000000000000001e-06,
168
- "loss": 0.37575048208236694,
169
- "step": 20,
170
- "token_acc": 0.8746806805808569
171
- },
172
- {
173
- "epoch": 0.08135593220338982,
174
- "grad_norm": 0.3177695870399475,
175
- "learning_rate": 4.2000000000000004e-06,
176
- "loss": 0.3877210021018982,
177
- "step": 21,
178
- "token_acc": 0.8709381583839385
179
- },
180
- {
181
- "epoch": 0.08523002421307506,
182
- "grad_norm": 0.3101595640182495,
183
- "learning_rate": 4.4e-06,
184
- "loss": 0.35647860169410706,
185
- "step": 22,
186
- "token_acc": 0.8802609194999448
187
- },
188
- {
189
- "epoch": 0.0891041162227603,
190
- "grad_norm": 0.42295873165130615,
191
- "learning_rate": 4.600000000000001e-06,
192
- "loss": 0.34535130858421326,
193
- "step": 23,
194
- "token_acc": 0.8842312960154491
195
- },
196
- {
197
- "epoch": 0.09297820823244551,
198
- "grad_norm": 0.38459983468055725,
199
- "learning_rate": 4.800000000000001e-06,
200
- "loss": 0.3480440676212311,
201
- "step": 24,
202
- "token_acc": 0.8830844934941354
203
- },
204
- {
205
- "epoch": 0.09685230024213075,
206
- "grad_norm": 0.3167020082473755,
207
- "learning_rate": 5e-06,
208
- "loss": 0.3617573082447052,
209
- "step": 25,
210
- "token_acc": 0.8794729562611736
211
- },
212
- {
213
- "epoch": 0.10072639225181598,
214
- "grad_norm": 0.3235217332839966,
215
- "learning_rate": 5.2e-06,
216
- "loss": 0.34485846757888794,
217
- "step": 26,
218
- "token_acc": 0.8849654381719892
219
- },
220
- {
221
- "epoch": 0.10460048426150122,
222
- "grad_norm": 0.33688801527023315,
223
- "learning_rate": 5.400000000000001e-06,
224
- "loss": 0.325369268655777,
225
- "step": 27,
226
- "token_acc": 0.8904570911619978
227
- },
228
- {
229
- "epoch": 0.10847457627118644,
230
- "grad_norm": 0.28384602069854736,
231
- "learning_rate": 5.600000000000001e-06,
232
- "loss": 0.3820268213748932,
233
- "step": 28,
234
- "token_acc": 0.8722670041260794
235
- },
236
- {
237
- "epoch": 0.11234866828087167,
238
- "grad_norm": 0.2726050019264221,
239
- "learning_rate": 5.8e-06,
240
- "loss": 0.34821516275405884,
241
- "step": 29,
242
- "token_acc": 0.8829695430808375
243
- },
244
- {
245
- "epoch": 0.1162227602905569,
246
- "grad_norm": 0.2613418698310852,
247
- "learning_rate": 6e-06,
248
- "loss": 0.3505156636238098,
249
- "step": 30,
250
- "token_acc": 0.8820213661332177
251
- },
252
- {
253
- "epoch": 0.12009685230024213,
254
- "grad_norm": 0.27066054940223694,
255
- "learning_rate": 6.200000000000001e-06,
256
- "loss": 0.3500295877456665,
257
- "step": 31,
258
- "token_acc": 0.8819775128328553
259
- },
260
- {
261
- "epoch": 0.12397094430992736,
262
- "grad_norm": 0.2605418562889099,
263
- "learning_rate": 6.4000000000000006e-06,
264
- "loss": 0.32833147048950195,
265
- "step": 32,
266
- "token_acc": 0.8892601629599358
267
- },
268
- {
269
- "epoch": 0.12784503631961258,
270
- "grad_norm": 0.2576088607311249,
271
- "learning_rate": 6.600000000000001e-06,
272
- "loss": 0.3447936475276947,
273
- "step": 33,
274
- "token_acc": 0.8835445537223737
275
- },
276
- {
277
- "epoch": 0.13171912832929783,
278
- "grad_norm": 0.2707255482673645,
279
- "learning_rate": 6.800000000000001e-06,
280
- "loss": 0.352622389793396,
281
- "step": 34,
282
- "token_acc": 0.8808578896779464
283
- },
284
- {
285
- "epoch": 0.13559322033898305,
286
- "grad_norm": 0.23704984784126282,
287
- "learning_rate": 7e-06,
288
- "loss": 0.34251606464385986,
289
- "step": 35,
290
- "token_acc": 0.8839590527934595
291
- },
292
- {
293
- "epoch": 0.13946731234866827,
294
- "grad_norm": 0.2552218735218048,
295
- "learning_rate": 7.2000000000000005e-06,
296
- "loss": 0.36937713623046875,
297
- "step": 36,
298
- "token_acc": 0.8746555562093041
299
- },
300
- {
301
- "epoch": 0.14334140435835352,
302
- "grad_norm": 0.25926339626312256,
303
- "learning_rate": 7.4e-06,
304
- "loss": 0.37243181467056274,
305
- "step": 37,
306
- "token_acc": 0.8742657147624016
307
- },
308
- {
309
- "epoch": 0.14721549636803874,
310
- "grad_norm": 0.25272250175476074,
311
- "learning_rate": 7.600000000000001e-06,
312
- "loss": 0.3371140956878662,
313
- "step": 38,
314
- "token_acc": 0.8851879286597788
315
- },
316
- {
317
- "epoch": 0.15108958837772396,
318
- "grad_norm": 0.2262120097875595,
319
- "learning_rate": 7.800000000000002e-06,
320
- "loss": 0.32758837938308716,
321
- "step": 39,
322
- "token_acc": 0.8883092864316684
323
- },
324
- {
325
- "epoch": 0.1549636803874092,
326
- "grad_norm": 0.26067835092544556,
327
- "learning_rate": 8.000000000000001e-06,
328
- "loss": 0.32051679491996765,
329
- "step": 40,
330
- "token_acc": 0.8908219532219895
331
- },
332
- {
333
- "epoch": 0.15883777239709443,
334
- "grad_norm": 0.22696885466575623,
335
- "learning_rate": 8.2e-06,
336
- "loss": 0.34018558263778687,
337
- "step": 41,
338
- "token_acc": 0.8843387459744694
339
- },
340
- {
341
- "epoch": 0.16271186440677965,
342
- "grad_norm": 0.2458319216966629,
343
- "learning_rate": 8.400000000000001e-06,
344
- "loss": 0.3157382607460022,
345
- "step": 42,
346
- "token_acc": 0.8923707458363505
347
- },
348
- {
349
- "epoch": 0.1665859564164649,
350
- "grad_norm": 0.23234310746192932,
351
- "learning_rate": 8.6e-06,
352
- "loss": 0.32486584782600403,
353
- "step": 43,
354
- "token_acc": 0.8898832391328527
355
- },
356
- {
357
- "epoch": 0.17046004842615012,
358
- "grad_norm": 0.24149972200393677,
359
- "learning_rate": 8.8e-06,
360
- "loss": 0.3565906286239624,
361
- "step": 44,
362
- "token_acc": 0.8786636478836652
363
- },
364
- {
365
- "epoch": 0.17433414043583534,
366
- "grad_norm": 0.23454472422599792,
367
- "learning_rate": 9e-06,
368
- "loss": 0.34243613481521606,
369
- "step": 45,
370
- "token_acc": 0.8836981353220466
371
- },
372
- {
373
- "epoch": 0.1782082324455206,
374
- "grad_norm": 0.22611235082149506,
375
- "learning_rate": 9.200000000000002e-06,
376
- "loss": 0.3169807493686676,
377
- "step": 46,
378
- "token_acc": 0.8913642111117898
379
- },
380
- {
381
- "epoch": 0.1820823244552058,
382
- "grad_norm": 0.2332201898097992,
383
- "learning_rate": 9.4e-06,
384
- "loss": 0.3335682153701782,
385
- "step": 47,
386
- "token_acc": 0.8853106607331619
387
- },
388
- {
389
- "epoch": 0.18595641646489103,
390
- "grad_norm": 0.26498886942863464,
391
- "learning_rate": 9.600000000000001e-06,
392
- "loss": 0.3396722674369812,
393
- "step": 48,
394
- "token_acc": 0.884526213547764
395
- },
396
- {
397
- "epoch": 0.18983050847457628,
398
- "grad_norm": 0.29751622676849365,
399
- "learning_rate": 9.800000000000001e-06,
400
- "loss": 0.3544740676879883,
401
- "step": 49,
402
- "token_acc": 0.8790830507178009
403
- },
404
- {
405
- "epoch": 0.1937046004842615,
406
- "grad_norm": 0.24125243723392487,
407
- "learning_rate": 1e-05,
408
- "loss": 0.3444434702396393,
409
- "step": 50,
410
- "token_acc": 0.8833375152943368
411
- },
412
- {
413
- "epoch": 0.19757869249394674,
414
- "grad_norm": 0.23450158536434174,
415
- "learning_rate": 9.999953315763929e-06,
416
- "loss": 0.34759777784347534,
417
- "step": 51,
418
- "token_acc": 0.8809645656414854
419
- },
420
- {
421
- "epoch": 0.20145278450363197,
422
- "grad_norm": 0.24415536224842072,
423
- "learning_rate": 9.999813263927483e-06,
424
- "loss": 0.3302762508392334,
425
- "step": 52,
426
- "token_acc": 0.8872595593874244
427
- },
428
- {
429
- "epoch": 0.20532687651331719,
430
- "grad_norm": 0.23792694509029388,
431
- "learning_rate": 9.999579847105947e-06,
432
- "loss": 0.3057291805744171,
433
- "step": 53,
434
- "token_acc": 0.8958096559669801
435
- },
436
- {
437
- "epoch": 0.20920096852300243,
438
- "grad_norm": 0.24918483197689056,
439
- "learning_rate": 9.999253069658074e-06,
440
- "loss": 0.3550814390182495,
441
- "step": 54,
442
- "token_acc": 0.8789014457104403
443
- },
444
- {
445
- "epoch": 0.21307506053268765,
446
- "grad_norm": 0.24681781232357025,
447
- "learning_rate": 9.99883293768601e-06,
448
- "loss": 0.329832524061203,
449
- "step": 55,
450
- "token_acc": 0.8862846605616155
451
- },
452
- {
453
- "epoch": 0.21694915254237288,
454
- "grad_norm": 0.25197944045066833,
455
- "learning_rate": 9.998319459035168e-06,
456
- "loss": 0.3133784532546997,
457
- "step": 56,
458
- "token_acc": 0.8929686000759445
459
- },
460
- {
461
- "epoch": 0.22082324455205812,
462
- "grad_norm": 0.29595333337783813,
463
- "learning_rate": 9.997712643294093e-06,
464
- "loss": 0.3238765597343445,
465
- "step": 57,
466
- "token_acc": 0.8900121095092376
467
- },
468
- {
469
- "epoch": 0.22469733656174334,
470
- "grad_norm": 0.2436024248600006,
471
- "learning_rate": 9.997012501794273e-06,
472
- "loss": 0.3283236622810364,
473
- "step": 58,
474
- "token_acc": 0.887762605178964
475
- },
476
- {
477
- "epoch": 0.22857142857142856,
478
- "grad_norm": 0.23041026294231415,
479
- "learning_rate": 9.996219047609943e-06,
480
- "loss": 0.3104722797870636,
481
- "step": 59,
482
- "token_acc": 0.8931121325749851
483
- },
484
- {
485
- "epoch": 0.2324455205811138,
486
- "grad_norm": 0.237432062625885,
487
- "learning_rate": 9.995332295557818e-06,
488
- "loss": 0.30940210819244385,
489
- "step": 60,
490
- "token_acc": 0.8942070394423697
491
- },
492
- {
493
- "epoch": 0.23631961259079903,
494
- "grad_norm": 0.23901380598545074,
495
- "learning_rate": 9.994352262196839e-06,
496
- "loss": 0.32523292303085327,
497
- "step": 61,
498
- "token_acc": 0.8885503611348168
499
- },
500
- {
501
- "epoch": 0.24019370460048425,
502
- "grad_norm": 0.27438339591026306,
503
- "learning_rate": 9.993278965827844e-06,
504
- "loss": 0.3501031994819641,
505
- "step": 62,
506
- "token_acc": 0.8796217252529412
507
- },
508
- {
509
- "epoch": 0.2440677966101695,
510
- "grad_norm": 0.23662753403186798,
511
- "learning_rate": 9.992112426493247e-06,
512
- "loss": 0.32605987787246704,
513
- "step": 63,
514
- "token_acc": 0.8890396653634925
515
- },
516
- {
517
- "epoch": 0.24794188861985472,
518
- "grad_norm": 0.2232031375169754,
519
- "learning_rate": 9.990852665976648e-06,
520
- "loss": 0.3196948170661926,
521
- "step": 64,
522
- "token_acc": 0.8907072739748918
523
- },
524
- {
525
- "epoch": 0.25181598062953997,
526
- "grad_norm": 0.2665523886680603,
527
- "learning_rate": 9.989499707802424e-06,
528
- "loss": 0.33278700709342957,
529
- "step": 65,
530
- "token_acc": 0.8863953116150797
531
- },
532
- {
533
- "epoch": 0.25569007263922516,
534
- "grad_norm": 0.23870785534381866,
535
- "learning_rate": 9.988053577235306e-06,
536
- "loss": 0.351688951253891,
537
- "step": 66,
538
- "token_acc": 0.879823584223047
539
- },
540
- {
541
- "epoch": 0.2595641646489104,
542
- "grad_norm": 0.24755656719207764,
543
- "learning_rate": 9.986514301279894e-06,
544
- "loss": 0.31553030014038086,
545
- "step": 67,
546
- "token_acc": 0.8921622627267041
547
- },
548
- {
549
- "epoch": 0.26343825665859566,
550
- "grad_norm": 0.23198164999485016,
551
- "learning_rate": 9.984881908680157e-06,
552
- "loss": 0.3355843424797058,
553
- "step": 68,
554
- "token_acc": 0.8848336232927391
555
- },
556
- {
557
- "epoch": 0.26731234866828085,
558
- "grad_norm": 0.2461438924074173,
559
- "learning_rate": 9.983156429918895e-06,
560
- "loss": 0.3341342508792877,
561
- "step": 69,
562
- "token_acc": 0.8856444439357114
563
- },
564
- {
565
- "epoch": 0.2711864406779661,
566
- "grad_norm": 0.22579748928546906,
567
- "learning_rate": 9.981337897217171e-06,
568
- "loss": 0.3188900947570801,
569
- "step": 70,
570
- "token_acc": 0.8906781387812132
571
- },
572
- {
573
- "epoch": 0.27506053268765135,
574
- "grad_norm": 0.24103567004203796,
575
- "learning_rate": 9.979426344533712e-06,
576
- "loss": 0.3240354061126709,
577
- "step": 71,
578
- "token_acc": 0.8889305949367731
579
- },
580
- {
581
- "epoch": 0.27893462469733654,
582
- "grad_norm": 0.23146985471248627,
583
- "learning_rate": 9.977421807564264e-06,
584
- "loss": 0.3256258964538574,
585
- "step": 72,
586
- "token_acc": 0.8886470476040884
587
- },
588
- {
589
- "epoch": 0.2828087167070218,
590
- "grad_norm": 0.2992205023765564,
591
- "learning_rate": 9.97532432374094e-06,
592
- "loss": 0.3162704110145569,
593
- "step": 73,
594
- "token_acc": 0.8912486582241803
595
- },
596
- {
597
- "epoch": 0.28668280871670704,
598
- "grad_norm": 0.2314625382423401,
599
- "learning_rate": 9.973133932231514e-06,
600
- "loss": 0.33748123049736023,
601
- "step": 74,
602
- "token_acc": 0.8834313251000246
603
- },
604
- {
605
- "epoch": 0.29055690072639223,
606
- "grad_norm": 0.23197512328624725,
607
- "learning_rate": 9.970850673938684e-06,
608
- "loss": 0.3105667233467102,
609
- "step": 75,
610
- "token_acc": 0.8935043208256486
611
- },
612
- {
613
- "epoch": 0.2944309927360775,
614
- "grad_norm": 0.2275022268295288,
615
- "learning_rate": 9.96847459149932e-06,
616
- "loss": 0.3327932357788086,
617
- "step": 76,
618
- "token_acc": 0.8861917159302386
619
- },
620
- {
621
- "epoch": 0.2983050847457627,
622
- "grad_norm": 0.2508430778980255,
623
- "learning_rate": 9.966005729283658e-06,
624
- "loss": 0.32548677921295166,
625
- "step": 77,
626
- "token_acc": 0.8882381273480396
627
- },
628
- {
629
- "epoch": 0.3021791767554479,
630
- "grad_norm": 0.5134550333023071,
631
- "learning_rate": 9.963444133394478e-06,
632
- "loss": 0.3120523691177368,
633
- "step": 78,
634
- "token_acc": 0.8919503736696569
635
- },
636
- {
637
- "epoch": 0.30605326876513317,
638
- "grad_norm": 0.21315379440784454,
639
- "learning_rate": 9.960789851666237e-06,
640
- "loss": 0.3215460181236267,
641
- "step": 79,
642
- "token_acc": 0.8896002985397907
643
- },
644
- {
645
- "epoch": 0.3099273607748184,
646
- "grad_norm": 0.23902781307697296,
647
- "learning_rate": 9.958042933664186e-06,
648
- "loss": 0.33162713050842285,
649
- "step": 80,
650
- "token_acc": 0.8866171518838251
651
- },
652
- {
653
- "epoch": 0.3138014527845036,
654
- "grad_norm": 0.24128590524196625,
655
- "learning_rate": 9.955203430683425e-06,
656
- "loss": 0.3268725574016571,
657
- "step": 81,
658
- "token_acc": 0.8882163748841388
659
- },
660
- {
661
- "epoch": 0.31767554479418886,
662
- "grad_norm": 0.24751782417297363,
663
- "learning_rate": 9.952271395747969e-06,
664
- "loss": 0.3100839853286743,
665
- "step": 82,
666
- "token_acc": 0.893085253361785
667
- },
668
- {
669
- "epoch": 0.3215496368038741,
670
- "grad_norm": 0.23644764721393585,
671
- "learning_rate": 9.949246883609743e-06,
672
- "loss": 0.32995104789733887,
673
- "step": 83,
674
- "token_acc": 0.8866222032237766
675
- },
676
- {
677
- "epoch": 0.3254237288135593,
678
- "grad_norm": 0.232451930642128,
679
- "learning_rate": 9.94612995074756e-06,
680
- "loss": 0.31272488832473755,
681
- "step": 84,
682
- "token_acc": 0.8926265473810503
683
- },
684
- {
685
- "epoch": 0.32929782082324455,
686
- "grad_norm": 0.21610639989376068,
687
- "learning_rate": 9.942920655366075e-06,
688
- "loss": 0.302722692489624,
689
- "step": 85,
690
- "token_acc": 0.8952372082627079
691
- },
692
- {
693
- "epoch": 0.3331719128329298,
694
- "grad_norm": 0.24474947154521942,
695
- "learning_rate": 9.939619057394687e-06,
696
- "loss": 0.31238657236099243,
697
- "step": 86,
698
- "token_acc": 0.8932181956136864
699
- },
700
- {
701
- "epoch": 0.337046004842615,
702
- "grad_norm": 0.22313052415847778,
703
- "learning_rate": 9.936225218486428e-06,
704
- "loss": 0.30595749616622925,
705
- "step": 87,
706
- "token_acc": 0.8942476419229949
707
- },
708
- {
709
- "epoch": 0.34092009685230024,
710
- "grad_norm": 0.25018593668937683,
711
- "learning_rate": 9.93273920201681e-06,
712
- "loss": 0.34218600392341614,
713
- "step": 88,
714
- "token_acc": 0.8826220754003523
715
- },
716
- {
717
- "epoch": 0.3447941888619855,
718
- "grad_norm": 0.21603761613368988,
719
- "learning_rate": 9.929161073082636e-06,
720
- "loss": 0.26845768094062805,
721
- "step": 89,
722
- "token_acc": 0.9068716054841073
723
- },
724
- {
725
- "epoch": 0.3486682808716707,
726
- "grad_norm": 0.22996748983860016,
727
- "learning_rate": 9.925490898500796e-06,
728
- "loss": 0.32508569955825806,
729
- "step": 90,
730
- "token_acc": 0.8884358725254423
731
- },
732
- {
733
- "epoch": 0.3525423728813559,
734
- "grad_norm": 0.3635949194431305,
735
- "learning_rate": 9.921728746807008e-06,
736
- "loss": 0.34217730164527893,
737
- "step": 91,
738
- "token_acc": 0.8833008019688547
739
- },
740
- {
741
- "epoch": 0.3564164648910412,
742
- "grad_norm": 0.22128325700759888,
743
- "learning_rate": 9.917874688254542e-06,
744
- "loss": 0.32345396280288696,
745
- "step": 92,
746
- "token_acc": 0.8889643834760571
747
- },
748
- {
749
- "epoch": 0.36029055690072637,
750
- "grad_norm": 0.24601417779922485,
751
- "learning_rate": 9.913928794812909e-06,
752
- "loss": 0.3252776265144348,
753
- "step": 93,
754
- "token_acc": 0.8881070006006884
755
- },
756
- {
757
- "epoch": 0.3641646489104116,
758
- "grad_norm": 0.23473182320594788,
759
- "learning_rate": 9.90989114016652e-06,
760
- "loss": 0.33626118302345276,
761
- "step": 94,
762
- "token_acc": 0.8841867411739727
763
- },
764
- {
765
- "epoch": 0.36803874092009686,
766
- "grad_norm": 0.22333025932312012,
767
- "learning_rate": 9.905761799713302e-06,
768
- "loss": 0.34545931220054626,
769
- "step": 95,
770
- "token_acc": 0.8803537032594166
771
- },
772
- {
773
- "epoch": 0.37191283292978206,
774
- "grad_norm": 0.21172457933425903,
775
- "learning_rate": 9.901540850563295e-06,
776
- "loss": 0.3074107766151428,
777
- "step": 96,
778
- "token_acc": 0.8944196156632918
779
- },
780
- {
781
- "epoch": 0.3757869249394673,
782
- "grad_norm": 0.2134028971195221,
783
- "learning_rate": 9.89722837153722e-06,
784
- "loss": 0.2957490086555481,
785
- "step": 97,
786
- "token_acc": 0.8978778618134635
787
- },
788
- {
789
- "epoch": 0.37966101694915255,
790
- "grad_norm": 0.2610202729701996,
791
- "learning_rate": 9.892824443164987e-06,
792
- "loss": 0.3412560224533081,
793
- "step": 98,
794
- "token_acc": 0.8829380073969748
795
- },
796
- {
797
- "epoch": 0.38353510895883774,
798
- "grad_norm": 0.25488367676734924,
799
- "learning_rate": 9.88832914768421e-06,
800
- "loss": 0.3430347442626953,
801
- "step": 99,
802
- "token_acc": 0.8815459290145207
803
- },
804
- {
805
- "epoch": 0.387409200968523,
806
- "grad_norm": 0.22882606089115143,
807
- "learning_rate": 9.883742569038663e-06,
808
- "loss": 0.33350762724876404,
809
- "step": 100,
810
- "token_acc": 0.8861422500817198
811
- },
812
- {
813
- "epoch": 0.39128329297820824,
814
- "grad_norm": 0.304647833108902,
815
- "learning_rate": 9.879064792876717e-06,
816
- "loss": 0.31420135498046875,
817
- "step": 101,
818
- "token_acc": 0.8915588172822687
819
- },
820
- {
821
- "epoch": 0.3951573849878935,
822
- "grad_norm": 0.22871072590351105,
823
- "learning_rate": 9.874295906549728e-06,
824
- "loss": 0.3116581439971924,
825
- "step": 102,
826
- "token_acc": 0.8917020548921253
827
- },
828
- {
829
- "epoch": 0.3990314769975787,
830
- "grad_norm": 0.2979466915130615,
831
- "learning_rate": 9.869435999110428e-06,
832
- "loss": 0.3145788013935089,
833
- "step": 103,
834
- "token_acc": 0.8916011830301528
835
- },
836
- {
837
- "epoch": 0.40290556900726393,
838
- "grad_norm": 0.20779502391815186,
839
- "learning_rate": 9.864485161311242e-06,
840
- "loss": 0.3070036768913269,
841
- "step": 104,
842
- "token_acc": 0.8938107647266995
843
- },
844
- {
845
- "epoch": 0.4067796610169492,
846
- "grad_norm": 0.2354535311460495,
847
- "learning_rate": 9.859443485602603e-06,
848
- "loss": 0.32298558950424194,
849
- "step": 105,
850
- "token_acc": 0.8882189451059107
851
- },
852
- {
853
- "epoch": 0.41065375302663437,
854
- "grad_norm": 0.22240500152111053,
855
- "learning_rate": 9.85431106613122e-06,
856
- "loss": 0.3104989528656006,
857
- "step": 106,
858
- "token_acc": 0.8923007628162216
859
- },
860
- {
861
- "epoch": 0.4145278450363196,
862
- "grad_norm": 0.21981710195541382,
863
- "learning_rate": 9.849087998738328e-06,
864
- "loss": 0.3237101435661316,
865
- "step": 107,
866
- "token_acc": 0.8879955719309623
867
- },
868
- {
869
- "epoch": 0.41840193704600487,
870
- "grad_norm": 0.2649724781513214,
871
- "learning_rate": 9.84377438095789e-06,
872
- "loss": 0.323306679725647,
873
- "step": 108,
874
- "token_acc": 0.8889382382835521
875
- },
876
- {
877
- "epoch": 0.42227602905569006,
878
- "grad_norm": 0.2193301022052765,
879
- "learning_rate": 9.838370312014783e-06,
880
- "loss": 0.31488102674484253,
881
- "step": 109,
882
- "token_acc": 0.8910646836196473
883
- },
884
- {
885
- "epoch": 0.4261501210653753,
886
- "grad_norm": 0.21842491626739502,
887
- "learning_rate": 9.832875892822937e-06,
888
- "loss": 0.3206183910369873,
889
- "step": 110,
890
- "token_acc": 0.8890832728771944
891
- },
892
- {
893
- "epoch": 0.43002421307506056,
894
- "grad_norm": 0.2456243336200714,
895
- "learning_rate": 9.827291225983458e-06,
896
- "loss": 0.3201240301132202,
897
- "step": 111,
898
- "token_acc": 0.8904148288428204
899
- },
900
- {
901
- "epoch": 0.43389830508474575,
902
- "grad_norm": 0.21340763568878174,
903
- "learning_rate": 9.821616415782708e-06,
904
- "loss": 0.29961007833480835,
905
- "step": 112,
906
- "token_acc": 0.8965660205577574
907
- },
908
- {
909
- "epoch": 0.437772397094431,
910
- "grad_norm": 0.2308902144432068,
911
- "learning_rate": 9.815851568190358e-06,
912
- "loss": 0.3107410669326782,
913
- "step": 113,
914
- "token_acc": 0.8927536025516888
915
- },
916
- {
917
- "epoch": 0.44164648910411625,
918
- "grad_norm": 0.2292374223470688,
919
- "learning_rate": 9.80999679085741e-06,
920
- "loss": 0.3277205228805542,
921
- "step": 114,
922
- "token_acc": 0.886787084498464
923
- },
924
- {
925
- "epoch": 0.44552058111380144,
926
- "grad_norm": 0.21509671211242676,
927
- "learning_rate": 9.80405219311419e-06,
928
- "loss": 0.3161908984184265,
929
- "step": 115,
930
- "token_acc": 0.8916077261448497
931
- },
932
- {
933
- "epoch": 0.4493946731234867,
934
- "grad_norm": 0.20529279112815857,
935
- "learning_rate": 9.798017885968295e-06,
936
- "loss": 0.29131007194519043,
937
- "step": 116,
938
- "token_acc": 0.8990066361086406
939
- },
940
- {
941
- "epoch": 0.45326876513317194,
942
- "grad_norm": 0.24888373911380768,
943
- "learning_rate": 9.791893982102537e-06,
944
- "loss": 0.31967025995254517,
945
- "step": 117,
946
- "token_acc": 0.8899925908756566
947
- },
948
- {
949
- "epoch": 0.45714285714285713,
950
- "grad_norm": 0.22014780342578888,
951
- "learning_rate": 9.785680595872824e-06,
952
- "loss": 0.31103435158729553,
953
- "step": 118,
954
- "token_acc": 0.8928936680571538
955
- },
956
- {
957
- "epoch": 0.4610169491525424,
958
- "grad_norm": 0.21783359348773956,
959
- "learning_rate": 9.77937784330603e-06,
960
- "loss": 0.307749480009079,
961
- "step": 119,
962
- "token_acc": 0.8931600584652736
963
- },
964
- {
965
- "epoch": 0.4648910411622276,
966
- "grad_norm": 0.2104286551475525,
967
- "learning_rate": 9.772985842097832e-06,
968
- "loss": 0.31199365854263306,
969
- "step": 120,
970
- "token_acc": 0.8926850259294361
971
- },
972
- {
973
- "epoch": 0.4687651331719128,
974
- "grad_norm": 0.21124128997325897,
975
- "learning_rate": 9.766504711610507e-06,
976
- "loss": 0.3170176148414612,
977
- "step": 121,
978
- "token_acc": 0.8906264477918435
979
- },
980
- {
981
- "epoch": 0.47263922518159807,
982
- "grad_norm": 0.23777632415294647,
983
- "learning_rate": 9.759934572870706e-06,
984
- "loss": 0.3052697777748108,
985
- "step": 122,
986
- "token_acc": 0.894442848003123
987
- },
988
- {
989
- "epoch": 0.4765133171912833,
990
- "grad_norm": 0.2527632713317871,
991
- "learning_rate": 9.753275548567192e-06,
992
- "loss": 0.3045836091041565,
993
- "step": 123,
994
- "token_acc": 0.8951105518605069
995
- },
996
- {
997
- "epoch": 0.4803874092009685,
998
- "grad_norm": 0.20530211925506592,
999
- "learning_rate": 9.74652776304855e-06,
1000
- "loss": 0.3366113305091858,
1001
- "step": 124,
1002
- "token_acc": 0.8836434912892324
1003
- },
1004
- {
1005
- "epoch": 0.48426150121065376,
1006
- "grad_norm": 0.26673150062561035,
1007
- "learning_rate": 9.739691342320866e-06,
1008
- "loss": 0.311764121055603,
1009
- "step": 125,
1010
- "token_acc": 0.8910826454277961
1011
- },
1012
- {
1013
- "epoch": 0.488135593220339,
1014
- "grad_norm": 0.2245185822248459,
1015
- "learning_rate": 9.732766414045368e-06,
1016
- "loss": 0.31055164337158203,
1017
- "step": 126,
1018
- "token_acc": 0.8926098098046538
1019
- },
1020
- {
1021
- "epoch": 0.4920096852300242,
1022
- "grad_norm": 0.2143883854150772,
1023
- "learning_rate": 9.725753107536053e-06,
1024
- "loss": 0.33499595522880554,
1025
- "step": 127,
1026
- "token_acc": 0.8840534260641282
1027
- },
1028
- {
1029
- "epoch": 0.49588377723970944,
1030
- "grad_norm": 0.22163285315036774,
1031
- "learning_rate": 9.718651553757266e-06,
1032
- "loss": 0.31920328736305237,
1033
- "step": 128,
1034
- "token_acc": 0.8901271163419964
1035
- },
1036
- {
1037
- "epoch": 0.4997578692493947,
1038
- "grad_norm": 0.2143898904323578,
1039
- "learning_rate": 9.711461885321247e-06,
1040
- "loss": 0.3301286995410919,
1041
- "step": 129,
1042
- "token_acc": 0.8853363916795757
1043
- },
1044
- {
1045
- "epoch": 0.5036319612590799,
1046
- "grad_norm": 0.24990734457969666,
1047
- "learning_rate": 9.704184236485672e-06,
1048
- "loss": 0.3278159201145172,
1049
- "step": 130,
1050
- "token_acc": 0.8874620923082561
1051
- },
1052
- {
1053
- "epoch": 0.5075060532687651,
1054
- "grad_norm": 0.22136539220809937,
1055
- "learning_rate": 9.696818743151128e-06,
1056
- "loss": 0.3319326937198639,
1057
- "step": 131,
1058
- "token_acc": 0.885009570455441
1059
- },
1060
- {
1061
- "epoch": 0.5113801452784503,
1062
- "grad_norm": 0.2669275999069214,
1063
- "learning_rate": 9.68936554285859e-06,
1064
- "loss": 0.3023684620857239,
1065
- "step": 132,
1066
- "token_acc": 0.8951259709956582
1067
- },
1068
- {
1069
- "epoch": 0.5152542372881356,
1070
- "grad_norm": 0.21833708882331848,
1071
- "learning_rate": 9.68182477478684e-06,
1072
- "loss": 0.3089104890823364,
1073
- "step": 133,
1074
- "token_acc": 0.8930920187299416
1075
- },
1076
- {
1077
- "epoch": 0.5191283292978208,
1078
- "grad_norm": 0.21197167038917542,
1079
- "learning_rate": 9.67419657974988e-06,
1080
- "loss": 0.3144392967224121,
1081
- "step": 134,
1082
- "token_acc": 0.8910884224709107
1083
- },
1084
- {
1085
- "epoch": 0.5230024213075061,
1086
- "grad_norm": 0.21434499323368073,
1087
- "learning_rate": 9.66648110019429e-06,
1088
- "loss": 0.3246540427207947,
1089
- "step": 135,
1090
- "token_acc": 0.8876412650671648
1091
- },
1092
- {
1093
- "epoch": 0.5268765133171913,
1094
- "grad_norm": 0.20343148708343506,
1095
- "learning_rate": 9.658678480196579e-06,
1096
- "loss": 0.315585196018219,
1097
- "step": 136,
1098
- "token_acc": 0.8905443269970013
1099
- },
1100
- {
1101
- "epoch": 0.5307506053268766,
1102
- "grad_norm": 0.23613257706165314,
1103
- "learning_rate": 9.650788865460487e-06,
1104
- "loss": 0.3131225109100342,
1105
- "step": 137,
1106
- "token_acc": 0.8912192170846405
1107
- },
1108
- {
1109
- "epoch": 0.5346246973365617,
1110
- "grad_norm": 0.4212075471878052,
1111
- "learning_rate": 9.642812403314272e-06,
1112
- "loss": 0.29884475469589233,
1113
- "step": 138,
1114
- "token_acc": 0.8966553773404051
1115
- },
1116
- {
1117
- "epoch": 0.538498789346247,
1118
- "grad_norm": 0.20193685591220856,
1119
- "learning_rate": 9.634749242707948e-06,
1120
- "loss": 0.26036083698272705,
1121
- "step": 139,
1122
- "token_acc": 0.9091038865111504
1123
- },
1124
- {
1125
- "epoch": 0.5423728813559322,
1126
- "grad_norm": 0.2208104431629181,
1127
- "learning_rate": 9.626599534210514e-06,
1128
- "loss": 0.33184394240379333,
1129
- "step": 140,
1130
- "token_acc": 0.8853617134142299
1131
- },
1132
- {
1133
- "epoch": 0.5462469733656174,
1134
- "grad_norm": 0.22493727505207062,
1135
- "learning_rate": 9.618363430007134e-06,
1136
- "loss": 0.31208667159080505,
1137
- "step": 141,
1138
- "token_acc": 0.8917024215686027
1139
- },
1140
- {
1141
- "epoch": 0.5501210653753027,
1142
- "grad_norm": 0.23963193595409393,
1143
- "learning_rate": 9.610041083896304e-06,
1144
- "loss": 0.33588868379592896,
1145
- "step": 142,
1146
- "token_acc": 0.883973627021253
1147
- },
1148
- {
1149
- "epoch": 0.553995157384988,
1150
- "grad_norm": 0.21784453094005585,
1151
- "learning_rate": 9.60163265128697e-06,
1152
- "loss": 0.3231375813484192,
1153
- "step": 143,
1154
- "token_acc": 0.8887875239014834
1155
- },
1156
- {
1157
- "epoch": 0.5578692493946731,
1158
- "grad_norm": 0.22835847735404968,
1159
- "learning_rate": 9.593138289195634e-06,
1160
- "loss": 0.3210199773311615,
1161
- "step": 144,
1162
- "token_acc": 0.8890582816354493
1163
- },
1164
- {
1165
- "epoch": 0.5617433414043583,
1166
- "grad_norm": 0.2136555314064026,
1167
- "learning_rate": 9.584558156243418e-06,
1168
- "loss": 0.3372665047645569,
1169
- "step": 145,
1170
- "token_acc": 0.8839793357706921
1171
- },
1172
- {
1173
- "epoch": 0.5656174334140436,
1174
- "grad_norm": 0.20598500967025757,
1175
- "learning_rate": 9.575892412653102e-06,
1176
- "loss": 0.30844664573669434,
1177
- "step": 146,
1178
- "token_acc": 0.8926156654585412
1179
- },
1180
- {
1181
- "epoch": 0.5694915254237288,
1182
- "grad_norm": 0.2522714138031006,
1183
- "learning_rate": 9.567141220246136e-06,
1184
- "loss": 0.36702272295951843,
1185
- "step": 147,
1186
- "token_acc": 0.8734296301671142
1187
- },
1188
- {
1189
- "epoch": 0.5733656174334141,
1190
- "grad_norm": 0.21975038945674896,
1191
- "learning_rate": 9.55830474243961e-06,
1192
- "loss": 0.32784411311149597,
1193
- "step": 148,
1194
- "token_acc": 0.8871756189192851
1195
- },
1196
- {
1197
- "epoch": 0.5772397094430993,
1198
- "grad_norm": 0.21233901381492615,
1199
- "learning_rate": 9.549383144243213e-06,
1200
- "loss": 0.2944122850894928,
1201
- "step": 149,
1202
- "token_acc": 0.8987453672884691
1203
- },
1204
- {
1205
- "epoch": 0.5811138014527845,
1206
- "grad_norm": 0.2199799120426178,
1207
- "learning_rate": 9.540376592256142e-06,
1208
- "loss": 0.3299463987350464,
1209
- "step": 150,
1210
- "token_acc": 0.8859144839374592
1211
- },
1212
- {
1213
- "epoch": 0.5849878934624697,
1214
- "grad_norm": 0.19698019325733185,
1215
- "learning_rate": 9.531285254663997e-06,
1216
- "loss": 0.3030051589012146,
1217
- "step": 151,
1218
- "token_acc": 0.8951707294894029
1219
- },
1220
- {
1221
- "epoch": 0.588861985472155,
1222
- "grad_norm": 0.22306668758392334,
1223
- "learning_rate": 9.522109301235637e-06,
1224
- "loss": 0.29752516746520996,
1225
- "step": 152,
1226
- "token_acc": 0.8966012679857996
1227
- },
1228
- {
1229
- "epoch": 0.5927360774818402,
1230
- "grad_norm": 0.21317337453365326,
1231
- "learning_rate": 9.512848903320017e-06,
1232
- "loss": 0.3052118122577667,
1233
- "step": 153,
1234
- "token_acc": 0.8944324633814714
1235
- },
1236
- {
1237
- "epoch": 0.5966101694915255,
1238
- "grad_norm": 0.2120915800333023,
1239
- "learning_rate": 9.503504233842973e-06,
1240
- "loss": 0.29761528968811035,
1241
- "step": 154,
1242
- "token_acc": 0.8966406260468731
1243
- },
1244
- {
1245
- "epoch": 0.6004842615012107,
1246
- "grad_norm": 0.23525090515613556,
1247
- "learning_rate": 9.494075467304007e-06,
1248
- "loss": 0.3034532070159912,
1249
- "step": 155,
1250
- "token_acc": 0.8944926637860167
1251
- },
1252
- {
1253
- "epoch": 0.6043583535108958,
1254
- "grad_norm": 0.2095353752374649,
1255
- "learning_rate": 9.484562779773027e-06,
1256
- "loss": 0.2903788089752197,
1257
- "step": 156,
1258
- "token_acc": 0.8990560027078014
1259
- },
1260
- {
1261
- "epoch": 0.6082324455205811,
1262
- "grad_norm": 0.23741677403450012,
1263
- "learning_rate": 9.474966348887055e-06,
1264
- "loss": 0.31467512249946594,
1265
- "step": 157,
1266
- "token_acc": 0.8904583329757747
1267
- },
1268
- {
1269
- "epoch": 0.6121065375302663,
1270
- "grad_norm": 0.2259555608034134,
1271
- "learning_rate": 9.465286353846905e-06,
1272
- "loss": 0.3404577374458313,
1273
- "step": 158,
1274
- "token_acc": 0.8826165622063978
1275
- },
1276
- {
1277
- "epoch": 0.6159806295399516,
1278
- "grad_norm": 0.2183879017829895,
1279
- "learning_rate": 9.455522975413846e-06,
1280
- "loss": 0.2766571640968323,
1281
- "step": 159,
1282
- "token_acc": 0.9038809421418853
1283
- },
1284
- {
1285
- "epoch": 0.6198547215496368,
1286
- "grad_norm": 0.22651784121990204,
1287
- "learning_rate": 9.445676395906226e-06,
1288
- "loss": 0.29638129472732544,
1289
- "step": 160,
1290
- "token_acc": 0.8970113168662065
1291
- },
1292
- {
1293
- "epoch": 0.6237288135593221,
1294
- "grad_norm": 0.22088395059108734,
1295
- "learning_rate": 9.435746799196061e-06,
1296
- "loss": 0.3023075759410858,
1297
- "step": 161,
1298
- "token_acc": 0.8946665593674712
1299
- },
1300
- {
1301
- "epoch": 0.6276029055690072,
1302
- "grad_norm": 0.21526560187339783,
1303
- "learning_rate": 9.425734370705606e-06,
1304
- "loss": 0.28661438822746277,
1305
- "step": 162,
1306
- "token_acc": 0.9002787847728345
1307
- },
1308
- {
1309
- "epoch": 0.6314769975786925,
1310
- "grad_norm": 0.23334769904613495,
1311
- "learning_rate": 9.415639297403891e-06,
1312
- "loss": 0.31685301661491394,
1313
- "step": 163,
1314
- "token_acc": 0.890886748080584
1315
- },
1316
- {
1317
- "epoch": 0.6353510895883777,
1318
- "grad_norm": 0.200165793299675,
1319
- "learning_rate": 9.40546176780323e-06,
1320
- "loss": 0.30981898307800293,
1321
- "step": 164,
1322
- "token_acc": 0.8924871164982372
1323
- },
1324
- {
1325
- "epoch": 0.639225181598063,
1326
- "grad_norm": 0.20800836384296417,
1327
- "learning_rate": 9.395201971955701e-06,
1328
- "loss": 0.3162352740764618,
1329
- "step": 165,
1330
- "token_acc": 0.8910434805285766
1331
- },
1332
- {
1333
- "epoch": 0.6430992736077482,
1334
- "grad_norm": 0.20923736691474915,
1335
- "learning_rate": 9.384860101449598e-06,
1336
- "loss": 0.32208406925201416,
1337
- "step": 166,
1338
- "token_acc": 0.8880633815629819
1339
- },
1340
- {
1341
- "epoch": 0.6469733656174335,
1342
- "grad_norm": 0.1986808031797409,
1343
- "learning_rate": 9.374436349405847e-06,
1344
- "loss": 0.28397923707962036,
1345
- "step": 167,
1346
- "token_acc": 0.9012052212352475
1347
- },
1348
- {
1349
- "epoch": 0.6508474576271186,
1350
- "grad_norm": 0.21215273439884186,
1351
- "learning_rate": 9.36393091047441e-06,
1352
- "loss": 0.3066609799861908,
1353
- "step": 168,
1354
- "token_acc": 0.894593303584187
1355
- },
1356
- {
1357
- "epoch": 0.6547215496368038,
1358
- "grad_norm": 0.20804037153720856,
1359
- "learning_rate": 9.353343980830644e-06,
1360
- "loss": 0.3097017705440521,
1361
- "step": 169,
1362
- "token_acc": 0.8926308156125992
1363
- },
1364
- {
1365
- "epoch": 0.6585956416464891,
1366
- "grad_norm": 0.20328834652900696,
1367
- "learning_rate": 9.342675758171638e-06,
1368
- "loss": 0.3010105490684509,
1369
- "step": 170,
1370
- "token_acc": 0.8950560660129195
1371
- },
1372
- {
1373
- "epoch": 0.6624697336561743,
1374
- "grad_norm": 0.2051060050725937,
1375
- "learning_rate": 9.331926441712522e-06,
1376
- "loss": 0.3019353151321411,
1377
- "step": 171,
1378
- "token_acc": 0.8949745506999682
1379
- },
1380
- {
1381
- "epoch": 0.6663438256658596,
1382
- "grad_norm": 0.24043123424053192,
1383
- "learning_rate": 9.32109623218275e-06,
1384
- "loss": 0.3116442859172821,
1385
- "step": 172,
1386
- "token_acc": 0.8915558784861239
1387
- },
1388
- {
1389
- "epoch": 0.6702179176755448,
1390
- "grad_norm": 0.21520181000232697,
1391
- "learning_rate": 9.310185331822338e-06,
1392
- "loss": 0.31186142563819885,
1393
- "step": 173,
1394
- "token_acc": 0.8917585320277845
1395
- },
1396
- {
1397
- "epoch": 0.67409200968523,
1398
- "grad_norm": 0.21344298124313354,
1399
- "learning_rate": 9.299193944378112e-06,
1400
- "loss": 0.3273160755634308,
1401
- "step": 174,
1402
- "token_acc": 0.886418268420563
1403
- },
1404
- {
1405
- "epoch": 0.6779661016949152,
1406
- "grad_norm": 0.20224156975746155,
1407
- "learning_rate": 9.28812227509988e-06,
1408
- "loss": 0.31608837842941284,
1409
- "step": 175,
1410
- "token_acc": 0.8894536504933755
1411
- },
1412
- {
1413
- "epoch": 0.6818401937046005,
1414
- "grad_norm": 0.2154257595539093,
1415
- "learning_rate": 9.27697053073661e-06,
1416
- "loss": 0.34367692470550537,
1417
- "step": 176,
1418
- "token_acc": 0.8811017511710314
1419
- },
1420
- {
1421
- "epoch": 0.6857142857142857,
1422
- "grad_norm": 0.22003678977489471,
1423
- "learning_rate": 9.26573891953257e-06,
1424
- "loss": 0.3205263018608093,
1425
- "step": 177,
1426
- "token_acc": 0.8893210947921869
1427
- },
1428
- {
1429
- "epoch": 0.689588377723971,
1430
- "grad_norm": 0.21449677646160126,
1431
- "learning_rate": 9.254427651223434e-06,
1432
- "loss": 0.28666430711746216,
1433
- "step": 178,
1434
- "token_acc": 0.9003720788020833
1435
- },
1436
- {
1437
- "epoch": 0.6934624697336562,
1438
- "grad_norm": 0.22110596299171448,
1439
- "learning_rate": 9.243036937032373e-06,
1440
- "loss": 0.3156067728996277,
1441
- "step": 179,
1442
- "token_acc": 0.8902597783694092
1443
- },
1444
- {
1445
- "epoch": 0.6973365617433414,
1446
- "grad_norm": 0.19700580835342407,
1447
- "learning_rate": 9.2315669896661e-06,
1448
- "loss": 0.28897273540496826,
1449
- "step": 180,
1450
- "token_acc": 0.8994499889336349
1451
- },
1452
- {
1453
- "epoch": 0.7012106537530266,
1454
- "grad_norm": 0.21460606157779694,
1455
- "learning_rate": 9.220018023310908e-06,
1456
- "loss": 0.31268295645713806,
1457
- "step": 181,
1458
- "token_acc": 0.8918378520876847
1459
- },
1460
- {
1461
- "epoch": 0.7050847457627119,
1462
- "grad_norm": 0.21692436933517456,
1463
- "learning_rate": 9.208390253628667e-06,
1464
- "loss": 0.28844964504241943,
1465
- "step": 182,
1466
- "token_acc": 0.8997311485616448
1467
- },
1468
- {
1469
- "epoch": 0.7089588377723971,
1470
- "grad_norm": 0.201703280210495,
1471
- "learning_rate": 9.196683897752794e-06,
1472
- "loss": 0.32861441373825073,
1473
- "step": 183,
1474
- "token_acc": 0.8854774295445417
1475
- },
1476
- {
1477
- "epoch": 0.7128329297820823,
1478
- "grad_norm": 3.976747751235962,
1479
- "learning_rate": 9.184899174284201e-06,
1480
- "loss": 0.33475255966186523,
1481
- "step": 184,
1482
- "token_acc": 0.8836819705392365
1483
- },
1484
- {
1485
- "epoch": 0.7167070217917676,
1486
- "grad_norm": 0.24247053265571594,
1487
- "learning_rate": 9.173036303287215e-06,
1488
- "loss": 0.3366454243659973,
1489
- "step": 185,
1490
- "token_acc": 0.8833432089980459
1491
- },
1492
- {
1493
- "epoch": 0.7205811138014527,
1494
- "grad_norm": 0.2282845675945282,
1495
- "learning_rate": 9.16109550628546e-06,
1496
- "loss": 0.2812536656856537,
1497
- "step": 186,
1498
- "token_acc": 0.9027706860502607
1499
- },
1500
- {
1501
- "epoch": 0.724455205811138,
1502
- "grad_norm": 0.2282128632068634,
1503
- "learning_rate": 9.149077006257734e-06,
1504
- "loss": 0.3136906027793884,
1505
- "step": 187,
1506
- "token_acc": 0.8912536222754189
1507
- },
1508
- {
1509
- "epoch": 0.7283292978208232,
1510
- "grad_norm": 0.20751290023326874,
1511
- "learning_rate": 9.136981027633834e-06,
1512
- "loss": 0.29636135697364807,
1513
- "step": 188,
1514
- "token_acc": 0.8974463288547996
1515
- },
1516
- {
1517
- "epoch": 0.7322033898305085,
1518
- "grad_norm": 0.23192144930362701,
1519
- "learning_rate": 9.124807796290366e-06,
1520
- "loss": 0.3046882152557373,
1521
- "step": 189,
1522
- "token_acc": 0.8943812414560115
1523
- },
1524
- {
1525
- "epoch": 0.7360774818401937,
1526
- "grad_norm": 0.221333310008049,
1527
- "learning_rate": 9.112557539546535e-06,
1528
- "loss": 0.32960376143455505,
1529
- "step": 190,
1530
- "token_acc": 0.8860915000599271
1531
- },
1532
- {
1533
- "epoch": 0.739951573849879,
1534
- "grad_norm": 0.1981872171163559,
1535
- "learning_rate": 9.100230486159893e-06,
1536
- "loss": 0.32151421904563904,
1537
- "step": 191,
1538
- "token_acc": 0.888598638535205
1539
- },
1540
- {
1541
- "epoch": 0.7438256658595641,
1542
- "grad_norm": 0.2172573357820511,
1543
- "learning_rate": 9.087826866322065e-06,
1544
- "loss": 0.3255336880683899,
1545
- "step": 192,
1546
- "token_acc": 0.8864367509340579
1547
- },
1548
- {
1549
- "epoch": 0.7476997578692494,
1550
- "grad_norm": 0.21215571463108063,
1551
- "learning_rate": 9.075346911654456e-06,
1552
- "loss": 0.30505236983299255,
1553
- "step": 193,
1554
- "token_acc": 0.8936060377931436
1555
- },
1556
- {
1557
- "epoch": 0.7515738498789346,
1558
- "grad_norm": 0.21355277299880981,
1559
- "learning_rate": 9.062790855203932e-06,
1560
- "loss": 0.3349328637123108,
1561
- "step": 194,
1562
- "token_acc": 0.8847527625851099
1563
- },
1564
- {
1565
- "epoch": 0.7554479418886199,
1566
- "grad_norm": 0.20415301620960236,
1567
- "learning_rate": 9.050158931438451e-06,
1568
- "loss": 0.3010273873806,
1569
- "step": 195,
1570
- "token_acc": 0.8946901896914337
1571
- },
1572
- {
1573
- "epoch": 0.7593220338983051,
1574
- "grad_norm": 0.2100018560886383,
1575
- "learning_rate": 9.037451376242696e-06,
1576
- "loss": 0.3295148015022278,
1577
- "step": 196,
1578
- "token_acc": 0.8861214255925314
1579
- },
1580
- {
1581
- "epoch": 0.7631961259079904,
1582
- "grad_norm": 0.21248096227645874,
1583
- "learning_rate": 9.024668426913671e-06,
1584
- "loss": 0.2901475727558136,
1585
- "step": 197,
1586
- "token_acc": 0.8984891018269412
1587
- },
1588
- {
1589
- "epoch": 0.7670702179176755,
1590
- "grad_norm": 0.20735451579093933,
1591
- "learning_rate": 9.011810322156269e-06,
1592
- "loss": 0.3123668134212494,
1593
- "step": 198,
1594
- "token_acc": 0.8911118341790296
1595
- },
1596
- {
1597
- "epoch": 0.7709443099273607,
1598
- "grad_norm": 0.2119433879852295,
1599
- "learning_rate": 8.998877302078803e-06,
1600
- "loss": 0.30766892433166504,
1601
- "step": 199,
1602
- "token_acc": 0.8930650097673094
1603
- },
1604
- {
1605
- "epoch": 0.774818401937046,
1606
- "grad_norm": 0.20151817798614502,
1607
- "learning_rate": 8.985869608188545e-06,
1608
- "loss": 0.294528067111969,
1609
- "step": 200,
1610
- "token_acc": 0.8973507748438794
1611
- },
1612
- {
1613
- "epoch": 0.7786924939467312,
1614
- "grad_norm": 0.20979715883731842,
1615
- "learning_rate": 8.97278748338719e-06,
1616
- "loss": 0.3116077184677124,
1617
- "step": 201,
1618
- "token_acc": 0.8916578293780434
1619
- },
1620
- {
1621
- "epoch": 0.7825665859564165,
1622
- "grad_norm": 0.21114560961723328,
1623
- "learning_rate": 8.95963117196634e-06,
1624
- "loss": 0.31117022037506104,
1625
- "step": 202,
1626
- "token_acc": 0.8922739117136779
1627
- },
1628
- {
1629
- "epoch": 0.7864406779661017,
1630
- "grad_norm": 0.2028111070394516,
1631
- "learning_rate": 8.946400919602933e-06,
1632
- "loss": 0.2925041913986206,
1633
- "step": 203,
1634
- "token_acc": 0.8979599612123477
1635
- },
1636
- {
1637
- "epoch": 0.790314769975787,
1638
- "grad_norm": 0.19873376190662384,
1639
- "learning_rate": 8.933096973354665e-06,
1640
- "loss": 0.3335387706756592,
1641
- "step": 204,
1642
- "token_acc": 0.8845781124549695
1643
- },
1644
- {
1645
- "epoch": 0.7941888619854721,
1646
- "grad_norm": 0.20865830779075623,
1647
- "learning_rate": 8.919719581655357e-06,
1648
- "loss": 0.3048374652862549,
1649
- "step": 205,
1650
- "token_acc": 0.8941424666394205
1651
- },
1652
- {
1653
- "epoch": 0.7980629539951574,
1654
- "grad_norm": 0.21847450733184814,
1655
- "learning_rate": 8.906268994310339e-06,
1656
- "loss": 0.30148929357528687,
1657
- "step": 206,
1658
- "token_acc": 0.8948231645494126
1659
- },
1660
- {
1661
- "epoch": 0.8019370460048426,
1662
- "grad_norm": 0.23447921872138977,
1663
- "learning_rate": 8.892745462491763e-06,
1664
- "loss": 0.3076891005039215,
1665
- "step": 207,
1666
- "token_acc": 0.8940680143003497
1667
- },
1668
- {
1669
- "epoch": 0.8058111380145279,
1670
- "grad_norm": 0.2047218531370163,
1671
- "learning_rate": 8.879149238733932e-06,
1672
- "loss": 0.2903471291065216,
1673
- "step": 208,
1674
- "token_acc": 0.8996930000967329
1675
- },
1676
- {
1677
- "epoch": 0.8096852300242131,
1678
- "grad_norm": 0.3560882806777954,
1679
- "learning_rate": 8.865480576928578e-06,
1680
- "loss": 0.2734353840351105,
1681
- "step": 209,
1682
- "token_acc": 0.9038816908230364
1683
- },
1684
- {
1685
- "epoch": 0.8135593220338984,
1686
- "grad_norm": 0.22588837146759033,
1687
- "learning_rate": 8.851739732320109e-06,
1688
- "loss": 0.30820316076278687,
1689
- "step": 210,
1690
- "token_acc": 0.8928903081404425
1691
- },
1692
- {
1693
- "epoch": 0.8174334140435835,
1694
- "grad_norm": 0.19928814470767975,
1695
- "learning_rate": 8.83792696150086e-06,
1696
- "loss": 0.30705487728118896,
1697
- "step": 211,
1698
- "token_acc": 0.8931717351449738
1699
- },
1700
- {
1701
- "epoch": 0.8213075060532687,
1702
- "grad_norm": 0.23134565353393555,
1703
- "learning_rate": 8.824042522406295e-06,
1704
- "loss": 0.3144133687019348,
1705
- "step": 212,
1706
- "token_acc": 0.8904542748607169
1707
- },
1708
- {
1709
- "epoch": 0.825181598062954,
1710
- "grad_norm": 0.20952780544757843,
1711
- "learning_rate": 8.810086674310184e-06,
1712
- "loss": 0.3166520595550537,
1713
- "step": 213,
1714
- "token_acc": 0.8902617260259249
1715
- },
1716
- {
1717
- "epoch": 0.8290556900726392,
1718
- "grad_norm": 0.21133121848106384,
1719
- "learning_rate": 8.796059677819773e-06,
1720
- "loss": 0.31384018063545227,
1721
- "step": 214,
1722
- "token_acc": 0.8909493414116798
1723
- },
1724
- {
1725
- "epoch": 0.8329297820823245,
1726
- "grad_norm": 0.3206462264060974,
1727
- "learning_rate": 8.781961794870903e-06,
1728
- "loss": 0.30939990282058716,
1729
- "step": 215,
1730
- "token_acc": 0.8926290243396312
1731
- },
1732
- {
1733
- "epoch": 0.8368038740920097,
1734
- "grad_norm": 0.21380406618118286,
1735
- "learning_rate": 8.767793288723137e-06,
1736
- "loss": 0.3126541078090668,
1737
- "step": 216,
1738
- "token_acc": 0.8918149018414423
1739
- },
1740
- {
1741
- "epoch": 0.8406779661016949,
1742
- "grad_norm": 0.2241922914981842,
1743
- "learning_rate": 8.753554423954828e-06,
1744
- "loss": 0.32906076312065125,
1745
- "step": 217,
1746
- "token_acc": 0.8866828065863777
1747
- },
1748
- {
1749
- "epoch": 0.8445520581113801,
1750
- "grad_norm": 0.19776619970798492,
1751
- "learning_rate": 8.739245466458187e-06,
1752
- "loss": 0.28062158823013306,
1753
- "step": 218,
1754
- "token_acc": 0.9022684784065322
1755
- },
1756
- {
1757
- "epoch": 0.8484261501210654,
1758
- "grad_norm": 0.2141999900341034,
1759
- "learning_rate": 8.72486668343431e-06,
1760
- "loss": 0.3276277184486389,
1761
- "step": 219,
1762
- "token_acc": 0.8861141792995992
1763
- },
1764
- {
1765
- "epoch": 0.8523002421307506,
1766
- "grad_norm": 0.2332129180431366,
1767
- "learning_rate": 8.7104183433882e-06,
1768
- "loss": 0.3168509304523468,
1769
- "step": 220,
1770
- "token_acc": 0.8899989570826125
1771
- },
1772
- {
1773
- "epoch": 0.8561743341404359,
1774
- "grad_norm": 0.2141677886247635,
1775
- "learning_rate": 8.695900716123744e-06,
1776
- "loss": 0.3259914219379425,
1777
- "step": 221,
1778
- "token_acc": 0.8866733094194235
1779
- },
1780
- {
1781
- "epoch": 0.8600484261501211,
1782
- "grad_norm": 0.20929858088493347,
1783
- "learning_rate": 8.681314072738678e-06,
1784
- "loss": 0.2776751220226288,
1785
- "step": 222,
1786
- "token_acc": 0.9029569916163804
1787
- },
1788
- {
1789
- "epoch": 0.8639225181598063,
1790
- "grad_norm": 0.26802197098731995,
1791
- "learning_rate": 8.666658685619523e-06,
1792
- "loss": 0.3192378282546997,
1793
- "step": 223,
1794
- "token_acc": 0.8888524656782731
1795
- },
1796
- {
1797
- "epoch": 0.8677966101694915,
1798
- "grad_norm": 0.19303195178508759,
1799
- "learning_rate": 8.651934828436497e-06,
1800
- "loss": 0.2820873260498047,
1801
- "step": 224,
1802
- "token_acc": 0.9010663601046539
1803
- },
1804
- {
1805
- "epoch": 0.8716707021791767,
1806
- "grad_norm": 0.20784462988376617,
1807
- "learning_rate": 8.637142776138415e-06,
1808
- "loss": 0.2850268483161926,
1809
- "step": 225,
1810
- "token_acc": 0.9003609394726915
1811
- },
1812
- {
1813
- "epoch": 0.875544794188862,
1814
- "grad_norm": 0.2194257229566574,
1815
- "learning_rate": 8.622282804947537e-06,
1816
- "loss": 0.31484997272491455,
1817
- "step": 226,
1818
- "token_acc": 0.8909253202507496
1819
- },
1820
- {
1821
- "epoch": 0.8794188861985472,
1822
- "grad_norm": 0.21197804808616638,
1823
- "learning_rate": 8.607355192354425e-06,
1824
- "loss": 0.3072202801704407,
1825
- "step": 227,
1826
- "token_acc": 0.8929364556285221
1827
- },
1828
- {
1829
- "epoch": 0.8832929782082325,
1830
- "grad_norm": 0.19514977931976318,
1831
- "learning_rate": 8.592360217112759e-06,
1832
- "loss": 0.31343895196914673,
1833
- "step": 228,
1834
- "token_acc": 0.8909144611151198
1835
- },
1836
- {
1837
- "epoch": 0.8871670702179176,
1838
- "grad_norm": 0.2198445200920105,
1839
- "learning_rate": 8.57729815923412e-06,
1840
- "loss": 0.31176120042800903,
1841
- "step": 229,
1842
- "token_acc": 0.8916788161998124
1843
- },
1844
- {
1845
- "epoch": 0.8910411622276029,
1846
- "grad_norm": 0.20297633111476898,
1847
- "learning_rate": 8.562169299982776e-06,
1848
- "loss": 0.30840498208999634,
1849
- "step": 230,
1850
- "token_acc": 0.8921534903182912
1851
- },
1852
- {
1853
- "epoch": 0.8949152542372881,
1854
- "grad_norm": 0.21356205642223358,
1855
- "learning_rate": 8.546973921870421e-06,
1856
- "loss": 0.3210839629173279,
1857
- "step": 231,
1858
- "token_acc": 0.8882864775840541
1859
- },
1860
- {
1861
- "epoch": 0.8987893462469734,
1862
- "grad_norm": 0.21405935287475586,
1863
- "learning_rate": 8.531712308650904e-06,
1864
- "loss": 0.3006952702999115,
1865
- "step": 232,
1866
- "token_acc": 0.8953128142705267
1867
- },
1868
- {
1869
- "epoch": 0.9026634382566586,
1870
- "grad_norm": 0.21220295131206512,
1871
- "learning_rate": 8.516384745314926e-06,
1872
- "loss": 0.33272668719291687,
1873
- "step": 233,
1874
- "token_acc": 0.8845533899027282
1875
- },
1876
- {
1877
- "epoch": 0.9065375302663439,
1878
- "grad_norm": 0.19546008110046387,
1879
- "learning_rate": 8.50099151808472e-06,
1880
- "loss": 0.26581257581710815,
1881
- "step": 234,
1882
- "token_acc": 0.9067262813046539
1883
- },
1884
- {
1885
- "epoch": 0.910411622276029,
1886
- "grad_norm": 0.2057773917913437,
1887
- "learning_rate": 8.485532914408712e-06,
1888
- "loss": 0.2936754524707794,
1889
- "step": 235,
1890
- "token_acc": 0.8980145512690381
1891
- },
1892
- {
1893
- "epoch": 0.9142857142857143,
1894
- "grad_norm": 0.21968601644039154,
1895
- "learning_rate": 8.470009222956138e-06,
1896
- "loss": 0.2990136742591858,
1897
- "step": 236,
1898
- "token_acc": 0.8944779048351311
1899
- },
1900
- {
1901
- "epoch": 0.9181598062953995,
1902
- "grad_norm": 0.22149494290351868,
1903
- "learning_rate": 8.45442073361167e-06,
1904
- "loss": 0.29907599091529846,
1905
- "step": 237,
1906
- "token_acc": 0.8953804266415489
1907
- },
1908
- {
1909
- "epoch": 0.9220338983050848,
1910
- "grad_norm": 0.18807418644428253,
1911
- "learning_rate": 8.438767737469995e-06,
1912
- "loss": 0.2596169412136078,
1913
- "step": 238,
1914
- "token_acc": 0.9094668271985952
1915
- },
1916
- {
1917
- "epoch": 0.92590799031477,
1918
- "grad_norm": 0.2053857445716858,
1919
- "learning_rate": 8.42305052683038e-06,
1920
- "loss": 0.320443719625473,
1921
- "step": 239,
1922
- "token_acc": 0.8882472950063495
1923
- },
1924
- {
1925
- "epoch": 0.9297820823244553,
1926
- "grad_norm": 0.19474725425243378,
1927
- "learning_rate": 8.407269395191216e-06,
1928
- "loss": 0.29054853320121765,
1929
- "step": 240,
1930
- "token_acc": 0.8986681898213217
1931
- },
1932
- {
1933
- "epoch": 0.9336561743341404,
1934
- "grad_norm": 0.22415153682231903,
1935
- "learning_rate": 8.391424637244528e-06,
1936
- "loss": 0.29720863699913025,
1937
- "step": 241,
1938
- "token_acc": 0.8967865758573351
1939
- },
1940
- {
1941
- "epoch": 0.9375302663438256,
1942
- "grad_norm": 0.20295462012290955,
1943
- "learning_rate": 8.375516548870489e-06,
1944
- "loss": 0.3213497996330261,
1945
- "step": 242,
1946
- "token_acc": 0.8888211973402874
1947
- },
1948
- {
1949
- "epoch": 0.9414043583535109,
1950
- "grad_norm": 0.235239177942276,
1951
- "learning_rate": 8.359545427131876e-06,
1952
- "loss": 0.31140708923339844,
1953
- "step": 243,
1954
- "token_acc": 0.8917541696945803
1955
- },
1956
- {
1957
- "epoch": 0.9452784503631961,
1958
- "grad_norm": 0.21419954299926758,
1959
- "learning_rate": 8.343511570268541e-06,
1960
- "loss": 0.3142154812812805,
1961
- "step": 244,
1962
- "token_acc": 0.890589961402836
1963
- },
1964
- {
1965
- "epoch": 0.9491525423728814,
1966
- "grad_norm": 0.20498663187026978,
1967
- "learning_rate": 8.327415277691824e-06,
1968
- "loss": 0.3464815020561218,
1969
- "step": 245,
1970
- "token_acc": 0.8797665540392294
1971
- },
1972
- {
1973
- "epoch": 0.9530266343825666,
1974
- "grad_norm": 0.20611073076725006,
1975
- "learning_rate": 8.311256849978974e-06,
1976
- "loss": 0.31497207283973694,
1977
- "step": 246,
1978
- "token_acc": 0.889790752866034
1979
- },
1980
- {
1981
- "epoch": 0.9569007263922518,
1982
- "grad_norm": 0.21447882056236267,
1983
- "learning_rate": 8.295036588867533e-06,
1984
- "loss": 0.28588759899139404,
1985
- "step": 247,
1986
- "token_acc": 0.8993494375908707
1987
- },
1988
- {
1989
- "epoch": 0.960774818401937,
1990
- "grad_norm": 0.21430622041225433,
1991
- "learning_rate": 8.278754797249702e-06,
1992
- "loss": 0.3209206461906433,
1993
- "step": 248,
1994
- "token_acc": 0.8878057052632179
1995
- },
1996
- {
1997
- "epoch": 0.9646489104116223,
1998
- "grad_norm": 0.1971716433763504,
1999
- "learning_rate": 8.262411779166681e-06,
2000
- "loss": 0.29577910900115967,
2001
- "step": 249,
2002
- "token_acc": 0.8970768255184925
2003
- },
2004
- {
2005
- "epoch": 0.9685230024213075,
2006
- "grad_norm": 0.20728042721748352,
2007
- "learning_rate": 8.246007839802997e-06,
2008
- "loss": 0.3149109482765198,
2009
- "step": 250,
2010
- "token_acc": 0.8904120076852685
2011
- },
2012
- {
2013
- "epoch": 0.9723970944309928,
2014
- "grad_norm": 0.23157289624214172,
2015
- "learning_rate": 8.229543285480797e-06,
2016
- "loss": 0.3057391047477722,
2017
- "step": 251,
2018
- "token_acc": 0.8943966929583815
2019
- },
2020
- {
2021
- "epoch": 0.976271186440678,
2022
- "grad_norm": 0.21818409860134125,
2023
- "learning_rate": 8.213018423654144e-06,
2024
- "loss": 0.3090881109237671,
2025
- "step": 252,
2026
- "token_acc": 0.8931029437419457
2027
- },
2028
- {
2029
- "epoch": 0.9801452784503631,
2030
- "grad_norm": 0.20345434546470642,
2031
- "learning_rate": 8.196433562903252e-06,
2032
- "loss": 0.2966330051422119,
2033
- "step": 253,
2034
- "token_acc": 0.8959465166900704
2035
- },
2036
- {
2037
- "epoch": 0.9840193704600484,
2038
- "grad_norm": 0.203868567943573,
2039
- "learning_rate": 8.179789012928747e-06,
2040
- "loss": 0.2893424928188324,
2041
- "step": 254,
2042
- "token_acc": 0.8989887993032385
2043
- },
2044
- {
2045
- "epoch": 0.9878934624697336,
2046
- "grad_norm": 0.20835842192173004,
2047
- "learning_rate": 8.163085084545867e-06,
2048
- "loss": 0.29561957716941833,
2049
- "step": 255,
2050
- "token_acc": 0.897130295078995
2051
- },
2052
- {
2053
- "epoch": 0.9917675544794189,
2054
- "grad_norm": 0.2602974772453308,
2055
- "learning_rate": 8.146322089678668e-06,
2056
- "loss": 0.33309951424598694,
2057
- "step": 256,
2058
- "token_acc": 0.8842519179704944
2059
- },
2060
- {
2061
- "epoch": 0.9956416464891041,
2062
- "grad_norm": 0.1993730664253235,
2063
- "learning_rate": 8.129500341354192e-06,
2064
- "loss": 0.32513946294784546,
2065
- "step": 257,
2066
- "token_acc": 0.8869922494628838
2067
- },
2068
- {
2069
- "epoch": 0.9995157384987894,
2070
- "grad_norm": 0.2033330649137497,
2071
- "learning_rate": 8.11262015369663e-06,
2072
- "loss": 0.29512181878089905,
2073
- "step": 258,
2074
- "token_acc": 0.8968425014801387
2075
- },
2076
- {
2077
- "epoch": 1.0,
2078
- "grad_norm": 0.6673643589019775,
2079
- "learning_rate": 8.095681841921441e-06,
2080
- "loss": 0.28728920221328735,
2081
- "step": 259,
2082
- "token_acc": 0.9003083713758805
2083
- },
2084
- {
2085
- "epoch": 1.0038740920096851,
2086
- "grad_norm": 0.32744893431663513,
2087
- "learning_rate": 8.07868572232949e-06,
2088
- "loss": 0.269972562789917,
2089
- "step": 260,
2090
- "token_acc": 0.9038492097273063
2091
- },
2092
- {
2093
- "epoch": 1.0077481840193705,
2094
- "grad_norm": 0.2596898376941681,
2095
- "learning_rate": 8.061632112301122e-06,
2096
- "loss": 0.2655790150165558,
2097
- "step": 261,
2098
- "token_acc": 0.9053338855906853
2099
- },
2100
- {
2101
- "epoch": 1.0116222760290556,
2102
- "grad_norm": 0.2612839639186859,
2103
- "learning_rate": 8.044521330290235e-06,
2104
- "loss": 0.2887282967567444,
2105
- "step": 262,
2106
- "token_acc": 0.8971828029711167
2107
- },
2108
- {
2109
- "epoch": 1.015496368038741,
2110
- "grad_norm": 0.2769652009010315,
2111
- "learning_rate": 8.027353695818345e-06,
2112
- "loss": 0.26126527786254883,
2113
- "step": 263,
2114
- "token_acc": 0.9065780969019781
2115
- },
2116
- {
2117
- "epoch": 1.0193704600484261,
2118
- "grad_norm": 0.27929142117500305,
2119
- "learning_rate": 8.010129529468614e-06,
2120
- "loss": 0.27868735790252686,
2121
- "step": 264,
2122
- "token_acc": 0.9001419249114798
2123
- },
2124
- {
2125
- "epoch": 1.0232445520581113,
2126
- "grad_norm": 0.23997750878334045,
2127
- "learning_rate": 7.992849152879857e-06,
2128
- "loss": 0.2831759750843048,
2129
- "step": 265,
2130
- "token_acc": 0.899304001670737
2131
- },
2132
- {
2133
- "epoch": 1.0271186440677966,
2134
- "grad_norm": 0.25313815474510193,
2135
- "learning_rate": 7.97551288874055e-06,
2136
- "loss": 0.27934202551841736,
2137
- "step": 266,
2138
- "token_acc": 0.9004498805562496
2139
- },
2140
- {
2141
- "epoch": 1.0309927360774818,
2142
- "grad_norm": 0.23287494480609894,
2143
- "learning_rate": 7.95812106078279e-06,
2144
- "loss": 0.26112881302833557,
2145
- "step": 267,
2146
- "token_acc": 0.9065508038300509
2147
- },
2148
- {
2149
- "epoch": 1.0348668280871671,
2150
- "grad_norm": 0.22660091519355774,
2151
- "learning_rate": 7.940673993776258e-06,
2152
- "loss": 0.2504875063896179,
2153
- "step": 268,
2154
- "token_acc": 0.9097140867981872
2155
- },
2156
- {
2157
- "epoch": 1.0387409200968523,
2158
- "grad_norm": 0.2266615480184555,
2159
- "learning_rate": 7.923172013522153e-06,
2160
- "loss": 0.25760790705680847,
2161
- "step": 269,
2162
- "token_acc": 0.9073963735109954
2163
- },
2164
- {
2165
- "epoch": 1.0426150121065376,
2166
- "grad_norm": 0.22593924403190613,
2167
- "learning_rate": 7.905615446847107e-06,
2168
- "loss": 0.28686419129371643,
2169
- "step": 270,
2170
- "token_acc": 0.8976161305002275
2171
- },
2172
- {
2173
- "epoch": 1.0464891041162228,
2174
- "grad_norm": 0.2425071895122528,
2175
- "learning_rate": 7.888004621597079e-06,
2176
- "loss": 0.2573948800563812,
2177
- "step": 271,
2178
- "token_acc": 0.907380557815819
2179
- },
2180
- {
2181
- "epoch": 1.050363196125908,
2182
- "grad_norm": 0.23996935784816742,
2183
- "learning_rate": 7.87033986663124e-06,
2184
- "loss": 0.2808932065963745,
2185
- "step": 272,
2186
- "token_acc": 0.8994914728045711
2187
- },
2188
- {
2189
- "epoch": 1.0542372881355933,
2190
- "grad_norm": 0.25931164622306824,
2191
- "learning_rate": 7.852621511815825e-06,
2192
- "loss": 0.26375657320022583,
2193
- "step": 273,
2194
- "token_acc": 0.9051297163863579
2195
- },
2196
- {
2197
- "epoch": 1.0581113801452784,
2198
- "grad_norm": 0.20594951510429382,
2199
- "learning_rate": 7.834849888017979e-06,
2200
- "loss": 0.23789554834365845,
2201
- "step": 274,
2202
- "token_acc": 0.9142479611743739
2203
- },
2204
- {
2205
- "epoch": 1.0619854721549637,
2206
- "grad_norm": 0.23315519094467163,
2207
- "learning_rate": 7.817025327099574e-06,
2208
- "loss": 0.24684631824493408,
2209
- "step": 275,
2210
- "token_acc": 0.9110874200426439
2211
- },
2212
- {
2213
- "epoch": 1.0658595641646489,
2214
- "grad_norm": 0.2189839482307434,
2215
- "learning_rate": 7.799148161911013e-06,
2216
- "loss": 0.2684437334537506,
2217
- "step": 276,
2218
- "token_acc": 0.9041172254519392
2219
- },
2220
- {
2221
- "epoch": 1.0697336561743342,
2222
- "grad_norm": 0.21298226714134216,
2223
- "learning_rate": 7.781218726285014e-06,
2224
- "loss": 0.2720562815666199,
2225
- "step": 277,
2226
- "token_acc": 0.9027445373018297
2227
- },
2228
- {
2229
- "epoch": 1.0736077481840194,
2230
- "grad_norm": 0.21282611787319183,
2231
- "learning_rate": 7.763237355030384e-06,
2232
- "loss": 0.2579670548439026,
2233
- "step": 278,
2234
- "token_acc": 0.9080073119376767
2235
- },
2236
- {
2237
- "epoch": 1.0774818401937045,
2238
- "grad_norm": 0.21488887071609497,
2239
- "learning_rate": 7.745204383925753e-06,
2240
- "loss": 0.2742394804954529,
2241
- "step": 279,
2242
- "token_acc": 0.9015262545209174
2243
- },
2244
- {
2245
- "epoch": 1.0813559322033899,
2246
- "grad_norm": 0.19826629757881165,
2247
- "learning_rate": 7.727120149713313e-06,
2248
- "loss": 0.23731666803359985,
2249
- "step": 280,
2250
- "token_acc": 0.9146603883445988
2251
- },
2252
- {
2253
- "epoch": 1.085230024213075,
2254
- "grad_norm": 0.20840346813201904,
2255
- "learning_rate": 7.708984990092528e-06,
2256
- "loss": 0.22673961520195007,
2257
- "step": 281,
2258
- "token_acc": 0.9184409845576723
2259
- },
2260
- {
2261
- "epoch": 1.0891041162227604,
2262
- "grad_norm": 0.21199366450309753,
2263
- "learning_rate": 7.690799243713825e-06,
2264
- "loss": 0.2788952887058258,
2265
- "step": 282,
2266
- "token_acc": 0.9002122640890617
2267
- },
2268
- {
2269
- "epoch": 1.0929782082324455,
2270
- "grad_norm": 0.23963455855846405,
2271
- "learning_rate": 7.672563250172278e-06,
2272
- "loss": 0.2703215479850769,
2273
- "step": 283,
2274
- "token_acc": 0.902904561306835
2275
- },
2276
- {
2277
- "epoch": 1.0968523002421307,
2278
- "grad_norm": 0.20739565789699554,
2279
- "learning_rate": 7.654277350001255e-06,
2280
- "loss": 0.2556743621826172,
2281
- "step": 284,
2282
- "token_acc": 0.9087778504769448
2283
- },
2284
- {
2285
- "epoch": 1.100726392251816,
2286
- "grad_norm": 0.3205340504646301,
2287
- "learning_rate": 7.635941884666072e-06,
2288
- "loss": 0.2660865783691406,
2289
- "step": 285,
2290
- "token_acc": 0.9052546447746934
2291
- },
2292
- {
2293
- "epoch": 1.1046004842615011,
2294
- "grad_norm": 0.20611628890037537,
2295
- "learning_rate": 7.617557196557601e-06,
2296
- "loss": 0.2590142488479614,
2297
- "step": 286,
2298
- "token_acc": 0.9070821077566713
2299
- },
2300
- {
2301
- "epoch": 1.1084745762711865,
2302
- "grad_norm": 0.1932753622531891,
2303
- "learning_rate": 7.599123628985894e-06,
2304
- "loss": 0.2396095246076584,
2305
- "step": 287,
2306
- "token_acc": 0.9135842317299648
2307
- },
2308
- {
2309
- "epoch": 1.1123486682808716,
2310
- "grad_norm": 0.21151748299598694,
2311
- "learning_rate": 7.580641526173758e-06,
2312
- "loss": 0.2544936537742615,
2313
- "step": 288,
2314
- "token_acc": 0.9088854539111634
2315
- },
2316
- {
2317
- "epoch": 1.116222760290557,
2318
- "grad_norm": 0.1992950737476349,
2319
- "learning_rate": 7.5621112332503325e-06,
2320
- "loss": 0.2544850707054138,
2321
- "step": 289,
2322
- "token_acc": 0.9090426161294457
2323
- },
2324
- {
2325
- "epoch": 1.1200968523002421,
2326
- "grad_norm": 0.20908565819263458,
2327
- "learning_rate": 7.543533096244644e-06,
2328
- "loss": 0.2762412428855896,
2329
- "step": 290,
2330
- "token_acc": 0.9013541447063986
2331
- },
2332
- {
2333
- "epoch": 1.1239709443099273,
2334
- "grad_norm": 0.2157965451478958,
2335
- "learning_rate": 7.524907462079149e-06,
2336
- "loss": 0.25533056259155273,
2337
- "step": 291,
2338
- "token_acc": 0.9080176353704462
2339
- },
2340
- {
2341
- "epoch": 1.1278450363196126,
2342
- "grad_norm": 0.19141145050525665,
2343
- "learning_rate": 7.506234678563248e-06,
2344
- "loss": 0.2362717241048813,
2345
- "step": 292,
2346
- "token_acc": 0.9155038610363999
2347
- },
2348
- {
2349
- "epoch": 1.1317191283292978,
2350
- "grad_norm": 0.21533732116222382,
2351
- "learning_rate": 7.487515094386792e-06,
2352
- "loss": 0.23099368810653687,
2353
- "step": 293,
2354
- "token_acc": 0.9173202498403009
2355
- },
2356
- {
2357
- "epoch": 1.1355932203389831,
2358
- "grad_norm": 0.20129309594631195,
2359
- "learning_rate": 7.468749059113578e-06,
2360
- "loss": 0.26144838333129883,
2361
- "step": 294,
2362
- "token_acc": 0.9057641431815713
2363
- },
2364
- {
2365
- "epoch": 1.1394673123486683,
2366
- "grad_norm": 0.3953739404678345,
2367
- "learning_rate": 7.449936923174813e-06,
2368
- "loss": 0.2557257413864136,
2369
- "step": 295,
2370
- "token_acc": 0.9087617787160037
2371
- },
2372
- {
2373
- "epoch": 1.1433414043583534,
2374
- "grad_norm": 0.21214410662651062,
2375
- "learning_rate": 7.431079037862575e-06,
2376
- "loss": 0.27983057498931885,
2377
- "step": 296,
2378
- "token_acc": 0.8996573827559394
2379
- },
2380
- {
2381
- "epoch": 1.1472154963680388,
2382
- "grad_norm": 0.20280665159225464,
2383
- "learning_rate": 7.412175755323254e-06,
2384
- "loss": 0.2772400677204132,
2385
- "step": 297,
2386
- "token_acc": 0.9010093723967251
2387
- },
2388
- {
2389
- "epoch": 1.151089588377724,
2390
- "grad_norm": 0.21776501834392548,
2391
- "learning_rate": 7.39322742855097e-06,
2392
- "loss": 0.24517808854579926,
2393
- "step": 298,
2394
- "token_acc": 0.9120538077359621
2395
- },
2396
- {
2397
- "epoch": 1.1549636803874093,
2398
- "grad_norm": 0.21630938351154327,
2399
- "learning_rate": 7.374234411380987e-06,
2400
- "loss": 0.2736694812774658,
2401
- "step": 299,
2402
- "token_acc": 0.9020631116999458
2403
- },
2404
- {
2405
- "epoch": 1.1588377723970944,
2406
- "grad_norm": 0.19338402152061462,
2407
- "learning_rate": 7.355197058483103e-06,
2408
- "loss": 0.24092288315296173,
2409
- "step": 300,
2410
- "token_acc": 0.9133508019967492
2411
- }
2412
- ],
2413
- "logging_steps": 1,
2414
- "max_steps": 777,
2415
- "num_input_tokens_seen": 0,
2416
- "num_train_epochs": 3,
2417
- "save_steps": 50,
2418
- "stateful_callbacks": {
2419
- "TrainerControl": {
2420
- "args": {
2421
- "should_epoch_stop": false,
2422
- "should_evaluate": false,
2423
- "should_log": false,
2424
- "should_save": true,
2425
- "should_training_stop": false
2426
- },
2427
- "attributes": {}
2428
- }
2429
- },
2430
- "total_flos": 1.1233780174946304e+16,
2431
- "train_batch_size": 1,
2432
- "trial_name": null,
2433
- "trial_params": null
2434
- }