SpireLab commited on
Commit
905e107
·
verified ·
1 Parent(s): ee5e715

Delete train_data/muril_mag_domain/checkpoint-42618/trainer_state.json with huggingface_hub

Browse files
train_data/muril_mag_domain/checkpoint-42618/trainer_state.json DELETED
@@ -1,3610 +0,0 @@
1
- {
2
- "best_metric": null,
3
- "best_model_checkpoint": null,
4
- "epoch": 2.999894414528561,
5
- "eval_steps": 500,
6
- "global_step": 42618,
7
- "is_hyper_param_search": false,
8
- "is_local_process_zero": true,
9
- "is_world_process_zero": true,
10
- "log_history": [
11
- {
12
- "epoch": 0.007039031429275332,
13
- "grad_norm": 36.44467544555664,
14
- "learning_rate": 1.1731581417175035e-06,
15
- "loss": 6.2071,
16
- "step": 100
17
- },
18
- {
19
- "epoch": 0.014078062858550663,
20
- "grad_norm": 43.126609802246094,
21
- "learning_rate": 2.346316283435007e-06,
22
- "loss": 5.9231,
23
- "step": 200
24
- },
25
- {
26
- "epoch": 0.021117094287825995,
27
- "grad_norm": 29.171634674072266,
28
- "learning_rate": 3.5194744251525106e-06,
29
- "loss": 5.4145,
30
- "step": 300
31
- },
32
- {
33
- "epoch": 0.028156125717101327,
34
- "grad_norm": 38.217105865478516,
35
- "learning_rate": 4.692632566870014e-06,
36
- "loss": 4.9149,
37
- "step": 400
38
- },
39
- {
40
- "epoch": 0.03519515714637666,
41
- "grad_norm": 35.40254211425781,
42
- "learning_rate": 5.865790708587518e-06,
43
- "loss": 4.5052,
44
- "step": 500
45
- },
46
- {
47
- "epoch": 0.03519515714637666,
48
- "eval_runtime": 191.3754,
49
- "eval_samples_per_second": 148.467,
50
- "eval_steps_per_second": 18.56,
51
- "step": 500
52
- },
53
- {
54
- "epoch": 0.04223418857565199,
55
- "grad_norm": 36.732643127441406,
56
- "learning_rate": 7.038948850305021e-06,
57
- "loss": 4.3715,
58
- "step": 600
59
- },
60
- {
61
- "epoch": 0.04927322000492732,
62
- "grad_norm": 36.27021408081055,
63
- "learning_rate": 8.212106992022525e-06,
64
- "loss": 4.3269,
65
- "step": 700
66
- },
67
- {
68
- "epoch": 0.056312251434202654,
69
- "grad_norm": 42.45858383178711,
70
- "learning_rate": 9.385265133740028e-06,
71
- "loss": 4.0589,
72
- "step": 800
73
- },
74
- {
75
- "epoch": 0.06335128286347799,
76
- "grad_norm": 42.08483123779297,
77
- "learning_rate": 1.0558423275457532e-05,
78
- "loss": 4.1336,
79
- "step": 900
80
- },
81
- {
82
- "epoch": 0.07039031429275332,
83
- "grad_norm": 42.23253631591797,
84
- "learning_rate": 1.1731581417175035e-05,
85
- "loss": 4.0719,
86
- "step": 1000
87
- },
88
- {
89
- "epoch": 0.07039031429275332,
90
- "eval_runtime": 193.8033,
91
- "eval_samples_per_second": 146.607,
92
- "eval_steps_per_second": 18.328,
93
- "step": 1000
94
- },
95
- {
96
- "epoch": 0.07742934572202866,
97
- "grad_norm": 48.905662536621094,
98
- "learning_rate": 1.2904739558892539e-05,
99
- "loss": 3.8613,
100
- "step": 1100
101
- },
102
- {
103
- "epoch": 0.08446837715130398,
104
- "grad_norm": 37.9277458190918,
105
- "learning_rate": 1.4077897700610042e-05,
106
- "loss": 3.8424,
107
- "step": 1200
108
- },
109
- {
110
- "epoch": 0.09150740858057932,
111
- "grad_norm": 48.82701110839844,
112
- "learning_rate": 1.5251055842327546e-05,
113
- "loss": 3.7771,
114
- "step": 1300
115
- },
116
- {
117
- "epoch": 0.09854644000985464,
118
- "grad_norm": 33.38028335571289,
119
- "learning_rate": 1.642421398404505e-05,
120
- "loss": 3.8094,
121
- "step": 1400
122
- },
123
- {
124
- "epoch": 0.10558547143912998,
125
- "grad_norm": 61.35352325439453,
126
- "learning_rate": 1.7597372125762555e-05,
127
- "loss": 3.8331,
128
- "step": 1500
129
- },
130
- {
131
- "epoch": 0.10558547143912998,
132
- "eval_runtime": 191.8384,
133
- "eval_samples_per_second": 148.109,
134
- "eval_steps_per_second": 18.516,
135
- "step": 1500
136
- },
137
- {
138
- "epoch": 0.11262450286840531,
139
- "grad_norm": 46.74394226074219,
140
- "learning_rate": 1.8770530267480057e-05,
141
- "loss": 3.6822,
142
- "step": 1600
143
- },
144
- {
145
- "epoch": 0.11966353429768065,
146
- "grad_norm": 35.53325271606445,
147
- "learning_rate": 1.9943688409197562e-05,
148
- "loss": 3.6282,
149
- "step": 1700
150
- },
151
- {
152
- "epoch": 0.12670256572695598,
153
- "grad_norm": 37.73524856567383,
154
- "learning_rate": 2.1116846550915064e-05,
155
- "loss": 3.5722,
156
- "step": 1800
157
- },
158
- {
159
- "epoch": 0.1337415971562313,
160
- "grad_norm": 33.76814651489258,
161
- "learning_rate": 2.229000469263257e-05,
162
- "loss": 3.6086,
163
- "step": 1900
164
- },
165
- {
166
- "epoch": 0.14078062858550663,
167
- "grad_norm": 41.888282775878906,
168
- "learning_rate": 2.346316283435007e-05,
169
- "loss": 3.6142,
170
- "step": 2000
171
- },
172
- {
173
- "epoch": 0.14078062858550663,
174
- "eval_runtime": 191.5815,
175
- "eval_samples_per_second": 148.308,
176
- "eval_steps_per_second": 18.54,
177
- "step": 2000
178
- },
179
- {
180
- "epoch": 0.14781966001478197,
181
- "grad_norm": 39.62664031982422,
182
- "learning_rate": 2.4636320976067576e-05,
183
- "loss": 3.6029,
184
- "step": 2100
185
- },
186
- {
187
- "epoch": 0.1548586914440573,
188
- "grad_norm": 38.377532958984375,
189
- "learning_rate": 2.5809479117785078e-05,
190
- "loss": 3.4959,
191
- "step": 2200
192
- },
193
- {
194
- "epoch": 0.16189772287333262,
195
- "grad_norm": 32.66987991333008,
196
- "learning_rate": 2.698263725950258e-05,
197
- "loss": 3.5252,
198
- "step": 2300
199
- },
200
- {
201
- "epoch": 0.16893675430260796,
202
- "grad_norm": 39.213592529296875,
203
- "learning_rate": 2.8155795401220085e-05,
204
- "loss": 3.5859,
205
- "step": 2400
206
- },
207
- {
208
- "epoch": 0.1759757857318833,
209
- "grad_norm": 31.646276473999023,
210
- "learning_rate": 2.9328953542937587e-05,
211
- "loss": 3.4995,
212
- "step": 2500
213
- },
214
- {
215
- "epoch": 0.1759757857318833,
216
- "eval_runtime": 194.3308,
217
- "eval_samples_per_second": 146.209,
218
- "eval_steps_per_second": 18.278,
219
- "step": 2500
220
- },
221
- {
222
- "epoch": 0.18301481716115864,
223
- "grad_norm": 32.30677032470703,
224
- "learning_rate": 3.0502111684655092e-05,
225
- "loss": 3.5853,
226
- "step": 2600
227
- },
228
- {
229
- "epoch": 0.19005384859043395,
230
- "grad_norm": 31.175769805908203,
231
- "learning_rate": 3.1675269826372594e-05,
232
- "loss": 3.5134,
233
- "step": 2700
234
- },
235
- {
236
- "epoch": 0.1970928800197093,
237
- "grad_norm": 31.389162063598633,
238
- "learning_rate": 3.28484279680901e-05,
239
- "loss": 3.4909,
240
- "step": 2800
241
- },
242
- {
243
- "epoch": 0.20413191144898463,
244
- "grad_norm": 33.105369567871094,
245
- "learning_rate": 3.4021586109807604e-05,
246
- "loss": 3.4099,
247
- "step": 2900
248
- },
249
- {
250
- "epoch": 0.21117094287825997,
251
- "grad_norm": 25.477977752685547,
252
- "learning_rate": 3.519474425152511e-05,
253
- "loss": 3.3823,
254
- "step": 3000
255
- },
256
- {
257
- "epoch": 0.21117094287825997,
258
- "eval_runtime": 194.3983,
259
- "eval_samples_per_second": 146.159,
260
- "eval_steps_per_second": 18.272,
261
- "step": 3000
262
- },
263
- {
264
- "epoch": 0.21820997430753528,
265
- "grad_norm": 29.61454200744629,
266
- "learning_rate": 3.636790239324261e-05,
267
- "loss": 3.3476,
268
- "step": 3100
269
- },
270
- {
271
- "epoch": 0.22524900573681061,
272
- "grad_norm": 26.82366180419922,
273
- "learning_rate": 3.754106053496011e-05,
274
- "loss": 3.389,
275
- "step": 3200
276
- },
277
- {
278
- "epoch": 0.23228803716608595,
279
- "grad_norm": 26.6168155670166,
280
- "learning_rate": 3.871421867667762e-05,
281
- "loss": 3.3712,
282
- "step": 3300
283
- },
284
- {
285
- "epoch": 0.2393270685953613,
286
- "grad_norm": 24.504793167114258,
287
- "learning_rate": 3.9887376818395124e-05,
288
- "loss": 3.2693,
289
- "step": 3400
290
- },
291
- {
292
- "epoch": 0.2463661000246366,
293
- "grad_norm": 22.34451675415039,
294
- "learning_rate": 4.106053496011262e-05,
295
- "loss": 3.3719,
296
- "step": 3500
297
- },
298
- {
299
- "epoch": 0.2463661000246366,
300
- "eval_runtime": 192.2522,
301
- "eval_samples_per_second": 147.79,
302
- "eval_steps_per_second": 18.476,
303
- "step": 3500
304
- },
305
- {
306
- "epoch": 0.25340513145391197,
307
- "grad_norm": 30.370140075683594,
308
- "learning_rate": 4.223369310183013e-05,
309
- "loss": 3.3216,
310
- "step": 3600
311
- },
312
- {
313
- "epoch": 0.2604441628831873,
314
- "grad_norm": 29.111398696899414,
315
- "learning_rate": 4.340685124354763e-05,
316
- "loss": 3.3085,
317
- "step": 3700
318
- },
319
- {
320
- "epoch": 0.2674831943124626,
321
- "grad_norm": 29.50999641418457,
322
- "learning_rate": 4.458000938526514e-05,
323
- "loss": 3.2907,
324
- "step": 3800
325
- },
326
- {
327
- "epoch": 0.27452222574173796,
328
- "grad_norm": 21.999244689941406,
329
- "learning_rate": 4.5753167526982636e-05,
330
- "loss": 3.2173,
331
- "step": 3900
332
- },
333
- {
334
- "epoch": 0.28156125717101327,
335
- "grad_norm": 28.0905818939209,
336
- "learning_rate": 4.692632566870014e-05,
337
- "loss": 3.3431,
338
- "step": 4000
339
- },
340
- {
341
- "epoch": 0.28156125717101327,
342
- "eval_runtime": 192.489,
343
- "eval_samples_per_second": 147.608,
344
- "eval_steps_per_second": 18.453,
345
- "step": 4000
346
- },
347
- {
348
- "epoch": 0.2886002886002886,
349
- "grad_norm": 27.252222061157227,
350
- "learning_rate": 4.809948381041765e-05,
351
- "loss": 3.4265,
352
- "step": 4100
353
- },
354
- {
355
- "epoch": 0.29563932002956395,
356
- "grad_norm": 20.001508712768555,
357
- "learning_rate": 4.927264195213515e-05,
358
- "loss": 3.2489,
359
- "step": 4200
360
- },
361
- {
362
- "epoch": 0.30267835145883926,
363
- "grad_norm": 24.947546005249023,
364
- "learning_rate": 4.995046407341746e-05,
365
- "loss": 3.2957,
366
- "step": 4300
367
- },
368
- {
369
- "epoch": 0.3097173828881146,
370
- "grad_norm": 18.58955192565918,
371
- "learning_rate": 4.982010637188445e-05,
372
- "loss": 3.2328,
373
- "step": 4400
374
- },
375
- {
376
- "epoch": 0.31675641431738993,
377
- "grad_norm": 22.946285247802734,
378
- "learning_rate": 4.968974867035145e-05,
379
- "loss": 3.177,
380
- "step": 4500
381
- },
382
- {
383
- "epoch": 0.31675641431738993,
384
- "eval_runtime": 192.63,
385
- "eval_samples_per_second": 147.5,
386
- "eval_steps_per_second": 18.439,
387
- "step": 4500
388
- },
389
- {
390
- "epoch": 0.32379544574666524,
391
- "grad_norm": 20.17714500427246,
392
- "learning_rate": 4.955939096881844e-05,
393
- "loss": 3.3241,
394
- "step": 4600
395
- },
396
- {
397
- "epoch": 0.3308344771759406,
398
- "grad_norm": 18.580751419067383,
399
- "learning_rate": 4.9429033267285435e-05,
400
- "loss": 3.3084,
401
- "step": 4700
402
- },
403
- {
404
- "epoch": 0.3378735086052159,
405
- "grad_norm": 16.068750381469727,
406
- "learning_rate": 4.929867556575243e-05,
407
- "loss": 3.1674,
408
- "step": 4800
409
- },
410
- {
411
- "epoch": 0.34491254003449123,
412
- "grad_norm": 23.636520385742188,
413
- "learning_rate": 4.916831786421942e-05,
414
- "loss": 3.2698,
415
- "step": 4900
416
- },
417
- {
418
- "epoch": 0.3519515714637666,
419
- "grad_norm": 19.445907592773438,
420
- "learning_rate": 4.903796016268641e-05,
421
- "loss": 3.2824,
422
- "step": 5000
423
- },
424
- {
425
- "epoch": 0.3519515714637666,
426
- "eval_runtime": 192.779,
427
- "eval_samples_per_second": 147.386,
428
- "eval_steps_per_second": 18.425,
429
- "step": 5000
430
- },
431
- {
432
- "epoch": 0.3589906028930419,
433
- "grad_norm": 20.217737197875977,
434
- "learning_rate": 4.8907602461153405e-05,
435
- "loss": 3.3673,
436
- "step": 5100
437
- },
438
- {
439
- "epoch": 0.3660296343223173,
440
- "grad_norm": 16.03109359741211,
441
- "learning_rate": 4.87772447596204e-05,
442
- "loss": 3.3105,
443
- "step": 5200
444
- },
445
- {
446
- "epoch": 0.3730686657515926,
447
- "grad_norm": 21.388507843017578,
448
- "learning_rate": 4.864688705808739e-05,
449
- "loss": 3.1998,
450
- "step": 5300
451
- },
452
- {
453
- "epoch": 0.3801076971808679,
454
- "grad_norm": 16.931922912597656,
455
- "learning_rate": 4.851652935655439e-05,
456
- "loss": 3.2054,
457
- "step": 5400
458
- },
459
- {
460
- "epoch": 0.38714672861014326,
461
- "grad_norm": 24.145727157592773,
462
- "learning_rate": 4.838617165502138e-05,
463
- "loss": 3.0802,
464
- "step": 5500
465
- },
466
- {
467
- "epoch": 0.38714672861014326,
468
- "eval_runtime": 192.7089,
469
- "eval_samples_per_second": 147.44,
470
- "eval_steps_per_second": 18.432,
471
- "step": 5500
472
- },
473
- {
474
- "epoch": 0.3941857600394186,
475
- "grad_norm": 24.199636459350586,
476
- "learning_rate": 4.8255813953488375e-05,
477
- "loss": 3.0688,
478
- "step": 5600
479
- },
480
- {
481
- "epoch": 0.4012247914686939,
482
- "grad_norm": 18.878751754760742,
483
- "learning_rate": 4.812545625195537e-05,
484
- "loss": 3.1667,
485
- "step": 5700
486
- },
487
- {
488
- "epoch": 0.40826382289796925,
489
- "grad_norm": 17.39369010925293,
490
- "learning_rate": 4.7995098550422363e-05,
491
- "loss": 3.1905,
492
- "step": 5800
493
- },
494
- {
495
- "epoch": 0.41530285432724456,
496
- "grad_norm": 19.778793334960938,
497
- "learning_rate": 4.786474084888936e-05,
498
- "loss": 3.0883,
499
- "step": 5900
500
- },
501
- {
502
- "epoch": 0.42234188575651993,
503
- "grad_norm": 14.80075454711914,
504
- "learning_rate": 4.773438314735635e-05,
505
- "loss": 3.0839,
506
- "step": 6000
507
- },
508
- {
509
- "epoch": 0.42234188575651993,
510
- "eval_runtime": 192.7353,
511
- "eval_samples_per_second": 147.42,
512
- "eval_steps_per_second": 18.429,
513
- "step": 6000
514
- },
515
- {
516
- "epoch": 0.42938091718579524,
517
- "grad_norm": 27.43608856201172,
518
- "learning_rate": 4.760402544582334e-05,
519
- "loss": 3.0621,
520
- "step": 6100
521
- },
522
- {
523
- "epoch": 0.43641994861507055,
524
- "grad_norm": 18.219221115112305,
525
- "learning_rate": 4.747366774429033e-05,
526
- "loss": 3.1461,
527
- "step": 6200
528
- },
529
- {
530
- "epoch": 0.4434589800443459,
531
- "grad_norm": 17.97977638244629,
532
- "learning_rate": 4.734331004275733e-05,
533
- "loss": 3.0795,
534
- "step": 6300
535
- },
536
- {
537
- "epoch": 0.45049801147362123,
538
- "grad_norm": 21.358592987060547,
539
- "learning_rate": 4.721295234122432e-05,
540
- "loss": 3.1361,
541
- "step": 6400
542
- },
543
- {
544
- "epoch": 0.45753704290289654,
545
- "grad_norm": 15.679008483886719,
546
- "learning_rate": 4.7082594639691315e-05,
547
- "loss": 3.0751,
548
- "step": 6500
549
- },
550
- {
551
- "epoch": 0.45753704290289654,
552
- "eval_runtime": 193.5839,
553
- "eval_samples_per_second": 146.774,
554
- "eval_steps_per_second": 18.349,
555
- "step": 6500
556
- },
557
- {
558
- "epoch": 0.4645760743321719,
559
- "grad_norm": 14.288241386413574,
560
- "learning_rate": 4.69522369381583e-05,
561
- "loss": 3.0886,
562
- "step": 6600
563
- },
564
- {
565
- "epoch": 0.4716151057614472,
566
- "grad_norm": 15.016201972961426,
567
- "learning_rate": 4.6821879236625304e-05,
568
- "loss": 2.973,
569
- "step": 6700
570
- },
571
- {
572
- "epoch": 0.4786541371907226,
573
- "grad_norm": 20.513479232788086,
574
- "learning_rate": 4.66915215350923e-05,
575
- "loss": 3.1001,
576
- "step": 6800
577
- },
578
- {
579
- "epoch": 0.4856931686199979,
580
- "grad_norm": 15.093891143798828,
581
- "learning_rate": 4.656116383355929e-05,
582
- "loss": 3.1073,
583
- "step": 6900
584
- },
585
- {
586
- "epoch": 0.4927322000492732,
587
- "grad_norm": 15.074331283569336,
588
- "learning_rate": 4.6430806132026286e-05,
589
- "loss": 3.1247,
590
- "step": 7000
591
- },
592
- {
593
- "epoch": 0.4927322000492732,
594
- "eval_runtime": 193.2211,
595
- "eval_samples_per_second": 147.049,
596
- "eval_steps_per_second": 18.383,
597
- "step": 7000
598
- },
599
- {
600
- "epoch": 0.49977123147854857,
601
- "grad_norm": 19.936674118041992,
602
- "learning_rate": 4.630044843049327e-05,
603
- "loss": 3.0331,
604
- "step": 7100
605
- },
606
- {
607
- "epoch": 0.5068102629078239,
608
- "grad_norm": 22.168909072875977,
609
- "learning_rate": 4.617009072896027e-05,
610
- "loss": 3.0611,
611
- "step": 7200
612
- },
613
- {
614
- "epoch": 0.5138492943370992,
615
- "grad_norm": 19.541671752929688,
616
- "learning_rate": 4.603973302742726e-05,
617
- "loss": 2.9378,
618
- "step": 7300
619
- },
620
- {
621
- "epoch": 0.5208883257663746,
622
- "grad_norm": 16.111820220947266,
623
- "learning_rate": 4.5909375325894256e-05,
624
- "loss": 3.0461,
625
- "step": 7400
626
- },
627
- {
628
- "epoch": 0.5279273571956499,
629
- "grad_norm": 18.88793182373047,
630
- "learning_rate": 4.577901762436125e-05,
631
- "loss": 3.0183,
632
- "step": 7500
633
- },
634
- {
635
- "epoch": 0.5279273571956499,
636
- "eval_runtime": 191.8152,
637
- "eval_samples_per_second": 148.127,
638
- "eval_steps_per_second": 18.518,
639
- "step": 7500
640
- },
641
- {
642
- "epoch": 0.5349663886249252,
643
- "grad_norm": 15.234626770019531,
644
- "learning_rate": 4.5648659922828244e-05,
645
- "loss": 3.0287,
646
- "step": 7600
647
- },
648
- {
649
- "epoch": 0.5420054200542005,
650
- "grad_norm": 16.836956024169922,
651
- "learning_rate": 4.551830222129523e-05,
652
- "loss": 2.9991,
653
- "step": 7700
654
- },
655
- {
656
- "epoch": 0.5490444514834759,
657
- "grad_norm": 20.984453201293945,
658
- "learning_rate": 4.538794451976223e-05,
659
- "loss": 3.0765,
660
- "step": 7800
661
- },
662
- {
663
- "epoch": 0.5560834829127512,
664
- "grad_norm": 18.113004684448242,
665
- "learning_rate": 4.5257586818229226e-05,
666
- "loss": 2.9704,
667
- "step": 7900
668
- },
669
- {
670
- "epoch": 0.5631225143420265,
671
- "grad_norm": 20.654579162597656,
672
- "learning_rate": 4.512722911669622e-05,
673
- "loss": 3.0604,
674
- "step": 8000
675
- },
676
- {
677
- "epoch": 0.5631225143420265,
678
- "eval_runtime": 193.8004,
679
- "eval_samples_per_second": 146.61,
680
- "eval_steps_per_second": 18.328,
681
- "step": 8000
682
- },
683
- {
684
- "epoch": 0.5701615457713018,
685
- "grad_norm": 18.815162658691406,
686
- "learning_rate": 4.499687141516321e-05,
687
- "loss": 2.8726,
688
- "step": 8100
689
- },
690
- {
691
- "epoch": 0.5772005772005772,
692
- "grad_norm": 15.366965293884277,
693
- "learning_rate": 4.48665137136302e-05,
694
- "loss": 3.0509,
695
- "step": 8200
696
- },
697
- {
698
- "epoch": 0.5842396086298526,
699
- "grad_norm": 17.203136444091797,
700
- "learning_rate": 4.4736156012097196e-05,
701
- "loss": 2.9718,
702
- "step": 8300
703
- },
704
- {
705
- "epoch": 0.5912786400591279,
706
- "grad_norm": 15.31092643737793,
707
- "learning_rate": 4.460579831056419e-05,
708
- "loss": 3.0382,
709
- "step": 8400
710
- },
711
- {
712
- "epoch": 0.5983176714884032,
713
- "grad_norm": 12.381194114685059,
714
- "learning_rate": 4.4475440609031184e-05,
715
- "loss": 2.9945,
716
- "step": 8500
717
- },
718
- {
719
- "epoch": 0.5983176714884032,
720
- "eval_runtime": 192.0001,
721
- "eval_samples_per_second": 147.984,
722
- "eval_steps_per_second": 18.5,
723
- "step": 8500
724
- },
725
- {
726
- "epoch": 0.6053567029176785,
727
- "grad_norm": 13.362293243408203,
728
- "learning_rate": 4.434508290749818e-05,
729
- "loss": 3.0427,
730
- "step": 8600
731
- },
732
- {
733
- "epoch": 0.6123957343469538,
734
- "grad_norm": 18.495370864868164,
735
- "learning_rate": 4.4214725205965165e-05,
736
- "loss": 3.0363,
737
- "step": 8700
738
- },
739
- {
740
- "epoch": 0.6194347657762292,
741
- "grad_norm": 16.98731803894043,
742
- "learning_rate": 4.408436750443216e-05,
743
- "loss": 2.9257,
744
- "step": 8800
745
- },
746
- {
747
- "epoch": 0.6264737972055046,
748
- "grad_norm": 18.510801315307617,
749
- "learning_rate": 4.395400980289916e-05,
750
- "loss": 3.0014,
751
- "step": 8900
752
- },
753
- {
754
- "epoch": 0.6335128286347799,
755
- "grad_norm": 15.024514198303223,
756
- "learning_rate": 4.3823652101366155e-05,
757
- "loss": 3.0435,
758
- "step": 9000
759
- },
760
- {
761
- "epoch": 0.6335128286347799,
762
- "eval_runtime": 192.5991,
763
- "eval_samples_per_second": 147.524,
764
- "eval_steps_per_second": 18.442,
765
- "step": 9000
766
- },
767
- {
768
- "epoch": 0.6405518600640552,
769
- "grad_norm": 19.763547897338867,
770
- "learning_rate": 4.369329439983315e-05,
771
- "loss": 2.9219,
772
- "step": 9100
773
- },
774
- {
775
- "epoch": 0.6475908914933305,
776
- "grad_norm": 19.536087036132812,
777
- "learning_rate": 4.3562936698300136e-05,
778
- "loss": 2.9749,
779
- "step": 9200
780
- },
781
- {
782
- "epoch": 0.6546299229226058,
783
- "grad_norm": 15.909920692443848,
784
- "learning_rate": 4.343257899676713e-05,
785
- "loss": 2.8793,
786
- "step": 9300
787
- },
788
- {
789
- "epoch": 0.6616689543518812,
790
- "grad_norm": 17.25068473815918,
791
- "learning_rate": 4.3302221295234124e-05,
792
- "loss": 2.9669,
793
- "step": 9400
794
- },
795
- {
796
- "epoch": 0.6687079857811565,
797
- "grad_norm": 14.32239055633545,
798
- "learning_rate": 4.317186359370112e-05,
799
- "loss": 2.9066,
800
- "step": 9500
801
- },
802
- {
803
- "epoch": 0.6687079857811565,
804
- "eval_runtime": 191.7836,
805
- "eval_samples_per_second": 148.151,
806
- "eval_steps_per_second": 18.521,
807
- "step": 9500
808
- },
809
- {
810
- "epoch": 0.6757470172104318,
811
- "grad_norm": 22.611879348754883,
812
- "learning_rate": 4.304150589216811e-05,
813
- "loss": 2.8333,
814
- "step": 9600
815
- },
816
- {
817
- "epoch": 0.6827860486397072,
818
- "grad_norm": 15.549399375915527,
819
- "learning_rate": 4.2911148190635107e-05,
820
- "loss": 2.9913,
821
- "step": 9700
822
- },
823
- {
824
- "epoch": 0.6898250800689825,
825
- "grad_norm": 13.915739059448242,
826
- "learning_rate": 4.2780790489102094e-05,
827
- "loss": 2.9524,
828
- "step": 9800
829
- },
830
- {
831
- "epoch": 0.6968641114982579,
832
- "grad_norm": 17.621822357177734,
833
- "learning_rate": 4.265043278756909e-05,
834
- "loss": 2.9844,
835
- "step": 9900
836
- },
837
- {
838
- "epoch": 0.7039031429275332,
839
- "grad_norm": 22.748342514038086,
840
- "learning_rate": 4.252007508603609e-05,
841
- "loss": 2.9083,
842
- "step": 10000
843
- },
844
- {
845
- "epoch": 0.7039031429275332,
846
- "eval_runtime": 193.6184,
847
- "eval_samples_per_second": 146.747,
848
- "eval_steps_per_second": 18.345,
849
- "step": 10000
850
- },
851
- {
852
- "epoch": 0.7109421743568085,
853
- "grad_norm": 18.526578903198242,
854
- "learning_rate": 4.238971738450308e-05,
855
- "loss": 2.9706,
856
- "step": 10100
857
- },
858
- {
859
- "epoch": 0.7179812057860838,
860
- "grad_norm": 15.676709175109863,
861
- "learning_rate": 4.225935968297007e-05,
862
- "loss": 2.8265,
863
- "step": 10200
864
- },
865
- {
866
- "epoch": 0.7250202372153591,
867
- "grad_norm": 18.21067237854004,
868
- "learning_rate": 4.2129001981437064e-05,
869
- "loss": 2.8592,
870
- "step": 10300
871
- },
872
- {
873
- "epoch": 0.7320592686446346,
874
- "grad_norm": 22.188024520874023,
875
- "learning_rate": 4.199864427990406e-05,
876
- "loss": 2.8158,
877
- "step": 10400
878
- },
879
- {
880
- "epoch": 0.7390983000739099,
881
- "grad_norm": 18.406801223754883,
882
- "learning_rate": 4.186828657837105e-05,
883
- "loss": 2.9264,
884
- "step": 10500
885
- },
886
- {
887
- "epoch": 0.7390983000739099,
888
- "eval_runtime": 193.9371,
889
- "eval_samples_per_second": 146.506,
890
- "eval_steps_per_second": 18.315,
891
- "step": 10500
892
- },
893
- {
894
- "epoch": 0.7461373315031852,
895
- "grad_norm": 20.65268325805664,
896
- "learning_rate": 4.173792887683805e-05,
897
- "loss": 2.9882,
898
- "step": 10600
899
- },
900
- {
901
- "epoch": 0.7531763629324605,
902
- "grad_norm": 12.223052024841309,
903
- "learning_rate": 4.160757117530504e-05,
904
- "loss": 2.8969,
905
- "step": 10700
906
- },
907
- {
908
- "epoch": 0.7602153943617358,
909
- "grad_norm": 18.218887329101562,
910
- "learning_rate": 4.147721347377203e-05,
911
- "loss": 2.8254,
912
- "step": 10800
913
- },
914
- {
915
- "epoch": 0.7672544257910111,
916
- "grad_norm": 19.009950637817383,
917
- "learning_rate": 4.134685577223902e-05,
918
- "loss": 2.9361,
919
- "step": 10900
920
- },
921
- {
922
- "epoch": 0.7742934572202865,
923
- "grad_norm": 33.881927490234375,
924
- "learning_rate": 4.1216498070706016e-05,
925
- "loss": 2.8528,
926
- "step": 11000
927
- },
928
- {
929
- "epoch": 0.7742934572202865,
930
- "eval_runtime": 194.2555,
931
- "eval_samples_per_second": 146.266,
932
- "eval_steps_per_second": 18.285,
933
- "step": 11000
934
- },
935
- {
936
- "epoch": 0.7813324886495618,
937
- "grad_norm": 19.02928924560547,
938
- "learning_rate": 4.108614036917302e-05,
939
- "loss": 2.9383,
940
- "step": 11100
941
- },
942
- {
943
- "epoch": 0.7883715200788372,
944
- "grad_norm": 18.154483795166016,
945
- "learning_rate": 4.095578266764001e-05,
946
- "loss": 2.7691,
947
- "step": 11200
948
- },
949
- {
950
- "epoch": 0.7954105515081125,
951
- "grad_norm": 13.669476509094238,
952
- "learning_rate": 4.0825424966107e-05,
953
- "loss": 2.8306,
954
- "step": 11300
955
- },
956
- {
957
- "epoch": 0.8024495829373878,
958
- "grad_norm": 16.23528289794922,
959
- "learning_rate": 4.069506726457399e-05,
960
- "loss": 2.8588,
961
- "step": 11400
962
- },
963
- {
964
- "epoch": 0.8094886143666632,
965
- "grad_norm": 16.63111686706543,
966
- "learning_rate": 4.056470956304099e-05,
967
- "loss": 2.91,
968
- "step": 11500
969
- },
970
- {
971
- "epoch": 0.8094886143666632,
972
- "eval_runtime": 193.9814,
973
- "eval_samples_per_second": 146.473,
974
- "eval_steps_per_second": 18.311,
975
- "step": 11500
976
- },
977
- {
978
- "epoch": 0.8165276457959385,
979
- "grad_norm": 19.989736557006836,
980
- "learning_rate": 4.043435186150798e-05,
981
- "loss": 2.8754,
982
- "step": 11600
983
- },
984
- {
985
- "epoch": 0.8235666772252138,
986
- "grad_norm": 15.608553886413574,
987
- "learning_rate": 4.0303994159974975e-05,
988
- "loss": 2.8896,
989
- "step": 11700
990
- },
991
- {
992
- "epoch": 0.8306057086544891,
993
- "grad_norm": 17.236600875854492,
994
- "learning_rate": 4.017363645844196e-05,
995
- "loss": 2.8897,
996
- "step": 11800
997
- },
998
- {
999
- "epoch": 0.8376447400837644,
1000
- "grad_norm": 16.03377342224121,
1001
- "learning_rate": 4.0043278756908957e-05,
1002
- "loss": 2.8936,
1003
- "step": 11900
1004
- },
1005
- {
1006
- "epoch": 0.8446837715130399,
1007
- "grad_norm": 25.3082332611084,
1008
- "learning_rate": 3.991292105537595e-05,
1009
- "loss": 2.8939,
1010
- "step": 12000
1011
- },
1012
- {
1013
- "epoch": 0.8446837715130399,
1014
- "eval_runtime": 192.6986,
1015
- "eval_samples_per_second": 147.448,
1016
- "eval_steps_per_second": 18.433,
1017
- "step": 12000
1018
- },
1019
- {
1020
- "epoch": 0.8517228029423152,
1021
- "grad_norm": 18.766387939453125,
1022
- "learning_rate": 3.9782563353842945e-05,
1023
- "loss": 2.9448,
1024
- "step": 12100
1025
- },
1026
- {
1027
- "epoch": 0.8587618343715905,
1028
- "grad_norm": 17.019485473632812,
1029
- "learning_rate": 3.9652205652309946e-05,
1030
- "loss": 2.7899,
1031
- "step": 12200
1032
- },
1033
- {
1034
- "epoch": 0.8658008658008658,
1035
- "grad_norm": 15.821990966796875,
1036
- "learning_rate": 3.952184795077693e-05,
1037
- "loss": 2.8069,
1038
- "step": 12300
1039
- },
1040
- {
1041
- "epoch": 0.8728398972301411,
1042
- "grad_norm": 19.532939910888672,
1043
- "learning_rate": 3.939149024924393e-05,
1044
- "loss": 2.8797,
1045
- "step": 12400
1046
- },
1047
- {
1048
- "epoch": 0.8798789286594164,
1049
- "grad_norm": 15.134819030761719,
1050
- "learning_rate": 3.926113254771092e-05,
1051
- "loss": 2.89,
1052
- "step": 12500
1053
- },
1054
- {
1055
- "epoch": 0.8798789286594164,
1056
- "eval_runtime": 191.6112,
1057
- "eval_samples_per_second": 148.285,
1058
- "eval_steps_per_second": 18.538,
1059
- "step": 12500
1060
- },
1061
- {
1062
- "epoch": 0.8869179600886918,
1063
- "grad_norm": 15.088654518127441,
1064
- "learning_rate": 3.9130774846177915e-05,
1065
- "loss": 2.8519,
1066
- "step": 12600
1067
- },
1068
- {
1069
- "epoch": 0.8939569915179671,
1070
- "grad_norm": 16.8511962890625,
1071
- "learning_rate": 3.900041714464491e-05,
1072
- "loss": 2.7864,
1073
- "step": 12700
1074
- },
1075
- {
1076
- "epoch": 0.9009960229472425,
1077
- "grad_norm": 19.323467254638672,
1078
- "learning_rate": 3.8870059443111903e-05,
1079
- "loss": 2.9093,
1080
- "step": 12800
1081
- },
1082
- {
1083
- "epoch": 0.9080350543765178,
1084
- "grad_norm": 16.42205810546875,
1085
- "learning_rate": 3.873970174157889e-05,
1086
- "loss": 2.8426,
1087
- "step": 12900
1088
- },
1089
- {
1090
- "epoch": 0.9150740858057931,
1091
- "grad_norm": 16.470041275024414,
1092
- "learning_rate": 3.8609344040045885e-05,
1093
- "loss": 2.9097,
1094
- "step": 13000
1095
- },
1096
- {
1097
- "epoch": 0.9150740858057931,
1098
- "eval_runtime": 192.3033,
1099
- "eval_samples_per_second": 147.751,
1100
- "eval_steps_per_second": 18.471,
1101
- "step": 13000
1102
- },
1103
- {
1104
- "epoch": 0.9221131172350685,
1105
- "grad_norm": 15.83154296875,
1106
- "learning_rate": 3.847898633851288e-05,
1107
- "loss": 2.8286,
1108
- "step": 13100
1109
- },
1110
- {
1111
- "epoch": 0.9291521486643438,
1112
- "grad_norm": 14.70117473602295,
1113
- "learning_rate": 3.834862863697987e-05,
1114
- "loss": 2.7575,
1115
- "step": 13200
1116
- },
1117
- {
1118
- "epoch": 0.9361911800936191,
1119
- "grad_norm": 12.124509811401367,
1120
- "learning_rate": 3.821827093544687e-05,
1121
- "loss": 2.8621,
1122
- "step": 13300
1123
- },
1124
- {
1125
- "epoch": 0.9432302115228944,
1126
- "grad_norm": 16.292022705078125,
1127
- "learning_rate": 3.808791323391386e-05,
1128
- "loss": 2.8619,
1129
- "step": 13400
1130
- },
1131
- {
1132
- "epoch": 0.9502692429521697,
1133
- "grad_norm": 18.999359130859375,
1134
- "learning_rate": 3.7957555532380855e-05,
1135
- "loss": 2.8396,
1136
- "step": 13500
1137
- },
1138
- {
1139
- "epoch": 0.9502692429521697,
1140
- "eval_runtime": 192.4881,
1141
- "eval_samples_per_second": 147.609,
1142
- "eval_steps_per_second": 18.453,
1143
- "step": 13500
1144
- },
1145
- {
1146
- "epoch": 0.9573082743814452,
1147
- "grad_norm": 17.78417205810547,
1148
- "learning_rate": 3.782719783084785e-05,
1149
- "loss": 2.842,
1150
- "step": 13600
1151
- },
1152
- {
1153
- "epoch": 0.9643473058107205,
1154
- "grad_norm": 16.267335891723633,
1155
- "learning_rate": 3.7696840129314844e-05,
1156
- "loss": 2.8373,
1157
- "step": 13700
1158
- },
1159
- {
1160
- "epoch": 0.9713863372399958,
1161
- "grad_norm": 15.711287498474121,
1162
- "learning_rate": 3.756648242778184e-05,
1163
- "loss": 2.9,
1164
- "step": 13800
1165
- },
1166
- {
1167
- "epoch": 0.9784253686692711,
1168
- "grad_norm": 26.957563400268555,
1169
- "learning_rate": 3.7436124726248825e-05,
1170
- "loss": 2.9499,
1171
- "step": 13900
1172
- },
1173
- {
1174
- "epoch": 0.9854644000985464,
1175
- "grad_norm": 11.875740051269531,
1176
- "learning_rate": 3.730576702471582e-05,
1177
- "loss": 2.7061,
1178
- "step": 14000
1179
- },
1180
- {
1181
- "epoch": 0.9854644000985464,
1182
- "eval_runtime": 192.2723,
1183
- "eval_samples_per_second": 147.775,
1184
- "eval_steps_per_second": 18.474,
1185
- "step": 14000
1186
- },
1187
- {
1188
- "epoch": 0.9925034315278217,
1189
- "grad_norm": 21.315786361694336,
1190
- "learning_rate": 3.717540932318281e-05,
1191
- "loss": 2.8303,
1192
- "step": 14100
1193
- },
1194
- {
1195
- "epoch": 0.9995424629570971,
1196
- "grad_norm": 13.713945388793945,
1197
- "learning_rate": 3.704505162164981e-05,
1198
- "loss": 2.7378,
1199
- "step": 14200
1200
- },
1201
- {
1202
- "epoch": 1.0065814943863725,
1203
- "grad_norm": 16.48957633972168,
1204
- "learning_rate": 3.69146939201168e-05,
1205
- "loss": 2.7313,
1206
- "step": 14300
1207
- },
1208
- {
1209
- "epoch": 1.0136205258156479,
1210
- "grad_norm": 19.689464569091797,
1211
- "learning_rate": 3.6784336218583796e-05,
1212
- "loss": 2.7612,
1213
- "step": 14400
1214
- },
1215
- {
1216
- "epoch": 1.020659557244923,
1217
- "grad_norm": 19.848342895507812,
1218
- "learning_rate": 3.665397851705079e-05,
1219
- "loss": 2.8205,
1220
- "step": 14500
1221
- },
1222
- {
1223
- "epoch": 1.020659557244923,
1224
- "eval_runtime": 193.2356,
1225
- "eval_samples_per_second": 147.038,
1226
- "eval_steps_per_second": 18.382,
1227
- "step": 14500
1228
- },
1229
- {
1230
- "epoch": 1.0276985886741985,
1231
- "grad_norm": 23.994319915771484,
1232
- "learning_rate": 3.6523620815517784e-05,
1233
- "loss": 2.7362,
1234
- "step": 14600
1235
- },
1236
- {
1237
- "epoch": 1.0347376201034737,
1238
- "grad_norm": 16.060455322265625,
1239
- "learning_rate": 3.639326311398478e-05,
1240
- "loss": 2.8076,
1241
- "step": 14700
1242
- },
1243
- {
1244
- "epoch": 1.0417766515327491,
1245
- "grad_norm": 17.874704360961914,
1246
- "learning_rate": 3.626290541245177e-05,
1247
- "loss": 2.6903,
1248
- "step": 14800
1249
- },
1250
- {
1251
- "epoch": 1.0488156829620243,
1252
- "grad_norm": 15.949551582336426,
1253
- "learning_rate": 3.613254771091876e-05,
1254
- "loss": 2.7765,
1255
- "step": 14900
1256
- },
1257
- {
1258
- "epoch": 1.0558547143912997,
1259
- "grad_norm": 17.344772338867188,
1260
- "learning_rate": 3.6002190009385754e-05,
1261
- "loss": 2.9004,
1262
- "step": 15000
1263
- },
1264
- {
1265
- "epoch": 1.0558547143912997,
1266
- "eval_runtime": 194.9228,
1267
- "eval_samples_per_second": 145.765,
1268
- "eval_steps_per_second": 18.223,
1269
- "step": 15000
1270
- },
1271
- {
1272
- "epoch": 1.0628937458205752,
1273
- "grad_norm": 13.621015548706055,
1274
- "learning_rate": 3.587183230785275e-05,
1275
- "loss": 2.7808,
1276
- "step": 15100
1277
- },
1278
- {
1279
- "epoch": 1.0699327772498504,
1280
- "grad_norm": 21.069551467895508,
1281
- "learning_rate": 3.574147460631974e-05,
1282
- "loss": 2.7421,
1283
- "step": 15200
1284
- },
1285
- {
1286
- "epoch": 1.0769718086791258,
1287
- "grad_norm": 15.854650497436523,
1288
- "learning_rate": 3.5611116904786736e-05,
1289
- "loss": 2.7086,
1290
- "step": 15300
1291
- },
1292
- {
1293
- "epoch": 1.084010840108401,
1294
- "grad_norm": 19.217153549194336,
1295
- "learning_rate": 3.548075920325373e-05,
1296
- "loss": 2.7957,
1297
- "step": 15400
1298
- },
1299
- {
1300
- "epoch": 1.0910498715376764,
1301
- "grad_norm": 20.781291961669922,
1302
- "learning_rate": 3.5350401501720724e-05,
1303
- "loss": 2.8039,
1304
- "step": 15500
1305
- },
1306
- {
1307
- "epoch": 1.0910498715376764,
1308
- "eval_runtime": 192.3565,
1309
- "eval_samples_per_second": 147.71,
1310
- "eval_steps_per_second": 18.466,
1311
- "step": 15500
1312
- },
1313
- {
1314
- "epoch": 1.0980889029669518,
1315
- "grad_norm": 15.170364379882812,
1316
- "learning_rate": 3.522004380018772e-05,
1317
- "loss": 2.7992,
1318
- "step": 15600
1319
- },
1320
- {
1321
- "epoch": 1.105127934396227,
1322
- "grad_norm": 18.8775634765625,
1323
- "learning_rate": 3.508968609865471e-05,
1324
- "loss": 2.7919,
1325
- "step": 15700
1326
- },
1327
- {
1328
- "epoch": 1.1121669658255025,
1329
- "grad_norm": 19.009754180908203,
1330
- "learning_rate": 3.4959328397121706e-05,
1331
- "loss": 2.7361,
1332
- "step": 15800
1333
- },
1334
- {
1335
- "epoch": 1.1192059972547777,
1336
- "grad_norm": 14.632086753845215,
1337
- "learning_rate": 3.48289706955887e-05,
1338
- "loss": 2.8233,
1339
- "step": 15900
1340
- },
1341
- {
1342
- "epoch": 1.126245028684053,
1343
- "grad_norm": 20.006601333618164,
1344
- "learning_rate": 3.469861299405569e-05,
1345
- "loss": 2.738,
1346
- "step": 16000
1347
- },
1348
- {
1349
- "epoch": 1.126245028684053,
1350
- "eval_runtime": 194.9892,
1351
- "eval_samples_per_second": 145.716,
1352
- "eval_steps_per_second": 18.216,
1353
- "step": 16000
1354
- },
1355
- {
1356
- "epoch": 1.1332840601133285,
1357
- "grad_norm": 14.096820831298828,
1358
- "learning_rate": 3.456825529252268e-05,
1359
- "loss": 2.8094,
1360
- "step": 16100
1361
- },
1362
- {
1363
- "epoch": 1.1403230915426037,
1364
- "grad_norm": 17.261428833007812,
1365
- "learning_rate": 3.4437897590989676e-05,
1366
- "loss": 2.8108,
1367
- "step": 16200
1368
- },
1369
- {
1370
- "epoch": 1.1473621229718791,
1371
- "grad_norm": 14.972962379455566,
1372
- "learning_rate": 3.430753988945667e-05,
1373
- "loss": 2.8055,
1374
- "step": 16300
1375
- },
1376
- {
1377
- "epoch": 1.1544011544011543,
1378
- "grad_norm": 14.345026969909668,
1379
- "learning_rate": 3.4177182187923664e-05,
1380
- "loss": 2.7363,
1381
- "step": 16400
1382
- },
1383
- {
1384
- "epoch": 1.1614401858304297,
1385
- "grad_norm": 15.4429292678833,
1386
- "learning_rate": 3.404682448639065e-05,
1387
- "loss": 2.8642,
1388
- "step": 16500
1389
- },
1390
- {
1391
- "epoch": 1.1614401858304297,
1392
- "eval_runtime": 194.218,
1393
- "eval_samples_per_second": 146.294,
1394
- "eval_steps_per_second": 18.289,
1395
- "step": 16500
1396
- },
1397
- {
1398
- "epoch": 1.1684792172597052,
1399
- "grad_norm": 13.637730598449707,
1400
- "learning_rate": 3.391646678485765e-05,
1401
- "loss": 2.6802,
1402
- "step": 16600
1403
- },
1404
- {
1405
- "epoch": 1.1755182486889804,
1406
- "grad_norm": 16.29159927368164,
1407
- "learning_rate": 3.3786109083324647e-05,
1408
- "loss": 2.7796,
1409
- "step": 16700
1410
- },
1411
- {
1412
- "epoch": 1.1825572801182558,
1413
- "grad_norm": 16.733455657958984,
1414
- "learning_rate": 3.365575138179164e-05,
1415
- "loss": 2.8216,
1416
- "step": 16800
1417
- },
1418
- {
1419
- "epoch": 1.189596311547531,
1420
- "grad_norm": 20.382347106933594,
1421
- "learning_rate": 3.3525393680258635e-05,
1422
- "loss": 2.6737,
1423
- "step": 16900
1424
- },
1425
- {
1426
- "epoch": 1.1966353429768064,
1427
- "grad_norm": 18.506607055664062,
1428
- "learning_rate": 3.339503597872562e-05,
1429
- "loss": 2.7144,
1430
- "step": 17000
1431
- },
1432
- {
1433
- "epoch": 1.1966353429768064,
1434
- "eval_runtime": 192.527,
1435
- "eval_samples_per_second": 147.579,
1436
- "eval_steps_per_second": 18.449,
1437
- "step": 17000
1438
- },
1439
- {
1440
- "epoch": 1.2036743744060816,
1441
- "grad_norm": 16.080759048461914,
1442
- "learning_rate": 3.3264678277192616e-05,
1443
- "loss": 2.8013,
1444
- "step": 17100
1445
- },
1446
- {
1447
- "epoch": 1.210713405835357,
1448
- "grad_norm": 16.123552322387695,
1449
- "learning_rate": 3.313432057565961e-05,
1450
- "loss": 2.7444,
1451
- "step": 17200
1452
- },
1453
- {
1454
- "epoch": 1.2177524372646324,
1455
- "grad_norm": 16.878711700439453,
1456
- "learning_rate": 3.3003962874126604e-05,
1457
- "loss": 2.8241,
1458
- "step": 17300
1459
- },
1460
- {
1461
- "epoch": 1.2247914686939076,
1462
- "grad_norm": 18.64569854736328,
1463
- "learning_rate": 3.28736051725936e-05,
1464
- "loss": 2.684,
1465
- "step": 17400
1466
- },
1467
- {
1468
- "epoch": 1.231830500123183,
1469
- "grad_norm": 16.125022888183594,
1470
- "learning_rate": 3.274324747106059e-05,
1471
- "loss": 2.8633,
1472
- "step": 17500
1473
- },
1474
- {
1475
- "epoch": 1.231830500123183,
1476
- "eval_runtime": 193.4861,
1477
- "eval_samples_per_second": 146.848,
1478
- "eval_steps_per_second": 18.358,
1479
- "step": 17500
1480
- },
1481
- {
1482
- "epoch": 1.2388695315524583,
1483
- "grad_norm": 13.886027336120605,
1484
- "learning_rate": 3.261288976952758e-05,
1485
- "loss": 2.7536,
1486
- "step": 17600
1487
- },
1488
- {
1489
- "epoch": 1.2459085629817337,
1490
- "grad_norm": 15.769869804382324,
1491
- "learning_rate": 3.248253206799458e-05,
1492
- "loss": 2.7889,
1493
- "step": 17700
1494
- },
1495
- {
1496
- "epoch": 1.2529475944110091,
1497
- "grad_norm": 19.419034957885742,
1498
- "learning_rate": 3.2352174366461575e-05,
1499
- "loss": 2.7672,
1500
- "step": 17800
1501
- },
1502
- {
1503
- "epoch": 1.2599866258402843,
1504
- "grad_norm": 18.742015838623047,
1505
- "learning_rate": 3.222181666492857e-05,
1506
- "loss": 2.7427,
1507
- "step": 17900
1508
- },
1509
- {
1510
- "epoch": 1.2670256572695597,
1511
- "grad_norm": 18.40927505493164,
1512
- "learning_rate": 3.209145896339556e-05,
1513
- "loss": 2.7596,
1514
- "step": 18000
1515
- },
1516
- {
1517
- "epoch": 1.2670256572695597,
1518
- "eval_runtime": 194.1066,
1519
- "eval_samples_per_second": 146.378,
1520
- "eval_steps_per_second": 18.299,
1521
- "step": 18000
1522
- },
1523
- {
1524
- "epoch": 1.274064688698835,
1525
- "grad_norm": 21.27202033996582,
1526
- "learning_rate": 3.196110126186255e-05,
1527
- "loss": 2.7888,
1528
- "step": 18100
1529
- },
1530
- {
1531
- "epoch": 1.2811037201281104,
1532
- "grad_norm": 13.953824043273926,
1533
- "learning_rate": 3.1830743560329545e-05,
1534
- "loss": 2.7347,
1535
- "step": 18200
1536
- },
1537
- {
1538
- "epoch": 1.2881427515573858,
1539
- "grad_norm": 16.453821182250977,
1540
- "learning_rate": 3.170038585879654e-05,
1541
- "loss": 2.8821,
1542
- "step": 18300
1543
- },
1544
- {
1545
- "epoch": 1.295181782986661,
1546
- "grad_norm": 16.67236328125,
1547
- "learning_rate": 3.157002815726353e-05,
1548
- "loss": 2.8431,
1549
- "step": 18400
1550
- },
1551
- {
1552
- "epoch": 1.3022208144159364,
1553
- "grad_norm": 13.558029174804688,
1554
- "learning_rate": 3.143967045573053e-05,
1555
- "loss": 2.7499,
1556
- "step": 18500
1557
- },
1558
- {
1559
- "epoch": 1.3022208144159364,
1560
- "eval_runtime": 192.4964,
1561
- "eval_samples_per_second": 147.603,
1562
- "eval_steps_per_second": 18.452,
1563
- "step": 18500
1564
- },
1565
- {
1566
- "epoch": 1.3092598458452116,
1567
- "grad_norm": 15.34234619140625,
1568
- "learning_rate": 3.1309312754197514e-05,
1569
- "loss": 2.8225,
1570
- "step": 18600
1571
- },
1572
- {
1573
- "epoch": 1.316298877274487,
1574
- "grad_norm": 14.304731369018555,
1575
- "learning_rate": 3.117895505266451e-05,
1576
- "loss": 2.8369,
1577
- "step": 18700
1578
- },
1579
- {
1580
- "epoch": 1.3233379087037624,
1581
- "grad_norm": 17.345626831054688,
1582
- "learning_rate": 3.104859735113151e-05,
1583
- "loss": 2.6865,
1584
- "step": 18800
1585
- },
1586
- {
1587
- "epoch": 1.3303769401330376,
1588
- "grad_norm": 16.954349517822266,
1589
- "learning_rate": 3.09182396495985e-05,
1590
- "loss": 2.741,
1591
- "step": 18900
1592
- },
1593
- {
1594
- "epoch": 1.337415971562313,
1595
- "grad_norm": 17.250642776489258,
1596
- "learning_rate": 3.07878819480655e-05,
1597
- "loss": 2.8111,
1598
- "step": 19000
1599
- },
1600
- {
1601
- "epoch": 1.337415971562313,
1602
- "eval_runtime": 192.7825,
1603
- "eval_samples_per_second": 147.384,
1604
- "eval_steps_per_second": 18.425,
1605
- "step": 19000
1606
- },
1607
- {
1608
- "epoch": 1.3444550029915883,
1609
- "grad_norm": 17.004776000976562,
1610
- "learning_rate": 3.0657524246532485e-05,
1611
- "loss": 2.7594,
1612
- "step": 19100
1613
- },
1614
- {
1615
- "epoch": 1.3514940344208637,
1616
- "grad_norm": 15.450813293457031,
1617
- "learning_rate": 3.052716654499948e-05,
1618
- "loss": 2.6522,
1619
- "step": 19200
1620
- },
1621
- {
1622
- "epoch": 1.358533065850139,
1623
- "grad_norm": 15.58588981628418,
1624
- "learning_rate": 3.0396808843466473e-05,
1625
- "loss": 2.8346,
1626
- "step": 19300
1627
- },
1628
- {
1629
- "epoch": 1.3655720972794143,
1630
- "grad_norm": 23.079944610595703,
1631
- "learning_rate": 3.0266451141933467e-05,
1632
- "loss": 2.829,
1633
- "step": 19400
1634
- },
1635
- {
1636
- "epoch": 1.3726111287086897,
1637
- "grad_norm": 23.278108596801758,
1638
- "learning_rate": 3.0136093440400458e-05,
1639
- "loss": 2.809,
1640
- "step": 19500
1641
- },
1642
- {
1643
- "epoch": 1.3726111287086897,
1644
- "eval_runtime": 191.7654,
1645
- "eval_samples_per_second": 148.165,
1646
- "eval_steps_per_second": 18.523,
1647
- "step": 19500
1648
- },
1649
- {
1650
- "epoch": 1.379650160137965,
1651
- "grad_norm": 12.144103050231934,
1652
- "learning_rate": 3.0005735738867452e-05,
1653
- "loss": 2.5999,
1654
- "step": 19600
1655
- },
1656
- {
1657
- "epoch": 1.3866891915672404,
1658
- "grad_norm": 18.378664016723633,
1659
- "learning_rate": 2.9875378037334446e-05,
1660
- "loss": 2.8226,
1661
- "step": 19700
1662
- },
1663
- {
1664
- "epoch": 1.3937282229965158,
1665
- "grad_norm": 15.180033683776855,
1666
- "learning_rate": 2.9745020335801437e-05,
1667
- "loss": 2.7773,
1668
- "step": 19800
1669
- },
1670
- {
1671
- "epoch": 1.400767254425791,
1672
- "grad_norm": 16.611019134521484,
1673
- "learning_rate": 2.9614662634268438e-05,
1674
- "loss": 2.7171,
1675
- "step": 19900
1676
- },
1677
- {
1678
- "epoch": 1.4078062858550664,
1679
- "grad_norm": 14.491551399230957,
1680
- "learning_rate": 2.948430493273543e-05,
1681
- "loss": 2.7234,
1682
- "step": 20000
1683
- },
1684
- {
1685
- "epoch": 1.4078062858550664,
1686
- "eval_runtime": 192.2568,
1687
- "eval_samples_per_second": 147.787,
1688
- "eval_steps_per_second": 18.475,
1689
- "step": 20000
1690
- },
1691
- {
1692
- "epoch": 1.4148453172843416,
1693
- "grad_norm": 15.652689933776855,
1694
- "learning_rate": 2.9353947231202422e-05,
1695
- "loss": 2.6237,
1696
- "step": 20100
1697
- },
1698
- {
1699
- "epoch": 1.421884348713617,
1700
- "grad_norm": 16.404693603515625,
1701
- "learning_rate": 2.9223589529669417e-05,
1702
- "loss": 2.7363,
1703
- "step": 20200
1704
- },
1705
- {
1706
- "epoch": 1.4289233801428924,
1707
- "grad_norm": 13.620403289794922,
1708
- "learning_rate": 2.9093231828136407e-05,
1709
- "loss": 2.7651,
1710
- "step": 20300
1711
- },
1712
- {
1713
- "epoch": 1.4359624115721676,
1714
- "grad_norm": 16.975452423095703,
1715
- "learning_rate": 2.89628741266034e-05,
1716
- "loss": 2.8431,
1717
- "step": 20400
1718
- },
1719
- {
1720
- "epoch": 1.443001443001443,
1721
- "grad_norm": 16.957857131958008,
1722
- "learning_rate": 2.8832516425070395e-05,
1723
- "loss": 2.7442,
1724
- "step": 20500
1725
- },
1726
- {
1727
- "epoch": 1.443001443001443,
1728
- "eval_runtime": 192.9203,
1729
- "eval_samples_per_second": 147.278,
1730
- "eval_steps_per_second": 18.412,
1731
- "step": 20500
1732
- },
1733
- {
1734
- "epoch": 1.4500404744307183,
1735
- "grad_norm": 15.085665702819824,
1736
- "learning_rate": 2.8702158723537386e-05,
1737
- "loss": 2.6764,
1738
- "step": 20600
1739
- },
1740
- {
1741
- "epoch": 1.4570795058599937,
1742
- "grad_norm": 15.870454788208008,
1743
- "learning_rate": 2.857180102200438e-05,
1744
- "loss": 2.6491,
1745
- "step": 20700
1746
- },
1747
- {
1748
- "epoch": 1.464118537289269,
1749
- "grad_norm": 15.54505729675293,
1750
- "learning_rate": 2.844144332047137e-05,
1751
- "loss": 2.785,
1752
- "step": 20800
1753
- },
1754
- {
1755
- "epoch": 1.4711575687185443,
1756
- "grad_norm": 16.786861419677734,
1757
- "learning_rate": 2.8311085618938365e-05,
1758
- "loss": 2.656,
1759
- "step": 20900
1760
- },
1761
- {
1762
- "epoch": 1.4781966001478197,
1763
- "grad_norm": 17.03700828552246,
1764
- "learning_rate": 2.8180727917405363e-05,
1765
- "loss": 2.7337,
1766
- "step": 21000
1767
- },
1768
- {
1769
- "epoch": 1.4781966001478197,
1770
- "eval_runtime": 194.0613,
1771
- "eval_samples_per_second": 146.413,
1772
- "eval_steps_per_second": 18.303,
1773
- "step": 21000
1774
- },
1775
- {
1776
- "epoch": 1.485235631577095,
1777
- "grad_norm": 15.768280982971191,
1778
- "learning_rate": 2.8050370215872357e-05,
1779
- "loss": 2.7419,
1780
- "step": 21100
1781
- },
1782
- {
1783
- "epoch": 1.4922746630063703,
1784
- "grad_norm": 16.70868682861328,
1785
- "learning_rate": 2.792001251433935e-05,
1786
- "loss": 2.7001,
1787
- "step": 21200
1788
- },
1789
- {
1790
- "epoch": 1.4993136944356458,
1791
- "grad_norm": 15.003129005432129,
1792
- "learning_rate": 2.778965481280634e-05,
1793
- "loss": 2.6372,
1794
- "step": 21300
1795
- },
1796
- {
1797
- "epoch": 1.506352725864921,
1798
- "grad_norm": 15.321432113647461,
1799
- "learning_rate": 2.7659297111273336e-05,
1800
- "loss": 2.734,
1801
- "step": 21400
1802
- },
1803
- {
1804
- "epoch": 1.5133917572941962,
1805
- "grad_norm": 12.912035942077637,
1806
- "learning_rate": 2.752893940974033e-05,
1807
- "loss": 2.6854,
1808
- "step": 21500
1809
- },
1810
- {
1811
- "epoch": 1.5133917572941962,
1812
- "eval_runtime": 192.892,
1813
- "eval_samples_per_second": 147.3,
1814
- "eval_steps_per_second": 18.414,
1815
- "step": 21500
1816
- },
1817
- {
1818
- "epoch": 1.5204307887234716,
1819
- "grad_norm": 17.954883575439453,
1820
- "learning_rate": 2.739858170820732e-05,
1821
- "loss": 2.6843,
1822
- "step": 21600
1823
- },
1824
- {
1825
- "epoch": 1.527469820152747,
1826
- "grad_norm": 20.32744026184082,
1827
- "learning_rate": 2.7268224006674315e-05,
1828
- "loss": 2.6404,
1829
- "step": 21700
1830
- },
1831
- {
1832
- "epoch": 1.5345088515820224,
1833
- "grad_norm": 14.839242935180664,
1834
- "learning_rate": 2.713786630514131e-05,
1835
- "loss": 2.7235,
1836
- "step": 21800
1837
- },
1838
- {
1839
- "epoch": 1.5415478830112976,
1840
- "grad_norm": 15.594539642333984,
1841
- "learning_rate": 2.70075086036083e-05,
1842
- "loss": 2.6163,
1843
- "step": 21900
1844
- },
1845
- {
1846
- "epoch": 1.5485869144405728,
1847
- "grad_norm": 14.877588272094727,
1848
- "learning_rate": 2.6877150902075294e-05,
1849
- "loss": 2.6331,
1850
- "step": 22000
1851
- },
1852
- {
1853
- "epoch": 1.5485869144405728,
1854
- "eval_runtime": 191.7338,
1855
- "eval_samples_per_second": 148.19,
1856
- "eval_steps_per_second": 18.526,
1857
- "step": 22000
1858
- },
1859
- {
1860
- "epoch": 1.5556259458698483,
1861
- "grad_norm": 14.593866348266602,
1862
- "learning_rate": 2.674679320054229e-05,
1863
- "loss": 2.7969,
1864
- "step": 22100
1865
- },
1866
- {
1867
- "epoch": 1.5626649772991237,
1868
- "grad_norm": 22.533540725708008,
1869
- "learning_rate": 2.6616435499009285e-05,
1870
- "loss": 2.7841,
1871
- "step": 22200
1872
- },
1873
- {
1874
- "epoch": 1.569704008728399,
1875
- "grad_norm": 16.089982986450195,
1876
- "learning_rate": 2.6486077797476276e-05,
1877
- "loss": 2.667,
1878
- "step": 22300
1879
- },
1880
- {
1881
- "epoch": 1.5767430401576743,
1882
- "grad_norm": 17.49601936340332,
1883
- "learning_rate": 2.635572009594327e-05,
1884
- "loss": 2.7963,
1885
- "step": 22400
1886
- },
1887
- {
1888
- "epoch": 1.5837820715869495,
1889
- "grad_norm": 17.72164535522461,
1890
- "learning_rate": 2.6225362394410264e-05,
1891
- "loss": 2.7669,
1892
- "step": 22500
1893
- },
1894
- {
1895
- "epoch": 1.5837820715869495,
1896
- "eval_runtime": 191.4019,
1897
- "eval_samples_per_second": 148.447,
1898
- "eval_steps_per_second": 18.558,
1899
- "step": 22500
1900
- },
1901
- {
1902
- "epoch": 1.590821103016225,
1903
- "grad_norm": 14.238466262817383,
1904
- "learning_rate": 2.6095004692877255e-05,
1905
- "loss": 2.732,
1906
- "step": 22600
1907
- },
1908
- {
1909
- "epoch": 1.5978601344455003,
1910
- "grad_norm": 19.9779052734375,
1911
- "learning_rate": 2.596464699134425e-05,
1912
- "loss": 2.7353,
1913
- "step": 22700
1914
- },
1915
- {
1916
- "epoch": 1.6048991658747758,
1917
- "grad_norm": 16.89205551147461,
1918
- "learning_rate": 2.5834289289811243e-05,
1919
- "loss": 2.7311,
1920
- "step": 22800
1921
- },
1922
- {
1923
- "epoch": 1.611938197304051,
1924
- "grad_norm": 13.072985649108887,
1925
- "learning_rate": 2.5703931588278234e-05,
1926
- "loss": 2.6313,
1927
- "step": 22900
1928
- },
1929
- {
1930
- "epoch": 1.6189772287333262,
1931
- "grad_norm": 22.408113479614258,
1932
- "learning_rate": 2.5573573886745228e-05,
1933
- "loss": 2.605,
1934
- "step": 23000
1935
- },
1936
- {
1937
- "epoch": 1.6189772287333262,
1938
- "eval_runtime": 192.574,
1939
- "eval_samples_per_second": 147.543,
1940
- "eval_steps_per_second": 18.445,
1941
- "step": 23000
1942
- },
1943
- {
1944
- "epoch": 1.6260162601626016,
1945
- "grad_norm": 21.51888084411621,
1946
- "learning_rate": 2.5443216185212222e-05,
1947
- "loss": 2.5964,
1948
- "step": 23100
1949
- },
1950
- {
1951
- "epoch": 1.633055291591877,
1952
- "grad_norm": 20.486024856567383,
1953
- "learning_rate": 2.531285848367922e-05,
1954
- "loss": 2.6883,
1955
- "step": 23200
1956
- },
1957
- {
1958
- "epoch": 1.6400943230211524,
1959
- "grad_norm": 17.860441207885742,
1960
- "learning_rate": 2.5182500782146214e-05,
1961
- "loss": 2.7572,
1962
- "step": 23300
1963
- },
1964
- {
1965
- "epoch": 1.6471333544504276,
1966
- "grad_norm": 19.4054012298584,
1967
- "learning_rate": 2.5052143080613204e-05,
1968
- "loss": 2.7643,
1969
- "step": 23400
1970
- },
1971
- {
1972
- "epoch": 1.6541723858797028,
1973
- "grad_norm": 15.56551742553711,
1974
- "learning_rate": 2.49217853790802e-05,
1975
- "loss": 2.6638,
1976
- "step": 23500
1977
- },
1978
- {
1979
- "epoch": 1.6541723858797028,
1980
- "eval_runtime": 192.3359,
1981
- "eval_samples_per_second": 147.726,
1982
- "eval_steps_per_second": 18.468,
1983
- "step": 23500
1984
- },
1985
- {
1986
- "epoch": 1.6612114173089783,
1987
- "grad_norm": 22.051755905151367,
1988
- "learning_rate": 2.4791427677547192e-05,
1989
- "loss": 2.6905,
1990
- "step": 23600
1991
- },
1992
- {
1993
- "epoch": 1.6682504487382537,
1994
- "grad_norm": 19.55982208251953,
1995
- "learning_rate": 2.4661069976014183e-05,
1996
- "loss": 2.7178,
1997
- "step": 23700
1998
- },
1999
- {
2000
- "epoch": 1.675289480167529,
2001
- "grad_norm": 14.777819633483887,
2002
- "learning_rate": 2.4530712274481177e-05,
2003
- "loss": 2.6219,
2004
- "step": 23800
2005
- },
2006
- {
2007
- "epoch": 1.6823285115968043,
2008
- "grad_norm": 15.4576997756958,
2009
- "learning_rate": 2.440035457294817e-05,
2010
- "loss": 2.6425,
2011
- "step": 23900
2012
- },
2013
- {
2014
- "epoch": 1.6893675430260795,
2015
- "grad_norm": 18.520376205444336,
2016
- "learning_rate": 2.4269996871415165e-05,
2017
- "loss": 2.6541,
2018
- "step": 24000
2019
- },
2020
- {
2021
- "epoch": 1.6893675430260795,
2022
- "eval_runtime": 191.8782,
2023
- "eval_samples_per_second": 148.078,
2024
- "eval_steps_per_second": 18.512,
2025
- "step": 24000
2026
- },
2027
- {
2028
- "epoch": 1.696406574455355,
2029
- "grad_norm": 18.677989959716797,
2030
- "learning_rate": 2.413963916988216e-05,
2031
- "loss": 2.7502,
2032
- "step": 24100
2033
- },
2034
- {
2035
- "epoch": 1.7034456058846303,
2036
- "grad_norm": 19.01474380493164,
2037
- "learning_rate": 2.400928146834915e-05,
2038
- "loss": 2.5849,
2039
- "step": 24200
2040
- },
2041
- {
2042
- "epoch": 1.7104846373139058,
2043
- "grad_norm": 14.854390144348145,
2044
- "learning_rate": 2.3878923766816144e-05,
2045
- "loss": 2.6224,
2046
- "step": 24300
2047
- },
2048
- {
2049
- "epoch": 1.717523668743181,
2050
- "grad_norm": 16.40928077697754,
2051
- "learning_rate": 2.374856606528314e-05,
2052
- "loss": 2.6996,
2053
- "step": 24400
2054
- },
2055
- {
2056
- "epoch": 1.7245627001724562,
2057
- "grad_norm": 14.962175369262695,
2058
- "learning_rate": 2.3618208363750133e-05,
2059
- "loss": 2.6928,
2060
- "step": 24500
2061
- },
2062
- {
2063
- "epoch": 1.7245627001724562,
2064
- "eval_runtime": 194.5262,
2065
- "eval_samples_per_second": 146.063,
2066
- "eval_steps_per_second": 18.26,
2067
- "step": 24500
2068
- },
2069
- {
2070
- "epoch": 1.7316017316017316,
2071
- "grad_norm": 19.39845085144043,
2072
- "learning_rate": 2.3487850662217127e-05,
2073
- "loss": 2.7458,
2074
- "step": 24600
2075
- },
2076
- {
2077
- "epoch": 1.738640763031007,
2078
- "grad_norm": 16.46622085571289,
2079
- "learning_rate": 2.3357492960684117e-05,
2080
- "loss": 2.7465,
2081
- "step": 24700
2082
- },
2083
- {
2084
- "epoch": 1.7456797944602824,
2085
- "grad_norm": 17.756010055541992,
2086
- "learning_rate": 2.322713525915111e-05,
2087
- "loss": 2.7617,
2088
- "step": 24800
2089
- },
2090
- {
2091
- "epoch": 1.7527188258895576,
2092
- "grad_norm": 17.55894660949707,
2093
- "learning_rate": 2.3096777557618106e-05,
2094
- "loss": 2.6085,
2095
- "step": 24900
2096
- },
2097
- {
2098
- "epoch": 1.7597578573188328,
2099
- "grad_norm": 16.707901000976562,
2100
- "learning_rate": 2.2966419856085096e-05,
2101
- "loss": 2.6305,
2102
- "step": 25000
2103
- },
2104
- {
2105
- "epoch": 1.7597578573188328,
2106
- "eval_runtime": 191.9564,
2107
- "eval_samples_per_second": 148.018,
2108
- "eval_steps_per_second": 18.504,
2109
- "step": 25000
2110
- },
2111
- {
2112
- "epoch": 1.7667968887481083,
2113
- "grad_norm": 21.4102840423584,
2114
- "learning_rate": 2.2836062154552094e-05,
2115
- "loss": 2.7509,
2116
- "step": 25100
2117
- },
2118
- {
2119
- "epoch": 1.7738359201773837,
2120
- "grad_norm": 21.17198944091797,
2121
- "learning_rate": 2.2705704453019085e-05,
2122
- "loss": 2.6602,
2123
- "step": 25200
2124
- },
2125
- {
2126
- "epoch": 1.7808749516066589,
2127
- "grad_norm": 14.665617942810059,
2128
- "learning_rate": 2.257534675148608e-05,
2129
- "loss": 2.6576,
2130
- "step": 25300
2131
- },
2132
- {
2133
- "epoch": 1.7879139830359343,
2134
- "grad_norm": 18.63422393798828,
2135
- "learning_rate": 2.2444989049953073e-05,
2136
- "loss": 2.6682,
2137
- "step": 25400
2138
- },
2139
- {
2140
- "epoch": 1.7949530144652095,
2141
- "grad_norm": 20.478769302368164,
2142
- "learning_rate": 2.2314631348420064e-05,
2143
- "loss": 2.5733,
2144
- "step": 25500
2145
- },
2146
- {
2147
- "epoch": 1.7949530144652095,
2148
- "eval_runtime": 194.5888,
2149
- "eval_samples_per_second": 146.016,
2150
- "eval_steps_per_second": 18.254,
2151
- "step": 25500
2152
- },
2153
- {
2154
- "epoch": 1.801992045894485,
2155
- "grad_norm": 13.144091606140137,
2156
- "learning_rate": 2.218427364688706e-05,
2157
- "loss": 2.5946,
2158
- "step": 25600
2159
- },
2160
- {
2161
- "epoch": 1.8090310773237603,
2162
- "grad_norm": 22.20168685913086,
2163
- "learning_rate": 2.2053915945354052e-05,
2164
- "loss": 2.7029,
2165
- "step": 25700
2166
- },
2167
- {
2168
- "epoch": 1.8160701087530355,
2169
- "grad_norm": 17.39105796813965,
2170
- "learning_rate": 2.1923558243821046e-05,
2171
- "loss": 2.7007,
2172
- "step": 25800
2173
- },
2174
- {
2175
- "epoch": 1.823109140182311,
2176
- "grad_norm": 16.70639419555664,
2177
- "learning_rate": 2.179320054228804e-05,
2178
- "loss": 2.6011,
2179
- "step": 25900
2180
- },
2181
- {
2182
- "epoch": 1.8301481716115862,
2183
- "grad_norm": 23.447750091552734,
2184
- "learning_rate": 2.166284284075503e-05,
2185
- "loss": 2.682,
2186
- "step": 26000
2187
- },
2188
- {
2189
- "epoch": 1.8301481716115862,
2190
- "eval_runtime": 192.912,
2191
- "eval_samples_per_second": 147.285,
2192
- "eval_steps_per_second": 18.413,
2193
- "step": 26000
2194
- },
2195
- {
2196
- "epoch": 1.8371872030408616,
2197
- "grad_norm": 20.410226821899414,
2198
- "learning_rate": 2.1532485139222025e-05,
2199
- "loss": 2.7565,
2200
- "step": 26100
2201
- },
2202
- {
2203
- "epoch": 1.844226234470137,
2204
- "grad_norm": 20.64243507385254,
2205
- "learning_rate": 2.1402127437689022e-05,
2206
- "loss": 2.566,
2207
- "step": 26200
2208
- },
2209
- {
2210
- "epoch": 1.8512652658994122,
2211
- "grad_norm": 15.64989185333252,
2212
- "learning_rate": 2.1271769736156013e-05,
2213
- "loss": 2.7237,
2214
- "step": 26300
2215
- },
2216
- {
2217
- "epoch": 1.8583042973286876,
2218
- "grad_norm": 13.451628684997559,
2219
- "learning_rate": 2.1141412034623007e-05,
2220
- "loss": 2.6218,
2221
- "step": 26400
2222
- },
2223
- {
2224
- "epoch": 1.8653433287579628,
2225
- "grad_norm": 16.763063430786133,
2226
- "learning_rate": 2.1011054333089998e-05,
2227
- "loss": 2.5891,
2228
- "step": 26500
2229
- },
2230
- {
2231
- "epoch": 1.8653433287579628,
2232
- "eval_runtime": 191.7231,
2233
- "eval_samples_per_second": 148.198,
2234
- "eval_steps_per_second": 18.527,
2235
- "step": 26500
2236
- },
2237
- {
2238
- "epoch": 1.8723823601872382,
2239
- "grad_norm": 16.719079971313477,
2240
- "learning_rate": 2.0880696631556992e-05,
2241
- "loss": 2.6568,
2242
- "step": 26600
2243
- },
2244
- {
2245
- "epoch": 1.8794213916165137,
2246
- "grad_norm": 20.345216751098633,
2247
- "learning_rate": 2.075033893002399e-05,
2248
- "loss": 2.5889,
2249
- "step": 26700
2250
- },
2251
- {
2252
- "epoch": 1.8864604230457889,
2253
- "grad_norm": 13.290498733520508,
2254
- "learning_rate": 2.061998122849098e-05,
2255
- "loss": 2.6915,
2256
- "step": 26800
2257
- },
2258
- {
2259
- "epoch": 1.893499454475064,
2260
- "grad_norm": 26.90572738647461,
2261
- "learning_rate": 2.0489623526957974e-05,
2262
- "loss": 2.6635,
2263
- "step": 26900
2264
- },
2265
- {
2266
- "epoch": 1.9005384859043395,
2267
- "grad_norm": 12.706587791442871,
2268
- "learning_rate": 2.035926582542497e-05,
2269
- "loss": 2.6886,
2270
- "step": 27000
2271
- },
2272
- {
2273
- "epoch": 1.9005384859043395,
2274
- "eval_runtime": 193.671,
2275
- "eval_samples_per_second": 146.708,
2276
- "eval_steps_per_second": 18.34,
2277
- "step": 27000
2278
- },
2279
- {
2280
- "epoch": 1.907577517333615,
2281
- "grad_norm": 19.337390899658203,
2282
- "learning_rate": 2.022890812389196e-05,
2283
- "loss": 2.5446,
2284
- "step": 27100
2285
- },
2286
- {
2287
- "epoch": 1.9146165487628903,
2288
- "grad_norm": 16.442127227783203,
2289
- "learning_rate": 2.0098550422358953e-05,
2290
- "loss": 2.6562,
2291
- "step": 27200
2292
- },
2293
- {
2294
- "epoch": 1.9216555801921655,
2295
- "grad_norm": 17.196496963500977,
2296
- "learning_rate": 1.9968192720825947e-05,
2297
- "loss": 2.5869,
2298
- "step": 27300
2299
- },
2300
- {
2301
- "epoch": 1.9286946116214407,
2302
- "grad_norm": 15.884928703308105,
2303
- "learning_rate": 1.983783501929294e-05,
2304
- "loss": 2.6127,
2305
- "step": 27400
2306
- },
2307
- {
2308
- "epoch": 1.9357336430507162,
2309
- "grad_norm": 15.426615715026855,
2310
- "learning_rate": 1.9707477317759935e-05,
2311
- "loss": 2.6043,
2312
- "step": 27500
2313
- },
2314
- {
2315
- "epoch": 1.9357336430507162,
2316
- "eval_runtime": 193.4431,
2317
- "eval_samples_per_second": 146.88,
2318
- "eval_steps_per_second": 18.362,
2319
- "step": 27500
2320
- },
2321
- {
2322
- "epoch": 1.9427726744799916,
2323
- "grad_norm": 20.6138858795166,
2324
- "learning_rate": 1.9577119616226926e-05,
2325
- "loss": 2.6387,
2326
- "step": 27600
2327
- },
2328
- {
2329
- "epoch": 1.949811705909267,
2330
- "grad_norm": 14.545782089233398,
2331
- "learning_rate": 1.944676191469392e-05,
2332
- "loss": 2.7687,
2333
- "step": 27700
2334
- },
2335
- {
2336
- "epoch": 1.9568507373385422,
2337
- "grad_norm": 15.325973510742188,
2338
- "learning_rate": 1.9316404213160914e-05,
2339
- "loss": 2.6876,
2340
- "step": 27800
2341
- },
2342
- {
2343
- "epoch": 1.9638897687678174,
2344
- "grad_norm": 16.72733497619629,
2345
- "learning_rate": 1.918604651162791e-05,
2346
- "loss": 2.6131,
2347
- "step": 27900
2348
- },
2349
- {
2350
- "epoch": 1.9709288001970928,
2351
- "grad_norm": 22.076963424682617,
2352
- "learning_rate": 1.9055688810094903e-05,
2353
- "loss": 2.7044,
2354
- "step": 28000
2355
- },
2356
- {
2357
- "epoch": 1.9709288001970928,
2358
- "eval_runtime": 192.6043,
2359
- "eval_samples_per_second": 147.52,
2360
- "eval_steps_per_second": 18.442,
2361
- "step": 28000
2362
- },
2363
- {
2364
- "epoch": 1.9779678316263682,
2365
- "grad_norm": 17.05091094970703,
2366
- "learning_rate": 1.8925331108561893e-05,
2367
- "loss": 2.6023,
2368
- "step": 28100
2369
- },
2370
- {
2371
- "epoch": 1.9850068630556437,
2372
- "grad_norm": 17.847782135009766,
2373
- "learning_rate": 1.8794973407028887e-05,
2374
- "loss": 2.5752,
2375
- "step": 28200
2376
- },
2377
- {
2378
- "epoch": 1.9920458944849189,
2379
- "grad_norm": 18.966585159301758,
2380
- "learning_rate": 1.866461570549588e-05,
2381
- "loss": 2.6339,
2382
- "step": 28300
2383
- },
2384
- {
2385
- "epoch": 1.999084925914194,
2386
- "grad_norm": 18.27726173400879,
2387
- "learning_rate": 1.8534258003962876e-05,
2388
- "loss": 2.6527,
2389
- "step": 28400
2390
- },
2391
- {
2392
- "epoch": 2.0061239573434695,
2393
- "grad_norm": 16.40408706665039,
2394
- "learning_rate": 1.840390030242987e-05,
2395
- "loss": 2.6285,
2396
- "step": 28500
2397
- },
2398
- {
2399
- "epoch": 2.0061239573434695,
2400
- "eval_runtime": 192.5321,
2401
- "eval_samples_per_second": 147.575,
2402
- "eval_steps_per_second": 18.449,
2403
- "step": 28500
2404
- },
2405
- {
2406
- "epoch": 2.013162988772745,
2407
- "grad_norm": 13.299867630004883,
2408
- "learning_rate": 1.827354260089686e-05,
2409
- "loss": 2.6401,
2410
- "step": 28600
2411
- },
2412
- {
2413
- "epoch": 2.0202020202020203,
2414
- "grad_norm": 16.995622634887695,
2415
- "learning_rate": 1.8143184899363855e-05,
2416
- "loss": 2.5986,
2417
- "step": 28700
2418
- },
2419
- {
2420
- "epoch": 2.0272410516312958,
2421
- "grad_norm": 18.69041633605957,
2422
- "learning_rate": 1.801282719783085e-05,
2423
- "loss": 2.644,
2424
- "step": 28800
2425
- },
2426
- {
2427
- "epoch": 2.0342800830605707,
2428
- "grad_norm": 20.12238883972168,
2429
- "learning_rate": 1.7882469496297843e-05,
2430
- "loss": 2.6802,
2431
- "step": 28900
2432
- },
2433
- {
2434
- "epoch": 2.041319114489846,
2435
- "grad_norm": 14.631281852722168,
2436
- "learning_rate": 1.7752111794764837e-05,
2437
- "loss": 2.6362,
2438
- "step": 29000
2439
- },
2440
- {
2441
- "epoch": 2.041319114489846,
2442
- "eval_runtime": 193.4189,
2443
- "eval_samples_per_second": 146.899,
2444
- "eval_steps_per_second": 18.364,
2445
- "step": 29000
2446
- },
2447
- {
2448
- "epoch": 2.0483581459191216,
2449
- "grad_norm": 16.323118209838867,
2450
- "learning_rate": 1.7621754093231828e-05,
2451
- "loss": 2.6249,
2452
- "step": 29100
2453
- },
2454
- {
2455
- "epoch": 2.055397177348397,
2456
- "grad_norm": 14.623433113098145,
2457
- "learning_rate": 1.7491396391698822e-05,
2458
- "loss": 2.6324,
2459
- "step": 29200
2460
- },
2461
- {
2462
- "epoch": 2.062436208777672,
2463
- "grad_norm": 19.917098999023438,
2464
- "learning_rate": 1.7361038690165816e-05,
2465
- "loss": 2.6893,
2466
- "step": 29300
2467
- },
2468
- {
2469
- "epoch": 2.0694752402069474,
2470
- "grad_norm": 14.357760429382324,
2471
- "learning_rate": 1.7230680988632807e-05,
2472
- "loss": 2.5841,
2473
- "step": 29400
2474
- },
2475
- {
2476
- "epoch": 2.076514271636223,
2477
- "grad_norm": 15.798065185546875,
2478
- "learning_rate": 1.7100323287099804e-05,
2479
- "loss": 2.6374,
2480
- "step": 29500
2481
- },
2482
- {
2483
- "epoch": 2.076514271636223,
2484
- "eval_runtime": 191.4525,
2485
- "eval_samples_per_second": 148.408,
2486
- "eval_steps_per_second": 18.553,
2487
- "step": 29500
2488
- },
2489
- {
2490
- "epoch": 2.0835533030654982,
2491
- "grad_norm": 19.128459930419922,
2492
- "learning_rate": 1.6969965585566798e-05,
2493
- "loss": 2.6451,
2494
- "step": 29600
2495
- },
2496
- {
2497
- "epoch": 2.0905923344947737,
2498
- "grad_norm": 22.39739990234375,
2499
- "learning_rate": 1.683960788403379e-05,
2500
- "loss": 2.6732,
2501
- "step": 29700
2502
- },
2503
- {
2504
- "epoch": 2.0976313659240486,
2505
- "grad_norm": 21.8306827545166,
2506
- "learning_rate": 1.6709250182500783e-05,
2507
- "loss": 2.6381,
2508
- "step": 29800
2509
- },
2510
- {
2511
- "epoch": 2.104670397353324,
2512
- "grad_norm": 16.79404640197754,
2513
- "learning_rate": 1.6578892480967774e-05,
2514
- "loss": 2.6643,
2515
- "step": 29900
2516
- },
2517
- {
2518
- "epoch": 2.1117094287825995,
2519
- "grad_norm": 20.273427963256836,
2520
- "learning_rate": 1.644853477943477e-05,
2521
- "loss": 2.6409,
2522
- "step": 30000
2523
- },
2524
- {
2525
- "epoch": 2.1117094287825995,
2526
- "eval_runtime": 192.3103,
2527
- "eval_samples_per_second": 147.746,
2528
- "eval_steps_per_second": 18.47,
2529
- "step": 30000
2530
- },
2531
- {
2532
- "epoch": 2.118748460211875,
2533
- "grad_norm": 16.260501861572266,
2534
- "learning_rate": 1.6318177077901765e-05,
2535
- "loss": 2.6085,
2536
- "step": 30100
2537
- },
2538
- {
2539
- "epoch": 2.1257874916411503,
2540
- "grad_norm": 17.500699996948242,
2541
- "learning_rate": 1.6187819376368756e-05,
2542
- "loss": 2.5923,
2543
- "step": 30200
2544
- },
2545
- {
2546
- "epoch": 2.1328265230704253,
2547
- "grad_norm": 19.523569107055664,
2548
- "learning_rate": 1.605746167483575e-05,
2549
- "loss": 2.562,
2550
- "step": 30300
2551
- },
2552
- {
2553
- "epoch": 2.1398655544997007,
2554
- "grad_norm": 16.805545806884766,
2555
- "learning_rate": 1.5927103973302744e-05,
2556
- "loss": 2.632,
2557
- "step": 30400
2558
- },
2559
- {
2560
- "epoch": 2.146904585928976,
2561
- "grad_norm": 14.419663429260254,
2562
- "learning_rate": 1.5796746271769735e-05,
2563
- "loss": 2.6956,
2564
- "step": 30500
2565
- },
2566
- {
2567
- "epoch": 2.146904585928976,
2568
- "eval_runtime": 192.4179,
2569
- "eval_samples_per_second": 147.663,
2570
- "eval_steps_per_second": 18.46,
2571
- "step": 30500
2572
- },
2573
- {
2574
- "epoch": 2.1539436173582516,
2575
- "grad_norm": 14.469121932983398,
2576
- "learning_rate": 1.5666388570236732e-05,
2577
- "loss": 2.6734,
2578
- "step": 30600
2579
- },
2580
- {
2581
- "epoch": 2.160982648787527,
2582
- "grad_norm": 14.521267890930176,
2583
- "learning_rate": 1.5536030868703723e-05,
2584
- "loss": 2.6272,
2585
- "step": 30700
2586
- },
2587
- {
2588
- "epoch": 2.168021680216802,
2589
- "grad_norm": 15.565622329711914,
2590
- "learning_rate": 1.5405673167170717e-05,
2591
- "loss": 2.5995,
2592
- "step": 30800
2593
- },
2594
- {
2595
- "epoch": 2.1750607116460774,
2596
- "grad_norm": 18.500350952148438,
2597
- "learning_rate": 1.527531546563771e-05,
2598
- "loss": 2.7019,
2599
- "step": 30900
2600
- },
2601
- {
2602
- "epoch": 2.182099743075353,
2603
- "grad_norm": 18.180660247802734,
2604
- "learning_rate": 1.5144957764104704e-05,
2605
- "loss": 2.6347,
2606
- "step": 31000
2607
- },
2608
- {
2609
- "epoch": 2.182099743075353,
2610
- "eval_runtime": 194.3931,
2611
- "eval_samples_per_second": 146.163,
2612
- "eval_steps_per_second": 18.272,
2613
- "step": 31000
2614
- },
2615
- {
2616
- "epoch": 2.1891387745046282,
2617
- "grad_norm": 15.68535041809082,
2618
- "learning_rate": 1.5014600062571698e-05,
2619
- "loss": 2.6679,
2620
- "step": 31100
2621
- },
2622
- {
2623
- "epoch": 2.1961778059339037,
2624
- "grad_norm": 18.195068359375,
2625
- "learning_rate": 1.4884242361038692e-05,
2626
- "loss": 2.6152,
2627
- "step": 31200
2628
- },
2629
- {
2630
- "epoch": 2.2032168373631786,
2631
- "grad_norm": 19.41796875,
2632
- "learning_rate": 1.4753884659505684e-05,
2633
- "loss": 2.6453,
2634
- "step": 31300
2635
- },
2636
- {
2637
- "epoch": 2.210255868792454,
2638
- "grad_norm": 16.178791046142578,
2639
- "learning_rate": 1.4623526957972677e-05,
2640
- "loss": 2.6175,
2641
- "step": 31400
2642
- },
2643
- {
2644
- "epoch": 2.2172949002217295,
2645
- "grad_norm": 17.970273971557617,
2646
- "learning_rate": 1.4493169256439671e-05,
2647
- "loss": 2.5633,
2648
- "step": 31500
2649
- },
2650
- {
2651
- "epoch": 2.2172949002217295,
2652
- "eval_runtime": 194.9638,
2653
- "eval_samples_per_second": 145.735,
2654
- "eval_steps_per_second": 18.219,
2655
- "step": 31500
2656
- },
2657
- {
2658
- "epoch": 2.224333931651005,
2659
- "grad_norm": 13.679678916931152,
2660
- "learning_rate": 1.4362811554906663e-05,
2661
- "loss": 2.5433,
2662
- "step": 31600
2663
- },
2664
- {
2665
- "epoch": 2.2313729630802803,
2666
- "grad_norm": 18.061559677124023,
2667
- "learning_rate": 1.423245385337366e-05,
2668
- "loss": 2.6087,
2669
- "step": 31700
2670
- },
2671
- {
2672
- "epoch": 2.2384119945095553,
2673
- "grad_norm": 20.32142448425293,
2674
- "learning_rate": 1.4102096151840652e-05,
2675
- "loss": 2.6079,
2676
- "step": 31800
2677
- },
2678
- {
2679
- "epoch": 2.2454510259388307,
2680
- "grad_norm": 16.483491897583008,
2681
- "learning_rate": 1.3971738450307646e-05,
2682
- "loss": 2.6693,
2683
- "step": 31900
2684
- },
2685
- {
2686
- "epoch": 2.252490057368106,
2687
- "grad_norm": 16.30838394165039,
2688
- "learning_rate": 1.3841380748774638e-05,
2689
- "loss": 2.5292,
2690
- "step": 32000
2691
- },
2692
- {
2693
- "epoch": 2.252490057368106,
2694
- "eval_runtime": 192.332,
2695
- "eval_samples_per_second": 147.729,
2696
- "eval_steps_per_second": 18.468,
2697
- "step": 32000
2698
- },
2699
- {
2700
- "epoch": 2.2595290887973816,
2701
- "grad_norm": 18.353946685791016,
2702
- "learning_rate": 1.371102304724163e-05,
2703
- "loss": 2.7162,
2704
- "step": 32100
2705
- },
2706
- {
2707
- "epoch": 2.266568120226657,
2708
- "grad_norm": 14.376470565795898,
2709
- "learning_rate": 1.3580665345708626e-05,
2710
- "loss": 2.6581,
2711
- "step": 32200
2712
- },
2713
- {
2714
- "epoch": 2.273607151655932,
2715
- "grad_norm": 16.625110626220703,
2716
- "learning_rate": 1.3450307644175619e-05,
2717
- "loss": 2.5675,
2718
- "step": 32300
2719
- },
2720
- {
2721
- "epoch": 2.2806461830852074,
2722
- "grad_norm": 17.9268798828125,
2723
- "learning_rate": 1.3319949942642613e-05,
2724
- "loss": 2.6001,
2725
- "step": 32400
2726
- },
2727
- {
2728
- "epoch": 2.287685214514483,
2729
- "grad_norm": 23.196901321411133,
2730
- "learning_rate": 1.3189592241109605e-05,
2731
- "loss": 2.6545,
2732
- "step": 32500
2733
- },
2734
- {
2735
- "epoch": 2.287685214514483,
2736
- "eval_runtime": 192.8693,
2737
- "eval_samples_per_second": 147.317,
2738
- "eval_steps_per_second": 18.417,
2739
- "step": 32500
2740
- },
2741
- {
2742
- "epoch": 2.2947242459437582,
2743
- "grad_norm": 18.328662872314453,
2744
- "learning_rate": 1.3059234539576598e-05,
2745
- "loss": 2.5476,
2746
- "step": 32600
2747
- },
2748
- {
2749
- "epoch": 2.3017632773730337,
2750
- "grad_norm": 16.62209129333496,
2751
- "learning_rate": 1.2928876838043592e-05,
2752
- "loss": 2.6778,
2753
- "step": 32700
2754
- },
2755
- {
2756
- "epoch": 2.3088023088023086,
2757
- "grad_norm": 15.676456451416016,
2758
- "learning_rate": 1.2798519136510586e-05,
2759
- "loss": 2.6188,
2760
- "step": 32800
2761
- },
2762
- {
2763
- "epoch": 2.315841340231584,
2764
- "grad_norm": 21.3188533782959,
2765
- "learning_rate": 1.266816143497758e-05,
2766
- "loss": 2.5383,
2767
- "step": 32900
2768
- },
2769
- {
2770
- "epoch": 2.3228803716608595,
2771
- "grad_norm": 15.253218650817871,
2772
- "learning_rate": 1.2537803733444572e-05,
2773
- "loss": 2.6703,
2774
- "step": 33000
2775
- },
2776
- {
2777
- "epoch": 2.3228803716608595,
2778
- "eval_runtime": 192.6821,
2779
- "eval_samples_per_second": 147.46,
2780
- "eval_steps_per_second": 18.435,
2781
- "step": 33000
2782
- },
2783
- {
2784
- "epoch": 2.329919403090135,
2785
- "grad_norm": 17.341787338256836,
2786
- "learning_rate": 1.2407446031911565e-05,
2787
- "loss": 2.6903,
2788
- "step": 33100
2789
- },
2790
- {
2791
- "epoch": 2.3369584345194103,
2792
- "grad_norm": 14.856354713439941,
2793
- "learning_rate": 1.2277088330378559e-05,
2794
- "loss": 2.5655,
2795
- "step": 33200
2796
- },
2797
- {
2798
- "epoch": 2.3439974659486853,
2799
- "grad_norm": 17.669092178344727,
2800
- "learning_rate": 1.2146730628845553e-05,
2801
- "loss": 2.6723,
2802
- "step": 33300
2803
- },
2804
- {
2805
- "epoch": 2.3510364973779607,
2806
- "grad_norm": 18.183189392089844,
2807
- "learning_rate": 1.2016372927312545e-05,
2808
- "loss": 2.6732,
2809
- "step": 33400
2810
- },
2811
- {
2812
- "epoch": 2.358075528807236,
2813
- "grad_norm": 20.30499267578125,
2814
- "learning_rate": 1.188601522577954e-05,
2815
- "loss": 2.6527,
2816
- "step": 33500
2817
- },
2818
- {
2819
- "epoch": 2.358075528807236,
2820
- "eval_runtime": 192.939,
2821
- "eval_samples_per_second": 147.264,
2822
- "eval_steps_per_second": 18.41,
2823
- "step": 33500
2824
- },
2825
- {
2826
- "epoch": 2.3651145602365116,
2827
- "grad_norm": 22.915029525756836,
2828
- "learning_rate": 1.1755657524246532e-05,
2829
- "loss": 2.4797,
2830
- "step": 33600
2831
- },
2832
- {
2833
- "epoch": 2.3721535916657865,
2834
- "grad_norm": 16.179378509521484,
2835
- "learning_rate": 1.1625299822713528e-05,
2836
- "loss": 2.5854,
2837
- "step": 33700
2838
- },
2839
- {
2840
- "epoch": 2.379192623095062,
2841
- "grad_norm": 14.764082908630371,
2842
- "learning_rate": 1.149494212118052e-05,
2843
- "loss": 2.4972,
2844
- "step": 33800
2845
- },
2846
- {
2847
- "epoch": 2.3862316545243374,
2848
- "grad_norm": 21.402334213256836,
2849
- "learning_rate": 1.1364584419647513e-05,
2850
- "loss": 2.5752,
2851
- "step": 33900
2852
- },
2853
- {
2854
- "epoch": 2.393270685953613,
2855
- "grad_norm": 19.00446891784668,
2856
- "learning_rate": 1.1234226718114507e-05,
2857
- "loss": 2.4806,
2858
- "step": 34000
2859
- },
2860
- {
2861
- "epoch": 2.393270685953613,
2862
- "eval_runtime": 192.7281,
2863
- "eval_samples_per_second": 147.425,
2864
- "eval_steps_per_second": 18.43,
2865
- "step": 34000
2866
- },
2867
- {
2868
- "epoch": 2.4003097173828882,
2869
- "grad_norm": 21.23725700378418,
2870
- "learning_rate": 1.11038690165815e-05,
2871
- "loss": 2.5424,
2872
- "step": 34100
2873
- },
2874
- {
2875
- "epoch": 2.407348748812163,
2876
- "grad_norm": 14.942157745361328,
2877
- "learning_rate": 1.0973511315048493e-05,
2878
- "loss": 2.5926,
2879
- "step": 34200
2880
- },
2881
- {
2882
- "epoch": 2.4143877802414386,
2883
- "grad_norm": 17.429502487182617,
2884
- "learning_rate": 1.0843153613515487e-05,
2885
- "loss": 2.5892,
2886
- "step": 34300
2887
- },
2888
- {
2889
- "epoch": 2.421426811670714,
2890
- "grad_norm": 15.42565631866455,
2891
- "learning_rate": 1.071279591198248e-05,
2892
- "loss": 2.5758,
2893
- "step": 34400
2894
- },
2895
- {
2896
- "epoch": 2.4284658430999895,
2897
- "grad_norm": 20.9206600189209,
2898
- "learning_rate": 1.0582438210449474e-05,
2899
- "loss": 2.6666,
2900
- "step": 34500
2901
- },
2902
- {
2903
- "epoch": 2.4284658430999895,
2904
- "eval_runtime": 192.2868,
2905
- "eval_samples_per_second": 147.764,
2906
- "eval_steps_per_second": 18.472,
2907
- "step": 34500
2908
- },
2909
- {
2910
- "epoch": 2.435504874529265,
2911
- "grad_norm": 16.189416885375977,
2912
- "learning_rate": 1.0452080508916468e-05,
2913
- "loss": 2.5727,
2914
- "step": 34600
2915
- },
2916
- {
2917
- "epoch": 2.44254390595854,
2918
- "grad_norm": 17.95191192626953,
2919
- "learning_rate": 1.032172280738346e-05,
2920
- "loss": 2.6171,
2921
- "step": 34700
2922
- },
2923
- {
2924
- "epoch": 2.4495829373878153,
2925
- "grad_norm": 15.953314781188965,
2926
- "learning_rate": 1.0191365105850454e-05,
2927
- "loss": 2.5181,
2928
- "step": 34800
2929
- },
2930
- {
2931
- "epoch": 2.4566219688170907,
2932
- "grad_norm": 20.293758392333984,
2933
- "learning_rate": 1.0061007404317447e-05,
2934
- "loss": 2.6109,
2935
- "step": 34900
2936
- },
2937
- {
2938
- "epoch": 2.463661000246366,
2939
- "grad_norm": 13.837769508361816,
2940
- "learning_rate": 9.930649702784441e-06,
2941
- "loss": 2.6833,
2942
- "step": 35000
2943
- },
2944
- {
2945
- "epoch": 2.463661000246366,
2946
- "eval_runtime": 192.3429,
2947
- "eval_samples_per_second": 147.721,
2948
- "eval_steps_per_second": 18.467,
2949
- "step": 35000
2950
- },
2951
- {
2952
- "epoch": 2.4707000316756416,
2953
- "grad_norm": 15.594371795654297,
2954
- "learning_rate": 9.800292001251435e-06,
2955
- "loss": 2.6111,
2956
- "step": 35100
2957
- },
2958
- {
2959
- "epoch": 2.4777390631049165,
2960
- "grad_norm": 18.549043655395508,
2961
- "learning_rate": 9.669934299718427e-06,
2962
- "loss": 2.5622,
2963
- "step": 35200
2964
- },
2965
- {
2966
- "epoch": 2.484778094534192,
2967
- "grad_norm": 15.56165599822998,
2968
- "learning_rate": 9.53957659818542e-06,
2969
- "loss": 2.5254,
2970
- "step": 35300
2971
- },
2972
- {
2973
- "epoch": 2.4918171259634674,
2974
- "grad_norm": 14.361612319946289,
2975
- "learning_rate": 9.409218896652416e-06,
2976
- "loss": 2.5388,
2977
- "step": 35400
2978
- },
2979
- {
2980
- "epoch": 2.498856157392743,
2981
- "grad_norm": 17.944364547729492,
2982
- "learning_rate": 9.278861195119408e-06,
2983
- "loss": 2.5671,
2984
- "step": 35500
2985
- },
2986
- {
2987
- "epoch": 2.498856157392743,
2988
- "eval_runtime": 193.2612,
2989
- "eval_samples_per_second": 147.019,
2990
- "eval_steps_per_second": 18.379,
2991
- "step": 35500
2992
- },
2993
- {
2994
- "epoch": 2.5058951888220182,
2995
- "grad_norm": 15.994379043579102,
2996
- "learning_rate": 9.1485034935864e-06,
2997
- "loss": 2.4926,
2998
- "step": 35600
2999
- },
3000
- {
3001
- "epoch": 2.512934220251293,
3002
- "grad_norm": 15.721161842346191,
3003
- "learning_rate": 9.018145792053395e-06,
3004
- "loss": 2.5956,
3005
- "step": 35700
3006
- },
3007
- {
3008
- "epoch": 2.5199732516805686,
3009
- "grad_norm": 21.510955810546875,
3010
- "learning_rate": 8.887788090520389e-06,
3011
- "loss": 2.6592,
3012
- "step": 35800
3013
- },
3014
- {
3015
- "epoch": 2.527012283109844,
3016
- "grad_norm": 16.77272605895996,
3017
- "learning_rate": 8.757430388987383e-06,
3018
- "loss": 2.655,
3019
- "step": 35900
3020
- },
3021
- {
3022
- "epoch": 2.5340513145391195,
3023
- "grad_norm": 18.944421768188477,
3024
- "learning_rate": 8.627072687454375e-06,
3025
- "loss": 2.5201,
3026
- "step": 36000
3027
- },
3028
- {
3029
- "epoch": 2.5340513145391195,
3030
- "eval_runtime": 192.2731,
3031
- "eval_samples_per_second": 147.774,
3032
- "eval_steps_per_second": 18.474,
3033
- "step": 36000
3034
- },
3035
- {
3036
- "epoch": 2.541090345968395,
3037
- "grad_norm": 19.00555992126465,
3038
- "learning_rate": 8.496714985921368e-06,
3039
- "loss": 2.5858,
3040
- "step": 36100
3041
- },
3042
- {
3043
- "epoch": 2.54812937739767,
3044
- "grad_norm": 16.338956832885742,
3045
- "learning_rate": 8.366357284388362e-06,
3046
- "loss": 2.5963,
3047
- "step": 36200
3048
- },
3049
- {
3050
- "epoch": 2.5551684088269453,
3051
- "grad_norm": 15.704483032226562,
3052
- "learning_rate": 8.235999582855356e-06,
3053
- "loss": 2.5504,
3054
- "step": 36300
3055
- },
3056
- {
3057
- "epoch": 2.5622074402562207,
3058
- "grad_norm": 17.013628005981445,
3059
- "learning_rate": 8.105641881322348e-06,
3060
- "loss": 2.6663,
3061
- "step": 36400
3062
- },
3063
- {
3064
- "epoch": 2.569246471685496,
3065
- "grad_norm": 16.901050567626953,
3066
- "learning_rate": 7.975284179789342e-06,
3067
- "loss": 2.5827,
3068
- "step": 36500
3069
- },
3070
- {
3071
- "epoch": 2.569246471685496,
3072
- "eval_runtime": 192.5506,
3073
- "eval_samples_per_second": 147.561,
3074
- "eval_steps_per_second": 18.447,
3075
- "step": 36500
3076
- },
3077
- {
3078
- "epoch": 2.5762855031147716,
3079
- "grad_norm": 16.243534088134766,
3080
- "learning_rate": 7.844926478256335e-06,
3081
- "loss": 2.6065,
3082
- "step": 36600
3083
- },
3084
- {
3085
- "epoch": 2.5833245345440465,
3086
- "grad_norm": 17.0561580657959,
3087
- "learning_rate": 7.714568776723329e-06,
3088
- "loss": 2.5166,
3089
- "step": 36700
3090
- },
3091
- {
3092
- "epoch": 2.590363565973322,
3093
- "grad_norm": 14.800107955932617,
3094
- "learning_rate": 7.584211075190323e-06,
3095
- "loss": 2.6966,
3096
- "step": 36800
3097
- },
3098
- {
3099
- "epoch": 2.5974025974025974,
3100
- "grad_norm": 17.22756576538086,
3101
- "learning_rate": 7.453853373657315e-06,
3102
- "loss": 2.5921,
3103
- "step": 36900
3104
- },
3105
- {
3106
- "epoch": 2.604441628831873,
3107
- "grad_norm": 16.94314956665039,
3108
- "learning_rate": 7.32349567212431e-06,
3109
- "loss": 2.7039,
3110
- "step": 37000
3111
- },
3112
- {
3113
- "epoch": 2.604441628831873,
3114
- "eval_runtime": 192.7372,
3115
- "eval_samples_per_second": 147.418,
3116
- "eval_steps_per_second": 18.429,
3117
- "step": 37000
3118
- },
3119
- {
3120
- "epoch": 2.6114806602611482,
3121
- "grad_norm": 15.262337684631348,
3122
- "learning_rate": 7.193137970591303e-06,
3123
- "loss": 2.5053,
3124
- "step": 37100
3125
- },
3126
- {
3127
- "epoch": 2.618519691690423,
3128
- "grad_norm": 16.485326766967773,
3129
- "learning_rate": 7.062780269058296e-06,
3130
- "loss": 2.6282,
3131
- "step": 37200
3132
- },
3133
- {
3134
- "epoch": 2.6255587231196986,
3135
- "grad_norm": 23.574670791625977,
3136
- "learning_rate": 6.93242256752529e-06,
3137
- "loss": 2.5995,
3138
- "step": 37300
3139
- },
3140
- {
3141
- "epoch": 2.632597754548974,
3142
- "grad_norm": 16.39130973815918,
3143
- "learning_rate": 6.802064865992283e-06,
3144
- "loss": 2.5522,
3145
- "step": 37400
3146
- },
3147
- {
3148
- "epoch": 2.6396367859782495,
3149
- "grad_norm": 20.67544174194336,
3150
- "learning_rate": 6.671707164459276e-06,
3151
- "loss": 2.6411,
3152
- "step": 37500
3153
- },
3154
- {
3155
- "epoch": 2.6396367859782495,
3156
- "eval_runtime": 192.0322,
3157
- "eval_samples_per_second": 147.96,
3158
- "eval_steps_per_second": 18.497,
3159
- "step": 37500
3160
- },
3161
- {
3162
- "epoch": 2.646675817407525,
3163
- "grad_norm": 17.333271026611328,
3164
- "learning_rate": 6.54134946292627e-06,
3165
- "loss": 2.6002,
3166
- "step": 37600
3167
- },
3168
- {
3169
- "epoch": 2.6537148488368,
3170
- "grad_norm": 17.444929122924805,
3171
- "learning_rate": 6.410991761393263e-06,
3172
- "loss": 2.4802,
3173
- "step": 37700
3174
- },
3175
- {
3176
- "epoch": 2.6607538802660753,
3177
- "grad_norm": 17.62455940246582,
3178
- "learning_rate": 6.2806340598602564e-06,
3179
- "loss": 2.5169,
3180
- "step": 37800
3181
- },
3182
- {
3183
- "epoch": 2.6677929116953507,
3184
- "grad_norm": 23.869504928588867,
3185
- "learning_rate": 6.1502763583272506e-06,
3186
- "loss": 2.6119,
3187
- "step": 37900
3188
- },
3189
- {
3190
- "epoch": 2.674831943124626,
3191
- "grad_norm": 14.378959655761719,
3192
- "learning_rate": 6.019918656794243e-06,
3193
- "loss": 2.6484,
3194
- "step": 38000
3195
- },
3196
- {
3197
- "epoch": 2.674831943124626,
3198
- "eval_runtime": 191.7147,
3199
- "eval_samples_per_second": 148.205,
3200
- "eval_steps_per_second": 18.528,
3201
- "step": 38000
3202
- },
3203
- {
3204
- "epoch": 2.6818709745539016,
3205
- "grad_norm": 13.199753761291504,
3206
- "learning_rate": 5.889560955261237e-06,
3207
- "loss": 2.5929,
3208
- "step": 38100
3209
- },
3210
- {
3211
- "epoch": 2.6889100059831765,
3212
- "grad_norm": 19.931673049926758,
3213
- "learning_rate": 5.75920325372823e-06,
3214
- "loss": 2.5691,
3215
- "step": 38200
3216
- },
3217
- {
3218
- "epoch": 2.695949037412452,
3219
- "grad_norm": 16.0571346282959,
3220
- "learning_rate": 5.6288455521952244e-06,
3221
- "loss": 2.5593,
3222
- "step": 38300
3223
- },
3224
- {
3225
- "epoch": 2.7029880688417274,
3226
- "grad_norm": 14.09821605682373,
3227
- "learning_rate": 5.498487850662217e-06,
3228
- "loss": 2.5663,
3229
- "step": 38400
3230
- },
3231
- {
3232
- "epoch": 2.710027100271003,
3233
- "grad_norm": 16.2088680267334,
3234
- "learning_rate": 5.368130149129211e-06,
3235
- "loss": 2.5763,
3236
- "step": 38500
3237
- },
3238
- {
3239
- "epoch": 2.710027100271003,
3240
- "eval_runtime": 192.1652,
3241
- "eval_samples_per_second": 147.857,
3242
- "eval_steps_per_second": 18.484,
3243
- "step": 38500
3244
- },
3245
- {
3246
- "epoch": 2.717066131700278,
3247
- "grad_norm": 26.869508743286133,
3248
- "learning_rate": 5.237772447596204e-06,
3249
- "loss": 2.7026,
3250
- "step": 38600
3251
- },
3252
- {
3253
- "epoch": 2.724105163129553,
3254
- "grad_norm": 17.842239379882812,
3255
- "learning_rate": 5.107414746063198e-06,
3256
- "loss": 2.69,
3257
- "step": 38700
3258
- },
3259
- {
3260
- "epoch": 2.7311441945588286,
3261
- "grad_norm": 17.31543731689453,
3262
- "learning_rate": 4.977057044530191e-06,
3263
- "loss": 2.6621,
3264
- "step": 38800
3265
- },
3266
- {
3267
- "epoch": 2.738183225988104,
3268
- "grad_norm": 15.826437950134277,
3269
- "learning_rate": 4.846699342997185e-06,
3270
- "loss": 2.7036,
3271
- "step": 38900
3272
- },
3273
- {
3274
- "epoch": 2.7452222574173795,
3275
- "grad_norm": 16.656599044799805,
3276
- "learning_rate": 4.716341641464178e-06,
3277
- "loss": 2.5078,
3278
- "step": 39000
3279
- },
3280
- {
3281
- "epoch": 2.7452222574173795,
3282
- "eval_runtime": 191.4626,
3283
- "eval_samples_per_second": 148.4,
3284
- "eval_steps_per_second": 18.552,
3285
- "step": 39000
3286
- },
3287
- {
3288
- "epoch": 2.752261288846655,
3289
- "grad_norm": 16.501192092895508,
3290
- "learning_rate": 4.585983939931171e-06,
3291
- "loss": 2.5312,
3292
- "step": 39100
3293
- },
3294
- {
3295
- "epoch": 2.75930032027593,
3296
- "grad_norm": 17.555389404296875,
3297
- "learning_rate": 4.455626238398165e-06,
3298
- "loss": 2.5059,
3299
- "step": 39200
3300
- },
3301
- {
3302
- "epoch": 2.7663393517052053,
3303
- "grad_norm": 18.289548873901367,
3304
- "learning_rate": 4.325268536865158e-06,
3305
- "loss": 2.6702,
3306
- "step": 39300
3307
- },
3308
- {
3309
- "epoch": 2.7733783831344807,
3310
- "grad_norm": 15.688879013061523,
3311
- "learning_rate": 4.194910835332152e-06,
3312
- "loss": 2.5357,
3313
- "step": 39400
3314
- },
3315
- {
3316
- "epoch": 2.780417414563756,
3317
- "grad_norm": 14.281635284423828,
3318
- "learning_rate": 4.064553133799144e-06,
3319
- "loss": 2.6129,
3320
- "step": 39500
3321
- },
3322
- {
3323
- "epoch": 2.780417414563756,
3324
- "eval_runtime": 193.5326,
3325
- "eval_samples_per_second": 146.812,
3326
- "eval_steps_per_second": 18.353,
3327
- "step": 39500
3328
- },
3329
- {
3330
- "epoch": 2.7874564459930316,
3331
- "grad_norm": 22.23700523376465,
3332
- "learning_rate": 3.9341954322661385e-06,
3333
- "loss": 2.5335,
3334
- "step": 39600
3335
- },
3336
- {
3337
- "epoch": 2.7944954774223065,
3338
- "grad_norm": 17.91628074645996,
3339
- "learning_rate": 3.803837730733132e-06,
3340
- "loss": 2.5757,
3341
- "step": 39700
3342
- },
3343
- {
3344
- "epoch": 2.801534508851582,
3345
- "grad_norm": 16.670568466186523,
3346
- "learning_rate": 3.6734800292001254e-06,
3347
- "loss": 2.5679,
3348
- "step": 39800
3349
- },
3350
- {
3351
- "epoch": 2.8085735402808574,
3352
- "grad_norm": 17.128202438354492,
3353
- "learning_rate": 3.5431223276671187e-06,
3354
- "loss": 2.5285,
3355
- "step": 39900
3356
- },
3357
- {
3358
- "epoch": 2.815612571710133,
3359
- "grad_norm": 14.024889945983887,
3360
- "learning_rate": 3.412764626134112e-06,
3361
- "loss": 2.515,
3362
- "step": 40000
3363
- },
3364
- {
3365
- "epoch": 2.815612571710133,
3366
- "eval_runtime": 192.6175,
3367
- "eval_samples_per_second": 147.51,
3368
- "eval_steps_per_second": 18.441,
3369
- "step": 40000
3370
- },
3371
- {
3372
- "epoch": 2.822651603139408,
3373
- "grad_norm": 16.694087982177734,
3374
- "learning_rate": 3.2824069246011056e-06,
3375
- "loss": 2.625,
3376
- "step": 40100
3377
- },
3378
- {
3379
- "epoch": 2.829690634568683,
3380
- "grad_norm": 22.633140563964844,
3381
- "learning_rate": 3.1520492230680985e-06,
3382
- "loss": 2.5637,
3383
- "step": 40200
3384
- },
3385
- {
3386
- "epoch": 2.8367296659979586,
3387
- "grad_norm": 18.231454849243164,
3388
- "learning_rate": 3.021691521535092e-06,
3389
- "loss": 2.5672,
3390
- "step": 40300
3391
- },
3392
- {
3393
- "epoch": 2.843768697427234,
3394
- "grad_norm": 15.228378295898438,
3395
- "learning_rate": 2.891333820002086e-06,
3396
- "loss": 2.6203,
3397
- "step": 40400
3398
- },
3399
- {
3400
- "epoch": 2.8508077288565095,
3401
- "grad_norm": 19.437833786010742,
3402
- "learning_rate": 2.760976118469079e-06,
3403
- "loss": 2.626,
3404
- "step": 40500
3405
- },
3406
- {
3407
- "epoch": 2.8508077288565095,
3408
- "eval_runtime": 192.5589,
3409
- "eval_samples_per_second": 147.555,
3410
- "eval_steps_per_second": 18.446,
3411
- "step": 40500
3412
- },
3413
- {
3414
- "epoch": 2.857846760285785,
3415
- "grad_norm": 16.506317138671875,
3416
- "learning_rate": 2.630618416936073e-06,
3417
- "loss": 2.5625,
3418
- "step": 40600
3419
- },
3420
- {
3421
- "epoch": 2.86488579171506,
3422
- "grad_norm": 25.00144386291504,
3423
- "learning_rate": 2.500260715403066e-06,
3424
- "loss": 2.616,
3425
- "step": 40700
3426
- },
3427
- {
3428
- "epoch": 2.8719248231443353,
3429
- "grad_norm": 16.7978572845459,
3430
- "learning_rate": 2.3699030138700597e-06,
3431
- "loss": 2.5959,
3432
- "step": 40800
3433
- },
3434
- {
3435
- "epoch": 2.8789638545736107,
3436
- "grad_norm": 15.795037269592285,
3437
- "learning_rate": 2.239545312337053e-06,
3438
- "loss": 2.6245,
3439
- "step": 40900
3440
- },
3441
- {
3442
- "epoch": 2.886002886002886,
3443
- "grad_norm": 16.411415100097656,
3444
- "learning_rate": 2.1091876108040467e-06,
3445
- "loss": 2.6368,
3446
- "step": 41000
3447
- },
3448
- {
3449
- "epoch": 2.886002886002886,
3450
- "eval_runtime": 192.2421,
3451
- "eval_samples_per_second": 147.798,
3452
- "eval_steps_per_second": 18.477,
3453
- "step": 41000
3454
- },
3455
- {
3456
- "epoch": 2.8930419174321615,
3457
- "grad_norm": 16.8485050201416,
3458
- "learning_rate": 1.97882990927104e-06,
3459
- "loss": 2.5946,
3460
- "step": 41100
3461
- },
3462
- {
3463
- "epoch": 2.9000809488614365,
3464
- "grad_norm": 15.294781684875488,
3465
- "learning_rate": 1.8484722077380334e-06,
3466
- "loss": 2.6035,
3467
- "step": 41200
3468
- },
3469
- {
3470
- "epoch": 2.907119980290712,
3471
- "grad_norm": 26.89401626586914,
3472
- "learning_rate": 1.7181145062050267e-06,
3473
- "loss": 2.6353,
3474
- "step": 41300
3475
- },
3476
- {
3477
- "epoch": 2.9141590117199874,
3478
- "grad_norm": 13.004213333129883,
3479
- "learning_rate": 1.58775680467202e-06,
3480
- "loss": 2.6299,
3481
- "step": 41400
3482
- },
3483
- {
3484
- "epoch": 2.921198043149263,
3485
- "grad_norm": 17.197162628173828,
3486
- "learning_rate": 1.4573991031390136e-06,
3487
- "loss": 2.6031,
3488
- "step": 41500
3489
- },
3490
- {
3491
- "epoch": 2.921198043149263,
3492
- "eval_runtime": 192.0882,
3493
- "eval_samples_per_second": 147.916,
3494
- "eval_steps_per_second": 18.492,
3495
- "step": 41500
3496
- },
3497
- {
3498
- "epoch": 2.928237074578538,
3499
- "grad_norm": 16.04857063293457,
3500
- "learning_rate": 1.327041401606007e-06,
3501
- "loss": 2.5869,
3502
- "step": 41600
3503
- },
3504
- {
3505
- "epoch": 2.935276106007813,
3506
- "grad_norm": 14.147359848022461,
3507
- "learning_rate": 1.1966837000730005e-06,
3508
- "loss": 2.5851,
3509
- "step": 41700
3510
- },
3511
- {
3512
- "epoch": 2.9423151374370886,
3513
- "grad_norm": 17.802715301513672,
3514
- "learning_rate": 1.0663259985399938e-06,
3515
- "loss": 2.4637,
3516
- "step": 41800
3517
- },
3518
- {
3519
- "epoch": 2.949354168866364,
3520
- "grad_norm": 20.130615234375,
3521
- "learning_rate": 9.359682970069872e-07,
3522
- "loss": 2.5418,
3523
- "step": 41900
3524
- },
3525
- {
3526
- "epoch": 2.9563932002956395,
3527
- "grad_norm": 18.158117294311523,
3528
- "learning_rate": 8.056105954739805e-07,
3529
- "loss": 2.5456,
3530
- "step": 42000
3531
- },
3532
- {
3533
- "epoch": 2.9563932002956395,
3534
- "eval_runtime": 192.2432,
3535
- "eval_samples_per_second": 147.797,
3536
- "eval_steps_per_second": 18.477,
3537
- "step": 42000
3538
- },
3539
- {
3540
- "epoch": 2.963432231724915,
3541
- "grad_norm": 15.710502624511719,
3542
- "learning_rate": 6.75252893940974e-07,
3543
- "loss": 2.6082,
3544
- "step": 42100
3545
- },
3546
- {
3547
- "epoch": 2.97047126315419,
3548
- "grad_norm": 15.316740989685059,
3549
- "learning_rate": 5.448951924079675e-07,
3550
- "loss": 2.5569,
3551
- "step": 42200
3552
- },
3553
- {
3554
- "epoch": 2.9775102945834653,
3555
- "grad_norm": 17.120691299438477,
3556
- "learning_rate": 4.145374908749609e-07,
3557
- "loss": 2.6017,
3558
- "step": 42300
3559
- },
3560
- {
3561
- "epoch": 2.9845493260127407,
3562
- "grad_norm": 14.475923538208008,
3563
- "learning_rate": 2.841797893419543e-07,
3564
- "loss": 2.5627,
3565
- "step": 42400
3566
- },
3567
- {
3568
- "epoch": 2.991588357442016,
3569
- "grad_norm": 16.937416076660156,
3570
- "learning_rate": 1.5382208780894776e-07,
3571
- "loss": 2.557,
3572
- "step": 42500
3573
- },
3574
- {
3575
- "epoch": 2.991588357442016,
3576
- "eval_runtime": 192.9822,
3577
- "eval_samples_per_second": 147.231,
3578
- "eval_steps_per_second": 18.406,
3579
- "step": 42500
3580
- },
3581
- {
3582
- "epoch": 2.9986273888712915,
3583
- "grad_norm": 25.87067413330078,
3584
- "learning_rate": 2.3464386275941184e-08,
3585
- "loss": 2.5282,
3586
- "step": 42600
3587
- }
3588
- ],
3589
- "logging_steps": 100,
3590
- "max_steps": 42618,
3591
- "num_input_tokens_seen": 0,
3592
- "num_train_epochs": 3,
3593
- "save_steps": 500,
3594
- "stateful_callbacks": {
3595
- "TrainerControl": {
3596
- "args": {
3597
- "should_epoch_stop": false,
3598
- "should_evaluate": false,
3599
- "should_log": false,
3600
- "should_save": true,
3601
- "should_training_stop": true
3602
- },
3603
- "attributes": {}
3604
- }
3605
- },
3606
- "total_flos": 1.1152556854519332e+16,
3607
- "train_batch_size": 8,
3608
- "trial_name": null,
3609
- "trial_params": null
3610
- }