Boyue27 commited on
Commit
30cddce
·
verified ·
1 Parent(s): bfb8bac

Delete trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +0 -1238
trainer_state.json DELETED
@@ -1,1238 +0,0 @@
1
- {
2
- "best_metric": 0.9822313785552979,
3
- "best_model_checkpoint": "/proj/berzelius-2023-338/users/x_boyji/model_hf/neft_medico-mixtral/checkpoint-500",
4
- "epoch": 0.9995841995841996,
5
- "eval_steps": 500,
6
- "global_step": 601,
7
- "is_hyper_param_search": false,
8
- "is_local_process_zero": true,
9
- "is_world_process_zero": true,
10
- "log_history": [
11
- {
12
- "epoch": 0.0,
13
- "learning_rate": 6.000000000000001e-07,
14
- "loss": 0.5838,
15
- "step": 3
16
- },
17
- {
18
- "epoch": 0.01,
19
- "learning_rate": 1.2000000000000002e-06,
20
- "loss": 0.6053,
21
- "step": 6
22
- },
23
- {
24
- "epoch": 0.01,
25
- "learning_rate": 1.8000000000000001e-06,
26
- "loss": 0.6054,
27
- "step": 9
28
- },
29
- {
30
- "epoch": 0.02,
31
- "learning_rate": 2.4000000000000003e-06,
32
- "loss": 0.5732,
33
- "step": 12
34
- },
35
- {
36
- "epoch": 0.02,
37
- "learning_rate": 3e-06,
38
- "loss": 0.5784,
39
- "step": 15
40
- },
41
- {
42
- "epoch": 0.03,
43
- "learning_rate": 3.6000000000000003e-06,
44
- "loss": 0.5493,
45
- "step": 18
46
- },
47
- {
48
- "epoch": 0.03,
49
- "learning_rate": 4.2000000000000004e-06,
50
- "loss": 0.5446,
51
- "step": 21
52
- },
53
- {
54
- "epoch": 0.04,
55
- "learning_rate": 4.800000000000001e-06,
56
- "loss": 0.5137,
57
- "step": 24
58
- },
59
- {
60
- "epoch": 0.04,
61
- "learning_rate": 5.400000000000001e-06,
62
- "loss": 0.5384,
63
- "step": 27
64
- },
65
- {
66
- "epoch": 0.05,
67
- "learning_rate": 6e-06,
68
- "loss": 0.5515,
69
- "step": 30
70
- },
71
- {
72
- "epoch": 0.05,
73
- "learning_rate": 6.600000000000001e-06,
74
- "loss": 0.499,
75
- "step": 33
76
- },
77
- {
78
- "epoch": 0.06,
79
- "learning_rate": 7.2000000000000005e-06,
80
- "loss": 0.5197,
81
- "step": 36
82
- },
83
- {
84
- "epoch": 0.06,
85
- "learning_rate": 7.800000000000002e-06,
86
- "loss": 0.4922,
87
- "step": 39
88
- },
89
- {
90
- "epoch": 0.07,
91
- "learning_rate": 8.400000000000001e-06,
92
- "loss": 0.477,
93
- "step": 42
94
- },
95
- {
96
- "epoch": 0.07,
97
- "learning_rate": 9e-06,
98
- "loss": 0.4598,
99
- "step": 45
100
- },
101
- {
102
- "epoch": 0.08,
103
- "learning_rate": 9.600000000000001e-06,
104
- "loss": 0.4465,
105
- "step": 48
106
- },
107
- {
108
- "epoch": 0.08,
109
- "learning_rate": 1.02e-05,
110
- "loss": 0.4188,
111
- "step": 51
112
- },
113
- {
114
- "epoch": 0.09,
115
- "learning_rate": 1.0800000000000002e-05,
116
- "loss": 0.4152,
117
- "step": 54
118
- },
119
- {
120
- "epoch": 0.09,
121
- "learning_rate": 1.14e-05,
122
- "loss": 0.4168,
123
- "step": 57
124
- },
125
- {
126
- "epoch": 0.1,
127
- "learning_rate": 1.2e-05,
128
- "loss": 0.4185,
129
- "step": 60
130
- },
131
- {
132
- "epoch": 0.1,
133
- "learning_rate": 1.2600000000000001e-05,
134
- "loss": 0.4046,
135
- "step": 63
136
- },
137
- {
138
- "epoch": 0.11,
139
- "learning_rate": 1.3200000000000002e-05,
140
- "loss": 0.4202,
141
- "step": 66
142
- },
143
- {
144
- "epoch": 0.11,
145
- "learning_rate": 1.38e-05,
146
- "loss": 0.3991,
147
- "step": 69
148
- },
149
- {
150
- "epoch": 0.12,
151
- "learning_rate": 1.4400000000000001e-05,
152
- "loss": 0.4156,
153
- "step": 72
154
- },
155
- {
156
- "epoch": 0.12,
157
- "learning_rate": 1.5000000000000002e-05,
158
- "loss": 0.409,
159
- "step": 75
160
- },
161
- {
162
- "epoch": 0.13,
163
- "learning_rate": 1.5600000000000003e-05,
164
- "loss": 0.3675,
165
- "step": 78
166
- },
167
- {
168
- "epoch": 0.13,
169
- "learning_rate": 1.62e-05,
170
- "loss": 0.3818,
171
- "step": 81
172
- },
173
- {
174
- "epoch": 0.14,
175
- "learning_rate": 1.6800000000000002e-05,
176
- "loss": 0.3978,
177
- "step": 84
178
- },
179
- {
180
- "epoch": 0.14,
181
- "learning_rate": 1.7400000000000003e-05,
182
- "loss": 0.4231,
183
- "step": 87
184
- },
185
- {
186
- "epoch": 0.15,
187
- "learning_rate": 1.8e-05,
188
- "loss": 0.4763,
189
- "step": 90
190
- },
191
- {
192
- "epoch": 0.15,
193
- "learning_rate": 1.86e-05,
194
- "loss": 0.5048,
195
- "step": 93
196
- },
197
- {
198
- "epoch": 0.16,
199
- "learning_rate": 1.9200000000000003e-05,
200
- "loss": 0.5014,
201
- "step": 96
202
- },
203
- {
204
- "epoch": 0.16,
205
- "learning_rate": 1.98e-05,
206
- "loss": 0.5062,
207
- "step": 99
208
- },
209
- {
210
- "epoch": 0.17,
211
- "learning_rate": 1.9999213590779245e-05,
212
- "loss": 0.3718,
213
- "step": 102
214
- },
215
- {
216
- "epoch": 0.17,
217
- "learning_rate": 1.999508528057537e-05,
218
- "loss": 0.3792,
219
- "step": 105
220
- },
221
- {
222
- "epoch": 0.18,
223
- "learning_rate": 1.998741992607014e-05,
224
- "loss": 0.3934,
225
- "step": 108
226
- },
227
- {
228
- "epoch": 0.18,
229
- "learning_rate": 1.9976220239866563e-05,
230
- "loss": 0.3623,
231
- "step": 111
232
- },
233
- {
234
- "epoch": 0.19,
235
- "learning_rate": 1.9961490185291128e-05,
236
- "loss": 0.3572,
237
- "step": 114
238
- },
239
- {
240
- "epoch": 0.19,
241
- "learning_rate": 1.994323497499127e-05,
242
- "loss": 0.3458,
243
- "step": 117
244
- },
245
- {
246
- "epoch": 0.2,
247
- "learning_rate": 1.9921461069090715e-05,
248
- "loss": 0.3684,
249
- "step": 120
250
- },
251
- {
252
- "epoch": 0.2,
253
- "learning_rate": 1.9896176172903384e-05,
254
- "loss": 0.3509,
255
- "step": 123
256
- },
257
- {
258
- "epoch": 0.21,
259
- "learning_rate": 1.9867389234206655e-05,
260
- "loss": 0.3355,
261
- "step": 126
262
- },
263
- {
264
- "epoch": 0.21,
265
- "learning_rate": 1.9835110440074938e-05,
266
- "loss": 0.343,
267
- "step": 129
268
- },
269
- {
270
- "epoch": 0.22,
271
- "learning_rate": 1.9799351213274688e-05,
272
- "loss": 0.3523,
273
- "step": 132
274
- },
275
- {
276
- "epoch": 0.22,
277
- "learning_rate": 1.9760124208222128e-05,
278
- "loss": 0.3373,
279
- "step": 135
280
- },
281
- {
282
- "epoch": 0.23,
283
- "learning_rate": 1.9717443306505143e-05,
284
- "loss": 0.3578,
285
- "step": 138
286
- },
287
- {
288
- "epoch": 0.23,
289
- "learning_rate": 1.9671323611970864e-05,
290
- "loss": 0.3636,
291
- "step": 141
292
- },
293
- {
294
- "epoch": 0.24,
295
- "learning_rate": 1.962178144538076e-05,
296
- "loss": 0.3512,
297
- "step": 144
298
- },
299
- {
300
- "epoch": 0.24,
301
- "learning_rate": 1.956883433863508e-05,
302
- "loss": 0.3416,
303
- "step": 147
304
- },
305
- {
306
- "epoch": 0.25,
307
- "learning_rate": 1.951250102856866e-05,
308
- "loss": 0.3323,
309
- "step": 150
310
- },
311
- {
312
- "epoch": 0.25,
313
- "learning_rate": 1.9452801450320398e-05,
314
- "loss": 0.3514,
315
- "step": 153
316
- },
317
- {
318
- "epoch": 0.26,
319
- "learning_rate": 1.9389756730278625e-05,
320
- "loss": 0.3597,
321
- "step": 156
322
- },
323
- {
324
- "epoch": 0.26,
325
- "learning_rate": 1.9323389178604942e-05,
326
- "loss": 0.3397,
327
- "step": 159
328
- },
329
- {
330
- "epoch": 0.27,
331
- "learning_rate": 1.925372228133912e-05,
332
- "loss": 0.3592,
333
- "step": 162
334
- },
335
- {
336
- "epoch": 0.27,
337
- "learning_rate": 1.9180780692087903e-05,
338
- "loss": 0.3443,
339
- "step": 165
340
- },
341
- {
342
- "epoch": 0.28,
343
- "learning_rate": 1.9104590223300656e-05,
344
- "loss": 0.3458,
345
- "step": 168
346
- },
347
- {
348
- "epoch": 0.28,
349
- "learning_rate": 1.9025177837134858e-05,
350
- "loss": 0.3472,
351
- "step": 171
352
- },
353
- {
354
- "epoch": 0.29,
355
- "learning_rate": 1.8942571635914814e-05,
356
- "loss": 0.3521,
357
- "step": 174
358
- },
359
- {
360
- "epoch": 0.29,
361
- "learning_rate": 1.8856800852186826e-05,
362
- "loss": 0.3542,
363
- "step": 177
364
- },
365
- {
366
- "epoch": 0.3,
367
- "learning_rate": 1.8767895838374438e-05,
368
- "loss": 0.3609,
369
- "step": 180
370
- },
371
- {
372
- "epoch": 0.3,
373
- "learning_rate": 1.8675888056037378e-05,
374
- "loss": 0.3543,
375
- "step": 183
376
- },
377
- {
378
- "epoch": 0.31,
379
- "learning_rate": 1.8580810064737967e-05,
380
- "loss": 0.3461,
381
- "step": 186
382
- },
383
- {
384
- "epoch": 0.31,
385
- "learning_rate": 1.8482695510519027e-05,
386
- "loss": 0.3458,
387
- "step": 189
388
- },
389
- {
390
- "epoch": 0.32,
391
- "learning_rate": 1.8381579113997255e-05,
392
- "loss": 0.3441,
393
- "step": 192
394
- },
395
- {
396
- "epoch": 0.32,
397
- "learning_rate": 1.827749665807635e-05,
398
- "loss": 0.342,
399
- "step": 195
400
- },
401
- {
402
- "epoch": 0.33,
403
- "learning_rate": 1.8170484975284233e-05,
404
- "loss": 0.3469,
405
- "step": 198
406
- },
407
- {
408
- "epoch": 0.33,
409
- "learning_rate": 1.8060581934738783e-05,
410
- "loss": 0.3537,
411
- "step": 201
412
- },
413
- {
414
- "epoch": 0.34,
415
- "learning_rate": 1.794782642874681e-05,
416
- "loss": 0.3649,
417
- "step": 204
418
- },
419
- {
420
- "epoch": 0.34,
421
- "learning_rate": 1.7832258359040885e-05,
422
- "loss": 0.3693,
423
- "step": 207
424
- },
425
- {
426
- "epoch": 0.35,
427
- "learning_rate": 1.771391862265899e-05,
428
- "loss": 0.3474,
429
- "step": 210
430
- },
431
- {
432
- "epoch": 0.35,
433
- "learning_rate": 1.759284909747195e-05,
434
- "loss": 0.3561,
435
- "step": 213
436
- },
437
- {
438
- "epoch": 0.36,
439
- "learning_rate": 1.7469092627363738e-05,
440
- "loss": 0.3549,
441
- "step": 216
442
- },
443
- {
444
- "epoch": 0.36,
445
- "learning_rate": 1.7342693007069974e-05,
446
- "loss": 0.3516,
447
- "step": 219
448
- },
449
- {
450
- "epoch": 0.37,
451
- "learning_rate": 1.7213694966679913e-05,
452
- "loss": 0.3489,
453
- "step": 222
454
- },
455
- {
456
- "epoch": 0.37,
457
- "learning_rate": 1.708214415580742e-05,
458
- "loss": 0.3596,
459
- "step": 225
460
- },
461
- {
462
- "epoch": 0.38,
463
- "learning_rate": 1.6948087127436573e-05,
464
- "loss": 0.3536,
465
- "step": 228
466
- },
467
- {
468
- "epoch": 0.38,
469
- "learning_rate": 1.6811571321447568e-05,
470
- "loss": 0.3481,
471
- "step": 231
472
- },
473
- {
474
- "epoch": 0.39,
475
- "learning_rate": 1.667264504782877e-05,
476
- "loss": 0.3246,
477
- "step": 234
478
- },
479
- {
480
- "epoch": 0.39,
481
- "learning_rate": 1.653135746958086e-05,
482
- "loss": 0.342,
483
- "step": 237
484
- },
485
- {
486
- "epoch": 0.4,
487
- "learning_rate": 1.63877585853191e-05,
488
- "loss": 0.3331,
489
- "step": 240
490
- },
491
- {
492
- "epoch": 0.4,
493
- "learning_rate": 1.624189921157992e-05,
494
- "loss": 0.358,
495
- "step": 243
496
- },
497
- {
498
- "epoch": 0.41,
499
- "learning_rate": 1.6093830964838036e-05,
500
- "loss": 0.3373,
501
- "step": 246
502
- },
503
- {
504
- "epoch": 0.41,
505
- "learning_rate": 1.5943606243240504e-05,
506
- "loss": 0.3473,
507
- "step": 249
508
- },
509
- {
510
- "epoch": 0.42,
511
- "learning_rate": 1.5791278208064142e-05,
512
- "loss": 0.3417,
513
- "step": 252
514
- },
515
- {
516
- "epoch": 0.42,
517
- "learning_rate": 1.5636900764902914e-05,
518
- "loss": 0.311,
519
- "step": 255
520
- },
521
- {
522
- "epoch": 0.43,
523
- "learning_rate": 1.5480528544591907e-05,
524
- "loss": 0.3262,
525
- "step": 258
526
- },
527
- {
528
- "epoch": 0.43,
529
- "learning_rate": 1.5322216883874643e-05,
530
- "loss": 0.3483,
531
- "step": 261
532
- },
533
- {
534
- "epoch": 0.44,
535
- "learning_rate": 1.5162021805820633e-05,
536
- "loss": 0.3525,
537
- "step": 264
538
- },
539
- {
540
- "epoch": 0.44,
541
- "learning_rate": 1.5000000000000002e-05,
542
- "loss": 0.35,
543
- "step": 267
544
- },
545
- {
546
- "epoch": 0.45,
547
- "learning_rate": 1.4836208802422287e-05,
548
- "loss": 0.3567,
549
- "step": 270
550
- },
551
- {
552
- "epoch": 0.45,
553
- "learning_rate": 1.4670706175246463e-05,
554
- "loss": 0.3525,
555
- "step": 273
556
- },
557
- {
558
- "epoch": 0.46,
559
- "learning_rate": 1.450355068626939e-05,
560
- "loss": 0.3346,
561
- "step": 276
562
- },
563
- {
564
- "epoch": 0.46,
565
- "learning_rate": 1.4334801488199941e-05,
566
- "loss": 0.3263,
567
- "step": 279
568
- },
569
- {
570
- "epoch": 0.47,
571
- "learning_rate": 1.4164518297726129e-05,
572
- "loss": 0.3239,
573
- "step": 282
574
- },
575
- {
576
- "epoch": 0.47,
577
- "learning_rate": 1.399276137438267e-05,
578
- "loss": 0.3437,
579
- "step": 285
580
- },
581
- {
582
- "epoch": 0.48,
583
- "learning_rate": 1.3819591499226446e-05,
584
- "loss": 0.3259,
585
- "step": 288
586
- },
587
- {
588
- "epoch": 0.48,
589
- "learning_rate": 1.364506995332739e-05,
590
- "loss": 0.3395,
591
- "step": 291
592
- },
593
- {
594
- "epoch": 0.49,
595
- "learning_rate": 1.3469258496082452e-05,
596
- "loss": 0.3318,
597
- "step": 294
598
- },
599
- {
600
- "epoch": 0.49,
601
- "learning_rate": 1.329221934336028e-05,
602
- "loss": 0.3271,
603
- "step": 297
604
- },
605
- {
606
- "epoch": 0.5,
607
- "learning_rate": 1.311401514548438e-05,
608
- "loss": 0.3351,
609
- "step": 300
610
- },
611
- {
612
- "epoch": 0.5,
613
- "learning_rate": 1.2934708965062508e-05,
614
- "loss": 0.3045,
615
- "step": 303
616
- },
617
- {
618
- "epoch": 0.51,
619
- "learning_rate": 1.2754364254670192e-05,
620
- "loss": 0.3203,
621
- "step": 306
622
- },
623
- {
624
- "epoch": 0.51,
625
- "learning_rate": 1.257304483439623e-05,
626
- "loss": 0.3354,
627
- "step": 309
628
- },
629
- {
630
- "epoch": 0.52,
631
- "learning_rate": 1.2390814869258155e-05,
632
- "loss": 0.3396,
633
- "step": 312
634
- },
635
- {
636
- "epoch": 0.52,
637
- "learning_rate": 1.22077388464956e-05,
638
- "loss": 0.3288,
639
- "step": 315
640
- },
641
- {
642
- "epoch": 0.53,
643
- "learning_rate": 1.2023881552749693e-05,
644
- "loss": 0.3417,
645
- "step": 318
646
- },
647
- {
648
- "epoch": 0.53,
649
- "learning_rate": 1.1839308051136432e-05,
650
- "loss": 0.3115,
651
- "step": 321
652
- },
653
- {
654
- "epoch": 0.54,
655
- "learning_rate": 1.1654083658222287e-05,
656
- "loss": 0.3172,
657
- "step": 324
658
- },
659
- {
660
- "epoch": 0.54,
661
- "learning_rate": 1.146827392091007e-05,
662
- "loss": 0.3278,
663
- "step": 327
664
- },
665
- {
666
- "epoch": 0.55,
667
- "learning_rate": 1.1281944593243289e-05,
668
- "loss": 0.3036,
669
- "step": 330
670
- },
671
- {
672
- "epoch": 0.55,
673
- "learning_rate": 1.1095161613137242e-05,
674
- "loss": 0.3289,
675
- "step": 333
676
- },
677
- {
678
- "epoch": 0.56,
679
- "learning_rate": 1.0907991079045006e-05,
680
- "loss": 0.3106,
681
- "step": 336
682
- },
683
- {
684
- "epoch": 0.56,
685
- "learning_rate": 1.0720499226566616e-05,
686
- "loss": 0.3219,
687
- "step": 339
688
- },
689
- {
690
- "epoch": 0.57,
691
- "learning_rate": 1.0532752405009733e-05,
692
- "loss": 0.3192,
693
- "step": 342
694
- },
695
- {
696
- "epoch": 0.57,
697
- "learning_rate": 1.034481705391005e-05,
698
- "loss": 0.3253,
699
- "step": 345
700
- },
701
- {
702
- "epoch": 0.58,
703
- "learning_rate": 1.015675967951979e-05,
704
- "loss": 0.3095,
705
- "step": 348
706
- },
707
- {
708
- "epoch": 0.58,
709
- "learning_rate": 9.96864683127257e-06,
710
- "loss": 0.3196,
711
- "step": 351
712
- },
713
- {
714
- "epoch": 0.59,
715
- "learning_rate": 9.780545078233015e-06,
716
- "loss": 0.3088,
717
- "step": 354
718
- },
719
- {
720
- "epoch": 0.59,
721
- "learning_rate": 9.592520985539382e-06,
722
- "loss": 0.333,
723
- "step": 357
724
- },
725
- {
726
- "epoch": 0.6,
727
- "learning_rate": 9.404641090847628e-06,
728
- "loss": 0.3232,
729
- "step": 360
730
- },
731
- {
732
- "epoch": 0.6,
733
- "learning_rate": 9.21697188078514e-06,
734
- "loss": 0.309,
735
- "step": 363
736
- },
737
- {
738
- "epoch": 0.61,
739
- "learning_rate": 9.029579767422592e-06,
740
- "loss": 0.3295,
741
- "step": 366
742
- },
743
- {
744
- "epoch": 0.61,
745
- "learning_rate": 8.842531064772137e-06,
746
- "loss": 0.3019,
747
- "step": 369
748
- },
749
- {
750
- "epoch": 0.62,
751
- "learning_rate": 8.655891965320292e-06,
752
- "loss": 0.3022,
753
- "step": 372
754
- },
755
- {
756
- "epoch": 0.62,
757
- "learning_rate": 8.469728516603886e-06,
758
- "loss": 0.3172,
759
- "step": 375
760
- },
761
- {
762
- "epoch": 0.63,
763
- "learning_rate": 8.284106597837244e-06,
764
- "loss": 0.3092,
765
- "step": 378
766
- },
767
- {
768
- "epoch": 0.63,
769
- "learning_rate": 8.099091896598964e-06,
770
- "loss": 0.445,
771
- "step": 381
772
- },
773
- {
774
- "epoch": 0.64,
775
- "learning_rate": 7.914749885586517e-06,
776
- "loss": 0.4746,
777
- "step": 384
778
- },
779
- {
780
- "epoch": 0.64,
781
- "learning_rate": 7.731145799446846e-06,
782
- "loss": 0.5142,
783
- "step": 387
784
- },
785
- {
786
- "epoch": 0.65,
787
- "learning_rate": 7.548344611691272e-06,
788
- "loss": 0.4106,
789
- "step": 390
790
- },
791
- {
792
- "epoch": 0.65,
793
- "learning_rate": 7.366411011702754e-06,
794
- "loss": 0.3275,
795
- "step": 393
796
- },
797
- {
798
- "epoch": 0.66,
799
- "learning_rate": 7.185409381843727e-06,
800
- "loss": 0.3061,
801
- "step": 396
802
- },
803
- {
804
- "epoch": 0.66,
805
- "learning_rate": 7.00540377467256e-06,
806
- "loss": 0.3315,
807
- "step": 399
808
- },
809
- {
810
- "epoch": 0.67,
811
- "learning_rate": 6.826457890276773e-06,
812
- "loss": 0.2961,
813
- "step": 402
814
- },
815
- {
816
- "epoch": 0.67,
817
- "learning_rate": 6.648635053730931e-06,
818
- "loss": 0.3256,
819
- "step": 405
820
- },
821
- {
822
- "epoch": 0.68,
823
- "learning_rate": 6.471998192687266e-06,
824
- "loss": 0.3129,
825
- "step": 408
826
- },
827
- {
828
- "epoch": 0.68,
829
- "learning_rate": 6.296609815106971e-06,
830
- "loss": 0.2998,
831
- "step": 411
832
- },
833
- {
834
- "epoch": 0.69,
835
- "learning_rate": 6.122531987139955e-06,
836
- "loss": 0.2994,
837
- "step": 414
838
- },
839
- {
840
- "epoch": 0.69,
841
- "learning_rate": 5.949826311161007e-06,
842
- "loss": 0.3078,
843
- "step": 417
844
- },
845
- {
846
- "epoch": 0.7,
847
- "learning_rate": 5.77855390397001e-06,
848
- "loss": 0.3174,
849
- "step": 420
850
- },
851
- {
852
- "epoch": 0.7,
853
- "learning_rate": 5.6087753751640725e-06,
854
- "loss": 0.306,
855
- "step": 423
856
- },
857
- {
858
- "epoch": 0.71,
859
- "learning_rate": 5.440550805689075e-06,
860
- "loss": 0.3253,
861
- "step": 426
862
- },
863
- {
864
- "epoch": 0.71,
865
- "learning_rate": 5.273939726578328e-06,
866
- "loss": 0.2987,
867
- "step": 429
868
- },
869
- {
870
- "epoch": 0.72,
871
- "learning_rate": 5.109001097885837e-06,
872
- "loss": 0.314,
873
- "step": 432
874
- },
875
- {
876
- "epoch": 0.72,
877
- "learning_rate": 4.945793287821604e-06,
878
- "loss": 0.3028,
879
- "step": 435
880
- },
881
- {
882
- "epoch": 0.73,
883
- "learning_rate": 4.784374052096396e-06,
884
- "loss": 0.2971,
885
- "step": 438
886
- },
887
- {
888
- "epoch": 0.73,
889
- "learning_rate": 4.62480051348324e-06,
890
- "loss": 0.305,
891
- "step": 441
892
- },
893
- {
894
- "epoch": 0.74,
895
- "learning_rate": 4.467129141602906e-06,
896
- "loss": 0.3213,
897
- "step": 444
898
- },
899
- {
900
- "epoch": 0.74,
901
- "learning_rate": 4.3114157329405494e-06,
902
- "loss": 0.3192,
903
- "step": 447
904
- },
905
- {
906
- "epoch": 0.75,
907
- "learning_rate": 4.157715391100509e-06,
908
- "loss": 0.2989,
909
- "step": 450
910
- },
911
- {
912
- "epoch": 0.75,
913
- "learning_rate": 4.006082507306379e-06,
914
- "loss": 0.3097,
915
- "step": 453
916
- },
917
- {
918
- "epoch": 0.76,
919
- "learning_rate": 3.856570741153087e-06,
920
- "loss": 0.3195,
921
- "step": 456
922
- },
923
- {
924
- "epoch": 0.76,
925
- "learning_rate": 3.709233001617961e-06,
926
- "loss": 0.2907,
927
- "step": 459
928
- },
929
- {
930
- "epoch": 0.77,
931
- "learning_rate": 3.564121428337369e-06,
932
- "loss": 0.3178,
933
- "step": 462
934
- },
935
- {
936
- "epoch": 0.77,
937
- "learning_rate": 3.4212873731556364e-06,
938
- "loss": 0.3083,
939
- "step": 465
940
- },
941
- {
942
- "epoch": 0.78,
943
- "learning_rate": 3.2807813819527377e-06,
944
- "loss": 0.2938,
945
- "step": 468
946
- },
947
- {
948
- "epoch": 0.78,
949
- "learning_rate": 3.1426531767572e-06,
950
- "loss": 0.2801,
951
- "step": 471
952
- },
953
- {
954
- "epoch": 0.79,
955
- "learning_rate": 3.006951638150567e-06,
956
- "loss": 0.3083,
957
- "step": 474
958
- },
959
- {
960
- "epoch": 0.79,
961
- "learning_rate": 2.8737247879696007e-06,
962
- "loss": 0.2814,
963
- "step": 477
964
- },
965
- {
966
- "epoch": 0.8,
967
- "learning_rate": 2.743019772312412e-06,
968
- "loss": 0.2997,
969
- "step": 480
970
- },
971
- {
972
- "epoch": 0.8,
973
- "learning_rate": 2.6148828448544574e-06,
974
- "loss": 0.288,
975
- "step": 483
976
- },
977
- {
978
- "epoch": 0.81,
979
- "learning_rate": 2.4893593504803824e-06,
980
- "loss": 0.2975,
981
- "step": 486
982
- },
983
- {
984
- "epoch": 0.81,
985
- "learning_rate": 2.3664937092374497e-06,
986
- "loss": 0.2885,
987
- "step": 489
988
- },
989
- {
990
- "epoch": 0.82,
991
- "learning_rate": 2.2463294006162185e-06,
992
- "loss": 0.2934,
993
- "step": 492
994
- },
995
- {
996
- "epoch": 0.82,
997
- "learning_rate": 2.128908948164128e-06,
998
- "loss": 0.2907,
999
- "step": 495
1000
- },
1001
- {
1002
- "epoch": 0.83,
1003
- "learning_rate": 2.0142739044373093e-06,
1004
- "loss": 0.2912,
1005
- "step": 498
1006
- },
1007
- {
1008
- "epoch": 0.83,
1009
- "eval_loss": 0.9822313785552979,
1010
- "eval_runtime": 60.9831,
1011
- "eval_samples_per_second": 7.592,
1012
- "eval_steps_per_second": 0.951,
1013
- "step": 500
1014
- },
1015
- {
1016
- "epoch": 0.83,
1017
- "learning_rate": 1.902464836296054e-06,
1018
- "loss": 0.3011,
1019
- "step": 501
1020
- },
1021
- {
1022
- "epoch": 0.84,
1023
- "learning_rate": 1.7935213105490722e-06,
1024
- "loss": 0.2984,
1025
- "step": 504
1026
- },
1027
- {
1028
- "epoch": 0.84,
1029
- "learning_rate": 1.6874818799516468e-06,
1030
- "loss": 0.2984,
1031
- "step": 507
1032
- },
1033
- {
1034
- "epoch": 0.85,
1035
- "learning_rate": 1.5843840695626666e-06,
1036
- "loss": 0.2819,
1037
- "step": 510
1038
- },
1039
- {
1040
- "epoch": 0.85,
1041
- "learning_rate": 1.4842643634653054e-06,
1042
- "loss": 0.296,
1043
- "step": 513
1044
- },
1045
- {
1046
- "epoch": 0.86,
1047
- "learning_rate": 1.387158191856105e-06,
1048
- "loss": 0.3125,
1049
- "step": 516
1050
- },
1051
- {
1052
- "epoch": 0.86,
1053
- "learning_rate": 1.2930999185069804e-06,
1054
- "loss": 0.3011,
1055
- "step": 519
1056
- },
1057
- {
1058
- "epoch": 0.87,
1059
- "learning_rate": 1.2021228286046461e-06,
1060
- "loss": 0.3203,
1061
- "step": 522
1062
- },
1063
- {
1064
- "epoch": 0.87,
1065
- "learning_rate": 1.1142591169716931e-06,
1066
- "loss": 0.3146,
1067
- "step": 525
1068
- },
1069
- {
1070
- "epoch": 0.88,
1071
- "learning_rate": 1.029539876673541e-06,
1072
- "loss": 0.3089,
1073
- "step": 528
1074
- },
1075
- {
1076
- "epoch": 0.88,
1077
- "learning_rate": 9.47995088015281e-07,
1078
- "loss": 0.3007,
1079
- "step": 531
1080
- },
1081
- {
1082
- "epoch": 0.89,
1083
- "learning_rate": 8.696536079322904e-07,
1084
- "loss": 0.2954,
1085
- "step": 534
1086
- },
1087
- {
1088
- "epoch": 0.89,
1089
- "learning_rate": 7.945431597783904e-07,
1090
- "loss": 0.2945,
1091
- "step": 537
1092
- },
1093
- {
1094
- "epoch": 0.9,
1095
- "learning_rate": 7.226903235151439e-07,
1096
- "loss": 0.2968,
1097
- "step": 540
1098
- },
1099
- {
1100
- "epoch": 0.9,
1101
- "learning_rate": 6.541205263057936e-07,
1102
- "loss": 0.2872,
1103
- "step": 543
1104
- },
1105
- {
1106
- "epoch": 0.91,
1107
- "learning_rate": 5.888580335171368e-07,
1108
- "loss": 0.3009,
1109
- "step": 546
1110
- },
1111
- {
1112
- "epoch": 0.91,
1113
- "learning_rate": 5.269259401325256e-07,
1114
- "loss": 0.2957,
1115
- "step": 549
1116
- },
1117
- {
1118
- "epoch": 0.92,
1119
- "learning_rate": 4.683461625790697e-07,
1120
- "loss": 0.3106,
1121
- "step": 552
1122
- },
1123
- {
1124
- "epoch": 0.92,
1125
- "learning_rate": 4.13139430971875e-07,
1126
- "loss": 0.2891,
1127
- "step": 555
1128
- },
1129
- {
1130
- "epoch": 0.93,
1131
- "learning_rate": 3.61325281778111e-07,
1132
- "loss": 0.2894,
1133
- "step": 558
1134
- },
1135
- {
1136
- "epoch": 0.93,
1137
- "learning_rate": 3.129220509034725e-07,
1138
- "loss": 0.3033,
1139
- "step": 561
1140
- },
1141
- {
1142
- "epoch": 0.94,
1143
- "learning_rate": 2.6794686720350573e-07,
1144
- "loss": 0.3038,
1145
- "step": 564
1146
- },
1147
- {
1148
- "epoch": 0.94,
1149
- "learning_rate": 2.2641564642207347e-07,
1150
- "loss": 0.3026,
1151
- "step": 567
1152
- },
1153
- {
1154
- "epoch": 0.95,
1155
- "learning_rate": 1.883430855591173e-07,
1156
- "loss": 0.3132,
1157
- "step": 570
1158
- },
1159
- {
1160
- "epoch": 0.95,
1161
- "learning_rate": 1.5374265766971097e-07,
1162
- "loss": 0.3041,
1163
- "step": 573
1164
- },
1165
- {
1166
- "epoch": 0.96,
1167
- "learning_rate": 1.2262660709623232e-07,
1168
- "loss": 0.2954,
1169
- "step": 576
1170
- },
1171
- {
1172
- "epoch": 0.96,
1173
- "learning_rate": 9.500594513536354e-08,
1174
- "loss": 0.2774,
1175
- "step": 579
1176
- },
1177
- {
1178
- "epoch": 0.97,
1179
- "learning_rate": 7.08904461414217e-08,
1180
- "loss": 0.3143,
1181
- "step": 582
1182
- },
1183
- {
1184
- "epoch": 0.97,
1185
- "learning_rate": 5.0288644067435496e-08,
1186
- "loss": 0.3013,
1187
- "step": 585
1188
- },
1189
- {
1190
- "epoch": 0.98,
1191
- "learning_rate": 3.3207829445159876e-08,
1192
- "loss": 0.293,
1193
- "step": 588
1194
- },
1195
- {
1196
- "epoch": 0.98,
1197
- "learning_rate": 1.965404680511207e-08,
1198
- "loss": 0.2989,
1199
- "step": 591
1200
- },
1201
- {
1202
- "epoch": 0.99,
1203
- "learning_rate": 9.632092537551485e-09,
1204
- "loss": 0.2904,
1205
- "step": 594
1206
- },
1207
- {
1208
- "epoch": 0.99,
1209
- "learning_rate": 3.1455131951285688e-09,
1210
- "loss": 0.304,
1211
- "step": 597
1212
- },
1213
- {
1214
- "epoch": 1.0,
1215
- "learning_rate": 1.9660423784984007e-10,
1216
- "loss": 0.316,
1217
- "step": 600
1218
- },
1219
- {
1220
- "epoch": 1.0,
1221
- "step": 601,
1222
- "total_flos": 6.6035761096408695e+19,
1223
- "train_loss": 0.35220495833335025,
1224
- "train_runtime": 148180.9169,
1225
- "train_samples_per_second": 3.116,
1226
- "train_steps_per_second": 0.004
1227
- }
1228
- ],
1229
- "logging_steps": 3,
1230
- "max_steps": 601,
1231
- "num_input_tokens_seen": 0,
1232
- "num_train_epochs": 1,
1233
- "save_steps": 500,
1234
- "total_flos": 6.6035761096408695e+19,
1235
- "train_batch_size": 96,
1236
- "trial_name": null,
1237
- "trial_params": null
1238
- }