Boyue27 commited on
Commit
5190479
·
verified ·
1 Parent(s): 6a89fc3

Delete trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +0 -3660
trainer_state.json DELETED
@@ -1,3660 +0,0 @@
1
- {
2
- "best_metric": 0.9312900900840759,
3
- "best_model_checkpoint": "/proj/berzelius-2023-338/users/x_boyji/model_hf/sft_medico-mixtral/checkpoint-1000",
4
- "epoch": 2.9991683991683993,
5
- "eval_steps": 500,
6
- "global_step": 1803,
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": 2.0272,
15
- "step": 3
16
- },
17
- {
18
- "epoch": 0.01,
19
- "learning_rate": 1.2000000000000002e-06,
20
- "loss": 1.921,
21
- "step": 6
22
- },
23
- {
24
- "epoch": 0.01,
25
- "learning_rate": 1.8000000000000001e-06,
26
- "loss": 1.6681,
27
- "step": 9
28
- },
29
- {
30
- "epoch": 0.02,
31
- "learning_rate": 2.4000000000000003e-06,
32
- "loss": 1.5087,
33
- "step": 12
34
- },
35
- {
36
- "epoch": 0.02,
37
- "learning_rate": 3e-06,
38
- "loss": 1.4299,
39
- "step": 15
40
- },
41
- {
42
- "epoch": 0.03,
43
- "learning_rate": 3.6000000000000003e-06,
44
- "loss": 1.3476,
45
- "step": 18
46
- },
47
- {
48
- "epoch": 0.03,
49
- "learning_rate": 4.2000000000000004e-06,
50
- "loss": 1.3558,
51
- "step": 21
52
- },
53
- {
54
- "epoch": 0.04,
55
- "learning_rate": 4.800000000000001e-06,
56
- "loss": 1.3113,
57
- "step": 24
58
- },
59
- {
60
- "epoch": 0.04,
61
- "learning_rate": 5.400000000000001e-06,
62
- "loss": 1.3135,
63
- "step": 27
64
- },
65
- {
66
- "epoch": 0.05,
67
- "learning_rate": 6e-06,
68
- "loss": 1.3068,
69
- "step": 30
70
- },
71
- {
72
- "epoch": 0.05,
73
- "learning_rate": 6.600000000000001e-06,
74
- "loss": 1.2869,
75
- "step": 33
76
- },
77
- {
78
- "epoch": 0.06,
79
- "learning_rate": 7.2000000000000005e-06,
80
- "loss": 1.2698,
81
- "step": 36
82
- },
83
- {
84
- "epoch": 0.06,
85
- "learning_rate": 7.800000000000002e-06,
86
- "loss": 1.2347,
87
- "step": 39
88
- },
89
- {
90
- "epoch": 0.07,
91
- "learning_rate": 8.400000000000001e-06,
92
- "loss": 1.2687,
93
- "step": 42
94
- },
95
- {
96
- "epoch": 0.07,
97
- "learning_rate": 9e-06,
98
- "loss": 1.2221,
99
- "step": 45
100
- },
101
- {
102
- "epoch": 0.08,
103
- "learning_rate": 9.600000000000001e-06,
104
- "loss": 1.2231,
105
- "step": 48
106
- },
107
- {
108
- "epoch": 0.08,
109
- "learning_rate": 1.02e-05,
110
- "loss": 1.1976,
111
- "step": 51
112
- },
113
- {
114
- "epoch": 0.09,
115
- "learning_rate": 1.0800000000000002e-05,
116
- "loss": 1.1833,
117
- "step": 54
118
- },
119
- {
120
- "epoch": 0.09,
121
- "learning_rate": 1.14e-05,
122
- "loss": 1.1798,
123
- "step": 57
124
- },
125
- {
126
- "epoch": 0.1,
127
- "learning_rate": 1.2e-05,
128
- "loss": 1.1878,
129
- "step": 60
130
- },
131
- {
132
- "epoch": 0.1,
133
- "learning_rate": 1.2600000000000001e-05,
134
- "loss": 1.1788,
135
- "step": 63
136
- },
137
- {
138
- "epoch": 0.11,
139
- "learning_rate": 1.3200000000000002e-05,
140
- "loss": 1.1948,
141
- "step": 66
142
- },
143
- {
144
- "epoch": 0.11,
145
- "learning_rate": 1.38e-05,
146
- "loss": 1.1836,
147
- "step": 69
148
- },
149
- {
150
- "epoch": 0.12,
151
- "learning_rate": 1.4400000000000001e-05,
152
- "loss": 1.1756,
153
- "step": 72
154
- },
155
- {
156
- "epoch": 0.12,
157
- "learning_rate": 1.5000000000000002e-05,
158
- "loss": 1.1868,
159
- "step": 75
160
- },
161
- {
162
- "epoch": 0.13,
163
- "learning_rate": 1.5600000000000003e-05,
164
- "loss": 1.1652,
165
- "step": 78
166
- },
167
- {
168
- "epoch": 0.13,
169
- "learning_rate": 1.62e-05,
170
- "loss": 1.1688,
171
- "step": 81
172
- },
173
- {
174
- "epoch": 0.14,
175
- "learning_rate": 1.6800000000000002e-05,
176
- "loss": 1.1783,
177
- "step": 84
178
- },
179
- {
180
- "epoch": 0.14,
181
- "learning_rate": 1.7400000000000003e-05,
182
- "loss": 1.1799,
183
- "step": 87
184
- },
185
- {
186
- "epoch": 0.15,
187
- "learning_rate": 1.8e-05,
188
- "loss": 1.3097,
189
- "step": 90
190
- },
191
- {
192
- "epoch": 0.15,
193
- "learning_rate": 1.86e-05,
194
- "loss": 1.2879,
195
- "step": 93
196
- },
197
- {
198
- "epoch": 0.16,
199
- "learning_rate": 1.9200000000000003e-05,
200
- "loss": 1.35,
201
- "step": 96
202
- },
203
- {
204
- "epoch": 0.16,
205
- "learning_rate": 1.98e-05,
206
- "loss": 1.2501,
207
- "step": 99
208
- },
209
- {
210
- "epoch": 0.17,
211
- "learning_rate": 1.9999931938745066e-05,
212
- "loss": 1.1856,
213
- "step": 102
214
- },
215
- {
216
- "epoch": 0.17,
217
- "learning_rate": 1.999957461968997e-05,
218
- "loss": 1.1927,
219
- "step": 105
220
- },
221
- {
222
- "epoch": 0.18,
223
- "learning_rate": 1.9998911038450304e-05,
224
- "loss": 1.2005,
225
- "step": 108
226
- },
227
- {
228
- "epoch": 0.18,
229
- "learning_rate": 1.999794121534991e-05,
230
- "loss": 1.1375,
231
- "step": 111
232
- },
233
- {
234
- "epoch": 0.19,
235
- "learning_rate": 1.9996665180092068e-05,
236
- "loss": 1.1645,
237
- "step": 114
238
- },
239
- {
240
- "epoch": 0.19,
241
- "learning_rate": 1.9995082971758574e-05,
242
- "loss": 1.1381,
243
- "step": 117
244
- },
245
- {
246
- "epoch": 0.2,
247
- "learning_rate": 1.999319463880855e-05,
248
- "loss": 1.189,
249
- "step": 120
250
- },
251
- {
252
- "epoch": 0.2,
253
- "learning_rate": 1.9991000239076952e-05,
254
- "loss": 1.1539,
255
- "step": 123
256
- },
257
- {
258
- "epoch": 0.21,
259
- "learning_rate": 1.9988499839772805e-05,
260
- "loss": 1.1454,
261
- "step": 126
262
- },
263
- {
264
- "epoch": 0.21,
265
- "learning_rate": 1.9985693517477142e-05,
266
- "loss": 1.1611,
267
- "step": 129
268
- },
269
- {
270
- "epoch": 0.22,
271
- "learning_rate": 1.9982581358140657e-05,
272
- "loss": 1.162,
273
- "step": 132
274
- },
275
- {
276
- "epoch": 0.22,
277
- "learning_rate": 1.997916345708108e-05,
278
- "loss": 1.1244,
279
- "step": 135
280
- },
281
- {
282
- "epoch": 0.23,
283
- "learning_rate": 1.9975439918980246e-05,
284
- "loss": 1.1738,
285
- "step": 138
286
- },
287
- {
288
- "epoch": 0.23,
289
- "learning_rate": 1.99714108578809e-05,
290
- "loss": 1.1746,
291
- "step": 141
292
- },
293
- {
294
- "epoch": 0.24,
295
- "learning_rate": 1.9967076397183187e-05,
296
- "loss": 1.1332,
297
- "step": 144
298
- },
299
- {
300
- "epoch": 0.24,
301
- "learning_rate": 1.9962436669640906e-05,
302
- "loss": 1.145,
303
- "step": 147
304
- },
305
- {
306
- "epoch": 0.25,
307
- "learning_rate": 1.9957491817357407e-05,
308
- "loss": 1.1246,
309
- "step": 150
310
- },
311
- {
312
- "epoch": 0.25,
313
- "learning_rate": 1.995224199178126e-05,
314
- "loss": 1.1561,
315
- "step": 153
316
- },
317
- {
318
- "epoch": 0.26,
319
- "learning_rate": 1.9946687353701607e-05,
320
- "loss": 1.1436,
321
- "step": 156
322
- },
323
- {
324
- "epoch": 0.26,
325
- "learning_rate": 1.9940828073243244e-05,
326
- "loss": 1.1285,
327
- "step": 159
328
- },
329
- {
330
- "epoch": 0.27,
331
- "learning_rate": 1.9934664329861412e-05,
332
- "loss": 1.1578,
333
- "step": 162
334
- },
335
- {
336
- "epoch": 0.27,
337
- "learning_rate": 1.9928196312336287e-05,
338
- "loss": 1.141,
339
- "step": 165
340
- },
341
- {
342
- "epoch": 0.28,
343
- "learning_rate": 1.9921424218767215e-05,
344
- "loss": 1.1133,
345
- "step": 168
346
- },
347
- {
348
- "epoch": 0.28,
349
- "learning_rate": 1.991434825656664e-05,
350
- "loss": 1.114,
351
- "step": 171
352
- },
353
- {
354
- "epoch": 0.29,
355
- "learning_rate": 1.9906968642453745e-05,
356
- "loss": 1.1049,
357
- "step": 174
358
- },
359
- {
360
- "epoch": 0.29,
361
- "learning_rate": 1.989928560244781e-05,
362
- "loss": 1.1417,
363
- "step": 177
364
- },
365
- {
366
- "epoch": 0.3,
367
- "learning_rate": 1.9891299371861312e-05,
368
- "loss": 1.1442,
369
- "step": 180
370
- },
371
- {
372
- "epoch": 0.3,
373
- "learning_rate": 1.9883010195292702e-05,
374
- "loss": 1.1164,
375
- "step": 183
376
- },
377
- {
378
- "epoch": 0.31,
379
- "learning_rate": 1.9874418326618906e-05,
380
- "loss": 1.1319,
381
- "step": 186
382
- },
383
- {
384
- "epoch": 0.31,
385
- "learning_rate": 1.9865524028987565e-05,
386
- "loss": 1.1265,
387
- "step": 189
388
- },
389
- {
390
- "epoch": 0.32,
391
- "learning_rate": 1.9856327574808974e-05,
392
- "loss": 1.1289,
393
- "step": 192
394
- },
395
- {
396
- "epoch": 0.32,
397
- "learning_rate": 1.9846829245747725e-05,
398
- "loss": 1.1381,
399
- "step": 195
400
- },
401
- {
402
- "epoch": 0.33,
403
- "learning_rate": 1.9837029332714094e-05,
404
- "loss": 1.13,
405
- "step": 198
406
- },
407
- {
408
- "epoch": 0.33,
409
- "learning_rate": 1.9826928135855128e-05,
410
- "loss": 1.134,
411
- "step": 201
412
- },
413
- {
414
- "epoch": 0.34,
415
- "learning_rate": 1.9816525964545447e-05,
416
- "loss": 1.1173,
417
- "step": 204
418
- },
419
- {
420
- "epoch": 0.34,
421
- "learning_rate": 1.9805823137377776e-05,
422
- "loss": 1.1364,
423
- "step": 207
424
- },
425
- {
426
- "epoch": 0.35,
427
- "learning_rate": 1.979481998215318e-05,
428
- "loss": 1.1241,
429
- "step": 210
430
- },
431
- {
432
- "epoch": 0.35,
433
- "learning_rate": 1.978351683587103e-05,
434
- "loss": 1.134,
435
- "step": 213
436
- },
437
- {
438
- "epoch": 0.36,
439
- "learning_rate": 1.9771914044718682e-05,
440
- "loss": 1.1064,
441
- "step": 216
442
- },
443
- {
444
- "epoch": 0.36,
445
- "learning_rate": 1.9760011964060863e-05,
446
- "loss": 1.1324,
447
- "step": 219
448
- },
449
- {
450
- "epoch": 0.37,
451
- "learning_rate": 1.974781095842881e-05,
452
- "loss": 1.1239,
453
- "step": 222
454
- },
455
- {
456
- "epoch": 0.37,
457
- "learning_rate": 1.9735311401509078e-05,
458
- "loss": 1.1348,
459
- "step": 225
460
- },
461
- {
462
- "epoch": 0.38,
463
- "learning_rate": 1.9722513676132112e-05,
464
- "loss": 1.1425,
465
- "step": 228
466
- },
467
- {
468
- "epoch": 0.38,
469
- "learning_rate": 1.9709418174260523e-05,
470
- "loss": 1.1224,
471
- "step": 231
472
- },
473
- {
474
- "epoch": 0.39,
475
- "learning_rate": 1.9696025296977067e-05,
476
- "loss": 1.0922,
477
- "step": 234
478
- },
479
- {
480
- "epoch": 0.39,
481
- "learning_rate": 1.9682335454472382e-05,
482
- "loss": 1.1123,
483
- "step": 237
484
- },
485
- {
486
- "epoch": 0.4,
487
- "learning_rate": 1.9668349066032412e-05,
488
- "loss": 1.0966,
489
- "step": 240
490
- },
491
- {
492
- "epoch": 0.4,
493
- "learning_rate": 1.9654066560025566e-05,
494
- "loss": 1.119,
495
- "step": 243
496
- },
497
- {
498
- "epoch": 0.41,
499
- "learning_rate": 1.9639488373889597e-05,
500
- "loss": 1.0947,
501
- "step": 246
502
- },
503
- {
504
- "epoch": 0.41,
505
- "learning_rate": 1.9624614954118214e-05,
506
- "loss": 1.0871,
507
- "step": 249
508
- },
509
- {
510
- "epoch": 0.42,
511
- "learning_rate": 1.9609446756247402e-05,
512
- "loss": 1.1088,
513
- "step": 252
514
- },
515
- {
516
- "epoch": 0.42,
517
- "learning_rate": 1.9593984244841455e-05,
518
- "loss": 1.0891,
519
- "step": 255
520
- },
521
- {
522
- "epoch": 0.43,
523
- "learning_rate": 1.957822789347878e-05,
524
- "loss": 1.1095,
525
- "step": 258
526
- },
527
- {
528
- "epoch": 0.43,
529
- "learning_rate": 1.9562178184737358e-05,
530
- "loss": 1.1145,
531
- "step": 261
532
- },
533
- {
534
- "epoch": 0.44,
535
- "learning_rate": 1.9545835610179985e-05,
536
- "loss": 1.1324,
537
- "step": 264
538
- },
539
- {
540
- "epoch": 0.44,
541
- "learning_rate": 1.952920067033921e-05,
542
- "loss": 1.1047,
543
- "step": 267
544
- },
545
- {
546
- "epoch": 0.45,
547
- "learning_rate": 1.9512273874702016e-05,
548
- "loss": 1.1257,
549
- "step": 270
550
- },
551
- {
552
- "epoch": 0.45,
553
- "learning_rate": 1.9495055741694188e-05,
554
- "loss": 1.1089,
555
- "step": 273
556
- },
557
- {
558
- "epoch": 0.46,
559
- "learning_rate": 1.9477546798664463e-05,
560
- "loss": 1.0767,
561
- "step": 276
562
- },
563
- {
564
- "epoch": 0.46,
565
- "learning_rate": 1.9459747581868366e-05,
566
- "loss": 1.0819,
567
- "step": 279
568
- },
569
- {
570
- "epoch": 0.47,
571
- "learning_rate": 1.9441658636451794e-05,
572
- "loss": 1.0888,
573
- "step": 282
574
- },
575
- {
576
- "epoch": 0.47,
577
- "learning_rate": 1.94232805164343e-05,
578
- "loss": 1.1043,
579
- "step": 285
580
- },
581
- {
582
- "epoch": 0.48,
583
- "learning_rate": 1.9404613784692152e-05,
584
- "loss": 1.0753,
585
- "step": 288
586
- },
587
- {
588
- "epoch": 0.48,
589
- "learning_rate": 1.9385659012941072e-05,
590
- "loss": 1.0885,
591
- "step": 291
592
- },
593
- {
594
- "epoch": 0.49,
595
- "learning_rate": 1.9366416781718733e-05,
596
- "loss": 1.08,
597
- "step": 294
598
- },
599
- {
600
- "epoch": 0.49,
601
- "learning_rate": 1.9346887680366984e-05,
602
- "loss": 1.0894,
603
- "step": 297
604
- },
605
- {
606
- "epoch": 0.5,
607
- "learning_rate": 1.9327072307013794e-05,
608
- "loss": 1.0798,
609
- "step": 300
610
- },
611
- {
612
- "epoch": 0.5,
613
- "learning_rate": 1.9306971268554927e-05,
614
- "loss": 1.0543,
615
- "step": 303
616
- },
617
- {
618
- "epoch": 0.51,
619
- "learning_rate": 1.9286585180635375e-05,
620
- "loss": 1.0734,
621
- "step": 306
622
- },
623
- {
624
- "epoch": 0.51,
625
- "learning_rate": 1.926591466763047e-05,
626
- "loss": 1.0822,
627
- "step": 309
628
- },
629
- {
630
- "epoch": 0.52,
631
- "learning_rate": 1.9244960362626795e-05,
632
- "loss": 1.0944,
633
- "step": 312
634
- },
635
- {
636
- "epoch": 0.52,
637
- "learning_rate": 1.922372290740277e-05,
638
- "loss": 1.0789,
639
- "step": 315
640
- },
641
- {
642
- "epoch": 0.53,
643
- "learning_rate": 1.9202202952409012e-05,
644
- "loss": 1.0973,
645
- "step": 318
646
- },
647
- {
648
- "epoch": 0.53,
649
- "learning_rate": 1.9180401156748397e-05,
650
- "loss": 1.0952,
651
- "step": 321
652
- },
653
- {
654
- "epoch": 0.54,
655
- "learning_rate": 1.9158318188155882e-05,
656
- "loss": 1.0775,
657
- "step": 324
658
- },
659
- {
660
- "epoch": 0.54,
661
- "learning_rate": 1.9135954722978062e-05,
662
- "loss": 1.0793,
663
- "step": 327
664
- },
665
- {
666
- "epoch": 0.55,
667
- "learning_rate": 1.9113311446152445e-05,
668
- "loss": 1.0439,
669
- "step": 330
670
- },
671
- {
672
- "epoch": 0.55,
673
- "learning_rate": 1.909038905118647e-05,
674
- "loss": 1.0889,
675
- "step": 333
676
- },
677
- {
678
- "epoch": 0.56,
679
- "learning_rate": 1.906718824013628e-05,
680
- "loss": 1.0371,
681
- "step": 336
682
- },
683
- {
684
- "epoch": 0.56,
685
- "learning_rate": 1.9043709723585207e-05,
686
- "loss": 1.093,
687
- "step": 339
688
- },
689
- {
690
- "epoch": 0.57,
691
- "learning_rate": 1.9019954220622018e-05,
692
- "loss": 1.0692,
693
- "step": 342
694
- },
695
- {
696
- "epoch": 0.57,
697
- "learning_rate": 1.8995922458818885e-05,
698
- "loss": 1.0658,
699
- "step": 345
700
- },
701
- {
702
- "epoch": 0.58,
703
- "learning_rate": 1.8971615174209104e-05,
704
- "loss": 1.0633,
705
- "step": 348
706
- },
707
- {
708
- "epoch": 0.58,
709
- "learning_rate": 1.8947033111264558e-05,
710
- "loss": 1.0732,
711
- "step": 351
712
- },
713
- {
714
- "epoch": 0.59,
715
- "learning_rate": 1.892217702287289e-05,
716
- "loss": 1.0536,
717
- "step": 354
718
- },
719
- {
720
- "epoch": 0.59,
721
- "learning_rate": 1.889704767031449e-05,
722
- "loss": 1.0752,
723
- "step": 357
724
- },
725
- {
726
- "epoch": 0.6,
727
- "learning_rate": 1.8871645823239128e-05,
728
- "loss": 1.0933,
729
- "step": 360
730
- },
731
- {
732
- "epoch": 0.6,
733
- "learning_rate": 1.8845972259642424e-05,
734
- "loss": 1.0703,
735
- "step": 363
736
- },
737
- {
738
- "epoch": 0.61,
739
- "learning_rate": 1.8820027765841993e-05,
740
- "loss": 1.0722,
741
- "step": 366
742
- },
743
- {
744
- "epoch": 0.61,
745
- "learning_rate": 1.8793813136453364e-05,
746
- "loss": 1.0367,
747
- "step": 369
748
- },
749
- {
750
- "epoch": 0.62,
751
- "learning_rate": 1.8767329174365664e-05,
752
- "loss": 1.0361,
753
- "step": 372
754
- },
755
- {
756
- "epoch": 0.62,
757
- "learning_rate": 1.8740576690717004e-05,
758
- "loss": 1.0454,
759
- "step": 375
760
- },
761
- {
762
- "epoch": 0.63,
763
- "learning_rate": 1.8713556504869644e-05,
764
- "loss": 1.0725,
765
- "step": 378
766
- },
767
- {
768
- "epoch": 0.63,
769
- "learning_rate": 1.8686269444384905e-05,
770
- "loss": 1.1055,
771
- "step": 381
772
- },
773
- {
774
- "epoch": 0.64,
775
- "learning_rate": 1.8658716344997812e-05,
776
- "loss": 1.1818,
777
- "step": 384
778
- },
779
- {
780
- "epoch": 0.64,
781
- "learning_rate": 1.8630898050591502e-05,
782
- "loss": 1.101,
783
- "step": 387
784
- },
785
- {
786
- "epoch": 0.65,
787
- "learning_rate": 1.860281541317138e-05,
788
- "loss": 1.0709,
789
- "step": 390
790
- },
791
- {
792
- "epoch": 0.65,
793
- "learning_rate": 1.857446929283902e-05,
794
- "loss": 1.0668,
795
- "step": 393
796
- },
797
- {
798
- "epoch": 0.66,
799
- "learning_rate": 1.8545860557765828e-05,
800
- "loss": 1.0396,
801
- "step": 396
802
- },
803
- {
804
- "epoch": 0.66,
805
- "learning_rate": 1.8516990084166444e-05,
806
- "loss": 1.0795,
807
- "step": 399
808
- },
809
- {
810
- "epoch": 0.67,
811
- "learning_rate": 1.8487858756271918e-05,
812
- "loss": 1.0401,
813
- "step": 402
814
- },
815
- {
816
- "epoch": 0.67,
817
- "learning_rate": 1.845846746630261e-05,
818
- "loss": 1.0679,
819
- "step": 405
820
- },
821
- {
822
- "epoch": 0.68,
823
- "learning_rate": 1.842881711444088e-05,
824
- "loss": 1.0702,
825
- "step": 408
826
- },
827
- {
828
- "epoch": 0.68,
829
- "learning_rate": 1.8398908608803518e-05,
830
- "loss": 1.0489,
831
- "step": 411
832
- },
833
- {
834
- "epoch": 0.69,
835
- "learning_rate": 1.8368742865413905e-05,
836
- "loss": 1.0245,
837
- "step": 414
838
- },
839
- {
840
- "epoch": 0.69,
841
- "learning_rate": 1.8338320808173998e-05,
842
- "loss": 1.0603,
843
- "step": 417
844
- },
845
- {
846
- "epoch": 0.7,
847
- "learning_rate": 1.8307643368836003e-05,
848
- "loss": 1.0493,
849
- "step": 420
850
- },
851
- {
852
- "epoch": 0.7,
853
- "learning_rate": 1.8276711486973838e-05,
854
- "loss": 1.048,
855
- "step": 423
856
- },
857
- {
858
- "epoch": 0.71,
859
- "learning_rate": 1.8245526109954385e-05,
860
- "loss": 1.078,
861
- "step": 426
862
- },
863
- {
864
- "epoch": 0.71,
865
- "learning_rate": 1.8214088192908433e-05,
866
- "loss": 1.0251,
867
- "step": 429
868
- },
869
- {
870
- "epoch": 0.72,
871
- "learning_rate": 1.818239869870146e-05,
872
- "loss": 1.0558,
873
- "step": 432
874
- },
875
- {
876
- "epoch": 0.72,
877
- "learning_rate": 1.815045859790413e-05,
878
- "loss": 1.0433,
879
- "step": 435
880
- },
881
- {
882
- "epoch": 0.73,
883
- "learning_rate": 1.811826886876255e-05,
884
- "loss": 1.0484,
885
- "step": 438
886
- },
887
- {
888
- "epoch": 0.73,
889
- "learning_rate": 1.8085830497168334e-05,
890
- "loss": 1.0559,
891
- "step": 441
892
- },
893
- {
894
- "epoch": 0.74,
895
- "learning_rate": 1.805314447662841e-05,
896
- "loss": 1.0715,
897
- "step": 444
898
- },
899
- {
900
- "epoch": 0.74,
901
- "learning_rate": 1.8020211808234556e-05,
902
- "loss": 1.0418,
903
- "step": 447
904
- },
905
- {
906
- "epoch": 0.75,
907
- "learning_rate": 1.7987033500632785e-05,
908
- "loss": 1.0166,
909
- "step": 450
910
- },
911
- {
912
- "epoch": 0.75,
913
- "learning_rate": 1.7953610569992415e-05,
914
- "loss": 1.0427,
915
- "step": 453
916
- },
917
- {
918
- "epoch": 0.76,
919
- "learning_rate": 1.7919944039974962e-05,
920
- "loss": 1.0463,
921
- "step": 456
922
- },
923
- {
924
- "epoch": 0.76,
925
- "learning_rate": 1.78860349417028e-05,
926
- "loss": 1.0035,
927
- "step": 459
928
- },
929
- {
930
- "epoch": 0.77,
931
- "learning_rate": 1.7851884313727547e-05,
932
- "loss": 1.047,
933
- "step": 462
934
- },
935
- {
936
- "epoch": 0.77,
937
- "learning_rate": 1.78174932019983e-05,
938
- "loss": 1.0241,
939
- "step": 465
940
- },
941
- {
942
- "epoch": 0.78,
943
- "learning_rate": 1.7782862659829553e-05,
944
- "loss": 1.0308,
945
- "step": 468
946
- },
947
- {
948
- "epoch": 0.78,
949
- "learning_rate": 1.7747993747868985e-05,
950
- "loss": 1.021,
951
- "step": 471
952
- },
953
- {
954
- "epoch": 0.79,
955
- "learning_rate": 1.7712887534064935e-05,
956
- "loss": 1.0216,
957
- "step": 474
958
- },
959
- {
960
- "epoch": 0.79,
961
- "learning_rate": 1.7677545093633713e-05,
962
- "loss": 1.0228,
963
- "step": 477
964
- },
965
- {
966
- "epoch": 0.8,
967
- "learning_rate": 1.7641967509026667e-05,
968
- "loss": 1.02,
969
- "step": 480
970
- },
971
- {
972
- "epoch": 0.8,
973
- "learning_rate": 1.7606155869897024e-05,
974
- "loss": 1.0065,
975
- "step": 483
976
- },
977
- {
978
- "epoch": 0.81,
979
- "learning_rate": 1.757011127306653e-05,
980
- "loss": 1.0353,
981
- "step": 486
982
- },
983
- {
984
- "epoch": 0.81,
985
- "learning_rate": 1.7533834822491832e-05,
986
- "loss": 1.0346,
987
- "step": 489
988
- },
989
- {
990
- "epoch": 0.82,
991
- "learning_rate": 1.7497327629230707e-05,
992
- "loss": 1.0241,
993
- "step": 492
994
- },
995
- {
996
- "epoch": 0.82,
997
- "learning_rate": 1.746059081140798e-05,
998
- "loss": 1.0036,
999
- "step": 495
1000
- },
1001
- {
1002
- "epoch": 0.83,
1003
- "learning_rate": 1.7423625494181334e-05,
1004
- "loss": 1.0195,
1005
- "step": 498
1006
- },
1007
- {
1008
- "epoch": 0.83,
1009
- "eval_loss": 0.9461708068847656,
1010
- "eval_runtime": 61.2738,
1011
- "eval_samples_per_second": 7.556,
1012
- "eval_steps_per_second": 0.947,
1013
- "step": 500
1014
- },
1015
- {
1016
- "epoch": 0.83,
1017
- "learning_rate": 1.7386432809706795e-05,
1018
- "loss": 1.0261,
1019
- "step": 501
1020
- },
1021
- {
1022
- "epoch": 0.84,
1023
- "learning_rate": 1.734901389710411e-05,
1024
- "loss": 1.0162,
1025
- "step": 504
1026
- },
1027
- {
1028
- "epoch": 0.84,
1029
- "learning_rate": 1.7311369902421814e-05,
1030
- "loss": 1.0056,
1031
- "step": 507
1032
- },
1033
- {
1034
- "epoch": 0.85,
1035
- "learning_rate": 1.727350197860216e-05,
1036
- "loss": 1.0025,
1037
- "step": 510
1038
- },
1039
- {
1040
- "epoch": 0.85,
1041
- "learning_rate": 1.723541128544579e-05,
1042
- "loss": 1.0088,
1043
- "step": 513
1044
- },
1045
- {
1046
- "epoch": 0.86,
1047
- "learning_rate": 1.7197098989576222e-05,
1048
- "loss": 1.0285,
1049
- "step": 516
1050
- },
1051
- {
1052
- "epoch": 0.86,
1053
- "learning_rate": 1.7158566264404112e-05,
1054
- "loss": 1.0105,
1055
- "step": 519
1056
- },
1057
- {
1058
- "epoch": 0.87,
1059
- "learning_rate": 1.7119814290091315e-05,
1060
- "loss": 1.0384,
1061
- "step": 522
1062
- },
1063
- {
1064
- "epoch": 0.87,
1065
- "learning_rate": 1.708084425351476e-05,
1066
- "loss": 1.0274,
1067
- "step": 525
1068
- },
1069
- {
1070
- "epoch": 0.88,
1071
- "learning_rate": 1.704165734823006e-05,
1072
- "loss": 1.0307,
1073
- "step": 528
1074
- },
1075
- {
1076
- "epoch": 0.88,
1077
- "learning_rate": 1.7002254774435e-05,
1078
- "loss": 1.0112,
1079
- "step": 531
1080
- },
1081
- {
1082
- "epoch": 0.89,
1083
- "learning_rate": 1.6962637738932746e-05,
1084
- "loss": 1.007,
1085
- "step": 534
1086
- },
1087
- {
1088
- "epoch": 0.89,
1089
- "learning_rate": 1.6922807455094898e-05,
1090
- "loss": 0.9999,
1091
- "step": 537
1092
- },
1093
- {
1094
- "epoch": 0.9,
1095
- "learning_rate": 1.688276514282431e-05,
1096
- "loss": 1.015,
1097
- "step": 540
1098
- },
1099
- {
1100
- "epoch": 0.9,
1101
- "learning_rate": 1.6842512028517773e-05,
1102
- "loss": 0.9912,
1103
- "step": 543
1104
- },
1105
- {
1106
- "epoch": 0.91,
1107
- "learning_rate": 1.6802049345028388e-05,
1108
- "loss": 1.0109,
1109
- "step": 546
1110
- },
1111
- {
1112
- "epoch": 0.91,
1113
- "learning_rate": 1.676137833162786e-05,
1114
- "loss": 1.0079,
1115
- "step": 549
1116
- },
1117
- {
1118
- "epoch": 0.92,
1119
- "learning_rate": 1.6720500233968522e-05,
1120
- "loss": 1.0352,
1121
- "step": 552
1122
- },
1123
- {
1124
- "epoch": 0.92,
1125
- "learning_rate": 1.6679416304045172e-05,
1126
- "loss": 1.0021,
1127
- "step": 555
1128
- },
1129
- {
1130
- "epoch": 0.93,
1131
- "learning_rate": 1.6638127800156754e-05,
1132
- "loss": 0.9867,
1133
- "step": 558
1134
- },
1135
- {
1136
- "epoch": 0.93,
1137
- "learning_rate": 1.65966359868678e-05,
1138
- "loss": 0.9974,
1139
- "step": 561
1140
- },
1141
- {
1142
- "epoch": 0.94,
1143
- "learning_rate": 1.65549421349697e-05,
1144
- "loss": 1.0049,
1145
- "step": 564
1146
- },
1147
- {
1148
- "epoch": 0.94,
1149
- "learning_rate": 1.6513047521441795e-05,
1150
- "loss": 1.0141,
1151
- "step": 567
1152
- },
1153
- {
1154
- "epoch": 0.95,
1155
- "learning_rate": 1.6470953429412257e-05,
1156
- "loss": 0.9925,
1157
- "step": 570
1158
- },
1159
- {
1160
- "epoch": 0.95,
1161
- "learning_rate": 1.6428661148118775e-05,
1162
- "loss": 1.0206,
1163
- "step": 573
1164
- },
1165
- {
1166
- "epoch": 0.96,
1167
- "learning_rate": 1.638617197286911e-05,
1168
- "loss": 1.0067,
1169
- "step": 576
1170
- },
1171
- {
1172
- "epoch": 0.96,
1173
- "learning_rate": 1.634348720500136e-05,
1174
- "loss": 0.9687,
1175
- "step": 579
1176
- },
1177
- {
1178
- "epoch": 0.97,
1179
- "learning_rate": 1.6300608151844184e-05,
1180
- "loss": 1.0217,
1181
- "step": 582
1182
- },
1183
- {
1184
- "epoch": 0.97,
1185
- "learning_rate": 1.6257536126676684e-05,
1186
- "loss": 0.9977,
1187
- "step": 585
1188
- },
1189
- {
1190
- "epoch": 0.98,
1191
- "learning_rate": 1.6214272448688228e-05,
1192
- "loss": 0.9951,
1193
- "step": 588
1194
- },
1195
- {
1196
- "epoch": 0.98,
1197
- "learning_rate": 1.6170818442938035e-05,
1198
- "loss": 0.994,
1199
- "step": 591
1200
- },
1201
- {
1202
- "epoch": 0.99,
1203
- "learning_rate": 1.6127175440314596e-05,
1204
- "loss": 0.9697,
1205
- "step": 594
1206
- },
1207
- {
1208
- "epoch": 0.99,
1209
- "learning_rate": 1.6083344777494896e-05,
1210
- "loss": 1.0084,
1211
- "step": 597
1212
- },
1213
- {
1214
- "epoch": 1.0,
1215
- "learning_rate": 1.6039327796903502e-05,
1216
- "loss": 0.9991,
1217
- "step": 600
1218
- },
1219
- {
1220
- "epoch": 1.0,
1221
- "learning_rate": 1.5995125846671417e-05,
1222
- "loss": 0.7961,
1223
- "step": 603
1224
- },
1225
- {
1226
- "epoch": 1.01,
1227
- "learning_rate": 1.5950740280594815e-05,
1228
- "loss": 0.6247,
1229
- "step": 606
1230
- },
1231
- {
1232
- "epoch": 1.01,
1233
- "learning_rate": 1.590617245809357e-05,
1234
- "loss": 0.6435,
1235
- "step": 609
1236
- },
1237
- {
1238
- "epoch": 1.02,
1239
- "learning_rate": 1.5861423744169608e-05,
1240
- "loss": 0.6282,
1241
- "step": 612
1242
- },
1243
- {
1244
- "epoch": 1.02,
1245
- "learning_rate": 1.5816495509365123e-05,
1246
- "loss": 0.6123,
1247
- "step": 615
1248
- },
1249
- {
1250
- "epoch": 1.03,
1251
- "learning_rate": 1.577138912972058e-05,
1252
- "loss": 0.6011,
1253
- "step": 618
1254
- },
1255
- {
1256
- "epoch": 1.03,
1257
- "learning_rate": 1.5726105986732582e-05,
1258
- "loss": 0.6344,
1259
- "step": 621
1260
- },
1261
- {
1262
- "epoch": 1.04,
1263
- "learning_rate": 1.568064746731156e-05,
1264
- "loss": 0.6108,
1265
- "step": 624
1266
- },
1267
- {
1268
- "epoch": 1.04,
1269
- "learning_rate": 1.5635014963739278e-05,
1270
- "loss": 0.6274,
1271
- "step": 627
1272
- },
1273
- {
1274
- "epoch": 1.05,
1275
- "learning_rate": 1.5589209873626215e-05,
1276
- "loss": 0.5958,
1277
- "step": 630
1278
- },
1279
- {
1280
- "epoch": 1.05,
1281
- "learning_rate": 1.5543233599868744e-05,
1282
- "loss": 0.6143,
1283
- "step": 633
1284
- },
1285
- {
1286
- "epoch": 1.06,
1287
- "learning_rate": 1.5497087550606167e-05,
1288
- "loss": 0.6181,
1289
- "step": 636
1290
- },
1291
- {
1292
- "epoch": 1.06,
1293
- "learning_rate": 1.5450773139177587e-05,
1294
- "loss": 0.6139,
1295
- "step": 639
1296
- },
1297
- {
1298
- "epoch": 1.07,
1299
- "learning_rate": 1.5404291784078633e-05,
1300
- "loss": 0.6214,
1301
- "step": 642
1302
- },
1303
- {
1304
- "epoch": 1.07,
1305
- "learning_rate": 1.5357644908917995e-05,
1306
- "loss": 0.5895,
1307
- "step": 645
1308
- },
1309
- {
1310
- "epoch": 1.08,
1311
- "learning_rate": 1.5310833942373833e-05,
1312
- "loss": 0.5976,
1313
- "step": 648
1314
- },
1315
- {
1316
- "epoch": 1.08,
1317
- "learning_rate": 1.526386031815003e-05,
1318
- "loss": 0.6034,
1319
- "step": 651
1320
- },
1321
- {
1322
- "epoch": 1.09,
1323
- "learning_rate": 1.5216725474932246e-05,
1324
- "loss": 0.6124,
1325
- "step": 654
1326
- },
1327
- {
1328
- "epoch": 1.09,
1329
- "learning_rate": 1.5169430856343909e-05,
1330
- "loss": 0.6143,
1331
- "step": 657
1332
- },
1333
- {
1334
- "epoch": 1.1,
1335
- "learning_rate": 1.5121977910901953e-05,
1336
- "loss": 0.6156,
1337
- "step": 660
1338
- },
1339
- {
1340
- "epoch": 1.1,
1341
- "learning_rate": 1.5074368091972476e-05,
1342
- "loss": 0.622,
1343
- "step": 663
1344
- },
1345
- {
1346
- "epoch": 1.11,
1347
- "learning_rate": 1.5026602857726216e-05,
1348
- "loss": 0.5983,
1349
- "step": 666
1350
- },
1351
- {
1352
- "epoch": 1.11,
1353
- "learning_rate": 1.4978683671093905e-05,
1354
- "loss": 0.601,
1355
- "step": 669
1356
- },
1357
- {
1358
- "epoch": 1.12,
1359
- "learning_rate": 1.4930611999721457e-05,
1360
- "loss": 0.5974,
1361
- "step": 672
1362
- },
1363
- {
1364
- "epoch": 1.12,
1365
- "learning_rate": 1.4882389315925002e-05,
1366
- "loss": 0.5977,
1367
- "step": 675
1368
- },
1369
- {
1370
- "epoch": 1.13,
1371
- "learning_rate": 1.483401709664582e-05,
1372
- "loss": 0.6019,
1373
- "step": 678
1374
- },
1375
- {
1376
- "epoch": 1.13,
1377
- "learning_rate": 1.4785496823405084e-05,
1378
- "loss": 0.608,
1379
- "step": 681
1380
- },
1381
- {
1382
- "epoch": 1.14,
1383
- "learning_rate": 1.4736829982258498e-05,
1384
- "loss": 0.6013,
1385
- "step": 684
1386
- },
1387
- {
1388
- "epoch": 1.14,
1389
- "learning_rate": 1.468801806375077e-05,
1390
- "loss": 0.5834,
1391
- "step": 687
1392
- },
1393
- {
1394
- "epoch": 1.15,
1395
- "learning_rate": 1.4639062562869986e-05,
1396
- "loss": 0.604,
1397
- "step": 690
1398
- },
1399
- {
1400
- "epoch": 1.15,
1401
- "learning_rate": 1.4589964979001779e-05,
1402
- "loss": 0.6028,
1403
- "step": 693
1404
- },
1405
- {
1406
- "epoch": 1.16,
1407
- "learning_rate": 1.4540726815883445e-05,
1408
- "loss": 0.5999,
1409
- "step": 696
1410
- },
1411
- {
1412
- "epoch": 1.16,
1413
- "learning_rate": 1.4491349581557884e-05,
1414
- "loss": 0.6015,
1415
- "step": 699
1416
- },
1417
- {
1418
- "epoch": 1.17,
1419
- "learning_rate": 1.4441834788327382e-05,
1420
- "loss": 0.6103,
1421
- "step": 702
1422
- },
1423
- {
1424
- "epoch": 1.17,
1425
- "learning_rate": 1.4392183952707324e-05,
1426
- "loss": 0.5958,
1427
- "step": 705
1428
- },
1429
- {
1430
- "epoch": 1.18,
1431
- "learning_rate": 1.4342398595379738e-05,
1432
- "loss": 0.6088,
1433
- "step": 708
1434
- },
1435
- {
1436
- "epoch": 1.18,
1437
- "learning_rate": 1.4292480241146715e-05,
1438
- "loss": 0.5966,
1439
- "step": 711
1440
- },
1441
- {
1442
- "epoch": 1.19,
1443
- "learning_rate": 1.4242430418883713e-05,
1444
- "loss": 0.5949,
1445
- "step": 714
1446
- },
1447
- {
1448
- "epoch": 1.19,
1449
- "learning_rate": 1.4192250661492724e-05,
1450
- "loss": 0.5962,
1451
- "step": 717
1452
- },
1453
- {
1454
- "epoch": 1.2,
1455
- "learning_rate": 1.414194250585534e-05,
1456
- "loss": 0.6081,
1457
- "step": 720
1458
- },
1459
- {
1460
- "epoch": 1.2,
1461
- "learning_rate": 1.409150749278567e-05,
1462
- "loss": 0.609,
1463
- "step": 723
1464
- },
1465
- {
1466
- "epoch": 1.21,
1467
- "learning_rate": 1.4040947166983147e-05,
1468
- "loss": 0.6169,
1469
- "step": 726
1470
- },
1471
- {
1472
- "epoch": 1.21,
1473
- "learning_rate": 1.3990263076985228e-05,
1474
- "loss": 0.6158,
1475
- "step": 729
1476
- },
1477
- {
1478
- "epoch": 1.22,
1479
- "learning_rate": 1.3939456775119962e-05,
1480
- "loss": 0.5725,
1481
- "step": 732
1482
- },
1483
- {
1484
- "epoch": 1.22,
1485
- "learning_rate": 1.3888529817458436e-05,
1486
- "loss": 0.6092,
1487
- "step": 735
1488
- },
1489
- {
1490
- "epoch": 1.23,
1491
- "learning_rate": 1.3837483763767128e-05,
1492
- "loss": 0.6091,
1493
- "step": 738
1494
- },
1495
- {
1496
- "epoch": 1.23,
1497
- "learning_rate": 1.3786320177460142e-05,
1498
- "loss": 0.629,
1499
- "step": 741
1500
- },
1501
- {
1502
- "epoch": 1.24,
1503
- "learning_rate": 1.3735040625551303e-05,
1504
- "loss": 0.5754,
1505
- "step": 744
1506
- },
1507
- {
1508
- "epoch": 1.24,
1509
- "learning_rate": 1.3683646678606176e-05,
1510
- "loss": 0.5961,
1511
- "step": 747
1512
- },
1513
- {
1514
- "epoch": 1.25,
1515
- "learning_rate": 1.3632139910693971e-05,
1516
- "loss": 0.5801,
1517
- "step": 750
1518
- },
1519
- {
1520
- "epoch": 1.25,
1521
- "learning_rate": 1.3580521899339316e-05,
1522
- "loss": 0.6093,
1523
- "step": 753
1524
- },
1525
- {
1526
- "epoch": 1.26,
1527
- "learning_rate": 1.3528794225473952e-05,
1528
- "loss": 0.5903,
1529
- "step": 756
1530
- },
1531
- {
1532
- "epoch": 1.26,
1533
- "learning_rate": 1.3476958473388315e-05,
1534
- "loss": 0.6145,
1535
- "step": 759
1536
- },
1537
- {
1538
- "epoch": 1.27,
1539
- "learning_rate": 1.3425016230683e-05,
1540
- "loss": 0.6015,
1541
- "step": 762
1542
- },
1543
- {
1544
- "epoch": 1.27,
1545
- "learning_rate": 1.3372969088220158e-05,
1546
- "loss": 0.5927,
1547
- "step": 765
1548
- },
1549
- {
1550
- "epoch": 1.28,
1551
- "learning_rate": 1.332081864007475e-05,
1552
- "loss": 0.6049,
1553
- "step": 768
1554
- },
1555
- {
1556
- "epoch": 1.28,
1557
- "learning_rate": 1.3268566483485734e-05,
1558
- "loss": 0.5824,
1559
- "step": 771
1560
- },
1561
- {
1562
- "epoch": 1.29,
1563
- "learning_rate": 1.3216214218807152e-05,
1564
- "loss": 0.593,
1565
- "step": 774
1566
- },
1567
- {
1568
- "epoch": 1.29,
1569
- "learning_rate": 1.3163763449459105e-05,
1570
- "loss": 0.5972,
1571
- "step": 777
1572
- },
1573
- {
1574
- "epoch": 1.3,
1575
- "learning_rate": 1.3111215781878648e-05,
1576
- "loss": 0.6,
1577
- "step": 780
1578
- },
1579
- {
1580
- "epoch": 1.3,
1581
- "learning_rate": 1.3058572825470588e-05,
1582
- "loss": 0.6063,
1583
- "step": 783
1584
- },
1585
- {
1586
- "epoch": 1.31,
1587
- "learning_rate": 1.3005836192558192e-05,
1588
- "loss": 0.6135,
1589
- "step": 786
1590
- },
1591
- {
1592
- "epoch": 1.31,
1593
- "learning_rate": 1.2953007498333807e-05,
1594
- "loss": 0.6082,
1595
- "step": 789
1596
- },
1597
- {
1598
- "epoch": 1.32,
1599
- "learning_rate": 1.2900088360809396e-05,
1600
- "loss": 0.5708,
1601
- "step": 792
1602
- },
1603
- {
1604
- "epoch": 1.32,
1605
- "learning_rate": 1.2847080400766962e-05,
1606
- "loss": 0.5901,
1607
- "step": 795
1608
- },
1609
- {
1610
- "epoch": 1.33,
1611
- "learning_rate": 1.2793985241708934e-05,
1612
- "loss": 0.612,
1613
- "step": 798
1614
- },
1615
- {
1616
- "epoch": 1.33,
1617
- "learning_rate": 1.274080450980843e-05,
1618
- "loss": 0.594,
1619
- "step": 801
1620
- },
1621
- {
1622
- "epoch": 1.34,
1623
- "learning_rate": 1.2687539833859444e-05,
1624
- "loss": 0.6105,
1625
- "step": 804
1626
- },
1627
- {
1628
- "epoch": 1.34,
1629
- "learning_rate": 1.2634192845226971e-05,
1630
- "loss": 0.6035,
1631
- "step": 807
1632
- },
1633
- {
1634
- "epoch": 1.35,
1635
- "learning_rate": 1.258076517779705e-05,
1636
- "loss": 0.5969,
1637
- "step": 810
1638
- },
1639
- {
1640
- "epoch": 1.35,
1641
- "learning_rate": 1.2527258467926688e-05,
1642
- "loss": 0.5916,
1643
- "step": 813
1644
- },
1645
- {
1646
- "epoch": 1.36,
1647
- "learning_rate": 1.2473674354393797e-05,
1648
- "loss": 0.6268,
1649
- "step": 816
1650
- },
1651
- {
1652
- "epoch": 1.36,
1653
- "learning_rate": 1.2420014478346944e-05,
1654
- "loss": 0.5937,
1655
- "step": 819
1656
- },
1657
- {
1658
- "epoch": 1.37,
1659
- "learning_rate": 1.2366280483255128e-05,
1660
- "loss": 0.5846,
1661
- "step": 822
1662
- },
1663
- {
1664
- "epoch": 1.37,
1665
- "learning_rate": 1.2312474014857432e-05,
1666
- "loss": 0.6201,
1667
- "step": 825
1668
- },
1669
- {
1670
- "epoch": 1.38,
1671
- "learning_rate": 1.2258596721112608e-05,
1672
- "loss": 0.5948,
1673
- "step": 828
1674
- },
1675
- {
1676
- "epoch": 1.38,
1677
- "learning_rate": 1.2204650252148615e-05,
1678
- "loss": 0.5824,
1679
- "step": 831
1680
- },
1681
- {
1682
- "epoch": 1.39,
1683
- "learning_rate": 1.2150636260212088e-05,
1684
- "loss": 0.5968,
1685
- "step": 834
1686
- },
1687
- {
1688
- "epoch": 1.39,
1689
- "learning_rate": 1.209655639961771e-05,
1690
- "loss": 0.6026,
1691
- "step": 837
1692
- },
1693
- {
1694
- "epoch": 1.4,
1695
- "learning_rate": 1.204241232669756e-05,
1696
- "loss": 0.6069,
1697
- "step": 840
1698
- },
1699
- {
1700
- "epoch": 1.4,
1701
- "learning_rate": 1.1988205699750391e-05,
1702
- "loss": 0.5881,
1703
- "step": 843
1704
- },
1705
- {
1706
- "epoch": 1.41,
1707
- "learning_rate": 1.1933938178990816e-05,
1708
- "loss": 0.5851,
1709
- "step": 846
1710
- },
1711
- {
1712
- "epoch": 1.41,
1713
- "learning_rate": 1.1879611426498488e-05,
1714
- "loss": 0.5966,
1715
- "step": 849
1716
- },
1717
- {
1718
- "epoch": 1.42,
1719
- "learning_rate": 1.1825227106167178e-05,
1720
- "loss": 0.5896,
1721
- "step": 852
1722
- },
1723
- {
1724
- "epoch": 1.42,
1725
- "learning_rate": 1.1770786883653806e-05,
1726
- "loss": 0.6103,
1727
- "step": 855
1728
- },
1729
- {
1730
- "epoch": 1.43,
1731
- "learning_rate": 1.1716292426327454e-05,
1732
- "loss": 0.6132,
1733
- "step": 858
1734
- },
1735
- {
1736
- "epoch": 1.43,
1737
- "learning_rate": 1.1661745403218276e-05,
1738
- "loss": 0.5875,
1739
- "step": 861
1740
- },
1741
- {
1742
- "epoch": 1.44,
1743
- "learning_rate": 1.1607147484966372e-05,
1744
- "loss": 0.5799,
1745
- "step": 864
1746
- },
1747
- {
1748
- "epoch": 1.44,
1749
- "learning_rate": 1.1552500343770658e-05,
1750
- "loss": 0.6214,
1751
- "step": 867
1752
- },
1753
- {
1754
- "epoch": 1.45,
1755
- "learning_rate": 1.149780565333761e-05,
1756
- "loss": 0.5999,
1757
- "step": 870
1758
- },
1759
- {
1760
- "epoch": 1.45,
1761
- "learning_rate": 1.1443065088830019e-05,
1762
- "loss": 0.5962,
1763
- "step": 873
1764
- },
1765
- {
1766
- "epoch": 1.46,
1767
- "learning_rate": 1.1388280326815697e-05,
1768
- "loss": 0.5969,
1769
- "step": 876
1770
- },
1771
- {
1772
- "epoch": 1.46,
1773
- "learning_rate": 1.1333453045216107e-05,
1774
- "loss": 0.5796,
1775
- "step": 879
1776
- },
1777
- {
1778
- "epoch": 1.47,
1779
- "learning_rate": 1.1278584923254981e-05,
1780
- "loss": 0.6056,
1781
- "step": 882
1782
- },
1783
- {
1784
- "epoch": 1.47,
1785
- "learning_rate": 1.1223677641406896e-05,
1786
- "loss": 0.5967,
1787
- "step": 885
1788
- },
1789
- {
1790
- "epoch": 1.48,
1791
- "learning_rate": 1.1168732881345794e-05,
1792
- "loss": 0.5935,
1793
- "step": 888
1794
- },
1795
- {
1796
- "epoch": 1.48,
1797
- "learning_rate": 1.1113752325893483e-05,
1798
- "loss": 0.602,
1799
- "step": 891
1800
- },
1801
- {
1802
- "epoch": 1.49,
1803
- "learning_rate": 1.1058737658968102e-05,
1804
- "loss": 0.5983,
1805
- "step": 894
1806
- },
1807
- {
1808
- "epoch": 1.49,
1809
- "learning_rate": 1.1003690565532522e-05,
1810
- "loss": 0.6014,
1811
- "step": 897
1812
- },
1813
- {
1814
- "epoch": 1.5,
1815
- "learning_rate": 1.0948612731542776e-05,
1816
- "loss": 0.5906,
1817
- "step": 900
1818
- },
1819
- {
1820
- "epoch": 1.5,
1821
- "learning_rate": 1.0893505843896403e-05,
1822
- "loss": 0.5803,
1823
- "step": 903
1824
- },
1825
- {
1826
- "epoch": 1.51,
1827
- "learning_rate": 1.0838371590380765e-05,
1828
- "loss": 0.5994,
1829
- "step": 906
1830
- },
1831
- {
1832
- "epoch": 1.51,
1833
- "learning_rate": 1.0783211659621396e-05,
1834
- "loss": 0.5852,
1835
- "step": 909
1836
- },
1837
- {
1838
- "epoch": 1.52,
1839
- "learning_rate": 1.0728027741030247e-05,
1840
- "loss": 0.6085,
1841
- "step": 912
1842
- },
1843
- {
1844
- "epoch": 1.52,
1845
- "learning_rate": 1.0672821524753953e-05,
1846
- "loss": 0.6101,
1847
- "step": 915
1848
- },
1849
- {
1850
- "epoch": 1.53,
1851
- "learning_rate": 1.061759470162209e-05,
1852
- "loss": 0.5844,
1853
- "step": 918
1854
- },
1855
- {
1856
- "epoch": 1.53,
1857
- "learning_rate": 1.056234896309535e-05,
1858
- "loss": 0.6055,
1859
- "step": 921
1860
- },
1861
- {
1862
- "epoch": 1.54,
1863
- "learning_rate": 1.050708600121377e-05,
1864
- "loss": 0.6062,
1865
- "step": 924
1866
- },
1867
- {
1868
- "epoch": 1.54,
1869
- "learning_rate": 1.0451807508544891e-05,
1870
- "loss": 0.5935,
1871
- "step": 927
1872
- },
1873
- {
1874
- "epoch": 1.55,
1875
- "learning_rate": 1.0396515178131925e-05,
1876
- "loss": 0.5854,
1877
- "step": 930
1878
- },
1879
- {
1880
- "epoch": 1.55,
1881
- "learning_rate": 1.0341210703441895e-05,
1882
- "loss": 0.5695,
1883
- "step": 933
1884
- },
1885
- {
1886
- "epoch": 1.56,
1887
- "learning_rate": 1.0285895778313783e-05,
1888
- "loss": 0.6004,
1889
- "step": 936
1890
- },
1891
- {
1892
- "epoch": 1.56,
1893
- "learning_rate": 1.0230572096906634e-05,
1894
- "loss": 0.5908,
1895
- "step": 939
1896
- },
1897
- {
1898
- "epoch": 1.57,
1899
- "learning_rate": 1.0175241353647678e-05,
1900
- "loss": 0.5852,
1901
- "step": 942
1902
- },
1903
- {
1904
- "epoch": 1.57,
1905
- "learning_rate": 1.0119905243180432e-05,
1906
- "loss": 0.5822,
1907
- "step": 945
1908
- },
1909
- {
1910
- "epoch": 1.58,
1911
- "learning_rate": 1.0064565460312794e-05,
1912
- "loss": 0.5818,
1913
- "step": 948
1914
- },
1915
- {
1916
- "epoch": 1.58,
1917
- "learning_rate": 1.000922369996515e-05,
1918
- "loss": 0.5944,
1919
- "step": 951
1920
- },
1921
- {
1922
- "epoch": 1.59,
1923
- "learning_rate": 9.953881657118438e-06,
1924
- "loss": 0.5785,
1925
- "step": 954
1926
- },
1927
- {
1928
- "epoch": 1.59,
1929
- "learning_rate": 9.898541026762252e-06,
1930
- "loss": 0.5894,
1931
- "step": 957
1932
- },
1933
- {
1934
- "epoch": 1.6,
1935
- "learning_rate": 9.843203503842928e-06,
1936
- "loss": 0.5947,
1937
- "step": 960
1938
- },
1939
- {
1940
- "epoch": 1.6,
1941
- "learning_rate": 9.787870783211633e-06,
1942
- "loss": 0.5969,
1943
- "step": 963
1944
- },
1945
- {
1946
- "epoch": 1.61,
1947
- "learning_rate": 9.732544559572443e-06,
1948
- "loss": 0.588,
1949
- "step": 966
1950
- },
1951
- {
1952
- "epoch": 1.61,
1953
- "learning_rate": 9.677226527430444e-06,
1954
- "loss": 0.5762,
1955
- "step": 969
1956
- },
1957
- {
1958
- "epoch": 1.62,
1959
- "learning_rate": 9.621918381039851e-06,
1960
- "loss": 0.5909,
1961
- "step": 972
1962
- },
1963
- {
1964
- "epoch": 1.62,
1965
- "learning_rate": 9.56662181435209e-06,
1966
- "loss": 0.5907,
1967
- "step": 975
1968
- },
1969
- {
1970
- "epoch": 1.63,
1971
- "learning_rate": 9.511338520963942e-06,
1972
- "loss": 0.5654,
1973
- "step": 978
1974
- },
1975
- {
1976
- "epoch": 1.63,
1977
- "learning_rate": 9.456070194065647e-06,
1978
- "loss": 0.5793,
1979
- "step": 981
1980
- },
1981
- {
1982
- "epoch": 1.64,
1983
- "learning_rate": 9.400818526389063e-06,
1984
- "loss": 0.5736,
1985
- "step": 984
1986
- },
1987
- {
1988
- "epoch": 1.64,
1989
- "learning_rate": 9.345585210155818e-06,
1990
- "loss": 0.5839,
1991
- "step": 987
1992
- },
1993
- {
1994
- "epoch": 1.65,
1995
- "learning_rate": 9.290371937025486e-06,
1996
- "loss": 0.5605,
1997
- "step": 990
1998
- },
1999
- {
2000
- "epoch": 1.65,
2001
- "learning_rate": 9.235180398043756e-06,
2002
- "loss": 0.5774,
2003
- "step": 993
2004
- },
2005
- {
2006
- "epoch": 1.66,
2007
- "learning_rate": 9.180012283590678e-06,
2008
- "loss": 0.5851,
2009
- "step": 996
2010
- },
2011
- {
2012
- "epoch": 1.66,
2013
- "learning_rate": 9.12486928332884e-06,
2014
- "loss": 0.5867,
2015
- "step": 999
2016
- },
2017
- {
2018
- "epoch": 1.66,
2019
- "eval_loss": 0.9312900900840759,
2020
- "eval_runtime": 61.0314,
2021
- "eval_samples_per_second": 7.586,
2022
- "eval_steps_per_second": 0.95,
2023
- "step": 1000
2024
- },
2025
- {
2026
- "epoch": 1.67,
2027
- "learning_rate": 9.069753086151657e-06,
2028
- "loss": 0.3263,
2029
- "step": 1002
2030
- },
2031
- {
2032
- "epoch": 1.67,
2033
- "learning_rate": 9.014665380131639e-06,
2034
- "loss": 0.3206,
2035
- "step": 1005
2036
- },
2037
- {
2038
- "epoch": 1.68,
2039
- "learning_rate": 8.959607852468667e-06,
2040
- "loss": 0.3393,
2041
- "step": 1008
2042
- },
2043
- {
2044
- "epoch": 1.68,
2045
- "learning_rate": 8.904582189438345e-06,
2046
- "loss": 0.3301,
2047
- "step": 1011
2048
- },
2049
- {
2050
- "epoch": 1.69,
2051
- "learning_rate": 8.849590076340344e-06,
2052
- "loss": 0.3327,
2053
- "step": 1014
2054
- },
2055
- {
2056
- "epoch": 1.69,
2057
- "learning_rate": 8.79463319744677e-06,
2058
- "loss": 0.3337,
2059
- "step": 1017
2060
- },
2061
- {
2062
- "epoch": 1.7,
2063
- "learning_rate": 8.739713235950608e-06,
2064
- "loss": 0.3504,
2065
- "step": 1020
2066
- },
2067
- {
2068
- "epoch": 1.7,
2069
- "learning_rate": 8.684831873914146e-06,
2070
- "loss": 0.3291,
2071
- "step": 1023
2072
- },
2073
- {
2074
- "epoch": 1.71,
2075
- "learning_rate": 8.629990792217464e-06,
2076
- "loss": 0.3301,
2077
- "step": 1026
2078
- },
2079
- {
2080
- "epoch": 1.71,
2081
- "learning_rate": 8.575191670506969e-06,
2082
- "loss": 0.3119,
2083
- "step": 1029
2084
- },
2085
- {
2086
- "epoch": 1.72,
2087
- "learning_rate": 8.520436187143921e-06,
2088
- "loss": 0.3326,
2089
- "step": 1032
2090
- },
2091
- {
2092
- "epoch": 1.72,
2093
- "learning_rate": 8.465726019153052e-06,
2094
- "loss": 0.3274,
2095
- "step": 1035
2096
- },
2097
- {
2098
- "epoch": 1.73,
2099
- "learning_rate": 8.411062842171198e-06,
2100
- "loss": 0.3294,
2101
- "step": 1038
2102
- },
2103
- {
2104
- "epoch": 1.73,
2105
- "learning_rate": 8.356448330395979e-06,
2106
- "loss": 0.3375,
2107
- "step": 1041
2108
- },
2109
- {
2110
- "epoch": 1.74,
2111
- "learning_rate": 8.301884156534511e-06,
2112
- "loss": 0.317,
2113
- "step": 1044
2114
- },
2115
- {
2116
- "epoch": 1.74,
2117
- "learning_rate": 8.247371991752201e-06,
2118
- "loss": 0.3359,
2119
- "step": 1047
2120
- },
2121
- {
2122
- "epoch": 1.75,
2123
- "learning_rate": 8.19291350562153e-06,
2124
- "loss": 0.3003,
2125
- "step": 1050
2126
- },
2127
- {
2128
- "epoch": 1.75,
2129
- "learning_rate": 8.138510366070937e-06,
2130
- "loss": 0.3194,
2131
- "step": 1053
2132
- },
2133
- {
2134
- "epoch": 1.76,
2135
- "learning_rate": 8.084164239333745e-06,
2136
- "loss": 0.3304,
2137
- "step": 1056
2138
- },
2139
- {
2140
- "epoch": 1.76,
2141
- "learning_rate": 8.029876789897102e-06,
2142
- "loss": 0.3219,
2143
- "step": 1059
2144
- },
2145
- {
2146
- "epoch": 1.77,
2147
- "learning_rate": 7.975649680451024e-06,
2148
- "loss": 0.3059,
2149
- "step": 1062
2150
- },
2151
- {
2152
- "epoch": 1.77,
2153
- "learning_rate": 7.921484571837465e-06,
2154
- "loss": 0.3342,
2155
- "step": 1065
2156
- },
2157
- {
2158
- "epoch": 1.78,
2159
- "learning_rate": 7.867383122999435e-06,
2160
- "loss": 0.3318,
2161
- "step": 1068
2162
- },
2163
- {
2164
- "epoch": 1.78,
2165
- "learning_rate": 7.813346990930209e-06,
2166
- "loss": 0.3303,
2167
- "step": 1071
2168
- },
2169
- {
2170
- "epoch": 1.79,
2171
- "learning_rate": 7.759377830622581e-06,
2172
- "loss": 0.3379,
2173
- "step": 1074
2174
- },
2175
- {
2176
- "epoch": 1.79,
2177
- "learning_rate": 7.705477295018148e-06,
2178
- "loss": 0.3267,
2179
- "step": 1077
2180
- },
2181
- {
2182
- "epoch": 1.8,
2183
- "learning_rate": 7.651647034956721e-06,
2184
- "loss": 0.3336,
2185
- "step": 1080
2186
- },
2187
- {
2188
- "epoch": 1.8,
2189
- "learning_rate": 7.597888699125731e-06,
2190
- "loss": 0.3346,
2191
- "step": 1083
2192
- },
2193
- {
2194
- "epoch": 1.81,
2195
- "learning_rate": 7.544203934009752e-06,
2196
- "loss": 0.3449,
2197
- "step": 1086
2198
- },
2199
- {
2200
- "epoch": 1.81,
2201
- "learning_rate": 7.4905943838400716e-06,
2202
- "loss": 0.3323,
2203
- "step": 1089
2204
- },
2205
- {
2206
- "epoch": 1.82,
2207
- "learning_rate": 7.437061690544331e-06,
2208
- "loss": 0.3114,
2209
- "step": 1092
2210
- },
2211
- {
2212
- "epoch": 1.82,
2213
- "learning_rate": 7.383607493696235e-06,
2214
- "loss": 0.3249,
2215
- "step": 1095
2216
- },
2217
- {
2218
- "epoch": 1.83,
2219
- "learning_rate": 7.330233430465334e-06,
2220
- "loss": 0.3248,
2221
- "step": 1098
2222
- },
2223
- {
2224
- "epoch": 1.83,
2225
- "learning_rate": 7.276941135566884e-06,
2226
- "loss": 0.3243,
2227
- "step": 1101
2228
- },
2229
- {
2230
- "epoch": 1.84,
2231
- "learning_rate": 7.223732241211783e-06,
2232
- "loss": 0.3255,
2233
- "step": 1104
2234
- },
2235
- {
2236
- "epoch": 1.84,
2237
- "learning_rate": 7.170608377056576e-06,
2238
- "loss": 0.3189,
2239
- "step": 1107
2240
- },
2241
- {
2242
- "epoch": 1.85,
2243
- "learning_rate": 7.117571170153543e-06,
2244
- "loss": 0.3145,
2245
- "step": 1110
2246
- },
2247
- {
2248
- "epoch": 1.85,
2249
- "learning_rate": 7.064622244900872e-06,
2250
- "loss": 0.3178,
2251
- "step": 1113
2252
- },
2253
- {
2254
- "epoch": 1.86,
2255
- "learning_rate": 7.011763222992891e-06,
2256
- "loss": 0.3007,
2257
- "step": 1116
2258
- },
2259
- {
2260
- "epoch": 1.86,
2261
- "learning_rate": 6.958995723370425e-06,
2262
- "loss": 0.3254,
2263
- "step": 1119
2264
- },
2265
- {
2266
- "epoch": 1.87,
2267
- "learning_rate": 6.906321362171189e-06,
2268
- "loss": 0.3263,
2269
- "step": 1122
2270
- },
2271
- {
2272
- "epoch": 1.87,
2273
- "learning_rate": 6.853741752680303e-06,
2274
- "loss": 0.3041,
2275
- "step": 1125
2276
- },
2277
- {
2278
- "epoch": 1.88,
2279
- "learning_rate": 6.801258505280876e-06,
2280
- "loss": 0.329,
2281
- "step": 1128
2282
- },
2283
- {
2284
- "epoch": 1.88,
2285
- "learning_rate": 6.74887322740469e-06,
2286
- "loss": 0.3324,
2287
- "step": 1131
2288
- },
2289
- {
2290
- "epoch": 1.89,
2291
- "learning_rate": 6.696587523482951e-06,
2292
- "loss": 0.3213,
2293
- "step": 1134
2294
- },
2295
- {
2296
- "epoch": 1.89,
2297
- "learning_rate": 6.644402994897171e-06,
2298
- "loss": 0.3221,
2299
- "step": 1137
2300
- },
2301
- {
2302
- "epoch": 1.9,
2303
- "learning_rate": 6.592321239930112e-06,
2304
- "loss": 0.3059,
2305
- "step": 1140
2306
- },
2307
- {
2308
- "epoch": 1.9,
2309
- "learning_rate": 6.540343853716834e-06,
2310
- "loss": 0.3248,
2311
- "step": 1143
2312
- },
2313
- {
2314
- "epoch": 1.91,
2315
- "learning_rate": 6.488472428195841e-06,
2316
- "loss": 0.3325,
2317
- "step": 1146
2318
- },
2319
- {
2320
- "epoch": 1.91,
2321
- "learning_rate": 6.436708552060317e-06,
2322
- "loss": 0.3132,
2323
- "step": 1149
2324
- },
2325
- {
2326
- "epoch": 1.92,
2327
- "learning_rate": 6.385053810709485e-06,
2328
- "loss": 0.3205,
2329
- "step": 1152
2330
- },
2331
- {
2332
- "epoch": 1.92,
2333
- "learning_rate": 6.333509786200032e-06,
2334
- "loss": 0.3095,
2335
- "step": 1155
2336
- },
2337
- {
2338
- "epoch": 1.93,
2339
- "learning_rate": 6.282078057197671e-06,
2340
- "loss": 0.3166,
2341
- "step": 1158
2342
- },
2343
- {
2344
- "epoch": 1.93,
2345
- "learning_rate": 6.230760198928774e-06,
2346
- "loss": 0.3172,
2347
- "step": 1161
2348
- },
2349
- {
2350
- "epoch": 1.94,
2351
- "learning_rate": 6.1795577831321405e-06,
2352
- "loss": 0.3251,
2353
- "step": 1164
2354
- },
2355
- {
2356
- "epoch": 1.94,
2357
- "learning_rate": 6.128472378010846e-06,
2358
- "loss": 0.3211,
2359
- "step": 1167
2360
- },
2361
- {
2362
- "epoch": 1.95,
2363
- "learning_rate": 6.077505548184226e-06,
2364
- "loss": 0.3091,
2365
- "step": 1170
2366
- },
2367
- {
2368
- "epoch": 1.95,
2369
- "learning_rate": 6.026658854639946e-06,
2370
- "loss": 0.3341,
2371
- "step": 1173
2372
- },
2373
- {
2374
- "epoch": 1.96,
2375
- "learning_rate": 5.975933854686196e-06,
2376
- "loss": 0.3016,
2377
- "step": 1176
2378
- },
2379
- {
2380
- "epoch": 1.96,
2381
- "learning_rate": 5.925332101903994e-06,
2382
- "loss": 0.333,
2383
- "step": 1179
2384
- },
2385
- {
2386
- "epoch": 1.97,
2387
- "learning_rate": 5.874855146099594e-06,
2388
- "loss": 0.3234,
2389
- "step": 1182
2390
- },
2391
- {
2392
- "epoch": 1.97,
2393
- "learning_rate": 5.824504533257042e-06,
2394
- "loss": 0.3256,
2395
- "step": 1185
2396
- },
2397
- {
2398
- "epoch": 1.98,
2399
- "learning_rate": 5.774281805490798e-06,
2400
- "loss": 0.3102,
2401
- "step": 1188
2402
- },
2403
- {
2404
- "epoch": 1.98,
2405
- "learning_rate": 5.724188500998531e-06,
2406
- "loss": 0.315,
2407
- "step": 1191
2408
- },
2409
- {
2410
- "epoch": 1.99,
2411
- "learning_rate": 5.674226154013988e-06,
2412
- "loss": 0.3095,
2413
- "step": 1194
2414
- },
2415
- {
2416
- "epoch": 1.99,
2417
- "learning_rate": 5.624396294760022e-06,
2418
- "loss": 0.3094,
2419
- "step": 1197
2420
- },
2421
- {
2422
- "epoch": 2.0,
2423
- "learning_rate": 5.574700449401702e-06,
2424
- "loss": 0.3182,
2425
- "step": 1200
2426
- },
2427
- {
2428
- "epoch": 2.0,
2429
- "learning_rate": 5.525140139999588e-06,
2430
- "loss": 0.2879,
2431
- "step": 1203
2432
- },
2433
- {
2434
- "epoch": 2.01,
2435
- "learning_rate": 5.475716884463113e-06,
2436
- "loss": 0.2313,
2437
- "step": 1206
2438
- },
2439
- {
2440
- "epoch": 2.01,
2441
- "learning_rate": 5.42643219650409e-06,
2442
- "loss": 0.2257,
2443
- "step": 1209
2444
- },
2445
- {
2446
- "epoch": 2.02,
2447
- "learning_rate": 5.377287585590352e-06,
2448
- "loss": 0.2328,
2449
- "step": 1212
2450
- },
2451
- {
2452
- "epoch": 2.02,
2453
- "learning_rate": 5.328284556899513e-06,
2454
- "loss": 0.239,
2455
- "step": 1215
2456
- },
2457
- {
2458
- "epoch": 2.03,
2459
- "learning_rate": 5.279424611272873e-06,
2460
- "loss": 0.2301,
2461
- "step": 1218
2462
- },
2463
- {
2464
- "epoch": 2.03,
2465
- "learning_rate": 5.230709245169461e-06,
2466
- "loss": 0.2333,
2467
- "step": 1221
2468
- },
2469
- {
2470
- "epoch": 2.04,
2471
- "learning_rate": 5.182139950620188e-06,
2472
- "loss": 0.2308,
2473
- "step": 1224
2474
- },
2475
- {
2476
- "epoch": 2.04,
2477
- "learning_rate": 5.133718215182162e-06,
2478
- "loss": 0.2562,
2479
- "step": 1227
2480
- },
2481
- {
2482
- "epoch": 2.05,
2483
- "learning_rate": 5.085445521893113e-06,
2484
- "loss": 0.2491,
2485
- "step": 1230
2486
- },
2487
- {
2488
- "epoch": 2.05,
2489
- "learning_rate": 5.037323349225984e-06,
2490
- "loss": 0.2534,
2491
- "step": 1233
2492
- },
2493
- {
2494
- "epoch": 2.06,
2495
- "learning_rate": 4.989353171043637e-06,
2496
- "loss": 0.2193,
2497
- "step": 1236
2498
- },
2499
- {
2500
- "epoch": 2.06,
2501
- "learning_rate": 4.94153645655373e-06,
2502
- "loss": 0.2435,
2503
- "step": 1239
2504
- },
2505
- {
2506
- "epoch": 2.07,
2507
- "learning_rate": 4.893874670263704e-06,
2508
- "loss": 0.2351,
2509
- "step": 1242
2510
- },
2511
- {
2512
- "epoch": 2.07,
2513
- "learning_rate": 4.846369271935932e-06,
2514
- "loss": 0.2399,
2515
- "step": 1245
2516
- },
2517
- {
2518
- "epoch": 2.08,
2519
- "learning_rate": 4.79902171654302e-06,
2520
- "loss": 0.2506,
2521
- "step": 1248
2522
- },
2523
- {
2524
- "epoch": 2.08,
2525
- "learning_rate": 4.7518334542232215e-06,
2526
- "loss": 0.239,
2527
- "step": 1251
2528
- },
2529
- {
2530
- "epoch": 2.09,
2531
- "learning_rate": 4.704805930236044e-06,
2532
- "loss": 0.2435,
2533
- "step": 1254
2534
- },
2535
- {
2536
- "epoch": 2.09,
2537
- "learning_rate": 4.657940584917983e-06,
2538
- "loss": 0.2505,
2539
- "step": 1257
2540
- },
2541
- {
2542
- "epoch": 2.1,
2543
- "learning_rate": 4.611238853638399e-06,
2544
- "loss": 0.2282,
2545
- "step": 1260
2546
- },
2547
- {
2548
- "epoch": 2.1,
2549
- "learning_rate": 4.564702166755565e-06,
2550
- "loss": 0.2302,
2551
- "step": 1263
2552
- },
2553
- {
2554
- "epoch": 2.11,
2555
- "learning_rate": 4.518331949572847e-06,
2556
- "loss": 0.2418,
2557
- "step": 1266
2558
- },
2559
- {
2560
- "epoch": 2.11,
2561
- "learning_rate": 4.472129622295055e-06,
2562
- "loss": 0.2361,
2563
- "step": 1269
2564
- },
2565
- {
2566
- "epoch": 2.12,
2567
- "learning_rate": 4.426096599984954e-06,
2568
- "loss": 0.2533,
2569
- "step": 1272
2570
- },
2571
- {
2572
- "epoch": 2.12,
2573
- "learning_rate": 4.380234292519917e-06,
2574
- "loss": 0.2389,
2575
- "step": 1275
2576
- },
2577
- {
2578
- "epoch": 2.13,
2579
- "learning_rate": 4.33454410454874e-06,
2580
- "loss": 0.2415,
2581
- "step": 1278
2582
- },
2583
- {
2584
- "epoch": 2.13,
2585
- "learning_rate": 4.289027435448633e-06,
2586
- "loss": 0.236,
2587
- "step": 1281
2588
- },
2589
- {
2590
- "epoch": 2.14,
2591
- "learning_rate": 4.243685679282343e-06,
2592
- "loss": 0.2363,
2593
- "step": 1284
2594
- },
2595
- {
2596
- "epoch": 2.14,
2597
- "learning_rate": 4.19852022475547e-06,
2598
- "loss": 0.2401,
2599
- "step": 1287
2600
- },
2601
- {
2602
- "epoch": 2.15,
2603
- "learning_rate": 4.153532455173939e-06,
2604
- "loss": 0.2428,
2605
- "step": 1290
2606
- },
2607
- {
2608
- "epoch": 2.15,
2609
- "learning_rate": 4.1087237484016185e-06,
2610
- "loss": 0.242,
2611
- "step": 1293
2612
- },
2613
- {
2614
- "epoch": 2.16,
2615
- "learning_rate": 4.064095476818133e-06,
2616
- "loss": 0.2352,
2617
- "step": 1296
2618
- },
2619
- {
2620
- "epoch": 2.16,
2621
- "learning_rate": 4.019649007276815e-06,
2622
- "loss": 0.2538,
2623
- "step": 1299
2624
- },
2625
- {
2626
- "epoch": 2.17,
2627
- "learning_rate": 3.9753857010628615e-06,
2628
- "loss": 0.2391,
2629
- "step": 1302
2630
- },
2631
- {
2632
- "epoch": 2.17,
2633
- "learning_rate": 3.931306913851621e-06,
2634
- "loss": 0.236,
2635
- "step": 1305
2636
- },
2637
- {
2638
- "epoch": 2.18,
2639
- "learning_rate": 3.887413995667091e-06,
2640
- "loss": 0.2344,
2641
- "step": 1308
2642
- },
2643
- {
2644
- "epoch": 2.18,
2645
- "learning_rate": 3.843708290840557e-06,
2646
- "loss": 0.2529,
2647
- "step": 1311
2648
- },
2649
- {
2650
- "epoch": 2.19,
2651
- "learning_rate": 3.8001911379694267e-06,
2652
- "loss": 0.2401,
2653
- "step": 1314
2654
- },
2655
- {
2656
- "epoch": 2.19,
2657
- "learning_rate": 3.756863869876224e-06,
2658
- "loss": 0.2384,
2659
- "step": 1317
2660
- },
2661
- {
2662
- "epoch": 2.2,
2663
- "learning_rate": 3.7137278135677714e-06,
2664
- "loss": 0.2271,
2665
- "step": 1320
2666
- },
2667
- {
2668
- "epoch": 2.2,
2669
- "learning_rate": 3.6707842901945546e-06,
2670
- "loss": 0.2273,
2671
- "step": 1323
2672
- },
2673
- {
2674
- "epoch": 2.21,
2675
- "learning_rate": 3.628034615010251e-06,
2676
- "loss": 0.2186,
2677
- "step": 1326
2678
- },
2679
- {
2680
- "epoch": 2.21,
2681
- "learning_rate": 3.58548009733145e-06,
2682
- "loss": 0.2526,
2683
- "step": 1329
2684
- },
2685
- {
2686
- "epoch": 2.22,
2687
- "learning_rate": 3.543122040497542e-06,
2688
- "loss": 0.2402,
2689
- "step": 1332
2690
- },
2691
- {
2692
- "epoch": 2.22,
2693
- "learning_rate": 3.5009617418308208e-06,
2694
- "loss": 0.2327,
2695
- "step": 1335
2696
- },
2697
- {
2698
- "epoch": 2.23,
2699
- "learning_rate": 3.4590004925967245e-06,
2700
- "loss": 0.2333,
2701
- "step": 1338
2702
- },
2703
- {
2704
- "epoch": 2.23,
2705
- "learning_rate": 3.417239577964312e-06,
2706
- "loss": 0.2345,
2707
- "step": 1341
2708
- },
2709
- {
2710
- "epoch": 2.24,
2711
- "learning_rate": 3.375680276966884e-06,
2712
- "loss": 0.2385,
2713
- "step": 1344
2714
- },
2715
- {
2716
- "epoch": 2.24,
2717
- "learning_rate": 3.3343238624628204e-06,
2718
- "loss": 0.2247,
2719
- "step": 1347
2720
- },
2721
- {
2722
- "epoch": 2.25,
2723
- "learning_rate": 3.2931716010965807e-06,
2724
- "loss": 0.2285,
2725
- "step": 1350
2726
- },
2727
- {
2728
- "epoch": 2.25,
2729
- "learning_rate": 3.2522247532599327e-06,
2730
- "loss": 0.251,
2731
- "step": 1353
2732
- },
2733
- {
2734
- "epoch": 2.26,
2735
- "learning_rate": 3.211484573053324e-06,
2736
- "loss": 0.2324,
2737
- "step": 1356
2738
- },
2739
- {
2740
- "epoch": 2.26,
2741
- "learning_rate": 3.1709523082474937e-06,
2742
- "loss": 0.2279,
2743
- "step": 1359
2744
- },
2745
- {
2746
- "epoch": 2.27,
2747
- "learning_rate": 3.130629200245249e-06,
2748
- "loss": 0.2438,
2749
- "step": 1362
2750
- },
2751
- {
2752
- "epoch": 2.27,
2753
- "learning_rate": 3.090516484043432e-06,
2754
- "loss": 0.2315,
2755
- "step": 1365
2756
- },
2757
- {
2758
- "epoch": 2.28,
2759
- "learning_rate": 3.0506153881951183e-06,
2760
- "loss": 0.2468,
2761
- "step": 1368
2762
- },
2763
- {
2764
- "epoch": 2.28,
2765
- "learning_rate": 3.010927134771966e-06,
2766
- "loss": 0.2221,
2767
- "step": 1371
2768
- },
2769
- {
2770
- "epoch": 2.29,
2771
- "learning_rate": 2.971452939326802e-06,
2772
- "loss": 0.2515,
2773
- "step": 1374
2774
- },
2775
- {
2776
- "epoch": 2.29,
2777
- "learning_rate": 2.9321940108563907e-06,
2778
- "loss": 0.2159,
2779
- "step": 1377
2780
- },
2781
- {
2782
- "epoch": 2.3,
2783
- "learning_rate": 2.8931515517644017e-06,
2784
- "loss": 0.2418,
2785
- "step": 1380
2786
- },
2787
- {
2788
- "epoch": 2.3,
2789
- "learning_rate": 2.854326757824576e-06,
2790
- "loss": 0.2311,
2791
- "step": 1383
2792
- },
2793
- {
2794
- "epoch": 2.31,
2795
- "learning_rate": 2.815720818144123e-06,
2796
- "loss": 0.2677,
2797
- "step": 1386
2798
- },
2799
- {
2800
- "epoch": 2.31,
2801
- "learning_rate": 2.7773349151272745e-06,
2802
- "loss": 0.2474,
2803
- "step": 1389
2804
- },
2805
- {
2806
- "epoch": 2.32,
2807
- "learning_rate": 2.7391702244390984e-06,
2808
- "loss": 0.2315,
2809
- "step": 1392
2810
- },
2811
- {
2812
- "epoch": 2.32,
2813
- "learning_rate": 2.7012279149694685e-06,
2814
- "loss": 0.2319,
2815
- "step": 1395
2816
- },
2817
- {
2818
- "epoch": 2.33,
2819
- "learning_rate": 2.663509148797281e-06,
2820
- "loss": 0.2272,
2821
- "step": 1398
2822
- },
2823
- {
2824
- "epoch": 2.33,
2825
- "learning_rate": 2.6260150811548434e-06,
2826
- "loss": 0.2451,
2827
- "step": 1401
2828
- },
2829
- {
2830
- "epoch": 2.34,
2831
- "learning_rate": 2.588746860392516e-06,
2832
- "loss": 0.2426,
2833
- "step": 1404
2834
- },
2835
- {
2836
- "epoch": 2.34,
2837
- "learning_rate": 2.5517056279435183e-06,
2838
- "loss": 0.238,
2839
- "step": 1407
2840
- },
2841
- {
2842
- "epoch": 2.35,
2843
- "learning_rate": 2.514892518288988e-06,
2844
- "loss": 0.2351,
2845
- "step": 1410
2846
- },
2847
- {
2848
- "epoch": 2.35,
2849
- "learning_rate": 2.47830865892323e-06,
2850
- "loss": 0.2278,
2851
- "step": 1413
2852
- },
2853
- {
2854
- "epoch": 2.36,
2855
- "learning_rate": 2.441955170319168e-06,
2856
- "loss": 0.2369,
2857
- "step": 1416
2858
- },
2859
- {
2860
- "epoch": 2.36,
2861
- "learning_rate": 2.405833165894055e-06,
2862
- "loss": 0.2356,
2863
- "step": 1419
2864
- },
2865
- {
2866
- "epoch": 2.37,
2867
- "learning_rate": 2.3699437519753444e-06,
2868
- "loss": 0.2281,
2869
- "step": 1422
2870
- },
2871
- {
2872
- "epoch": 2.37,
2873
- "learning_rate": 2.334288027766828e-06,
2874
- "loss": 0.2253,
2875
- "step": 1425
2876
- },
2877
- {
2878
- "epoch": 2.38,
2879
- "learning_rate": 2.2988670853149574e-06,
2880
- "loss": 0.2432,
2881
- "step": 1428
2882
- },
2883
- {
2884
- "epoch": 2.38,
2885
- "learning_rate": 2.2636820094754077e-06,
2886
- "loss": 0.2365,
2887
- "step": 1431
2888
- },
2889
- {
2890
- "epoch": 2.39,
2891
- "learning_rate": 2.2287338778798283e-06,
2892
- "loss": 0.2324,
2893
- "step": 1434
2894
- },
2895
- {
2896
- "epoch": 2.39,
2897
- "learning_rate": 2.1940237609028735e-06,
2898
- "loss": 0.2435,
2899
- "step": 1437
2900
- },
2901
- {
2902
- "epoch": 2.4,
2903
- "learning_rate": 2.159552721629381e-06,
2904
- "loss": 0.2275,
2905
- "step": 1440
2906
- },
2907
- {
2908
- "epoch": 2.4,
2909
- "learning_rate": 2.125321815821846e-06,
2910
- "loss": 0.2411,
2911
- "step": 1443
2912
- },
2913
- {
2914
- "epoch": 2.41,
2915
- "learning_rate": 2.0913320918880643e-06,
2916
- "loss": 0.2582,
2917
- "step": 1446
2918
- },
2919
- {
2920
- "epoch": 2.41,
2921
- "learning_rate": 2.0575845908490267e-06,
2922
- "loss": 0.2433,
2923
- "step": 1449
2924
- },
2925
- {
2926
- "epoch": 2.42,
2927
- "learning_rate": 2.0240803463070425e-06,
2928
- "loss": 0.2225,
2929
- "step": 1452
2930
- },
2931
- {
2932
- "epoch": 2.42,
2933
- "learning_rate": 1.990820384414075e-06,
2934
- "loss": 0.2155,
2935
- "step": 1455
2936
- },
2937
- {
2938
- "epoch": 2.43,
2939
- "learning_rate": 1.9578057238403124e-06,
2940
- "loss": 0.2397,
2941
- "step": 1458
2942
- },
2943
- {
2944
- "epoch": 2.43,
2945
- "learning_rate": 1.9250373757429753e-06,
2946
- "loss": 0.2453,
2947
- "step": 1461
2948
- },
2949
- {
2950
- "epoch": 2.44,
2951
- "learning_rate": 1.8925163437353465e-06,
2952
- "loss": 0.2407,
2953
- "step": 1464
2954
- },
2955
- {
2956
- "epoch": 2.44,
2957
- "learning_rate": 1.8602436238560218e-06,
2958
- "loss": 0.2512,
2959
- "step": 1467
2960
- },
2961
- {
2962
- "epoch": 2.45,
2963
- "learning_rate": 1.828220204538419e-06,
2964
- "loss": 0.2458,
2965
- "step": 1470
2966
- },
2967
- {
2968
- "epoch": 2.45,
2969
- "learning_rate": 1.7964470665804923e-06,
2970
- "loss": 0.2314,
2971
- "step": 1473
2972
- },
2973
- {
2974
- "epoch": 2.46,
2975
- "learning_rate": 1.7649251831147018e-06,
2976
- "loss": 0.2339,
2977
- "step": 1476
2978
- },
2979
- {
2980
- "epoch": 2.46,
2981
- "learning_rate": 1.733655519578209e-06,
2982
- "loss": 0.2383,
2983
- "step": 1479
2984
- },
2985
- {
2986
- "epoch": 2.47,
2987
- "learning_rate": 1.7026390336832933e-06,
2988
- "loss": 0.2307,
2989
- "step": 1482
2990
- },
2991
- {
2992
- "epoch": 2.47,
2993
- "learning_rate": 1.671876675388039e-06,
2994
- "loss": 0.2364,
2995
- "step": 1485
2996
- },
2997
- {
2998
- "epoch": 2.48,
2999
- "learning_rate": 1.6413693868672332e-06,
3000
- "loss": 0.2278,
3001
- "step": 1488
3002
- },
3003
- {
3004
- "epoch": 2.48,
3005
- "learning_rate": 1.6111181024834999e-06,
3006
- "loss": 0.2282,
3007
- "step": 1491
3008
- },
3009
- {
3010
- "epoch": 2.49,
3011
- "learning_rate": 1.5811237487586973e-06,
3012
- "loss": 0.2436,
3013
- "step": 1494
3014
- },
3015
- {
3016
- "epoch": 2.49,
3017
- "learning_rate": 1.5513872443455358e-06,
3018
- "loss": 0.2206,
3019
- "step": 1497
3020
- },
3021
- {
3022
- "epoch": 2.5,
3023
- "learning_rate": 1.5219094999994322e-06,
3024
- "loss": 0.2405,
3025
- "step": 1500
3026
- },
3027
- {
3028
- "epoch": 2.5,
3029
- "eval_loss": 0.9922696948051453,
3030
- "eval_runtime": 61.3926,
3031
- "eval_samples_per_second": 7.542,
3032
- "eval_steps_per_second": 0.945,
3033
- "step": 1500
3034
- },
3035
- {
3036
- "epoch": 2.5,
3037
- "learning_rate": 1.4926914185506315e-06,
3038
- "loss": 0.2409,
3039
- "step": 1503
3040
- },
3041
- {
3042
- "epoch": 2.51,
3043
- "learning_rate": 1.4637338948765467e-06,
3044
- "loss": 0.2313,
3045
- "step": 1506
3046
- },
3047
- {
3048
- "epoch": 2.51,
3049
- "learning_rate": 1.4350378158743484e-06,
3050
- "loss": 0.2355,
3051
- "step": 1509
3052
- },
3053
- {
3054
- "epoch": 2.52,
3055
- "learning_rate": 1.4066040604338104e-06,
3056
- "loss": 0.2309,
3057
- "step": 1512
3058
- },
3059
- {
3060
- "epoch": 2.52,
3061
- "learning_rate": 1.3784334994103787e-06,
3062
- "loss": 0.2225,
3063
- "step": 1515
3064
- },
3065
- {
3066
- "epoch": 2.53,
3067
- "learning_rate": 1.3505269955985122e-06,
3068
- "loss": 0.2376,
3069
- "step": 1518
3070
- },
3071
- {
3072
- "epoch": 2.53,
3073
- "learning_rate": 1.322885403705254e-06,
3074
- "loss": 0.2168,
3075
- "step": 1521
3076
- },
3077
- {
3078
- "epoch": 2.54,
3079
- "learning_rate": 1.2955095703240438e-06,
3080
- "loss": 0.2493,
3081
- "step": 1524
3082
- },
3083
- {
3084
- "epoch": 2.54,
3085
- "learning_rate": 1.2684003339088024e-06,
3086
- "loss": 0.2473,
3087
- "step": 1527
3088
- },
3089
- {
3090
- "epoch": 2.55,
3091
- "learning_rate": 1.2415585247482498e-06,
3092
- "loss": 0.2251,
3093
- "step": 1530
3094
- },
3095
- {
3096
- "epoch": 2.55,
3097
- "learning_rate": 1.21498496494046e-06,
3098
- "loss": 0.2266,
3099
- "step": 1533
3100
- },
3101
- {
3102
- "epoch": 2.56,
3103
- "learning_rate": 1.1886804683677068e-06,
3104
- "loss": 0.2196,
3105
- "step": 1536
3106
- },
3107
- {
3108
- "epoch": 2.56,
3109
- "learning_rate": 1.1626458406715192e-06,
3110
- "loss": 0.2298,
3111
- "step": 1539
3112
- },
3113
- {
3114
- "epoch": 2.57,
3115
- "learning_rate": 1.1368818792280089e-06,
3116
- "loss": 0.2332,
3117
- "step": 1542
3118
- },
3119
- {
3120
- "epoch": 2.57,
3121
- "learning_rate": 1.1113893731234537e-06,
3122
- "loss": 0.2366,
3123
- "step": 1545
3124
- },
3125
- {
3126
- "epoch": 2.58,
3127
- "learning_rate": 1.0861691031301313e-06,
3128
- "loss": 0.2327,
3129
- "step": 1548
3130
- },
3131
- {
3132
- "epoch": 2.58,
3133
- "learning_rate": 1.0612218416823926e-06,
3134
- "loss": 0.2244,
3135
- "step": 1551
3136
- },
3137
- {
3138
- "epoch": 2.59,
3139
- "learning_rate": 1.0365483528530206e-06,
3140
- "loss": 0.2349,
3141
- "step": 1554
3142
- },
3143
- {
3144
- "epoch": 2.59,
3145
- "learning_rate": 1.012149392329822e-06,
3146
- "loss": 0.248,
3147
- "step": 1557
3148
- },
3149
- {
3150
- "epoch": 2.6,
3151
- "learning_rate": 9.88025707392475e-07,
3152
- "loss": 0.226,
3153
- "step": 1560
3154
- },
3155
- {
3156
- "epoch": 2.6,
3157
- "learning_rate": 9.641780368896592e-07,
3158
- "loss": 0.2437,
3159
- "step": 1563
3160
- },
3161
- {
3162
- "epoch": 2.6,
3163
- "learning_rate": 9.406071112164061e-07,
3164
- "loss": 0.2188,
3165
- "step": 1566
3166
- },
3167
- {
3168
- "epoch": 2.61,
3169
- "learning_rate": 9.173136522917458e-07,
3170
- "loss": 0.2368,
3171
- "step": 1569
3172
- },
3173
- {
3174
- "epoch": 2.61,
3175
- "learning_rate": 8.942983735365885e-07,
3176
- "loss": 0.2409,
3177
- "step": 1572
3178
- },
3179
- {
3180
- "epoch": 2.62,
3181
- "learning_rate": 8.715619798518715e-07,
3182
- "loss": 0.2313,
3183
- "step": 1575
3184
- },
3185
- {
3186
- "epoch": 2.62,
3187
- "learning_rate": 8.491051675969774e-07,
3188
- "loss": 0.2467,
3189
- "step": 1578
3190
- },
3191
- {
3192
- "epoch": 2.63,
3193
- "learning_rate": 8.269286245684038e-07,
3194
- "loss": 0.2276,
3195
- "step": 1581
3196
- },
3197
- {
3198
- "epoch": 2.63,
3199
- "learning_rate": 8.050330299786913e-07,
3200
- "loss": 0.2377,
3201
- "step": 1584
3202
- },
3203
- {
3204
- "epoch": 2.64,
3205
- "learning_rate": 7.834190544356312e-07,
3206
- "loss": 0.2311,
3207
- "step": 1587
3208
- },
3209
- {
3210
- "epoch": 2.64,
3211
- "learning_rate": 7.620873599217215e-07,
3212
- "loss": 0.2366,
3213
- "step": 1590
3214
- },
3215
- {
3216
- "epoch": 2.65,
3217
- "learning_rate": 7.410385997738867e-07,
3218
- "loss": 0.2264,
3219
- "step": 1593
3220
- },
3221
- {
3222
- "epoch": 2.65,
3223
- "learning_rate": 7.202734186634785e-07,
3224
- "loss": 0.2235,
3225
- "step": 1596
3226
- },
3227
- {
3228
- "epoch": 2.66,
3229
- "learning_rate": 6.997924525765209e-07,
3230
- "loss": 0.2264,
3231
- "step": 1599
3232
- },
3233
- {
3234
- "epoch": 2.66,
3235
- "learning_rate": 6.79596328794242e-07,
3236
- "loss": 0.2267,
3237
- "step": 1602
3238
- },
3239
- {
3240
- "epoch": 2.67,
3241
- "learning_rate": 6.596856658738504e-07,
3242
- "loss": 0.223,
3243
- "step": 1605
3244
- },
3245
- {
3246
- "epoch": 2.67,
3247
- "learning_rate": 6.40061073629602e-07,
3248
- "loss": 0.2284,
3249
- "step": 1608
3250
- },
3251
- {
3252
- "epoch": 2.68,
3253
- "learning_rate": 6.20723153114109e-07,
3254
- "loss": 0.2485,
3255
- "step": 1611
3256
- },
3257
- {
3258
- "epoch": 2.68,
3259
- "learning_rate": 6.016724965999487e-07,
3260
- "loss": 0.2402,
3261
- "step": 1614
3262
- },
3263
- {
3264
- "epoch": 2.69,
3265
- "learning_rate": 5.829096875615048e-07,
3266
- "loss": 0.2168,
3267
- "step": 1617
3268
- },
3269
- {
3270
- "epoch": 2.69,
3271
- "learning_rate": 5.644353006571124e-07,
3272
- "loss": 0.2476,
3273
- "step": 1620
3274
- },
3275
- {
3276
- "epoch": 2.7,
3277
- "learning_rate": 5.462499017114508e-07,
3278
- "loss": 0.2362,
3279
- "step": 1623
3280
- },
3281
- {
3282
- "epoch": 2.7,
3283
- "learning_rate": 5.283540476982108e-07,
3284
- "loss": 0.2446,
3285
- "step": 1626
3286
- },
3287
- {
3288
- "epoch": 2.71,
3289
- "learning_rate": 5.107482867230462e-07,
3290
- "loss": 0.2244,
3291
- "step": 1629
3292
- },
3293
- {
3294
- "epoch": 2.71,
3295
- "learning_rate": 4.934331580067719e-07,
3296
- "loss": 0.2285,
3297
- "step": 1632
3298
- },
3299
- {
3300
- "epoch": 2.72,
3301
- "learning_rate": 4.7640919186886294e-07,
3302
- "loss": 0.2382,
3303
- "step": 1635
3304
- },
3305
- {
3306
- "epoch": 2.72,
3307
- "learning_rate": 4.596769097112042e-07,
3308
- "loss": 0.2369,
3309
- "step": 1638
3310
- },
3311
- {
3312
- "epoch": 2.73,
3313
- "learning_rate": 4.4323682400212675e-07,
3314
- "loss": 0.2276,
3315
- "step": 1641
3316
- },
3317
- {
3318
- "epoch": 2.73,
3319
- "learning_rate": 4.2708943826070157e-07,
3320
- "loss": 0.2217,
3321
- "step": 1644
3322
- },
3323
- {
3324
- "epoch": 2.74,
3325
- "learning_rate": 4.1123524704133276e-07,
3326
- "loss": 0.2245,
3327
- "step": 1647
3328
- },
3329
- {
3330
- "epoch": 2.74,
3331
- "learning_rate": 3.956747359185964e-07,
3332
- "loss": 0.251,
3333
- "step": 1650
3334
- },
3335
- {
3336
- "epoch": 2.75,
3337
- "learning_rate": 3.804083814723791e-07,
3338
- "loss": 0.2263,
3339
- "step": 1653
3340
- },
3341
- {
3342
- "epoch": 2.75,
3343
- "learning_rate": 3.6543665127327765e-07,
3344
- "loss": 0.2521,
3345
- "step": 1656
3346
- },
3347
- {
3348
- "epoch": 2.76,
3349
- "learning_rate": 3.5076000386827903e-07,
3350
- "loss": 0.2253,
3351
- "step": 1659
3352
- },
3353
- {
3354
- "epoch": 2.76,
3355
- "learning_rate": 3.3637888876671076e-07,
3356
- "loss": 0.2405,
3357
- "step": 1662
3358
- },
3359
- {
3360
- "epoch": 2.77,
3361
- "learning_rate": 3.2229374642648193e-07,
3362
- "loss": 0.2062,
3363
- "step": 1665
3364
- },
3365
- {
3366
- "epoch": 2.77,
3367
- "learning_rate": 3.085050082405916e-07,
3368
- "loss": 0.2409,
3369
- "step": 1668
3370
- },
3371
- {
3372
- "epoch": 2.78,
3373
- "learning_rate": 2.9501309652391063e-07,
3374
- "loss": 0.2245,
3375
- "step": 1671
3376
- },
3377
- {
3378
- "epoch": 2.78,
3379
- "learning_rate": 2.818184245002542e-07,
3380
- "loss": 0.2338,
3381
- "step": 1674
3382
- },
3383
- {
3384
- "epoch": 2.79,
3385
- "learning_rate": 2.689213962897186e-07,
3386
- "loss": 0.2306,
3387
- "step": 1677
3388
- },
3389
- {
3390
- "epoch": 2.79,
3391
- "learning_rate": 2.5632240689631216e-07,
3392
- "loss": 0.2255,
3393
- "step": 1680
3394
- },
3395
- {
3396
- "epoch": 2.8,
3397
- "learning_rate": 2.4402184219584735e-07,
3398
- "loss": 0.224,
3399
- "step": 1683
3400
- },
3401
- {
3402
- "epoch": 2.8,
3403
- "learning_rate": 2.3202007892413447e-07,
3404
- "loss": 0.2341,
3405
- "step": 1686
3406
- },
3407
- {
3408
- "epoch": 2.81,
3409
- "learning_rate": 2.2031748466543214e-07,
3410
- "loss": 0.244,
3411
- "step": 1689
3412
- },
3413
- {
3414
- "epoch": 2.81,
3415
- "learning_rate": 2.0891441784119725e-07,
3416
- "loss": 0.2269,
3417
- "step": 1692
3418
- },
3419
- {
3420
- "epoch": 2.82,
3421
- "learning_rate": 1.9781122769910065e-07,
3422
- "loss": 0.2213,
3423
- "step": 1695
3424
- },
3425
- {
3426
- "epoch": 2.82,
3427
- "learning_rate": 1.8700825430233216e-07,
3428
- "loss": 0.2259,
3429
- "step": 1698
3430
- },
3431
- {
3432
- "epoch": 2.83,
3433
- "learning_rate": 1.765058285191923e-07,
3434
- "loss": 0.2355,
3435
- "step": 1701
3436
- },
3437
- {
3438
- "epoch": 2.83,
3439
- "learning_rate": 1.663042720129504e-07,
3440
- "loss": 0.2335,
3441
- "step": 1704
3442
- },
3443
- {
3444
- "epoch": 2.84,
3445
- "learning_rate": 1.5640389723199478e-07,
3446
- "loss": 0.252,
3447
- "step": 1707
3448
- },
3449
- {
3450
- "epoch": 2.84,
3451
- "learning_rate": 1.468050074002658e-07,
3452
- "loss": 0.223,
3453
- "step": 1710
3454
- },
3455
- {
3456
- "epoch": 2.85,
3457
- "learning_rate": 1.375078965079657e-07,
3458
- "loss": 0.2249,
3459
- "step": 1713
3460
- },
3461
- {
3462
- "epoch": 2.85,
3463
- "learning_rate": 1.2851284930255447e-07,
3464
- "loss": 0.249,
3465
- "step": 1716
3466
- },
3467
- {
3468
- "epoch": 2.86,
3469
- "learning_rate": 1.1982014128003373e-07,
3470
- "loss": 0.2375,
3471
- "step": 1719
3472
- },
3473
- {
3474
- "epoch": 2.86,
3475
- "learning_rate": 1.1143003867650437e-07,
3476
- "loss": 0.2182,
3477
- "step": 1722
3478
- },
3479
- {
3480
- "epoch": 2.87,
3481
- "learning_rate": 1.0334279846001106e-07,
3482
- "loss": 0.2249,
3483
- "step": 1725
3484
- },
3485
- {
3486
- "epoch": 2.87,
3487
- "learning_rate": 9.555866832267502e-08,
3488
- "loss": 0.2259,
3489
- "step": 1728
3490
- },
3491
- {
3492
- "epoch": 2.88,
3493
- "learning_rate": 8.807788667311023e-08,
3494
- "loss": 0.2455,
3495
- "step": 1731
3496
- },
3497
- {
3498
- "epoch": 2.88,
3499
- "learning_rate": 8.090068262911366e-08,
3500
- "loss": 0.2314,
3501
- "step": 1734
3502
- },
3503
- {
3504
- "epoch": 2.89,
3505
- "learning_rate": 7.402727601065529e-08,
3506
- "loss": 0.2351,
3507
- "step": 1737
3508
- },
3509
- {
3510
- "epoch": 2.89,
3511
- "learning_rate": 6.745787733314469e-08,
3512
- "loss": 0.2203,
3513
- "step": 1740
3514
- },
3515
- {
3516
- "epoch": 2.9,
3517
- "learning_rate": 6.11926878009772e-08,
3518
- "loss": 0.2377,
3519
- "step": 1743
3520
- },
3521
- {
3522
- "epoch": 2.9,
3523
- "learning_rate": 5.5231899301380065e-08,
3524
- "loss": 0.23,
3525
- "step": 1746
3526
- },
3527
- {
3528
- "epoch": 2.91,
3529
- "learning_rate": 4.957569439853038e-08,
3530
- "loss": 0.2136,
3531
- "step": 1749
3532
- },
3533
- {
3534
- "epoch": 2.91,
3535
- "learning_rate": 4.422424632796185e-08,
3536
- "loss": 0.2231,
3537
- "step": 1752
3538
- },
3539
- {
3540
- "epoch": 2.92,
3541
- "learning_rate": 3.917771899126677e-08,
3542
- "loss": 0.2499,
3543
- "step": 1755
3544
- },
3545
- {
3546
- "epoch": 2.92,
3547
- "learning_rate": 3.4436266951067834e-08,
3548
- "loss": 0.2348,
3549
- "step": 1758
3550
- },
3551
- {
3552
- "epoch": 2.93,
3553
- "learning_rate": 3.000003542628749e-08,
3554
- "loss": 0.2341,
3555
- "step": 1761
3556
- },
3557
- {
3558
- "epoch": 2.93,
3559
- "learning_rate": 2.586916028770259e-08,
3560
- "loss": 0.2424,
3561
- "step": 1764
3562
- },
3563
- {
3564
- "epoch": 2.94,
3565
- "learning_rate": 2.2043768053777726e-08,
3566
- "loss": 0.2495,
3567
- "step": 1767
3568
- },
3569
- {
3570
- "epoch": 2.94,
3571
- "learning_rate": 1.852397588679722e-08,
3572
- "loss": 0.2498,
3573
- "step": 1770
3574
- },
3575
- {
3576
- "epoch": 2.95,
3577
- "learning_rate": 1.5309891589267988e-08,
3578
- "loss": 0.2203,
3579
- "step": 1773
3580
- },
3581
- {
3582
- "epoch": 2.95,
3583
- "learning_rate": 1.2401613600627749e-08,
3584
- "loss": 0.2216,
3585
- "step": 1776
3586
- },
3587
- {
3588
- "epoch": 2.96,
3589
- "learning_rate": 9.799230994220754e-09,
3590
- "loss": 0.2279,
3591
- "step": 1779
3592
- },
3593
- {
3594
- "epoch": 2.96,
3595
- "learning_rate": 7.502823474576648e-09,
3596
- "loss": 0.2379,
3597
- "step": 1782
3598
- },
3599
- {
3600
- "epoch": 2.97,
3601
- "learning_rate": 5.512461374964639e-09,
3602
- "loss": 0.2306,
3603
- "step": 1785
3604
- },
3605
- {
3606
- "epoch": 2.97,
3607
- "learning_rate": 3.828205655244111e-09,
3608
- "loss": 0.2546,
3609
- "step": 1788
3610
- },
3611
- {
3612
- "epoch": 2.98,
3613
- "learning_rate": 2.4501078999916716e-09,
3614
- "loss": 0.2396,
3615
- "step": 1791
3616
- },
3617
- {
3618
- "epoch": 2.98,
3619
- "learning_rate": 1.3782103169246441e-09,
3620
- "loss": 0.2226,
3621
- "step": 1794
3622
- },
3623
- {
3624
- "epoch": 2.99,
3625
- "learning_rate": 6.125457356120912e-10,
3626
- "loss": 0.2357,
3627
- "step": 1797
3628
- },
3629
- {
3630
- "epoch": 2.99,
3631
- "learning_rate": 1.5313760645896403e-10,
3632
- "loss": 0.2264,
3633
- "step": 1800
3634
- },
3635
- {
3636
- "epoch": 3.0,
3637
- "learning_rate": 0.0,
3638
- "loss": 0.236,
3639
- "step": 1803
3640
- },
3641
- {
3642
- "epoch": 3.0,
3643
- "step": 1803,
3644
- "total_flos": 1.9808467849371766e+20,
3645
- "train_loss": 0.11453570330466949,
3646
- "train_runtime": 197727.0226,
3647
- "train_samples_per_second": 7.005,
3648
- "train_steps_per_second": 0.009
3649
- }
3650
- ],
3651
- "logging_steps": 3,
3652
- "max_steps": 1803,
3653
- "num_input_tokens_seen": 0,
3654
- "num_train_epochs": 3,
3655
- "save_steps": 500,
3656
- "total_flos": 1.9808467849371766e+20,
3657
- "train_batch_size": 96,
3658
- "trial_name": null,
3659
- "trial_params": null
3660
- }