sanchit-gandhi HF staff commited on
Commit
1be04f9
1 Parent(s): c5516e0

End of training

Browse files
all_results.json CHANGED
@@ -1,14 +1,14 @@
1
  {
2
  "epoch": 1.0,
3
- "eval_loss": 4.851853370666504,
4
- "eval_runtime": 903.3793,
5
  "eval_samples": 2642,
6
- "eval_samples_per_second": 2.925,
7
  "eval_steps_per_second": 0.245,
8
- "eval_wer": 2.5405791352637843,
9
- "train_loss": 4.7614816479409985,
10
- "train_runtime": 5563.7628,
11
  "train_samples": 28538,
12
- "train_samples_per_second": 5.129,
13
- "train_steps_per_second": 0.107
14
  }
 
1
  {
2
  "epoch": 1.0,
3
+ "eval_loss": 4.804400444030762,
4
+ "eval_runtime": 902.131,
5
  "eval_samples": 2642,
6
+ "eval_samples_per_second": 2.929,
7
  "eval_steps_per_second": 0.245,
8
+ "eval_wer": 2.364081713605712,
9
+ "train_loss": 4.473529295487837,
10
+ "train_runtime": 4542.6558,
11
  "train_samples": 28538,
12
+ "train_samples_per_second": 6.282,
13
+ "train_steps_per_second": 0.065
14
  }
eval_results.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "epoch": 1.0,
3
- "eval_loss": 4.851853370666504,
4
- "eval_runtime": 903.3793,
5
  "eval_samples": 2642,
6
- "eval_samples_per_second": 2.925,
7
  "eval_steps_per_second": 0.245,
8
- "eval_wer": 2.5405791352637843
9
  }
 
1
  {
2
  "epoch": 1.0,
3
+ "eval_loss": 4.804400444030762,
4
+ "eval_runtime": 902.131,
5
  "eval_samples": 2642,
6
+ "eval_samples_per_second": 2.929,
7
  "eval_steps_per_second": 0.245,
8
+ "eval_wer": 2.364081713605712
9
  }
runs/Feb26_15-29-49_sanchit--v100/events.out.tfevents.1645895147.sanchit--v100.203992.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2620999f5dee571505940557287f2fa29c788fdddb627a3c79c88fcda3856ed9
3
+ size 358
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "epoch": 1.0,
3
- "train_loss": 4.7614816479409985,
4
- "train_runtime": 5563.7628,
5
  "train_samples": 28538,
6
- "train_samples_per_second": 5.129,
7
- "train_steps_per_second": 0.107
8
  }
 
1
  {
2
  "epoch": 1.0,
3
+ "train_loss": 4.473529295487837,
4
+ "train_runtime": 4542.6558,
5
  "train_samples": 28538,
6
+ "train_samples_per_second": 6.282,
7
+ "train_steps_per_second": 0.065
8
  }
trainer_state.json CHANGED
@@ -2,3595 +2,1804 @@
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
  "epoch": 0.9987389659520807,
5
- "global_step": 594,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
  "epoch": 0.0,
12
- "learning_rate": 0.0,
13
- "loss": 4.798,
14
  "step": 1
15
  },
16
  {
17
- "epoch": 0.0,
18
- "learning_rate": 2e-06,
19
- "loss": 4.9659,
20
  "step": 2
21
  },
22
  {
23
  "epoch": 0.01,
24
- "learning_rate": 4e-06,
25
- "loss": 4.8548,
26
  "step": 3
27
  },
28
  {
29
  "epoch": 0.01,
30
- "learning_rate": 6e-06,
31
- "loss": 4.7832,
32
  "step": 4
33
  },
34
  {
35
- "epoch": 0.01,
36
- "learning_rate": 8e-06,
37
- "loss": 4.7029,
38
  "step": 5
39
  },
40
  {
41
- "epoch": 0.01,
42
  "learning_rate": 1e-05,
43
- "loss": 4.6549,
44
  "step": 6
45
  },
46
  {
47
- "epoch": 0.01,
48
  "learning_rate": 1.2e-05,
49
- "loss": 4.6679,
50
  "step": 7
51
  },
52
  {
53
- "epoch": 0.01,
54
  "learning_rate": 1.4e-05,
55
- "loss": 4.5376,
56
  "step": 8
57
  },
58
  {
59
- "epoch": 0.02,
60
  "learning_rate": 1.6e-05,
61
- "loss": 4.6115,
62
  "step": 9
63
  },
64
  {
65
- "epoch": 0.02,
66
  "learning_rate": 1.8e-05,
67
- "loss": 4.4613,
68
  "step": 10
69
  },
70
  {
71
- "epoch": 0.02,
72
  "learning_rate": 2e-05,
73
- "loss": 4.4478,
74
  "step": 11
75
  },
76
  {
77
- "epoch": 0.02,
78
  "learning_rate": 2.2e-05,
79
- "loss": 4.4566,
80
  "step": 12
81
  },
82
  {
83
- "epoch": 0.02,
84
  "learning_rate": 2.4e-05,
85
- "loss": 4.3608,
86
  "step": 13
87
  },
88
  {
89
- "epoch": 0.02,
90
  "learning_rate": 2.6e-05,
91
- "loss": 4.3267,
92
  "step": 14
93
  },
94
  {
95
- "epoch": 0.03,
96
  "learning_rate": 2.8e-05,
97
- "loss": 4.237,
98
  "step": 15
99
  },
100
  {
101
- "epoch": 0.03,
102
  "learning_rate": 3e-05,
103
- "loss": 4.3687,
104
  "step": 16
105
  },
106
  {
107
- "epoch": 0.03,
108
  "learning_rate": 3.2e-05,
109
- "loss": 4.3606,
110
  "step": 17
111
  },
112
  {
113
- "epoch": 0.03,
114
  "learning_rate": 3.4000000000000007e-05,
115
- "loss": 4.4764,
116
  "step": 18
117
  },
118
  {
119
- "epoch": 0.03,
120
  "learning_rate": 3.6e-05,
121
- "loss": 4.2192,
122
  "step": 19
123
  },
124
  {
125
- "epoch": 0.03,
126
  "learning_rate": 3.8e-05,
127
- "loss": 4.4582,
128
  "step": 20
129
  },
130
  {
131
- "epoch": 0.04,
132
  "learning_rate": 4e-05,
133
- "loss": 4.3744,
134
  "step": 21
135
  },
136
  {
137
- "epoch": 0.04,
138
  "learning_rate": 4.2000000000000004e-05,
139
- "loss": 4.331,
140
  "step": 22
141
  },
142
  {
143
- "epoch": 0.04,
144
  "learning_rate": 4.4e-05,
145
- "loss": 4.3407,
146
  "step": 23
147
  },
148
  {
149
- "epoch": 0.04,
150
  "learning_rate": 4.6e-05,
151
- "loss": 4.3821,
152
  "step": 24
153
  },
154
  {
155
- "epoch": 0.04,
156
  "learning_rate": 4.8e-05,
157
- "loss": 4.1777,
158
  "step": 25
159
  },
160
  {
161
- "epoch": 0.04,
162
  "learning_rate": 5e-05,
163
- "loss": 4.5667,
164
  "step": 26
165
  },
166
  {
167
- "epoch": 0.05,
168
  "learning_rate": 5.2e-05,
169
- "loss": 4.2537,
170
  "step": 27
171
  },
172
  {
173
- "epoch": 0.05,
174
  "learning_rate": 5.4e-05,
175
- "loss": 4.1534,
176
  "step": 28
177
  },
178
  {
179
- "epoch": 0.05,
180
  "learning_rate": 5.6e-05,
181
- "loss": 4.1515,
182
  "step": 29
183
  },
184
  {
185
- "epoch": 0.05,
186
  "learning_rate": 5.800000000000001e-05,
187
- "loss": 4.3429,
188
  "step": 30
189
  },
190
  {
191
- "epoch": 0.05,
192
  "learning_rate": 6e-05,
193
- "loss": 4.348,
194
  "step": 31
195
  },
196
  {
197
- "epoch": 0.05,
198
  "learning_rate": 6.2e-05,
199
- "loss": 4.2301,
200
  "step": 32
201
  },
202
  {
203
- "epoch": 0.06,
204
  "learning_rate": 6.4e-05,
205
- "loss": 4.2124,
206
  "step": 33
207
  },
208
  {
209
- "epoch": 0.06,
210
  "learning_rate": 6.6e-05,
211
- "loss": 4.1963,
212
  "step": 34
213
  },
214
  {
215
- "epoch": 0.06,
216
  "learning_rate": 6.800000000000001e-05,
217
- "loss": 4.2277,
218
  "step": 35
219
  },
220
  {
221
- "epoch": 0.06,
222
  "learning_rate": 7.000000000000001e-05,
223
- "loss": 4.4154,
224
  "step": 36
225
  },
226
  {
227
- "epoch": 0.06,
228
  "learning_rate": 7.2e-05,
229
- "loss": 4.2038,
230
  "step": 37
231
  },
232
  {
233
- "epoch": 0.06,
234
  "learning_rate": 7.4e-05,
235
- "loss": 4.3656,
236
  "step": 38
237
  },
238
  {
239
- "epoch": 0.07,
240
  "learning_rate": 7.6e-05,
241
- "loss": 4.5172,
242
  "step": 39
243
  },
244
  {
245
- "epoch": 0.07,
246
  "learning_rate": 7.8e-05,
247
- "loss": 4.2148,
248
  "step": 40
249
  },
250
  {
251
- "epoch": 0.07,
252
  "learning_rate": 8e-05,
253
- "loss": 4.196,
254
  "step": 41
255
  },
256
  {
257
- "epoch": 0.07,
258
  "learning_rate": 8.2e-05,
259
- "loss": 4.4284,
260
  "step": 42
261
  },
262
  {
263
- "epoch": 0.07,
264
  "learning_rate": 8.400000000000001e-05,
265
- "loss": 4.1641,
266
  "step": 43
267
  },
268
  {
269
- "epoch": 0.07,
270
  "learning_rate": 8.599999999999999e-05,
271
- "loss": 4.3293,
272
  "step": 44
273
  },
274
  {
275
- "epoch": 0.08,
276
  "learning_rate": 8.8e-05,
277
- "loss": 4.4,
278
  "step": 45
279
  },
280
  {
281
- "epoch": 0.08,
282
  "learning_rate": 8.999999999999999e-05,
283
- "loss": 4.4597,
284
  "step": 46
285
  },
286
  {
287
- "epoch": 0.08,
288
  "learning_rate": 9.2e-05,
289
- "loss": 4.374,
290
  "step": 47
291
  },
292
  {
293
- "epoch": 0.08,
294
  "learning_rate": 9.400000000000001e-05,
295
- "loss": 4.7429,
296
  "step": 48
297
  },
298
  {
299
- "epoch": 0.08,
300
  "learning_rate": 9.6e-05,
301
- "loss": 4.6918,
302
  "step": 49
303
  },
304
  {
305
- "epoch": 0.08,
306
- "learning_rate": 9.6e-05,
307
- "loss": 4.4926,
308
  "step": 50
309
  },
310
  {
311
- "epoch": 0.09,
312
- "learning_rate": 9.800000000000001e-05,
313
- "loss": 4.2253,
314
  "step": 51
315
  },
316
  {
317
- "epoch": 0.09,
318
- "learning_rate": 0.0001,
319
- "loss": 4.2335,
320
  "step": 52
321
  },
322
  {
323
- "epoch": 0.09,
324
- "learning_rate": 0.000102,
325
- "loss": 4.2367,
326
  "step": 53
327
  },
328
  {
329
- "epoch": 0.09,
330
- "learning_rate": 0.000104,
331
- "loss": 4.3182,
332
  "step": 54
333
  },
334
  {
335
- "epoch": 0.09,
336
- "learning_rate": 0.000106,
337
- "loss": 4.1924,
338
  "step": 55
339
  },
340
  {
341
- "epoch": 0.09,
342
- "learning_rate": 0.000108,
343
- "loss": 4.3919,
344
  "step": 56
345
  },
346
  {
347
- "epoch": 0.1,
348
- "learning_rate": 0.00011,
349
- "loss": 4.1947,
350
  "step": 57
351
  },
352
  {
353
- "epoch": 0.1,
354
- "learning_rate": 0.000112,
355
- "loss": 4.5164,
356
  "step": 58
357
  },
358
  {
359
- "epoch": 0.1,
360
- "learning_rate": 0.000114,
361
- "loss": 4.2716,
362
  "step": 59
363
  },
364
  {
365
- "epoch": 0.1,
366
- "learning_rate": 0.00011600000000000001,
367
- "loss": 4.0629,
368
  "step": 60
369
  },
370
  {
371
- "epoch": 0.1,
372
- "learning_rate": 0.000118,
373
- "loss": 4.328,
374
  "step": 61
375
  },
376
  {
377
- "epoch": 0.1,
378
- "learning_rate": 0.00012,
379
- "loss": 4.3771,
380
  "step": 62
381
  },
382
  {
383
- "epoch": 0.11,
384
- "learning_rate": 0.000122,
385
- "loss": 4.2684,
386
  "step": 63
387
  },
388
  {
389
- "epoch": 0.11,
390
- "learning_rate": 0.000124,
391
- "loss": 4.2398,
392
  "step": 64
393
  },
394
  {
395
- "epoch": 0.11,
396
- "learning_rate": 0.000126,
397
- "loss": 4.2612,
398
  "step": 65
399
  },
400
  {
401
- "epoch": 0.11,
402
- "learning_rate": 0.000128,
403
- "loss": 4.1913,
404
  "step": 66
405
  },
406
  {
407
- "epoch": 0.11,
408
- "learning_rate": 0.00013000000000000002,
409
- "loss": 4.1965,
410
  "step": 67
411
  },
412
  {
413
- "epoch": 0.11,
414
- "learning_rate": 0.000132,
415
- "loss": 4.16,
416
  "step": 68
417
  },
418
  {
419
- "epoch": 0.12,
420
- "learning_rate": 0.000134,
421
- "loss": 4.2527,
422
  "step": 69
423
  },
424
  {
425
- "epoch": 0.12,
426
- "learning_rate": 0.00013600000000000003,
427
- "loss": 4.2567,
428
  "step": 70
429
  },
430
  {
431
- "epoch": 0.12,
432
- "learning_rate": 0.00013800000000000002,
433
- "loss": 4.3733,
434
  "step": 71
435
  },
436
  {
437
- "epoch": 0.12,
438
- "learning_rate": 0.00014000000000000001,
439
- "loss": 4.3124,
440
  "step": 72
441
  },
442
  {
443
- "epoch": 0.12,
444
- "learning_rate": 0.00014199999999999998,
445
- "loss": 4.1534,
446
  "step": 73
447
  },
448
  {
449
- "epoch": 0.12,
450
- "learning_rate": 0.000144,
451
- "loss": 4.3542,
452
  "step": 74
453
  },
454
  {
455
- "epoch": 0.13,
456
- "learning_rate": 0.000146,
457
- "loss": 4.3096,
458
  "step": 75
459
  },
460
  {
461
- "epoch": 0.13,
462
- "learning_rate": 0.000148,
463
- "loss": 4.3124,
464
  "step": 76
465
  },
466
  {
467
- "epoch": 0.13,
468
- "learning_rate": 0.00015,
469
- "loss": 4.2301,
470
  "step": 77
471
  },
472
  {
473
- "epoch": 0.13,
474
- "learning_rate": 0.000152,
475
- "loss": 4.274,
476
  "step": 78
477
  },
478
  {
479
- "epoch": 0.13,
480
- "learning_rate": 0.000154,
481
- "loss": 4.2065,
482
  "step": 79
483
  },
484
  {
485
- "epoch": 0.13,
486
- "learning_rate": 0.000156,
487
- "loss": 4.4052,
488
  "step": 80
489
  },
490
  {
491
- "epoch": 0.14,
492
- "learning_rate": 0.000158,
493
- "loss": 4.354,
494
  "step": 81
495
  },
496
  {
497
- "epoch": 0.14,
498
- "learning_rate": 0.00016,
499
- "loss": 4.3246,
500
  "step": 82
501
  },
502
  {
503
- "epoch": 0.14,
504
- "learning_rate": 0.000162,
505
- "loss": 4.4008,
506
  "step": 83
507
  },
508
  {
509
- "epoch": 0.14,
510
- "learning_rate": 0.000164,
511
- "loss": 4.3554,
512
  "step": 84
513
  },
514
  {
515
- "epoch": 0.14,
516
- "learning_rate": 0.00016600000000000002,
517
- "loss": 4.2769,
518
- "step": 85
519
- },
520
- {
521
- "epoch": 0.14,
522
  "learning_rate": 0.00016800000000000002,
523
- "loss": 4.2777,
524
- "step": 86
525
  },
526
  {
527
- "epoch": 0.15,
528
  "learning_rate": 0.00017,
529
- "loss": 4.5118,
530
- "step": 87
531
  },
532
  {
533
- "epoch": 0.15,
534
  "learning_rate": 0.00017199999999999998,
535
- "loss": 4.3368,
536
- "step": 88
537
  },
538
  {
539
- "epoch": 0.15,
540
  "learning_rate": 0.000174,
541
- "loss": 4.4394,
542
- "step": 89
543
  },
544
  {
545
- "epoch": 0.15,
546
  "learning_rate": 0.000176,
547
- "loss": 4.3865,
548
- "step": 90
549
  },
550
  {
551
- "epoch": 0.15,
552
  "learning_rate": 0.000178,
553
- "loss": 4.3442,
554
- "step": 91
555
  },
556
  {
557
- "epoch": 0.15,
558
  "learning_rate": 0.00017999999999999998,
559
- "loss": 4.4925,
560
- "step": 92
561
  },
562
  {
563
- "epoch": 0.16,
564
  "learning_rate": 0.000182,
565
- "loss": 4.3855,
566
- "step": 93
567
  },
568
  {
569
- "epoch": 0.16,
570
  "learning_rate": 0.000184,
571
- "loss": 4.5602,
572
- "step": 94
573
  },
574
  {
575
- "epoch": 0.16,
576
  "learning_rate": 0.000186,
577
- "loss": 4.4853,
578
- "step": 95
579
  },
580
  {
581
- "epoch": 0.16,
582
  "learning_rate": 0.00018800000000000002,
583
- "loss": 4.6151,
584
- "step": 96
585
  },
586
  {
587
- "epoch": 0.16,
588
  "learning_rate": 0.00019,
589
- "loss": 4.5962,
590
- "step": 97
591
  },
592
  {
593
- "epoch": 0.16,
594
  "learning_rate": 0.000192,
595
- "loss": 4.4133,
596
- "step": 98
597
  },
598
  {
599
- "epoch": 0.17,
600
  "learning_rate": 0.000194,
601
- "loss": 4.792,
602
- "step": 99
603
  },
604
  {
605
- "epoch": 0.17,
606
  "learning_rate": 0.00019600000000000002,
607
- "loss": 4.4569,
608
- "step": 100
609
  },
610
  {
611
- "epoch": 0.17,
612
  "learning_rate": 0.00019800000000000002,
613
- "loss": 4.5401,
614
- "step": 101
615
  },
616
  {
617
- "epoch": 0.17,
618
  "learning_rate": 0.0002,
619
- "loss": 4.3572,
620
- "step": 102
621
  },
622
  {
623
- "epoch": 0.17,
624
  "learning_rate": 0.000202,
625
- "loss": 4.4242,
626
- "step": 103
627
  },
628
  {
629
- "epoch": 0.17,
630
  "learning_rate": 0.000204,
631
- "loss": 4.3279,
632
- "step": 104
633
  },
634
  {
635
- "epoch": 0.18,
636
  "learning_rate": 0.000206,
637
- "loss": 4.3101,
638
- "step": 105
639
  },
640
  {
641
- "epoch": 0.18,
642
  "learning_rate": 0.000208,
643
- "loss": 4.2788,
644
- "step": 106
645
  },
646
  {
647
- "epoch": 0.18,
648
  "learning_rate": 0.00021,
649
- "loss": 4.4495,
650
- "step": 107
651
  },
652
  {
653
- "epoch": 0.18,
654
  "learning_rate": 0.000212,
655
- "loss": 4.4467,
656
- "step": 108
657
  },
658
  {
659
- "epoch": 0.18,
660
  "learning_rate": 0.000214,
661
- "loss": 4.5925,
662
- "step": 109
663
  },
664
  {
665
- "epoch": 0.18,
666
  "learning_rate": 0.000216,
667
- "loss": 4.4049,
668
- "step": 110
669
  },
670
  {
671
- "epoch": 0.19,
672
  "learning_rate": 0.000218,
673
- "loss": 4.3936,
674
- "step": 111
675
  },
676
  {
677
- "epoch": 0.19,
678
  "learning_rate": 0.00022,
679
- "loss": 4.4172,
680
- "step": 112
681
  },
682
  {
683
- "epoch": 0.19,
684
  "learning_rate": 0.000222,
685
- "loss": 4.3436,
686
- "step": 113
687
  },
688
  {
689
- "epoch": 0.19,
690
  "learning_rate": 0.000224,
691
- "loss": 4.4727,
692
- "step": 114
693
  },
694
  {
695
- "epoch": 0.19,
696
  "learning_rate": 0.00022600000000000002,
697
- "loss": 4.389,
698
- "step": 115
699
  },
700
  {
701
- "epoch": 0.2,
702
  "learning_rate": 0.000228,
703
- "loss": 4.4001,
704
- "step": 116
705
  },
706
  {
707
- "epoch": 0.2,
708
  "learning_rate": 0.00023,
709
- "loss": 4.5036,
710
- "step": 117
711
  },
712
  {
713
- "epoch": 0.2,
714
  "learning_rate": 0.00023200000000000003,
715
- "loss": 4.2843,
716
- "step": 118
717
  },
718
  {
719
- "epoch": 0.2,
720
  "learning_rate": 0.00023400000000000002,
721
- "loss": 4.3065,
722
- "step": 119
723
  },
724
  {
725
- "epoch": 0.2,
726
  "learning_rate": 0.000236,
727
- "loss": 4.5466,
728
- "step": 120
729
  },
730
  {
731
- "epoch": 0.2,
732
  "learning_rate": 0.00023799999999999998,
733
- "loss": 4.4443,
734
- "step": 121
735
  },
736
  {
737
- "epoch": 0.21,
738
  "learning_rate": 0.00024,
739
- "loss": 4.6387,
740
- "step": 122
741
  },
742
  {
743
- "epoch": 0.21,
744
  "learning_rate": 0.000242,
745
- "loss": 4.3381,
746
- "step": 123
747
  },
748
  {
749
- "epoch": 0.21,
750
  "learning_rate": 0.000244,
751
- "loss": 4.4811,
752
- "step": 124
753
  },
754
  {
755
- "epoch": 0.21,
756
  "learning_rate": 0.000246,
757
- "loss": 4.3945,
758
- "step": 125
759
  },
760
  {
761
- "epoch": 0.21,
762
  "learning_rate": 0.000248,
763
- "loss": 4.3425,
764
- "step": 126
765
  },
766
  {
767
- "epoch": 0.21,
768
  "learning_rate": 0.00025,
769
- "loss": 4.3311,
770
- "step": 127
771
  },
772
  {
773
- "epoch": 0.22,
774
  "learning_rate": 0.000252,
775
- "loss": 4.3136,
776
- "step": 128
777
  },
778
  {
779
- "epoch": 0.22,
780
  "learning_rate": 0.000254,
781
- "loss": 4.3739,
782
- "step": 129
783
  },
784
  {
785
- "epoch": 0.22,
786
  "learning_rate": 0.000256,
787
- "loss": 4.463,
788
- "step": 130
789
  },
790
  {
791
- "epoch": 0.22,
792
  "learning_rate": 0.00025800000000000004,
793
- "loss": 4.4122,
794
- "step": 131
795
  },
796
  {
797
- "epoch": 0.22,
798
  "learning_rate": 0.00026000000000000003,
799
- "loss": 4.4737,
800
- "step": 132
801
  },
802
  {
803
- "epoch": 0.22,
804
  "learning_rate": 0.000262,
805
- "loss": 4.3602,
806
- "step": 133
807
  },
808
  {
809
- "epoch": 0.23,
810
  "learning_rate": 0.000264,
811
- "loss": 4.7136,
812
- "step": 134
813
  },
814
  {
815
- "epoch": 0.23,
816
  "learning_rate": 0.000266,
817
- "loss": 4.4939,
818
- "step": 135
819
  },
820
  {
821
- "epoch": 0.23,
822
  "learning_rate": 0.000268,
823
- "loss": 4.383,
824
- "step": 136
825
  },
826
  {
827
- "epoch": 0.23,
828
  "learning_rate": 0.00027,
829
- "loss": 4.3695,
830
- "step": 137
831
  },
832
  {
833
- "epoch": 0.23,
834
  "learning_rate": 0.00027200000000000005,
835
- "loss": 4.3884,
836
- "step": 138
837
  },
838
  {
839
- "epoch": 0.23,
840
  "learning_rate": 0.00027400000000000005,
841
- "loss": 4.3642,
842
- "step": 139
843
  },
844
  {
845
- "epoch": 0.24,
846
  "learning_rate": 0.00027600000000000004,
847
- "loss": 4.4841,
848
- "step": 140
849
  },
850
  {
851
- "epoch": 0.24,
852
  "learning_rate": 0.00027800000000000004,
853
- "loss": 4.3954,
854
- "step": 141
855
  },
856
  {
857
- "epoch": 0.24,
858
  "learning_rate": 0.00028000000000000003,
859
- "loss": 4.62,
860
- "step": 142
861
  },
862
  {
863
- "epoch": 0.24,
864
  "learning_rate": 0.00028199999999999997,
865
- "loss": 4.9846,
866
- "step": 143
867
  },
868
  {
869
- "epoch": 0.24,
870
  "learning_rate": 0.00028399999999999996,
871
- "loss": 5.0245,
872
- "step": 144
873
  },
874
  {
875
- "epoch": 0.24,
876
  "learning_rate": 0.00028599999999999996,
877
- "loss": 4.5489,
878
- "step": 145
879
  },
880
  {
881
- "epoch": 0.25,
882
  "learning_rate": 0.000288,
883
- "loss": 5.0615,
884
- "step": 146
885
  },
886
  {
887
- "epoch": 0.25,
888
  "learning_rate": 0.00029,
889
- "loss": 4.8247,
890
- "step": 147
891
  },
892
  {
893
- "epoch": 0.25,
894
  "learning_rate": 0.000292,
895
- "loss": 5.7443,
896
- "step": 148
897
  },
898
  {
899
- "epoch": 0.25,
900
  "learning_rate": 0.000294,
901
- "loss": 4.9307,
902
- "step": 149
903
  },
904
  {
905
- "epoch": 0.25,
906
  "learning_rate": 0.000296,
907
- "loss": 4.8052,
908
- "step": 150
909
  },
910
  {
911
- "epoch": 0.25,
912
  "learning_rate": 0.000298,
913
- "loss": 4.8397,
914
- "step": 151
915
  },
916
  {
917
- "epoch": 0.26,
918
  "learning_rate": 0.0003,
919
- "loss": 4.5232,
920
- "step": 152
921
  },
922
  {
923
- "epoch": 0.26,
924
  "learning_rate": 0.000302,
925
- "loss": 4.7358,
926
- "step": 153
927
  },
928
  {
929
- "epoch": 0.26,
930
  "learning_rate": 0.000304,
931
- "loss": 4.4171,
932
- "step": 154
933
  },
934
  {
935
- "epoch": 0.26,
936
  "learning_rate": 0.000306,
937
- "loss": 4.4927,
938
- "step": 155
939
  },
940
  {
941
- "epoch": 0.26,
942
  "learning_rate": 0.000308,
943
- "loss": 4.5971,
944
- "step": 156
945
  },
946
  {
947
- "epoch": 0.26,
948
  "learning_rate": 0.00031,
949
- "loss": 4.5412,
950
- "step": 157
951
  },
952
  {
953
- "epoch": 0.27,
954
  "learning_rate": 0.000312,
955
- "loss": 4.3854,
956
- "step": 158
957
  },
958
  {
959
- "epoch": 0.27,
960
  "learning_rate": 0.000314,
961
- "loss": 4.3552,
962
- "step": 159
963
  },
964
  {
965
- "epoch": 0.27,
966
  "learning_rate": 0.000316,
967
- "loss": 4.466,
968
- "step": 160
969
  },
970
  {
971
- "epoch": 0.27,
972
  "learning_rate": 0.00031800000000000003,
973
- "loss": 4.4675,
974
- "step": 161
975
  },
976
  {
977
- "epoch": 0.27,
978
  "learning_rate": 0.00032,
979
- "loss": 4.3639,
980
- "step": 162
981
  },
982
  {
983
- "epoch": 0.27,
984
  "learning_rate": 0.000322,
985
- "loss": 4.4132,
986
- "step": 163
987
  },
988
  {
989
- "epoch": 0.28,
990
  "learning_rate": 0.000324,
991
- "loss": 4.411,
992
- "step": 164
993
  },
994
  {
995
- "epoch": 0.28,
996
  "learning_rate": 0.000326,
997
- "loss": 4.6485,
998
- "step": 165
999
  },
1000
  {
1001
- "epoch": 0.28,
1002
  "learning_rate": 0.000328,
1003
- "loss": 4.5554,
1004
- "step": 166
1005
  },
1006
  {
1007
- "epoch": 0.28,
1008
  "learning_rate": 0.00033,
1009
- "loss": 4.395,
1010
- "step": 167
1011
  },
1012
  {
1013
- "epoch": 0.28,
1014
  "learning_rate": 0.00033200000000000005,
1015
- "loss": 4.607,
1016
- "step": 168
1017
  },
1018
  {
1019
- "epoch": 0.28,
1020
  "learning_rate": 0.00033400000000000004,
1021
- "loss": 4.4171,
1022
- "step": 169
1023
  },
1024
  {
1025
- "epoch": 0.29,
1026
  "learning_rate": 0.00033600000000000004,
1027
- "loss": 4.4899,
1028
- "step": 170
1029
  },
1030
  {
1031
- "epoch": 0.29,
1032
  "learning_rate": 0.00033800000000000003,
1033
- "loss": 4.5043,
1034
- "step": 171
1035
  },
1036
  {
1037
- "epoch": 0.29,
1038
  "learning_rate": 0.00034,
1039
- "loss": 4.39,
1040
- "step": 172
1041
  },
1042
  {
1043
- "epoch": 0.29,
1044
  "learning_rate": 0.000342,
1045
- "loss": 4.4518,
1046
- "step": 173
1047
  },
1048
  {
1049
- "epoch": 0.29,
1050
  "learning_rate": 0.00034399999999999996,
1051
- "loss": 4.4439,
1052
- "step": 174
1053
  },
1054
  {
1055
- "epoch": 0.29,
1056
  "learning_rate": 0.000346,
1057
- "loss": 4.675,
1058
- "step": 175
1059
  },
1060
  {
1061
- "epoch": 0.3,
1062
  "learning_rate": 0.000348,
1063
- "loss": 4.6505,
1064
- "step": 176
1065
  },
1066
  {
1067
- "epoch": 0.3,
1068
  "learning_rate": 0.00035,
1069
- "loss": 4.5945,
1070
- "step": 177
1071
  },
1072
  {
1073
- "epoch": 0.3,
1074
  "learning_rate": 0.000352,
1075
- "loss": 4.395,
1076
- "step": 178
1077
  },
1078
  {
1079
- "epoch": 0.3,
1080
  "learning_rate": 0.000354,
1081
- "loss": 4.5278,
1082
- "step": 179
1083
  },
1084
  {
1085
- "epoch": 0.3,
1086
  "learning_rate": 0.000356,
1087
- "loss": 4.424,
1088
- "step": 180
1089
  },
1090
  {
1091
- "epoch": 0.3,
1092
  "learning_rate": 0.000358,
1093
- "loss": 4.8036,
1094
- "step": 181
1095
  },
1096
  {
1097
- "epoch": 0.31,
1098
  "learning_rate": 0.00035999999999999997,
1099
- "loss": 4.5129,
1100
- "step": 182
1101
  },
1102
  {
1103
- "epoch": 0.31,
1104
  "learning_rate": 0.000362,
1105
- "loss": 4.559,
1106
- "step": 183
1107
  },
1108
  {
1109
- "epoch": 0.31,
1110
  "learning_rate": 0.000364,
1111
- "loss": 4.3398,
1112
- "step": 184
1113
  },
1114
  {
1115
- "epoch": 0.31,
1116
  "learning_rate": 0.000366,
1117
- "loss": 4.9099,
1118
- "step": 185
1119
  },
1120
  {
1121
- "epoch": 0.31,
1122
  "learning_rate": 0.000368,
1123
- "loss": 4.5334,
1124
- "step": 186
1125
  },
1126
  {
1127
- "epoch": 0.31,
1128
  "learning_rate": 0.00037,
1129
- "loss": 4.6366,
1130
- "step": 187
1131
  },
1132
  {
1133
- "epoch": 0.32,
1134
  "learning_rate": 0.000372,
1135
- "loss": 10.5141,
1136
- "step": 188
1137
  },
1138
  {
1139
- "epoch": 0.32,
1140
  "learning_rate": 0.000374,
1141
- "loss": 5.2225,
1142
- "step": 189
1143
  },
1144
  {
1145
- "epoch": 0.32,
1146
  "learning_rate": 0.00037600000000000003,
1147
- "loss": 4.7898,
1148
- "step": 190
1149
  },
1150
  {
1151
- "epoch": 0.32,
1152
  "learning_rate": 0.000378,
1153
- "loss": 4.9854,
1154
- "step": 191
1155
  },
1156
  {
1157
- "epoch": 0.32,
1158
  "learning_rate": 0.00038,
1159
- "loss": 4.7617,
1160
- "step": 192
1161
  },
1162
  {
1163
- "epoch": 0.32,
1164
  "learning_rate": 0.000382,
1165
- "loss": 5.625,
1166
- "step": 193
1167
  },
1168
  {
1169
- "epoch": 0.33,
1170
  "learning_rate": 0.000384,
1171
- "loss": 4.8983,
1172
- "step": 194
1173
  },
1174
  {
1175
- "epoch": 0.33,
1176
  "learning_rate": 0.000386,
1177
- "loss": 4.7163,
1178
- "step": 195
1179
  },
1180
  {
1181
- "epoch": 0.33,
1182
  "learning_rate": 0.000388,
1183
- "loss": 4.709,
1184
- "step": 196
1185
  },
1186
  {
1187
- "epoch": 0.33,
1188
  "learning_rate": 0.00039000000000000005,
1189
- "loss": 5.4669,
1190
- "step": 197
1191
  },
1192
  {
1193
- "epoch": 0.33,
1194
  "learning_rate": 0.00039200000000000004,
1195
- "loss": 5.1185,
1196
- "step": 198
1197
  },
1198
  {
1199
- "epoch": 0.33,
1200
  "learning_rate": 0.00039400000000000004,
1201
- "loss": 5.4637,
1202
- "step": 199
1203
  },
1204
  {
1205
- "epoch": 0.34,
1206
  "learning_rate": 0.00039600000000000003,
1207
- "loss": 5.1612,
1208
- "step": 200
1209
  },
1210
  {
1211
- "epoch": 0.34,
1212
  "learning_rate": 0.000398,
1213
- "loss": 4.7685,
1214
- "step": 201
1215
  },
1216
  {
1217
- "epoch": 0.34,
1218
  "learning_rate": 0.0004,
1219
- "loss": 4.7321,
1220
- "step": 202
1221
  },
1222
  {
1223
- "epoch": 0.34,
1224
  "learning_rate": 0.000402,
1225
- "loss": 4.8251,
1226
- "step": 203
1227
  },
1228
  {
1229
- "epoch": 0.34,
1230
  "learning_rate": 0.000404,
1231
- "loss": 4.582,
1232
- "step": 204
1233
  },
1234
  {
1235
- "epoch": 0.34,
1236
  "learning_rate": 0.00040600000000000006,
1237
- "loss": 4.7087,
1238
- "step": 205
1239
  },
1240
  {
1241
- "epoch": 0.35,
1242
  "learning_rate": 0.000408,
1243
- "loss": 4.6998,
1244
- "step": 206
1245
  },
1246
  {
1247
- "epoch": 0.35,
1248
  "learning_rate": 0.00041,
1249
- "loss": 4.7384,
1250
- "step": 207
1251
  },
1252
  {
1253
- "epoch": 0.35,
1254
  "learning_rate": 0.000412,
1255
- "loss": 5.0725,
1256
- "step": 208
1257
  },
1258
  {
1259
- "epoch": 0.35,
1260
  "learning_rate": 0.000414,
1261
- "loss": 4.5905,
1262
- "step": 209
1263
  },
1264
  {
1265
- "epoch": 0.35,
1266
  "learning_rate": 0.000416,
1267
- "loss": 4.6788,
1268
- "step": 210
1269
  },
1270
  {
1271
- "epoch": 0.35,
1272
  "learning_rate": 0.00041799999999999997,
1273
- "loss": 4.7633,
1274
- "step": 211
1275
  },
1276
  {
1277
- "epoch": 0.36,
1278
  "learning_rate": 0.00042,
1279
- "loss": 4.5983,
1280
- "step": 212
1281
  },
1282
  {
1283
- "epoch": 0.36,
1284
  "learning_rate": 0.000422,
1285
- "loss": 4.7049,
1286
- "step": 213
1287
  },
1288
  {
1289
- "epoch": 0.36,
1290
  "learning_rate": 0.000424,
1291
- "loss": 4.7085,
1292
- "step": 214
1293
  },
1294
  {
1295
- "epoch": 0.36,
1296
  "learning_rate": 0.000426,
1297
- "loss": 4.7324,
1298
- "step": 215
1299
  },
1300
  {
1301
- "epoch": 0.36,
1302
  "learning_rate": 0.000428,
1303
- "loss": 4.7549,
1304
- "step": 216
1305
  },
1306
  {
1307
- "epoch": 0.36,
1308
  "learning_rate": 0.00043,
1309
- "loss": 4.5151,
1310
- "step": 217
1311
  },
1312
  {
1313
- "epoch": 0.37,
1314
  "learning_rate": 0.000432,
1315
- "loss": 4.737,
1316
- "step": 218
1317
  },
1318
  {
1319
- "epoch": 0.37,
1320
  "learning_rate": 0.00043400000000000003,
1321
- "loss": 5.079,
1322
- "step": 219
1323
  },
1324
  {
1325
- "epoch": 0.37,
1326
  "learning_rate": 0.000436,
1327
- "loss": 4.9301,
1328
- "step": 220
1329
  },
1330
  {
1331
- "epoch": 0.37,
1332
  "learning_rate": 0.000438,
1333
- "loss": 4.7236,
1334
- "step": 221
1335
  },
1336
  {
1337
- "epoch": 0.37,
1338
  "learning_rate": 0.00044,
1339
- "loss": 4.6851,
1340
- "step": 222
1341
  },
1342
  {
1343
- "epoch": 0.37,
1344
  "learning_rate": 0.000442,
1345
- "loss": 4.5696,
1346
- "step": 223
1347
  },
1348
  {
1349
- "epoch": 0.38,
1350
  "learning_rate": 0.000444,
1351
- "loss": 4.7287,
1352
- "step": 224
1353
  },
1354
  {
1355
- "epoch": 0.38,
1356
  "learning_rate": 0.000446,
1357
- "loss": 4.9131,
1358
- "step": 225
1359
  },
1360
  {
1361
- "epoch": 0.38,
1362
  "learning_rate": 0.000448,
1363
- "loss": 4.545,
1364
- "step": 226
1365
  },
1366
  {
1367
- "epoch": 0.38,
1368
  "learning_rate": 0.00045000000000000004,
1369
- "loss": 4.411,
1370
- "step": 227
1371
  },
1372
  {
1373
- "epoch": 0.38,
1374
  "learning_rate": 0.00045200000000000004,
1375
- "loss": 4.7007,
1376
- "step": 228
1377
  },
1378
  {
1379
- "epoch": 0.39,
1380
  "learning_rate": 0.00045400000000000003,
1381
- "loss": 4.8379,
1382
- "step": 229
1383
  },
1384
  {
1385
- "epoch": 0.39,
1386
  "learning_rate": 0.000456,
1387
- "loss": 4.6935,
1388
- "step": 230
1389
  },
1390
  {
1391
- "epoch": 0.39,
1392
  "learning_rate": 0.000458,
1393
- "loss": 4.6056,
1394
- "step": 231
1395
  },
1396
  {
1397
- "epoch": 0.39,
1398
  "learning_rate": 0.00046,
1399
- "loss": 4.695,
1400
- "step": 232
1401
  },
1402
  {
1403
- "epoch": 0.39,
1404
  "learning_rate": 0.000462,
1405
- "loss": 5.4116,
1406
- "step": 233
1407
  },
1408
  {
1409
- "epoch": 0.39,
1410
  "learning_rate": 0.00046400000000000006,
1411
- "loss": 4.7751,
1412
- "step": 234
1413
  },
1414
  {
1415
- "epoch": 0.4,
1416
  "learning_rate": 0.00046600000000000005,
1417
- "loss": 4.8383,
1418
- "step": 235
1419
  },
1420
  {
1421
- "epoch": 0.4,
1422
  "learning_rate": 0.00046800000000000005,
1423
- "loss": 4.5302,
1424
- "step": 236
1425
  },
1426
  {
1427
- "epoch": 0.4,
1428
  "learning_rate": 0.00047,
1429
- "loss": 4.4752,
1430
- "step": 237
1431
  },
1432
  {
1433
- "epoch": 0.4,
1434
  "learning_rate": 0.000472,
1435
- "loss": 4.4729,
1436
- "step": 238
1437
  },
1438
  {
1439
- "epoch": 0.4,
1440
  "learning_rate": 0.000474,
1441
- "loss": 4.6228,
1442
- "step": 239
1443
  },
1444
  {
1445
- "epoch": 0.4,
1446
  "learning_rate": 0.00047599999999999997,
1447
- "loss": 4.5783,
1448
- "step": 240
1449
  },
1450
  {
1451
- "epoch": 0.41,
1452
  "learning_rate": 0.00047799999999999996,
1453
- "loss": 4.9697,
1454
- "step": 241
1455
  },
1456
  {
1457
- "epoch": 0.41,
1458
  "learning_rate": 0.00048,
1459
- "loss": 4.8311,
1460
- "step": 242
1461
  },
1462
  {
1463
- "epoch": 0.41,
1464
  "learning_rate": 0.000482,
1465
- "loss": 4.7072,
1466
- "step": 243
1467
  },
1468
  {
1469
- "epoch": 0.41,
1470
  "learning_rate": 0.000484,
1471
- "loss": 4.7736,
1472
- "step": 244
1473
  },
1474
  {
1475
- "epoch": 0.41,
1476
  "learning_rate": 0.000486,
1477
- "loss": 5.3237,
1478
- "step": 245
1479
  },
1480
  {
1481
- "epoch": 0.41,
1482
  "learning_rate": 0.000488,
1483
- "loss": 4.9429,
1484
- "step": 246
1485
  },
1486
  {
1487
- "epoch": 0.42,
1488
  "learning_rate": 0.00049,
1489
- "loss": 5.2372,
1490
- "step": 247
1491
  },
1492
  {
1493
- "epoch": 0.42,
1494
  "learning_rate": 0.000492,
1495
- "loss": 5.3449,
1496
- "step": 248
1497
  },
1498
  {
1499
- "epoch": 0.42,
1500
  "learning_rate": 0.000494,
1501
- "loss": 5.7414,
1502
- "step": 249
1503
  },
1504
  {
1505
- "epoch": 0.42,
1506
  "learning_rate": 0.000496,
1507
- "loss": 5.3932,
1508
- "step": 250
1509
  },
1510
  {
1511
- "epoch": 0.42,
1512
  "learning_rate": 0.000498,
1513
- "loss": 5.1953,
1514
- "step": 251
1515
  },
1516
  {
1517
- "epoch": 0.42,
1518
  "learning_rate": 0.0005,
1519
- "loss": 4.9961,
1520
- "step": 252
1521
  },
1522
  {
1523
- "epoch": 0.43,
1524
  "learning_rate": 0.0005020000000000001,
1525
- "loss": 4.8454,
1526
- "step": 253
1527
  },
1528
  {
1529
- "epoch": 0.43,
1530
  "learning_rate": 0.000504,
1531
- "loss": 4.7123,
1532
- "step": 254
1533
  },
1534
  {
1535
- "epoch": 0.43,
1536
  "learning_rate": 0.000506,
1537
- "loss": 5.0182,
1538
- "step": 255
1539
  },
1540
  {
1541
- "epoch": 0.43,
1542
  "learning_rate": 0.000508,
1543
- "loss": 4.715,
1544
- "step": 256
1545
  },
1546
  {
1547
- "epoch": 0.43,
1548
  "learning_rate": 0.00051,
1549
- "loss": 4.7092,
1550
- "step": 257
1551
  },
1552
  {
1553
- "epoch": 0.43,
1554
  "learning_rate": 0.000512,
1555
- "loss": 4.6732,
1556
- "step": 258
1557
  },
1558
  {
1559
- "epoch": 0.44,
1560
  "learning_rate": 0.000514,
1561
- "loss": 4.7211,
1562
- "step": 259
1563
  },
1564
  {
1565
- "epoch": 0.44,
1566
  "learning_rate": 0.0005160000000000001,
1567
- "loss": 4.8059,
1568
- "step": 260
1569
  },
1570
  {
1571
- "epoch": 0.44,
1572
  "learning_rate": 0.000518,
1573
- "loss": 4.6999,
1574
- "step": 261
1575
  },
1576
  {
1577
- "epoch": 0.44,
1578
  "learning_rate": 0.0005200000000000001,
1579
- "loss": 4.7275,
1580
- "step": 262
1581
  },
1582
  {
1583
- "epoch": 0.44,
1584
  "learning_rate": 0.000522,
1585
- "loss": 5.2414,
1586
- "step": 263
1587
  },
1588
  {
1589
- "epoch": 0.44,
1590
  "learning_rate": 0.000524,
1591
- "loss": 4.4321,
1592
- "step": 264
1593
  },
1594
  {
1595
- "epoch": 0.45,
1596
  "learning_rate": 0.000526,
1597
- "loss": 4.6628,
1598
- "step": 265
1599
  },
1600
  {
1601
- "epoch": 0.45,
1602
  "learning_rate": 0.000528,
1603
- "loss": 4.6351,
1604
- "step": 266
1605
  },
1606
  {
1607
- "epoch": 0.45,
1608
  "learning_rate": 0.0005300000000000001,
1609
- "loss": 4.8597,
1610
- "step": 267
1611
  },
1612
  {
1613
- "epoch": 0.45,
1614
  "learning_rate": 0.000532,
1615
- "loss": 4.6425,
1616
- "step": 268
1617
  },
1618
  {
1619
- "epoch": 0.45,
1620
  "learning_rate": 0.0005340000000000001,
1621
- "loss": 4.7052,
1622
- "step": 269
1623
  },
1624
  {
1625
- "epoch": 0.45,
1626
  "learning_rate": 0.000536,
1627
- "loss": 4.7256,
1628
- "step": 270
1629
  },
1630
  {
1631
- "epoch": 0.46,
1632
  "learning_rate": 0.0005380000000000001,
1633
- "loss": 4.6104,
1634
- "step": 271
1635
  },
1636
  {
1637
- "epoch": 0.46,
1638
  "learning_rate": 0.00054,
1639
- "loss": 4.819,
1640
- "step": 272
1641
  },
1642
  {
1643
- "epoch": 0.46,
1644
  "learning_rate": 0.0005420000000000001,
1645
- "loss": 4.9087,
1646
- "step": 273
1647
  },
1648
  {
1649
- "epoch": 0.46,
1650
  "learning_rate": 0.0005440000000000001,
1651
- "loss": 4.7817,
1652
- "step": 274
1653
- },
1654
- {
1655
- "epoch": 0.46,
1656
- "learning_rate": 0.000546,
1657
- "loss": 4.7076,
1658
- "step": 275
1659
- },
1660
- {
1661
- "epoch": 0.46,
1662
- "learning_rate": 0.0005480000000000001,
1663
- "loss": 4.7856,
1664
- "step": 276
1665
- },
1666
- {
1667
- "epoch": 0.47,
1668
- "learning_rate": 0.00055,
1669
- "loss": 4.8639,
1670
- "step": 277
1671
- },
1672
- {
1673
- "epoch": 0.47,
1674
- "learning_rate": 0.0005520000000000001,
1675
- "loss": 4.9141,
1676
- "step": 278
1677
- },
1678
- {
1679
- "epoch": 0.47,
1680
- "learning_rate": 0.000554,
1681
- "loss": 4.7599,
1682
- "step": 279
1683
- },
1684
- {
1685
- "epoch": 0.47,
1686
- "learning_rate": 0.0005560000000000001,
1687
- "loss": 4.8849,
1688
- "step": 280
1689
- },
1690
- {
1691
- "epoch": 0.47,
1692
- "learning_rate": 0.000558,
1693
- "loss": 4.8114,
1694
- "step": 281
1695
- },
1696
- {
1697
- "epoch": 0.47,
1698
- "learning_rate": 0.0005600000000000001,
1699
- "loss": 4.8759,
1700
- "step": 282
1701
- },
1702
- {
1703
- "epoch": 0.48,
1704
- "learning_rate": 0.0005620000000000001,
1705
- "loss": 4.6076,
1706
- "step": 283
1707
- },
1708
- {
1709
- "epoch": 0.48,
1710
- "learning_rate": 0.0005639999999999999,
1711
- "loss": 4.6381,
1712
- "step": 284
1713
- },
1714
- {
1715
- "epoch": 0.48,
1716
- "learning_rate": 0.000566,
1717
- "loss": 4.8706,
1718
- "step": 285
1719
- },
1720
- {
1721
- "epoch": 0.48,
1722
- "learning_rate": 0.0005679999999999999,
1723
- "loss": 4.7356,
1724
- "step": 286
1725
- },
1726
- {
1727
- "epoch": 0.48,
1728
- "learning_rate": 0.00057,
1729
- "loss": 5.0577,
1730
- "step": 287
1731
- },
1732
- {
1733
- "epoch": 0.48,
1734
- "learning_rate": 0.0005719999999999999,
1735
- "loss": 4.5367,
1736
- "step": 288
1737
- },
1738
- {
1739
- "epoch": 0.49,
1740
- "learning_rate": 0.000574,
1741
- "loss": 4.7395,
1742
- "step": 289
1743
- },
1744
- {
1745
- "epoch": 0.49,
1746
- "learning_rate": 0.000576,
1747
- "loss": 4.7516,
1748
- "step": 290
1749
- },
1750
- {
1751
- "epoch": 0.49,
1752
- "learning_rate": 0.000578,
1753
- "loss": 4.8814,
1754
- "step": 291
1755
- },
1756
- {
1757
- "epoch": 0.49,
1758
- "learning_rate": 0.00058,
1759
- "loss": 4.7253,
1760
- "step": 292
1761
- },
1762
- {
1763
- "epoch": 0.49,
1764
- "learning_rate": 0.0005819999999999999,
1765
- "loss": 4.8064,
1766
- "step": 293
1767
- },
1768
- {
1769
- "epoch": 0.49,
1770
- "learning_rate": 0.000584,
1771
- "loss": 4.9757,
1772
- "step": 294
1773
- },
1774
- {
1775
- "epoch": 0.5,
1776
- "learning_rate": 0.0005859999999999999,
1777
- "loss": 4.8986,
1778
- "step": 295
1779
- },
1780
- {
1781
- "epoch": 0.5,
1782
- "learning_rate": 0.000588,
1783
- "loss": 4.7132,
1784
- "step": 296
1785
- },
1786
- {
1787
- "epoch": 0.5,
1788
- "learning_rate": 0.00059,
1789
- "loss": 5.6149,
1790
- "step": 297
1791
- },
1792
- {
1793
- "epoch": 0.5,
1794
- "learning_rate": 0.000592,
1795
- "loss": 6.4538,
1796
- "step": 298
1797
- },
1798
- {
1799
- "epoch": 0.5,
1800
- "learning_rate": 0.000594,
1801
- "loss": 5.2126,
1802
- "step": 299
1803
- },
1804
- {
1805
- "epoch": 0.5,
1806
- "learning_rate": 0.000596,
1807
- "loss": 5.7153,
1808
- "step": 300
1809
- },
1810
- {
1811
- "epoch": 0.51,
1812
- "learning_rate": 0.000598,
1813
- "loss": 5.235,
1814
- "step": 301
1815
- },
1816
- {
1817
- "epoch": 0.51,
1818
- "learning_rate": 0.0006,
1819
- "loss": 4.9152,
1820
- "step": 302
1821
- },
1822
- {
1823
- "epoch": 0.51,
1824
- "learning_rate": 0.000602,
1825
- "loss": 5.0532,
1826
- "step": 303
1827
- },
1828
- {
1829
- "epoch": 0.51,
1830
- "learning_rate": 0.000604,
1831
- "loss": 4.7904,
1832
- "step": 304
1833
- },
1834
- {
1835
- "epoch": 0.51,
1836
- "learning_rate": 0.000606,
1837
- "loss": 5.0049,
1838
- "step": 305
1839
- },
1840
- {
1841
- "epoch": 0.51,
1842
- "learning_rate": 0.000608,
1843
- "loss": 4.9579,
1844
- "step": 306
1845
- },
1846
- {
1847
- "epoch": 0.52,
1848
- "learning_rate": 0.00061,
1849
- "loss": 4.8652,
1850
- "step": 307
1851
- },
1852
- {
1853
- "epoch": 0.52,
1854
- "learning_rate": 0.000612,
1855
- "loss": 5.024,
1856
- "step": 308
1857
- },
1858
- {
1859
- "epoch": 0.52,
1860
- "learning_rate": 0.000614,
1861
- "loss": 4.9125,
1862
- "step": 309
1863
- },
1864
- {
1865
- "epoch": 0.52,
1866
- "learning_rate": 0.000616,
1867
- "loss": 4.8775,
1868
- "step": 310
1869
- },
1870
- {
1871
- "epoch": 0.52,
1872
- "learning_rate": 0.0006180000000000001,
1873
- "loss": 4.6219,
1874
- "step": 311
1875
- },
1876
- {
1877
- "epoch": 0.52,
1878
- "learning_rate": 0.00062,
1879
- "loss": 4.9048,
1880
- "step": 312
1881
- },
1882
- {
1883
- "epoch": 0.53,
1884
- "learning_rate": 0.000622,
1885
- "loss": 4.8739,
1886
- "step": 313
1887
- },
1888
- {
1889
- "epoch": 0.53,
1890
- "learning_rate": 0.000624,
1891
- "loss": 4.8017,
1892
- "step": 314
1893
- },
1894
- {
1895
- "epoch": 0.53,
1896
- "learning_rate": 0.000626,
1897
- "loss": 4.7513,
1898
- "step": 315
1899
- },
1900
- {
1901
- "epoch": 0.53,
1902
- "learning_rate": 0.000628,
1903
- "loss": 4.8819,
1904
- "step": 316
1905
- },
1906
- {
1907
- "epoch": 0.53,
1908
- "learning_rate": 0.00063,
1909
- "loss": 4.7052,
1910
- "step": 317
1911
- },
1912
- {
1913
- "epoch": 0.53,
1914
- "learning_rate": 0.000632,
1915
- "loss": 5.1057,
1916
- "step": 318
1917
- },
1918
- {
1919
- "epoch": 0.54,
1920
- "learning_rate": 0.000634,
1921
- "loss": 5.0369,
1922
- "step": 319
1923
- },
1924
- {
1925
- "epoch": 0.54,
1926
- "learning_rate": 0.0006360000000000001,
1927
- "loss": 4.8897,
1928
- "step": 320
1929
- },
1930
- {
1931
- "epoch": 0.54,
1932
- "learning_rate": 0.000638,
1933
- "loss": 4.6517,
1934
- "step": 321
1935
- },
1936
- {
1937
- "epoch": 0.54,
1938
- "learning_rate": 0.00064,
1939
- "loss": 4.662,
1940
- "step": 322
1941
- },
1942
- {
1943
- "epoch": 0.54,
1944
- "learning_rate": 0.000642,
1945
- "loss": 5.0655,
1946
- "step": 323
1947
- },
1948
- {
1949
- "epoch": 0.54,
1950
- "learning_rate": 0.000644,
1951
- "loss": 4.7497,
1952
- "step": 324
1953
- },
1954
- {
1955
- "epoch": 0.55,
1956
- "learning_rate": 0.000646,
1957
- "loss": 4.9575,
1958
- "step": 325
1959
- },
1960
- {
1961
- "epoch": 0.55,
1962
- "learning_rate": 0.000648,
1963
- "loss": 4.8104,
1964
- "step": 326
1965
- },
1966
- {
1967
- "epoch": 0.55,
1968
- "learning_rate": 0.0006500000000000001,
1969
- "loss": 5.0224,
1970
- "step": 327
1971
- },
1972
- {
1973
- "epoch": 0.55,
1974
- "learning_rate": 0.000652,
1975
- "loss": 4.7339,
1976
- "step": 328
1977
- },
1978
- {
1979
- "epoch": 0.55,
1980
- "learning_rate": 0.0006540000000000001,
1981
- "loss": 4.6435,
1982
- "step": 329
1983
- },
1984
- {
1985
- "epoch": 0.55,
1986
- "learning_rate": 0.000656,
1987
- "loss": 4.8785,
1988
- "step": 330
1989
- },
1990
- {
1991
- "epoch": 0.56,
1992
- "learning_rate": 0.0006580000000000001,
1993
- "loss": 4.8467,
1994
- "step": 331
1995
- },
1996
- {
1997
- "epoch": 0.56,
1998
- "learning_rate": 0.00066,
1999
- "loss": 4.6529,
2000
- "step": 332
2001
- },
2002
- {
2003
- "epoch": 0.56,
2004
- "learning_rate": 0.000662,
2005
- "loss": 5.1175,
2006
- "step": 333
2007
- },
2008
- {
2009
- "epoch": 0.56,
2010
- "learning_rate": 0.0006640000000000001,
2011
- "loss": 5.1239,
2012
- "step": 334
2013
- },
2014
- {
2015
- "epoch": 0.56,
2016
- "learning_rate": 0.000666,
2017
- "loss": 4.916,
2018
- "step": 335
2019
- },
2020
- {
2021
- "epoch": 0.56,
2022
- "learning_rate": 0.0006680000000000001,
2023
- "loss": 4.6657,
2024
- "step": 336
2025
- },
2026
- {
2027
- "epoch": 0.57,
2028
- "learning_rate": 0.00067,
2029
- "loss": 4.7292,
2030
- "step": 337
2031
- },
2032
- {
2033
- "epoch": 0.57,
2034
- "learning_rate": 0.0006720000000000001,
2035
- "loss": 4.6982,
2036
- "step": 338
2037
- },
2038
- {
2039
- "epoch": 0.57,
2040
- "learning_rate": 0.000674,
2041
- "loss": 4.689,
2042
- "step": 339
2043
- },
2044
- {
2045
- "epoch": 0.57,
2046
- "learning_rate": 0.0006760000000000001,
2047
- "loss": 5.0327,
2048
- "step": 340
2049
- },
2050
- {
2051
- "epoch": 0.57,
2052
- "learning_rate": 0.0006780000000000001,
2053
- "loss": 4.813,
2054
- "step": 341
2055
- },
2056
- {
2057
- "epoch": 0.58,
2058
- "learning_rate": 0.00068,
2059
- "loss": 4.8782,
2060
- "step": 342
2061
- },
2062
- {
2063
- "epoch": 0.58,
2064
- "learning_rate": 0.0006820000000000001,
2065
- "loss": 4.9557,
2066
- "step": 343
2067
- },
2068
- {
2069
- "epoch": 0.58,
2070
- "learning_rate": 0.000684,
2071
- "loss": 4.9666,
2072
- "step": 344
2073
- },
2074
- {
2075
- "epoch": 0.58,
2076
- "learning_rate": 0.0006860000000000001,
2077
- "loss": 5.353,
2078
- "step": 345
2079
- },
2080
- {
2081
- "epoch": 0.58,
2082
- "learning_rate": 0.0006879999999999999,
2083
- "loss": 4.8186,
2084
- "step": 346
2085
- },
2086
- {
2087
- "epoch": 0.58,
2088
- "learning_rate": 0.00069,
2089
- "loss": 4.7452,
2090
- "step": 347
2091
- },
2092
- {
2093
- "epoch": 0.59,
2094
- "learning_rate": 0.000692,
2095
- "loss": 5.1056,
2096
- "step": 348
2097
- },
2098
- {
2099
- "epoch": 0.59,
2100
- "learning_rate": 0.000694,
2101
- "loss": 4.8236,
2102
- "step": 349
2103
- },
2104
- {
2105
- "epoch": 0.59,
2106
- "learning_rate": 0.000696,
2107
- "loss": 4.9508,
2108
- "step": 350
2109
- },
2110
- {
2111
- "epoch": 0.59,
2112
- "learning_rate": 0.0006979999999999999,
2113
- "loss": 5.2547,
2114
- "step": 351
2115
- },
2116
- {
2117
- "epoch": 0.59,
2118
- "learning_rate": 0.0007,
2119
- "loss": 5.0417,
2120
- "step": 352
2121
- },
2122
- {
2123
- "epoch": 0.59,
2124
- "learning_rate": 0.0007019999999999999,
2125
- "loss": 4.9277,
2126
- "step": 353
2127
- },
2128
- {
2129
- "epoch": 0.6,
2130
- "learning_rate": 0.000704,
2131
- "loss": 4.7628,
2132
- "step": 354
2133
- },
2134
- {
2135
- "epoch": 0.6,
2136
- "learning_rate": 0.0007059999999999999,
2137
- "loss": 4.8104,
2138
- "step": 355
2139
- },
2140
- {
2141
- "epoch": 0.6,
2142
- "learning_rate": 0.000708,
2143
- "loss": 4.8422,
2144
- "step": 356
2145
- },
2146
- {
2147
- "epoch": 0.6,
2148
- "learning_rate": 0.00071,
2149
- "loss": 4.8045,
2150
- "step": 357
2151
- },
2152
- {
2153
- "epoch": 0.6,
2154
- "learning_rate": 0.000712,
2155
- "loss": 4.9178,
2156
- "step": 358
2157
- },
2158
- {
2159
- "epoch": 0.6,
2160
- "learning_rate": 0.000714,
2161
- "loss": 4.8115,
2162
- "step": 359
2163
- },
2164
- {
2165
- "epoch": 0.61,
2166
- "learning_rate": 0.000716,
2167
- "loss": 5.0046,
2168
- "step": 360
2169
- },
2170
- {
2171
- "epoch": 0.61,
2172
- "learning_rate": 0.000718,
2173
- "loss": 4.8284,
2174
- "step": 361
2175
- },
2176
- {
2177
- "epoch": 0.61,
2178
- "learning_rate": 0.0007199999999999999,
2179
- "loss": 4.746,
2180
- "step": 362
2181
- },
2182
- {
2183
- "epoch": 0.61,
2184
- "learning_rate": 0.000722,
2185
- "loss": 5.0836,
2186
- "step": 363
2187
- },
2188
- {
2189
- "epoch": 0.61,
2190
- "learning_rate": 0.000724,
2191
- "loss": 4.7086,
2192
- "step": 364
2193
- },
2194
- {
2195
- "epoch": 0.61,
2196
- "learning_rate": 0.000726,
2197
- "loss": 4.8806,
2198
- "step": 365
2199
- },
2200
- {
2201
- "epoch": 0.62,
2202
- "learning_rate": 0.000728,
2203
- "loss": 4.7059,
2204
- "step": 366
2205
- },
2206
- {
2207
- "epoch": 0.62,
2208
- "learning_rate": 0.00073,
2209
- "loss": 4.8378,
2210
- "step": 367
2211
- },
2212
- {
2213
- "epoch": 0.62,
2214
- "learning_rate": 0.000732,
2215
- "loss": 4.8017,
2216
- "step": 368
2217
- },
2218
- {
2219
- "epoch": 0.62,
2220
- "learning_rate": 0.000734,
2221
- "loss": 4.8343,
2222
- "step": 369
2223
- },
2224
- {
2225
- "epoch": 0.62,
2226
- "learning_rate": 0.000736,
2227
- "loss": 5.0083,
2228
- "step": 370
2229
- },
2230
- {
2231
- "epoch": 0.62,
2232
- "learning_rate": 0.000738,
2233
- "loss": 4.7563,
2234
- "step": 371
2235
- },
2236
- {
2237
- "epoch": 0.63,
2238
- "learning_rate": 0.00074,
2239
- "loss": 4.7571,
2240
- "step": 372
2241
- },
2242
- {
2243
- "epoch": 0.63,
2244
- "learning_rate": 0.000742,
2245
- "loss": 4.7098,
2246
- "step": 373
2247
- },
2248
- {
2249
- "epoch": 0.63,
2250
- "learning_rate": 0.000744,
2251
- "loss": 5.308,
2252
- "step": 374
2253
- },
2254
- {
2255
- "epoch": 0.63,
2256
- "learning_rate": 0.000746,
2257
- "loss": 5.0252,
2258
- "step": 375
2259
- },
2260
- {
2261
- "epoch": 0.63,
2262
- "learning_rate": 0.000748,
2263
- "loss": 5.1313,
2264
- "step": 376
2265
- },
2266
- {
2267
- "epoch": 0.63,
2268
- "learning_rate": 0.00075,
2269
- "loss": 4.8109,
2270
- "step": 377
2271
- },
2272
- {
2273
- "epoch": 0.64,
2274
- "learning_rate": 0.0007520000000000001,
2275
- "loss": 4.8508,
2276
- "step": 378
2277
- },
2278
- {
2279
- "epoch": 0.64,
2280
- "learning_rate": 0.000754,
2281
- "loss": 4.8471,
2282
- "step": 379
2283
- },
2284
- {
2285
- "epoch": 0.64,
2286
- "learning_rate": 0.000756,
2287
- "loss": 4.9023,
2288
- "step": 380
2289
- },
2290
- {
2291
- "epoch": 0.64,
2292
- "learning_rate": 0.000758,
2293
- "loss": 4.9553,
2294
- "step": 381
2295
- },
2296
- {
2297
- "epoch": 0.64,
2298
- "learning_rate": 0.00076,
2299
- "loss": 4.7834,
2300
- "step": 382
2301
- },
2302
- {
2303
- "epoch": 0.64,
2304
- "learning_rate": 0.000762,
2305
- "loss": 5.1141,
2306
- "step": 383
2307
- },
2308
- {
2309
- "epoch": 0.65,
2310
- "learning_rate": 0.000764,
2311
- "loss": 4.8394,
2312
- "step": 384
2313
- },
2314
- {
2315
- "epoch": 0.65,
2316
- "learning_rate": 0.0007660000000000001,
2317
- "loss": 4.8951,
2318
- "step": 385
2319
- },
2320
- {
2321
- "epoch": 0.65,
2322
- "learning_rate": 0.000768,
2323
- "loss": 4.9039,
2324
- "step": 386
2325
- },
2326
- {
2327
- "epoch": 0.65,
2328
- "learning_rate": 0.0007700000000000001,
2329
- "loss": 5.2676,
2330
- "step": 387
2331
- },
2332
- {
2333
- "epoch": 0.65,
2334
- "learning_rate": 0.000772,
2335
- "loss": 4.8199,
2336
- "step": 388
2337
- },
2338
- {
2339
- "epoch": 0.65,
2340
- "learning_rate": 0.0007740000000000001,
2341
- "loss": 4.7974,
2342
- "step": 389
2343
- },
2344
- {
2345
- "epoch": 0.66,
2346
- "learning_rate": 0.000776,
2347
- "loss": 4.8914,
2348
- "step": 390
2349
- },
2350
- {
2351
- "epoch": 0.66,
2352
- "learning_rate": 0.000778,
2353
- "loss": 4.965,
2354
- "step": 391
2355
- },
2356
- {
2357
- "epoch": 0.66,
2358
- "learning_rate": 0.0007800000000000001,
2359
- "loss": 5.3528,
2360
- "step": 392
2361
- },
2362
- {
2363
- "epoch": 0.66,
2364
- "learning_rate": 0.000782,
2365
- "loss": 5.4938,
2366
- "step": 393
2367
- },
2368
- {
2369
- "epoch": 0.66,
2370
- "learning_rate": 0.0007840000000000001,
2371
- "loss": 5.1061,
2372
- "step": 394
2373
- },
2374
- {
2375
- "epoch": 0.66,
2376
- "learning_rate": 0.000786,
2377
- "loss": 5.2562,
2378
- "step": 395
2379
- },
2380
- {
2381
- "epoch": 0.67,
2382
- "learning_rate": 0.0007880000000000001,
2383
- "loss": 5.2242,
2384
- "step": 396
2385
- },
2386
- {
2387
- "epoch": 0.67,
2388
- "learning_rate": 0.00079,
2389
- "loss": 5.0489,
2390
- "step": 397
2391
- },
2392
- {
2393
- "epoch": 0.67,
2394
- "learning_rate": 0.0007920000000000001,
2395
- "loss": 5.7935,
2396
- "step": 398
2397
- },
2398
- {
2399
- "epoch": 0.67,
2400
- "learning_rate": 0.0007940000000000001,
2401
- "loss": 5.1701,
2402
- "step": 399
2403
- },
2404
- {
2405
- "epoch": 0.67,
2406
- "learning_rate": 0.000796,
2407
- "loss": 4.8904,
2408
- "step": 400
2409
- },
2410
- {
2411
- "epoch": 0.67,
2412
- "learning_rate": 0.0007980000000000001,
2413
- "loss": 5.0654,
2414
- "step": 401
2415
- },
2416
- {
2417
- "epoch": 0.68,
2418
- "learning_rate": 0.0008,
2419
- "loss": 4.8698,
2420
- "step": 402
2421
- },
2422
- {
2423
- "epoch": 0.68,
2424
- "learning_rate": 0.0008020000000000001,
2425
- "loss": 4.9322,
2426
- "step": 403
2427
- },
2428
- {
2429
- "epoch": 0.68,
2430
- "learning_rate": 0.000804,
2431
- "loss": 4.8403,
2432
- "step": 404
2433
- },
2434
- {
2435
- "epoch": 0.68,
2436
- "learning_rate": 0.0008060000000000001,
2437
- "loss": 4.8574,
2438
- "step": 405
2439
- },
2440
- {
2441
- "epoch": 0.68,
2442
- "learning_rate": 0.000808,
2443
- "loss": 4.742,
2444
- "step": 406
2445
- },
2446
- {
2447
- "epoch": 0.68,
2448
- "learning_rate": 0.0008100000000000001,
2449
- "loss": 5.2273,
2450
- "step": 407
2451
- },
2452
- {
2453
- "epoch": 0.69,
2454
- "learning_rate": 0.0008120000000000001,
2455
- "loss": 4.7645,
2456
- "step": 408
2457
- },
2458
- {
2459
- "epoch": 0.69,
2460
- "learning_rate": 0.0008139999999999999,
2461
- "loss": 4.752,
2462
- "step": 409
2463
- },
2464
- {
2465
- "epoch": 0.69,
2466
- "learning_rate": 0.000816,
2467
- "loss": 4.7622,
2468
- "step": 410
2469
- },
2470
- {
2471
- "epoch": 0.69,
2472
- "learning_rate": 0.0008179999999999999,
2473
- "loss": 4.688,
2474
- "step": 411
2475
- },
2476
- {
2477
- "epoch": 0.69,
2478
- "learning_rate": 0.00082,
2479
- "loss": 4.929,
2480
- "step": 412
2481
- },
2482
- {
2483
- "epoch": 0.69,
2484
- "learning_rate": 0.0008219999999999999,
2485
- "loss": 4.886,
2486
- "step": 413
2487
- },
2488
- {
2489
- "epoch": 0.7,
2490
- "learning_rate": 0.000824,
2491
- "loss": 4.9073,
2492
- "step": 414
2493
- },
2494
- {
2495
- "epoch": 0.7,
2496
- "learning_rate": 0.000826,
2497
- "loss": 4.8768,
2498
- "step": 415
2499
- },
2500
- {
2501
- "epoch": 0.7,
2502
- "learning_rate": 0.000828,
2503
- "loss": 4.7801,
2504
- "step": 416
2505
- },
2506
- {
2507
- "epoch": 0.7,
2508
- "learning_rate": 0.00083,
2509
- "loss": 4.734,
2510
- "step": 417
2511
- },
2512
- {
2513
- "epoch": 0.7,
2514
- "learning_rate": 0.000832,
2515
- "loss": 4.4348,
2516
- "step": 418
2517
- },
2518
- {
2519
- "epoch": 0.7,
2520
- "learning_rate": 0.000834,
2521
- "loss": 4.8986,
2522
- "step": 419
2523
- },
2524
- {
2525
- "epoch": 0.71,
2526
- "learning_rate": 0.0008359999999999999,
2527
- "loss": 4.8039,
2528
- "step": 420
2529
- },
2530
- {
2531
- "epoch": 0.71,
2532
- "learning_rate": 0.000838,
2533
- "loss": 4.4695,
2534
- "step": 421
2535
- },
2536
- {
2537
- "epoch": 0.71,
2538
- "learning_rate": 0.00084,
2539
- "loss": 4.6531,
2540
- "step": 422
2541
- },
2542
- {
2543
- "epoch": 0.71,
2544
- "learning_rate": 0.000842,
2545
- "loss": 4.8529,
2546
- "step": 423
2547
- },
2548
- {
2549
- "epoch": 0.71,
2550
- "learning_rate": 0.000844,
2551
- "loss": 4.7331,
2552
- "step": 424
2553
- },
2554
- {
2555
- "epoch": 0.71,
2556
- "learning_rate": 0.000846,
2557
- "loss": 4.8075,
2558
- "step": 425
2559
- },
2560
- {
2561
- "epoch": 0.72,
2562
- "learning_rate": 0.000848,
2563
- "loss": 4.8201,
2564
- "step": 426
2565
- },
2566
- {
2567
- "epoch": 0.72,
2568
- "learning_rate": 0.00085,
2569
- "loss": 4.5947,
2570
- "step": 427
2571
- },
2572
- {
2573
- "epoch": 0.72,
2574
- "learning_rate": 0.000852,
2575
- "loss": 4.5512,
2576
- "step": 428
2577
- },
2578
- {
2579
- "epoch": 0.72,
2580
- "learning_rate": 0.000854,
2581
- "loss": 4.9711,
2582
- "step": 429
2583
- },
2584
- {
2585
- "epoch": 0.72,
2586
- "learning_rate": 0.000856,
2587
- "loss": 4.742,
2588
- "step": 430
2589
- },
2590
- {
2591
- "epoch": 0.72,
2592
- "learning_rate": 0.000858,
2593
- "loss": 5.1297,
2594
- "step": 431
2595
- },
2596
- {
2597
- "epoch": 0.73,
2598
- "learning_rate": 0.00086,
2599
- "loss": 4.6386,
2600
- "step": 432
2601
- },
2602
- {
2603
- "epoch": 0.73,
2604
- "learning_rate": 0.000862,
2605
- "loss": 4.9415,
2606
- "step": 433
2607
- },
2608
- {
2609
- "epoch": 0.73,
2610
- "learning_rate": 0.000864,
2611
- "loss": 4.808,
2612
- "step": 434
2613
- },
2614
- {
2615
- "epoch": 0.73,
2616
- "learning_rate": 0.000866,
2617
- "loss": 4.9157,
2618
- "step": 435
2619
- },
2620
- {
2621
- "epoch": 0.73,
2622
- "learning_rate": 0.0008680000000000001,
2623
- "loss": 4.7158,
2624
- "step": 436
2625
- },
2626
- {
2627
- "epoch": 0.73,
2628
- "learning_rate": 0.00087,
2629
- "loss": 4.8154,
2630
- "step": 437
2631
- },
2632
- {
2633
- "epoch": 0.74,
2634
- "learning_rate": 0.000872,
2635
- "loss": 4.7926,
2636
- "step": 438
2637
- },
2638
- {
2639
- "epoch": 0.74,
2640
- "learning_rate": 0.000874,
2641
- "loss": 4.936,
2642
- "step": 439
2643
- },
2644
- {
2645
- "epoch": 0.74,
2646
- "learning_rate": 0.000876,
2647
- "loss": 5.1139,
2648
- "step": 440
2649
- },
2650
- {
2651
- "epoch": 0.74,
2652
- "learning_rate": 0.000878,
2653
- "loss": 5.2978,
2654
- "step": 441
2655
- },
2656
- {
2657
- "epoch": 0.74,
2658
- "learning_rate": 0.00088,
2659
- "loss": 6.1523,
2660
- "step": 442
2661
- },
2662
- {
2663
- "epoch": 0.74,
2664
- "learning_rate": 0.000882,
2665
- "loss": 5.2146,
2666
- "step": 443
2667
- },
2668
- {
2669
- "epoch": 0.75,
2670
- "learning_rate": 0.000884,
2671
- "loss": 5.0577,
2672
- "step": 444
2673
- },
2674
- {
2675
- "epoch": 0.75,
2676
- "learning_rate": 0.0008860000000000001,
2677
- "loss": 4.7544,
2678
- "step": 445
2679
- },
2680
- {
2681
- "epoch": 0.75,
2682
- "learning_rate": 0.000888,
2683
- "loss": 4.9598,
2684
- "step": 446
2685
- },
2686
- {
2687
- "epoch": 0.75,
2688
- "learning_rate": 0.0008900000000000001,
2689
- "loss": 4.9329,
2690
- "step": 447
2691
- },
2692
- {
2693
- "epoch": 0.75,
2694
- "learning_rate": 0.000892,
2695
- "loss": 4.969,
2696
- "step": 448
2697
- },
2698
- {
2699
- "epoch": 0.75,
2700
- "learning_rate": 0.000894,
2701
- "loss": 5.5613,
2702
- "step": 449
2703
- },
2704
- {
2705
- "epoch": 0.76,
2706
- "learning_rate": 0.000896,
2707
- "loss": 5.0466,
2708
- "step": 450
2709
- },
2710
- {
2711
- "epoch": 0.76,
2712
- "learning_rate": 0.000898,
2713
- "loss": 5.0624,
2714
- "step": 451
2715
- },
2716
- {
2717
- "epoch": 0.76,
2718
- "learning_rate": 0.0009000000000000001,
2719
- "loss": 5.1022,
2720
- "step": 452
2721
- },
2722
- {
2723
- "epoch": 0.76,
2724
- "learning_rate": 0.000902,
2725
- "loss": 4.892,
2726
- "step": 453
2727
- },
2728
- {
2729
- "epoch": 0.76,
2730
- "learning_rate": 0.0009040000000000001,
2731
- "loss": 4.9988,
2732
- "step": 454
2733
- },
2734
- {
2735
- "epoch": 0.77,
2736
- "learning_rate": 0.000906,
2737
- "loss": 5.0485,
2738
- "step": 455
2739
- },
2740
- {
2741
- "epoch": 0.77,
2742
- "learning_rate": 0.0009080000000000001,
2743
- "loss": 4.9697,
2744
- "step": 456
2745
- },
2746
- {
2747
- "epoch": 0.77,
2748
- "learning_rate": 0.00091,
2749
- "loss": 5.1938,
2750
- "step": 457
2751
- },
2752
- {
2753
- "epoch": 0.77,
2754
- "learning_rate": 0.000912,
2755
- "loss": 5.1972,
2756
- "step": 458
2757
- },
2758
- {
2759
- "epoch": 0.77,
2760
- "learning_rate": 0.0009140000000000001,
2761
- "loss": 5.088,
2762
- "step": 459
2763
- },
2764
- {
2765
- "epoch": 0.77,
2766
- "learning_rate": 0.000916,
2767
- "loss": 4.9671,
2768
- "step": 460
2769
- },
2770
- {
2771
- "epoch": 0.78,
2772
- "learning_rate": 0.0009180000000000001,
2773
- "loss": 5.0251,
2774
- "step": 461
2775
- },
2776
- {
2777
- "epoch": 0.78,
2778
- "learning_rate": 0.00092,
2779
- "loss": 4.7734,
2780
- "step": 462
2781
- },
2782
- {
2783
- "epoch": 0.78,
2784
- "learning_rate": 0.0009220000000000001,
2785
- "loss": 4.8,
2786
- "step": 463
2787
- },
2788
- {
2789
- "epoch": 0.78,
2790
- "learning_rate": 0.000924,
2791
- "loss": 5.2001,
2792
- "step": 464
2793
- },
2794
- {
2795
- "epoch": 0.78,
2796
- "learning_rate": 0.0009260000000000001,
2797
- "loss": 5.2366,
2798
- "step": 465
2799
- },
2800
- {
2801
- "epoch": 0.78,
2802
- "learning_rate": 0.0009280000000000001,
2803
- "loss": 5.0479,
2804
- "step": 466
2805
- },
2806
- {
2807
- "epoch": 0.79,
2808
- "learning_rate": 0.00093,
2809
- "loss": 4.9839,
2810
- "step": 467
2811
- },
2812
- {
2813
- "epoch": 0.79,
2814
- "learning_rate": 0.0009320000000000001,
2815
- "loss": 4.8302,
2816
- "step": 468
2817
- },
2818
- {
2819
- "epoch": 0.79,
2820
- "learning_rate": 0.000934,
2821
- "loss": 5.0382,
2822
- "step": 469
2823
- },
2824
- {
2825
- "epoch": 0.79,
2826
- "learning_rate": 0.0009360000000000001,
2827
- "loss": 4.7301,
2828
- "step": 470
2829
- },
2830
- {
2831
- "epoch": 0.79,
2832
- "learning_rate": 0.0009379999999999999,
2833
- "loss": 4.8497,
2834
- "step": 471
2835
- },
2836
- {
2837
- "epoch": 0.79,
2838
- "learning_rate": 0.00094,
2839
- "loss": 4.819,
2840
- "step": 472
2841
- },
2842
- {
2843
- "epoch": 0.8,
2844
- "learning_rate": 0.000942,
2845
- "loss": 4.9396,
2846
- "step": 473
2847
- },
2848
- {
2849
- "epoch": 0.8,
2850
- "learning_rate": 0.000944,
2851
- "loss": 4.8436,
2852
- "step": 474
2853
- },
2854
- {
2855
- "epoch": 0.8,
2856
- "learning_rate": 0.000946,
2857
- "loss": 4.8317,
2858
- "step": 475
2859
- },
2860
- {
2861
- "epoch": 0.8,
2862
- "learning_rate": 0.000948,
2863
- "loss": 5.2336,
2864
- "step": 476
2865
- },
2866
- {
2867
- "epoch": 0.8,
2868
- "learning_rate": 0.00095,
2869
- "loss": 4.9405,
2870
- "step": 477
2871
- },
2872
- {
2873
- "epoch": 0.8,
2874
- "learning_rate": 0.0009519999999999999,
2875
- "loss": 4.927,
2876
- "step": 478
2877
- },
2878
- {
2879
- "epoch": 0.81,
2880
- "learning_rate": 0.000954,
2881
- "loss": 4.6957,
2882
- "step": 479
2883
- },
2884
- {
2885
- "epoch": 0.81,
2886
- "learning_rate": 0.0009559999999999999,
2887
- "loss": 4.625,
2888
- "step": 480
2889
- },
2890
- {
2891
- "epoch": 0.81,
2892
- "learning_rate": 0.000958,
2893
- "loss": 5.3568,
2894
- "step": 481
2895
- },
2896
- {
2897
- "epoch": 0.81,
2898
- "learning_rate": 0.00096,
2899
- "loss": 5.7506,
2900
- "step": 482
2901
- },
2902
- {
2903
- "epoch": 0.81,
2904
- "learning_rate": 0.000962,
2905
- "loss": 5.1788,
2906
- "step": 483
2907
- },
2908
- {
2909
- "epoch": 0.81,
2910
- "learning_rate": 0.000964,
2911
- "loss": 5.0361,
2912
- "step": 484
2913
- },
2914
- {
2915
- "epoch": 0.82,
2916
- "learning_rate": 0.000966,
2917
- "loss": 4.8903,
2918
- "step": 485
2919
- },
2920
- {
2921
- "epoch": 0.82,
2922
- "learning_rate": 0.000968,
2923
- "loss": 4.8917,
2924
- "step": 486
2925
- },
2926
- {
2927
- "epoch": 0.82,
2928
- "learning_rate": 0.0009699999999999999,
2929
- "loss": 4.8628,
2930
- "step": 487
2931
- },
2932
- {
2933
- "epoch": 0.82,
2934
- "learning_rate": 0.000972,
2935
- "loss": 4.8763,
2936
- "step": 488
2937
- },
2938
- {
2939
- "epoch": 0.82,
2940
- "learning_rate": 0.000974,
2941
- "loss": 5.002,
2942
- "step": 489
2943
- },
2944
- {
2945
- "epoch": 0.82,
2946
- "learning_rate": 0.000976,
2947
- "loss": 4.7335,
2948
- "step": 490
2949
- },
2950
- {
2951
- "epoch": 0.83,
2952
- "learning_rate": 0.000978,
2953
- "loss": 4.9556,
2954
- "step": 491
2955
- },
2956
- {
2957
- "epoch": 0.83,
2958
- "learning_rate": 0.00098,
2959
- "loss": 4.9359,
2960
- "step": 492
2961
- },
2962
- {
2963
- "epoch": 0.83,
2964
- "learning_rate": 0.000982,
2965
- "loss": 4.715,
2966
- "step": 493
2967
- },
2968
- {
2969
- "epoch": 0.83,
2970
- "learning_rate": 0.000984,
2971
- "loss": 4.8199,
2972
- "step": 494
2973
- },
2974
- {
2975
- "epoch": 0.83,
2976
- "learning_rate": 0.0009860000000000001,
2977
- "loss": 4.9607,
2978
- "step": 495
2979
- },
2980
- {
2981
- "epoch": 0.83,
2982
- "learning_rate": 0.000988,
2983
- "loss": 5.1614,
2984
- "step": 496
2985
- },
2986
- {
2987
- "epoch": 0.84,
2988
- "learning_rate": 0.00099,
2989
- "loss": 4.9319,
2990
- "step": 497
2991
- },
2992
- {
2993
- "epoch": 0.84,
2994
- "learning_rate": 0.000992,
2995
- "loss": 4.8599,
2996
- "step": 498
2997
- },
2998
- {
2999
- "epoch": 0.84,
3000
- "learning_rate": 0.000994,
3001
- "loss": 4.9016,
3002
- "step": 499
3003
- },
3004
- {
3005
- "epoch": 0.84,
3006
- "learning_rate": 0.000996,
3007
- "loss": 4.8633,
3008
- "step": 500
3009
- },
3010
- {
3011
- "epoch": 0.84,
3012
- "eval_loss": 5.217503070831299,
3013
- "eval_runtime": 761.5786,
3014
- "eval_samples_per_second": 3.469,
3015
- "eval_steps_per_second": 0.29,
3016
- "eval_wer": 2.0223324077746927,
3017
- "step": 500
3018
- },
3019
- {
3020
- "epoch": 0.84,
3021
- "learning_rate": 0.000998,
3022
- "loss": 5.0821,
3023
- "step": 501
3024
- },
3025
- {
3026
- "epoch": 0.84,
3027
- "learning_rate": 0.001,
3028
- "loss": 4.9581,
3029
- "step": 502
3030
- },
3031
- {
3032
- "epoch": 0.85,
3033
- "learning_rate": 0.0009893617021276596,
3034
- "loss": 4.9624,
3035
- "step": 503
3036
- },
3037
- {
3038
- "epoch": 0.85,
3039
- "learning_rate": 0.0009787234042553192,
3040
- "loss": 5.1349,
3041
- "step": 504
3042
- },
3043
- {
3044
- "epoch": 0.85,
3045
- "learning_rate": 0.0009680851063829787,
3046
- "loss": 5.0018,
3047
- "step": 505
3048
- },
3049
- {
3050
- "epoch": 0.85,
3051
- "learning_rate": 0.0009574468085106384,
3052
- "loss": 5.6814,
3053
- "step": 506
3054
- },
3055
- {
3056
- "epoch": 0.85,
3057
- "learning_rate": 0.0009468085106382979,
3058
- "loss": 4.8924,
3059
- "step": 507
3060
- },
3061
- {
3062
- "epoch": 0.85,
3063
- "learning_rate": 0.0009361702127659575,
3064
- "loss": 4.9127,
3065
- "step": 508
3066
- },
3067
- {
3068
- "epoch": 0.86,
3069
- "learning_rate": 0.000925531914893617,
3070
- "loss": 4.9308,
3071
- "step": 509
3072
- },
3073
- {
3074
- "epoch": 0.86,
3075
- "learning_rate": 0.0009148936170212766,
3076
- "loss": 5.075,
3077
- "step": 510
3078
- },
3079
- {
3080
- "epoch": 0.86,
3081
- "learning_rate": 0.0009042553191489362,
3082
- "loss": 5.0711,
3083
- "step": 511
3084
- },
3085
- {
3086
- "epoch": 0.86,
3087
- "learning_rate": 0.0008936170212765957,
3088
- "loss": 4.7922,
3089
- "step": 512
3090
- },
3091
- {
3092
- "epoch": 0.86,
3093
- "learning_rate": 0.0008829787234042553,
3094
- "loss": 4.8064,
3095
- "step": 513
3096
- },
3097
- {
3098
- "epoch": 0.86,
3099
- "learning_rate": 0.0008723404255319149,
3100
- "loss": 5.0308,
3101
- "step": 514
3102
- },
3103
- {
3104
- "epoch": 0.87,
3105
- "learning_rate": 0.0008617021276595744,
3106
- "loss": 4.9314,
3107
- "step": 515
3108
- },
3109
- {
3110
- "epoch": 0.87,
3111
- "learning_rate": 0.000851063829787234,
3112
- "loss": 4.8477,
3113
- "step": 516
3114
- },
3115
- {
3116
- "epoch": 0.87,
3117
- "learning_rate": 0.0008404255319148936,
3118
- "loss": 4.9286,
3119
- "step": 517
3120
- },
3121
- {
3122
- "epoch": 0.87,
3123
- "learning_rate": 0.0008297872340425531,
3124
- "loss": 5.4933,
3125
- "step": 518
3126
- },
3127
- {
3128
- "epoch": 0.87,
3129
- "learning_rate": 0.0008191489361702128,
3130
- "loss": 4.8794,
3131
- "step": 519
3132
- },
3133
- {
3134
- "epoch": 0.87,
3135
- "learning_rate": 0.0008085106382978723,
3136
- "loss": 4.8869,
3137
- "step": 520
3138
- },
3139
- {
3140
- "epoch": 0.88,
3141
- "learning_rate": 0.0007978723404255319,
3142
- "loss": 5.1175,
3143
- "step": 521
3144
- },
3145
- {
3146
- "epoch": 0.88,
3147
- "learning_rate": 0.0007872340425531915,
3148
- "loss": 4.9459,
3149
- "step": 522
3150
- },
3151
- {
3152
- "epoch": 0.88,
3153
- "learning_rate": 0.000776595744680851,
3154
- "loss": 4.7847,
3155
- "step": 523
3156
- },
3157
- {
3158
- "epoch": 0.88,
3159
- "learning_rate": 0.0007659574468085106,
3160
- "loss": 4.5425,
3161
- "step": 524
3162
- },
3163
- {
3164
- "epoch": 0.88,
3165
- "learning_rate": 0.0007553191489361702,
3166
- "loss": 4.8546,
3167
- "step": 525
3168
- },
3169
- {
3170
- "epoch": 0.88,
3171
- "learning_rate": 0.0007446808510638298,
3172
- "loss": 4.9989,
3173
- "step": 526
3174
- },
3175
- {
3176
- "epoch": 0.89,
3177
- "learning_rate": 0.0007340425531914894,
3178
- "loss": 5.0116,
3179
- "step": 527
3180
- },
3181
- {
3182
- "epoch": 0.89,
3183
- "learning_rate": 0.000723404255319149,
3184
- "loss": 4.7372,
3185
- "step": 528
3186
- },
3187
- {
3188
- "epoch": 0.89,
3189
- "learning_rate": 0.0007127659574468085,
3190
- "loss": 4.8166,
3191
- "step": 529
3192
- },
3193
- {
3194
- "epoch": 0.89,
3195
- "learning_rate": 0.0007021276595744682,
3196
- "loss": 4.6332,
3197
- "step": 530
3198
- },
3199
- {
3200
- "epoch": 0.89,
3201
- "learning_rate": 0.0006914893617021278,
3202
- "loss": 4.6867,
3203
- "step": 531
3204
- },
3205
- {
3206
- "epoch": 0.89,
3207
- "learning_rate": 0.0006808510638297873,
3208
- "loss": 4.987,
3209
- "step": 532
3210
- },
3211
- {
3212
- "epoch": 0.9,
3213
- "learning_rate": 0.0006702127659574469,
3214
- "loss": 4.8099,
3215
- "step": 533
3216
- },
3217
- {
3218
- "epoch": 0.9,
3219
- "learning_rate": 0.0006595744680851064,
3220
- "loss": 4.7489,
3221
- "step": 534
3222
- },
3223
- {
3224
- "epoch": 0.9,
3225
- "learning_rate": 0.000648936170212766,
3226
- "loss": 4.736,
3227
- "step": 535
3228
- },
3229
- {
3230
- "epoch": 0.9,
3231
- "learning_rate": 0.0006382978723404256,
3232
- "loss": 5.0293,
3233
- "step": 536
3234
- },
3235
- {
3236
- "epoch": 0.9,
3237
- "learning_rate": 0.0006276595744680851,
3238
- "loss": 4.9617,
3239
- "step": 537
3240
- },
3241
- {
3242
- "epoch": 0.9,
3243
- "learning_rate": 0.0006170212765957447,
3244
- "loss": 4.794,
3245
- "step": 538
3246
- },
3247
- {
3248
- "epoch": 0.91,
3249
- "learning_rate": 0.0006063829787234043,
3250
- "loss": 4.7831,
3251
- "step": 539
3252
- },
3253
- {
3254
- "epoch": 0.91,
3255
- "learning_rate": 0.0005957446808510638,
3256
- "loss": 4.718,
3257
- "step": 540
3258
- },
3259
- {
3260
- "epoch": 0.91,
3261
- "learning_rate": 0.0005851063829787234,
3262
- "loss": 4.6424,
3263
- "step": 541
3264
- },
3265
- {
3266
- "epoch": 0.91,
3267
- "learning_rate": 0.0005744680851063831,
3268
- "loss": 4.9381,
3269
- "step": 542
3270
- },
3271
- {
3272
- "epoch": 0.91,
3273
- "learning_rate": 0.0005638297872340426,
3274
- "loss": 4.7384,
3275
- "step": 543
3276
- },
3277
- {
3278
- "epoch": 0.91,
3279
- "learning_rate": 0.0005531914893617022,
3280
- "loss": 4.8799,
3281
- "step": 544
3282
- },
3283
- {
3284
- "epoch": 0.92,
3285
- "learning_rate": 0.0005425531914893617,
3286
- "loss": 4.8677,
3287
- "step": 545
3288
- },
3289
- {
3290
- "epoch": 0.92,
3291
- "learning_rate": 0.0005319148936170213,
3292
- "loss": 5.0294,
3293
- "step": 546
3294
- },
3295
- {
3296
- "epoch": 0.92,
3297
- "learning_rate": 0.0005212765957446809,
3298
- "loss": 4.6956,
3299
- "step": 547
3300
- },
3301
- {
3302
- "epoch": 0.92,
3303
- "learning_rate": 0.0005106382978723404,
3304
- "loss": 5.2215,
3305
- "step": 548
3306
  },
3307
  {
3308
  "epoch": 0.92,
3309
- "learning_rate": 0.0005,
3310
- "loss": 5.5748,
3311
- "step": 549
3312
  },
3313
  {
3314
  "epoch": 0.92,
3315
- "learning_rate": 0.0004893617021276596,
3316
- "loss": 4.9262,
3317
- "step": 550
3318
- },
3319
- {
3320
- "epoch": 0.93,
3321
- "learning_rate": 0.0004787234042553192,
3322
- "loss": 4.9315,
3323
- "step": 551
3324
- },
3325
- {
3326
- "epoch": 0.93,
3327
- "learning_rate": 0.00046808510638297874,
3328
- "loss": 5.0552,
3329
- "step": 552
3330
- },
3331
- {
3332
- "epoch": 0.93,
3333
- "learning_rate": 0.0004574468085106383,
3334
- "loss": 4.862,
3335
- "step": 553
3336
  },
3337
  {
3338
  "epoch": 0.93,
3339
- "learning_rate": 0.00044680851063829785,
3340
- "loss": 4.6946,
3341
- "step": 554
3342
  },
3343
  {
3344
  "epoch": 0.93,
3345
- "learning_rate": 0.00043617021276595746,
3346
- "loss": 4.7878,
3347
- "step": 555
3348
  },
3349
  {
3350
  "epoch": 0.93,
3351
- "learning_rate": 0.000425531914893617,
3352
- "loss": 4.8172,
3353
- "step": 556
3354
- },
3355
- {
3356
- "epoch": 0.94,
3357
- "learning_rate": 0.0004148936170212766,
3358
- "loss": 4.9713,
3359
- "step": 557
3360
- },
3361
- {
3362
- "epoch": 0.94,
3363
- "learning_rate": 0.00040425531914893613,
3364
- "loss": 4.7684,
3365
- "step": 558
3366
- },
3367
- {
3368
- "epoch": 0.94,
3369
- "learning_rate": 0.00039361702127659574,
3370
- "loss": 4.5917,
3371
- "step": 559
3372
  },
3373
  {
3374
  "epoch": 0.94,
3375
- "learning_rate": 0.0003829787234042553,
3376
- "loss": 4.5853,
3377
- "step": 560
3378
  },
3379
  {
3380
  "epoch": 0.94,
3381
- "learning_rate": 0.0003723404255319149,
3382
- "loss": 5.0219,
3383
- "step": 561
3384
  },
3385
  {
3386
  "epoch": 0.94,
3387
- "learning_rate": 0.0003617021276595745,
3388
- "loss": 4.6859,
3389
- "step": 562
3390
- },
3391
- {
3392
- "epoch": 0.95,
3393
- "learning_rate": 0.0003510638297872341,
3394
- "loss": 4.6475,
3395
- "step": 563
3396
- },
3397
- {
3398
- "epoch": 0.95,
3399
- "learning_rate": 0.00034042553191489364,
3400
- "loss": 5.0037,
3401
- "step": 564
3402
- },
3403
- {
3404
- "epoch": 0.95,
3405
- "learning_rate": 0.0003297872340425532,
3406
- "loss": 4.6778,
3407
- "step": 565
3408
  },
3409
  {
3410
  "epoch": 0.95,
3411
- "learning_rate": 0.0003191489361702128,
3412
- "loss": 5.0328,
3413
- "step": 566
3414
  },
3415
  {
3416
  "epoch": 0.95,
3417
- "learning_rate": 0.00030851063829787236,
3418
- "loss": 4.7141,
3419
- "step": 567
3420
- },
3421
- {
3422
- "epoch": 0.96,
3423
- "learning_rate": 0.0002978723404255319,
3424
- "loss": 4.7798,
3425
- "step": 568
3426
- },
3427
- {
3428
- "epoch": 0.96,
3429
- "learning_rate": 0.00028723404255319153,
3430
- "loss": 4.6515,
3431
- "step": 569
3432
- },
3433
- {
3434
- "epoch": 0.96,
3435
- "learning_rate": 0.0002765957446808511,
3436
- "loss": 4.7992,
3437
- "step": 570
3438
  },
3439
  {
3440
  "epoch": 0.96,
3441
- "learning_rate": 0.00026595744680851064,
3442
- "loss": 4.8235,
3443
- "step": 571
3444
  },
3445
  {
3446
  "epoch": 0.96,
3447
- "learning_rate": 0.0002553191489361702,
3448
- "loss": 4.7412,
3449
- "step": 572
3450
  },
3451
  {
3452
  "epoch": 0.96,
3453
- "learning_rate": 0.0002446808510638298,
3454
- "loss": 4.7087,
3455
- "step": 573
3456
- },
3457
- {
3458
- "epoch": 0.97,
3459
- "learning_rate": 0.00023404255319148937,
3460
- "loss": 4.5968,
3461
- "step": 574
3462
- },
3463
- {
3464
- "epoch": 0.97,
3465
- "learning_rate": 0.00022340425531914892,
3466
- "loss": 4.7226,
3467
- "step": 575
3468
- },
3469
- {
3470
- "epoch": 0.97,
3471
- "learning_rate": 0.0002127659574468085,
3472
- "loss": 4.7007,
3473
- "step": 576
3474
  },
3475
  {
3476
  "epoch": 0.97,
3477
- "learning_rate": 0.00020212765957446807,
3478
- "loss": 4.6527,
3479
- "step": 577
3480
  },
3481
  {
3482
  "epoch": 0.97,
3483
- "learning_rate": 0.00019148936170212765,
3484
- "loss": 4.4518,
3485
- "step": 578
3486
  },
3487
  {
3488
  "epoch": 0.97,
3489
- "learning_rate": 0.00018085106382978726,
3490
- "loss": 4.4938,
3491
- "step": 579
3492
- },
3493
- {
3494
- "epoch": 0.98,
3495
- "learning_rate": 0.00017021276595744682,
3496
- "loss": 4.6588,
3497
- "step": 580
3498
- },
3499
- {
3500
- "epoch": 0.98,
3501
- "learning_rate": 0.0001595744680851064,
3502
- "loss": 4.8716,
3503
- "step": 581
3504
- },
3505
- {
3506
- "epoch": 0.98,
3507
- "learning_rate": 0.00014893617021276596,
3508
- "loss": 4.6649,
3509
- "step": 582
3510
  },
3511
  {
3512
  "epoch": 0.98,
3513
- "learning_rate": 0.00013829787234042554,
3514
- "loss": 4.5137,
3515
- "step": 583
3516
  },
3517
  {
3518
  "epoch": 0.98,
3519
- "learning_rate": 0.0001276595744680851,
3520
- "loss": 4.6762,
3521
- "step": 584
3522
  },
3523
  {
3524
  "epoch": 0.98,
3525
- "learning_rate": 0.00011702127659574468,
3526
- "loss": 4.7224,
3527
- "step": 585
3528
- },
3529
- {
3530
- "epoch": 0.99,
3531
- "learning_rate": 0.00010638297872340425,
3532
- "loss": 4.5619,
3533
- "step": 586
3534
- },
3535
- {
3536
- "epoch": 0.99,
3537
- "learning_rate": 9.574468085106382e-05,
3538
- "loss": 4.6919,
3539
- "step": 587
3540
- },
3541
- {
3542
- "epoch": 0.99,
3543
- "learning_rate": 8.510638297872341e-05,
3544
- "loss": 4.7372,
3545
- "step": 588
3546
  },
3547
  {
3548
  "epoch": 0.99,
3549
- "learning_rate": 7.446808510638298e-05,
3550
- "loss": 4.749,
3551
- "step": 589
3552
  },
3553
  {
3554
  "epoch": 0.99,
3555
- "learning_rate": 6.382978723404255e-05,
3556
- "loss": 4.5827,
3557
- "step": 590
3558
  },
3559
  {
3560
  "epoch": 0.99,
3561
- "learning_rate": 5.319148936170213e-05,
3562
- "loss": 4.8585,
3563
- "step": 591
3564
- },
3565
- {
3566
- "epoch": 1.0,
3567
- "learning_rate": 4.2553191489361704e-05,
3568
- "loss": 6.0408,
3569
- "step": 592
3570
  },
3571
  {
3572
  "epoch": 1.0,
3573
- "learning_rate": 3.1914893617021275e-05,
3574
- "loss": 4.9389,
3575
- "step": 593
3576
  },
3577
  {
3578
  "epoch": 1.0,
3579
- "learning_rate": 2.1276595744680852e-05,
3580
- "loss": 4.8241,
3581
- "step": 594
3582
  },
3583
  {
3584
  "epoch": 1.0,
3585
- "step": 594,
3586
  "total_flos": 0.0,
3587
- "train_loss": 4.7614816479409985,
3588
- "train_runtime": 5563.7628,
3589
- "train_samples_per_second": 5.129,
3590
- "train_steps_per_second": 0.107
3591
  }
3592
  ],
3593
- "max_steps": 594,
3594
  "num_train_epochs": 1,
3595
  "total_flos": 0.0,
3596
  "trial_name": null,
 
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
  "epoch": 0.9987389659520807,
5
+ "global_step": 297,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
  "epoch": 0.0,
12
+ "learning_rate": 2e-06,
13
+ "loss": 4.7722,
14
  "step": 1
15
  },
16
  {
17
+ "epoch": 0.01,
18
+ "learning_rate": 4e-06,
19
+ "loss": 4.8471,
20
  "step": 2
21
  },
22
  {
23
  "epoch": 0.01,
24
+ "learning_rate": 6e-06,
25
+ "loss": 4.8901,
26
  "step": 3
27
  },
28
  {
29
  "epoch": 0.01,
30
+ "learning_rate": 8e-06,
31
+ "loss": 4.7936,
32
  "step": 4
33
  },
34
  {
35
+ "epoch": 0.02,
36
+ "learning_rate": 1e-05,
37
+ "loss": 4.6226,
38
  "step": 5
39
  },
40
  {
41
+ "epoch": 0.02,
42
  "learning_rate": 1e-05,
43
+ "loss": 4.6818,
44
  "step": 6
45
  },
46
  {
47
+ "epoch": 0.02,
48
  "learning_rate": 1.2e-05,
49
+ "loss": 4.6542,
50
  "step": 7
51
  },
52
  {
53
+ "epoch": 0.03,
54
  "learning_rate": 1.4e-05,
55
+ "loss": 4.6833,
56
  "step": 8
57
  },
58
  {
59
+ "epoch": 0.03,
60
  "learning_rate": 1.6e-05,
61
+ "loss": 4.5591,
62
  "step": 9
63
  },
64
  {
65
+ "epoch": 0.03,
66
  "learning_rate": 1.8e-05,
67
+ "loss": 4.3732,
68
  "step": 10
69
  },
70
  {
71
+ "epoch": 0.04,
72
  "learning_rate": 2e-05,
73
+ "loss": 4.4692,
74
  "step": 11
75
  },
76
  {
77
+ "epoch": 0.04,
78
  "learning_rate": 2.2e-05,
79
+ "loss": 4.4868,
80
  "step": 12
81
  },
82
  {
83
+ "epoch": 0.04,
84
  "learning_rate": 2.4e-05,
85
+ "loss": 4.3669,
86
  "step": 13
87
  },
88
  {
89
+ "epoch": 0.05,
90
  "learning_rate": 2.6e-05,
91
+ "loss": 4.3111,
92
  "step": 14
93
  },
94
  {
95
+ "epoch": 0.05,
96
  "learning_rate": 2.8e-05,
97
+ "loss": 4.3386,
98
  "step": 15
99
  },
100
  {
101
+ "epoch": 0.05,
102
  "learning_rate": 3e-05,
103
+ "loss": 4.2529,
104
  "step": 16
105
  },
106
  {
107
+ "epoch": 0.06,
108
  "learning_rate": 3.2e-05,
109
+ "loss": 4.3119,
110
  "step": 17
111
  },
112
  {
113
+ "epoch": 0.06,
114
  "learning_rate": 3.4000000000000007e-05,
115
+ "loss": 4.345,
116
  "step": 18
117
  },
118
  {
119
+ "epoch": 0.06,
120
  "learning_rate": 3.6e-05,
121
+ "loss": 4.3061,
122
  "step": 19
123
  },
124
  {
125
+ "epoch": 0.07,
126
  "learning_rate": 3.8e-05,
127
+ "loss": 4.3575,
128
  "step": 20
129
  },
130
  {
131
+ "epoch": 0.07,
132
  "learning_rate": 4e-05,
133
+ "loss": 4.3639,
134
  "step": 21
135
  },
136
  {
137
+ "epoch": 0.07,
138
  "learning_rate": 4.2000000000000004e-05,
139
+ "loss": 4.332,
140
  "step": 22
141
  },
142
  {
143
+ "epoch": 0.08,
144
  "learning_rate": 4.4e-05,
145
+ "loss": 4.2139,
146
  "step": 23
147
  },
148
  {
149
+ "epoch": 0.08,
150
  "learning_rate": 4.6e-05,
151
+ "loss": 4.342,
152
  "step": 24
153
  },
154
  {
155
+ "epoch": 0.08,
156
  "learning_rate": 4.8e-05,
157
+ "loss": 4.1966,
158
  "step": 25
159
  },
160
  {
161
+ "epoch": 0.09,
162
  "learning_rate": 5e-05,
163
+ "loss": 4.4635,
164
  "step": 26
165
  },
166
  {
167
+ "epoch": 0.09,
168
  "learning_rate": 5.2e-05,
169
+ "loss": 4.2323,
170
  "step": 27
171
  },
172
  {
173
+ "epoch": 0.09,
174
  "learning_rate": 5.4e-05,
175
+ "loss": 4.2192,
176
  "step": 28
177
  },
178
  {
179
+ "epoch": 0.1,
180
  "learning_rate": 5.6e-05,
181
+ "loss": 4.124,
182
  "step": 29
183
  },
184
  {
185
+ "epoch": 0.1,
186
  "learning_rate": 5.800000000000001e-05,
187
+ "loss": 4.3273,
188
  "step": 30
189
  },
190
  {
191
+ "epoch": 0.1,
192
  "learning_rate": 6e-05,
193
+ "loss": 4.3173,
194
  "step": 31
195
  },
196
  {
197
+ "epoch": 0.11,
198
  "learning_rate": 6.2e-05,
199
+ "loss": 4.2152,
200
  "step": 32
201
  },
202
  {
203
+ "epoch": 0.11,
204
  "learning_rate": 6.4e-05,
205
+ "loss": 4.2741,
206
  "step": 33
207
  },
208
  {
209
+ "epoch": 0.11,
210
  "learning_rate": 6.6e-05,
211
+ "loss": 4.2259,
212
  "step": 34
213
  },
214
  {
215
+ "epoch": 0.12,
216
  "learning_rate": 6.800000000000001e-05,
217
+ "loss": 4.2298,
218
  "step": 35
219
  },
220
  {
221
+ "epoch": 0.12,
222
  "learning_rate": 7.000000000000001e-05,
223
+ "loss": 4.2443,
224
  "step": 36
225
  },
226
  {
227
+ "epoch": 0.12,
228
  "learning_rate": 7.2e-05,
229
+ "loss": 4.3289,
230
  "step": 37
231
  },
232
  {
233
+ "epoch": 0.13,
234
  "learning_rate": 7.4e-05,
235
+ "loss": 4.2854,
236
  "step": 38
237
  },
238
  {
239
+ "epoch": 0.13,
240
  "learning_rate": 7.6e-05,
241
+ "loss": 4.3371,
242
  "step": 39
243
  },
244
  {
245
+ "epoch": 0.13,
246
  "learning_rate": 7.8e-05,
247
+ "loss": 4.2088,
248
  "step": 40
249
  },
250
  {
251
+ "epoch": 0.14,
252
  "learning_rate": 8e-05,
253
+ "loss": 4.2763,
254
  "step": 41
255
  },
256
  {
257
+ "epoch": 0.14,
258
  "learning_rate": 8.2e-05,
259
+ "loss": 4.3167,
260
  "step": 42
261
  },
262
  {
263
+ "epoch": 0.14,
264
  "learning_rate": 8.400000000000001e-05,
265
+ "loss": 4.1436,
266
  "step": 43
267
  },
268
  {
269
+ "epoch": 0.15,
270
  "learning_rate": 8.599999999999999e-05,
271
+ "loss": 4.4299,
272
  "step": 44
273
  },
274
  {
275
+ "epoch": 0.15,
276
  "learning_rate": 8.8e-05,
277
+ "loss": 4.3716,
278
  "step": 45
279
  },
280
  {
281
+ "epoch": 0.15,
282
  "learning_rate": 8.999999999999999e-05,
283
+ "loss": 4.4052,
284
  "step": 46
285
  },
286
  {
287
+ "epoch": 0.16,
288
  "learning_rate": 9.2e-05,
289
+ "loss": 4.3802,
290
  "step": 47
291
  },
292
  {
293
+ "epoch": 0.16,
294
  "learning_rate": 9.400000000000001e-05,
295
+ "loss": 4.4711,
296
  "step": 48
297
  },
298
  {
299
+ "epoch": 0.16,
300
  "learning_rate": 9.6e-05,
301
+ "loss": 4.5475,
302
  "step": 49
303
  },
304
  {
305
+ "epoch": 0.17,
306
+ "learning_rate": 9.800000000000001e-05,
307
+ "loss": 4.3163,
308
  "step": 50
309
  },
310
  {
311
+ "epoch": 0.17,
312
+ "learning_rate": 0.0001,
313
+ "loss": 4.2876,
314
  "step": 51
315
  },
316
  {
317
+ "epoch": 0.17,
318
+ "learning_rate": 0.000102,
319
+ "loss": 4.1673,
320
  "step": 52
321
  },
322
  {
323
+ "epoch": 0.18,
324
+ "learning_rate": 0.000104,
325
+ "loss": 4.1641,
326
  "step": 53
327
  },
328
  {
329
+ "epoch": 0.18,
330
+ "learning_rate": 0.000106,
331
+ "loss": 4.1324,
332
  "step": 54
333
  },
334
  {
335
+ "epoch": 0.18,
336
+ "learning_rate": 0.000108,
337
+ "loss": 4.1149,
338
  "step": 55
339
  },
340
  {
341
+ "epoch": 0.19,
342
+ "learning_rate": 0.00011,
343
+ "loss": 4.1346,
344
  "step": 56
345
  },
346
  {
347
+ "epoch": 0.19,
348
+ "learning_rate": 0.000112,
349
+ "loss": 4.2194,
350
  "step": 57
351
  },
352
  {
353
+ "epoch": 0.2,
354
+ "learning_rate": 0.000114,
355
+ "loss": 4.2206,
356
  "step": 58
357
  },
358
  {
359
+ "epoch": 0.2,
360
+ "learning_rate": 0.00011600000000000001,
361
+ "loss": 4.2098,
362
  "step": 59
363
  },
364
  {
365
+ "epoch": 0.2,
366
+ "learning_rate": 0.000118,
367
+ "loss": 4.1653,
368
  "step": 60
369
  },
370
  {
371
+ "epoch": 0.21,
372
+ "learning_rate": 0.00012,
373
+ "loss": 4.2034,
374
  "step": 61
375
  },
376
  {
377
+ "epoch": 0.21,
378
+ "learning_rate": 0.000122,
379
+ "loss": 4.1436,
380
  "step": 62
381
  },
382
  {
383
+ "epoch": 0.21,
384
+ "learning_rate": 0.000124,
385
+ "loss": 4.2173,
386
  "step": 63
387
  },
388
  {
389
+ "epoch": 0.22,
390
+ "learning_rate": 0.000126,
391
+ "loss": 4.1701,
392
  "step": 64
393
  },
394
  {
395
+ "epoch": 0.22,
396
+ "learning_rate": 0.000128,
397
+ "loss": 4.2817,
398
  "step": 65
399
  },
400
  {
401
+ "epoch": 0.22,
402
+ "learning_rate": 0.00013000000000000002,
403
+ "loss": 4.2719,
404
  "step": 66
405
  },
406
  {
407
+ "epoch": 0.23,
408
+ "learning_rate": 0.000132,
409
+ "loss": 4.2492,
410
  "step": 67
411
  },
412
  {
413
+ "epoch": 0.23,
414
+ "learning_rate": 0.000134,
415
+ "loss": 4.2513,
416
  "step": 68
417
  },
418
  {
419
+ "epoch": 0.23,
420
+ "learning_rate": 0.00013600000000000003,
421
+ "loss": 4.1976,
422
  "step": 69
423
  },
424
  {
425
+ "epoch": 0.24,
426
+ "learning_rate": 0.00013800000000000002,
427
+ "loss": 4.2182,
428
  "step": 70
429
  },
430
  {
431
+ "epoch": 0.24,
432
+ "learning_rate": 0.00014000000000000001,
433
+ "loss": 4.1483,
434
  "step": 71
435
  },
436
  {
437
+ "epoch": 0.24,
438
+ "learning_rate": 0.00014199999999999998,
439
+ "loss": 4.2812,
440
  "step": 72
441
  },
442
  {
443
+ "epoch": 0.25,
444
+ "learning_rate": 0.000144,
445
+ "loss": 4.224,
446
  "step": 73
447
  },
448
  {
449
+ "epoch": 0.25,
450
+ "learning_rate": 0.000146,
451
+ "loss": 4.2017,
452
  "step": 74
453
  },
454
  {
455
+ "epoch": 0.25,
456
+ "learning_rate": 0.000148,
457
+ "loss": 4.2882,
458
  "step": 75
459
  },
460
  {
461
+ "epoch": 0.26,
462
+ "learning_rate": 0.00015,
463
+ "loss": 4.204,
464
  "step": 76
465
  },
466
  {
467
+ "epoch": 0.26,
468
+ "learning_rate": 0.000152,
469
+ "loss": 4.1371,
470
  "step": 77
471
  },
472
  {
473
+ "epoch": 0.26,
474
+ "learning_rate": 0.000154,
475
+ "loss": 4.2038,
476
  "step": 78
477
  },
478
  {
479
+ "epoch": 0.27,
480
+ "learning_rate": 0.000156,
481
+ "loss": 4.0844,
482
  "step": 79
483
  },
484
  {
485
+ "epoch": 0.27,
486
+ "learning_rate": 0.000158,
487
+ "loss": 4.3707,
488
  "step": 80
489
  },
490
  {
491
+ "epoch": 0.27,
492
+ "learning_rate": 0.00016,
493
+ "loss": 4.2139,
494
  "step": 81
495
  },
496
  {
497
+ "epoch": 0.28,
498
+ "learning_rate": 0.000162,
499
+ "loss": 4.2325,
500
  "step": 82
501
  },
502
  {
503
+ "epoch": 0.28,
504
+ "learning_rate": 0.000164,
505
+ "loss": 4.1949,
506
  "step": 83
507
  },
508
  {
509
+ "epoch": 0.28,
510
+ "learning_rate": 0.00016600000000000002,
511
+ "loss": 4.2897,
512
  "step": 84
513
  },
514
  {
515
+ "epoch": 0.29,
 
 
 
 
 
 
516
  "learning_rate": 0.00016800000000000002,
517
+ "loss": 4.3493,
518
+ "step": 85
519
  },
520
  {
521
+ "epoch": 0.29,
522
  "learning_rate": 0.00017,
523
+ "loss": 4.2138,
524
+ "step": 86
525
  },
526
  {
527
+ "epoch": 0.29,
528
  "learning_rate": 0.00017199999999999998,
529
+ "loss": 4.2173,
530
+ "step": 87
531
  },
532
  {
533
+ "epoch": 0.3,
534
  "learning_rate": 0.000174,
535
+ "loss": 4.3482,
536
+ "step": 88
537
  },
538
  {
539
+ "epoch": 0.3,
540
  "learning_rate": 0.000176,
541
+ "loss": 4.2918,
542
+ "step": 89
543
  },
544
  {
545
+ "epoch": 0.3,
546
  "learning_rate": 0.000178,
547
+ "loss": 4.318,
548
+ "step": 90
549
  },
550
  {
551
+ "epoch": 0.31,
552
  "learning_rate": 0.00017999999999999998,
553
+ "loss": 4.3096,
554
+ "step": 91
555
  },
556
  {
557
+ "epoch": 0.31,
558
  "learning_rate": 0.000182,
559
+ "loss": 4.4027,
560
+ "step": 92
561
  },
562
  {
563
+ "epoch": 0.31,
564
  "learning_rate": 0.000184,
565
+ "loss": 4.5431,
566
+ "step": 93
567
  },
568
  {
569
+ "epoch": 0.32,
570
  "learning_rate": 0.000186,
571
+ "loss": 4.4248,
572
+ "step": 94
573
  },
574
  {
575
+ "epoch": 0.32,
576
  "learning_rate": 0.00018800000000000002,
577
+ "loss": 4.341,
578
+ "step": 95
579
  },
580
  {
581
+ "epoch": 0.32,
582
  "learning_rate": 0.00019,
583
+ "loss": 4.4704,
584
+ "step": 96
585
  },
586
  {
587
+ "epoch": 0.33,
588
  "learning_rate": 0.000192,
589
+ "loss": 4.6444,
590
+ "step": 97
591
  },
592
  {
593
+ "epoch": 0.33,
594
  "learning_rate": 0.000194,
595
+ "loss": 4.6959,
596
+ "step": 98
597
  },
598
  {
599
+ "epoch": 0.33,
600
  "learning_rate": 0.00019600000000000002,
601
+ "loss": 4.5906,
602
+ "step": 99
603
  },
604
  {
605
+ "epoch": 0.34,
606
  "learning_rate": 0.00019800000000000002,
607
+ "loss": 4.5786,
608
+ "step": 100
609
  },
610
  {
611
+ "epoch": 0.34,
612
  "learning_rate": 0.0002,
613
+ "loss": 4.2771,
614
+ "step": 101
615
  },
616
  {
617
+ "epoch": 0.34,
618
  "learning_rate": 0.000202,
619
+ "loss": 4.3638,
620
+ "step": 102
621
  },
622
  {
623
+ "epoch": 0.35,
624
  "learning_rate": 0.000204,
625
+ "loss": 4.3463,
626
+ "step": 103
627
  },
628
  {
629
+ "epoch": 0.35,
630
  "learning_rate": 0.000206,
631
+ "loss": 4.2307,
632
+ "step": 104
633
  },
634
  {
635
+ "epoch": 0.35,
636
  "learning_rate": 0.000208,
637
+ "loss": 4.3191,
638
+ "step": 105
639
  },
640
  {
641
+ "epoch": 0.36,
642
  "learning_rate": 0.00021,
643
+ "loss": 4.2077,
644
+ "step": 106
645
  },
646
  {
647
+ "epoch": 0.36,
648
  "learning_rate": 0.000212,
649
+ "loss": 4.171,
650
+ "step": 107
651
  },
652
  {
653
+ "epoch": 0.36,
654
  "learning_rate": 0.000214,
655
+ "loss": 4.28,
656
+ "step": 108
657
  },
658
  {
659
+ "epoch": 0.37,
660
  "learning_rate": 0.000216,
661
+ "loss": 4.3413,
662
+ "step": 109
663
  },
664
  {
665
+ "epoch": 0.37,
666
  "learning_rate": 0.000218,
667
+ "loss": 4.3297,
668
+ "step": 110
669
  },
670
  {
671
+ "epoch": 0.37,
672
  "learning_rate": 0.00022,
673
+ "loss": 4.3912,
674
+ "step": 111
675
  },
676
  {
677
+ "epoch": 0.38,
678
  "learning_rate": 0.000222,
679
+ "loss": 4.2747,
680
+ "step": 112
681
  },
682
  {
683
+ "epoch": 0.38,
684
  "learning_rate": 0.000224,
685
+ "loss": 4.2566,
686
+ "step": 113
687
  },
688
  {
689
+ "epoch": 0.38,
690
  "learning_rate": 0.00022600000000000002,
691
+ "loss": 4.2711,
692
+ "step": 114
693
  },
694
  {
695
+ "epoch": 0.39,
696
  "learning_rate": 0.000228,
697
+ "loss": 4.2383,
698
+ "step": 115
699
  },
700
  {
701
+ "epoch": 0.39,
702
  "learning_rate": 0.00023,
703
+ "loss": 4.2432,
704
+ "step": 116
705
  },
706
  {
707
+ "epoch": 0.39,
708
  "learning_rate": 0.00023200000000000003,
709
+ "loss": 4.1736,
710
+ "step": 117
711
  },
712
  {
713
+ "epoch": 0.4,
714
  "learning_rate": 0.00023400000000000002,
715
+ "loss": 4.1851,
716
+ "step": 118
717
  },
718
  {
719
+ "epoch": 0.4,
720
  "learning_rate": 0.000236,
721
+ "loss": 4.2587,
722
+ "step": 119
723
  },
724
  {
725
+ "epoch": 0.4,
726
  "learning_rate": 0.00023799999999999998,
727
+ "loss": 4.2478,
728
+ "step": 120
729
  },
730
  {
731
+ "epoch": 0.41,
732
  "learning_rate": 0.00024,
733
+ "loss": 4.2483,
734
+ "step": 121
735
  },
736
  {
737
+ "epoch": 0.41,
738
  "learning_rate": 0.000242,
739
+ "loss": 4.3501,
740
+ "step": 122
741
  },
742
  {
743
+ "epoch": 0.41,
744
  "learning_rate": 0.000244,
745
+ "loss": 4.3387,
746
+ "step": 123
747
  },
748
  {
749
+ "epoch": 0.42,
750
  "learning_rate": 0.000246,
751
+ "loss": 4.4114,
752
+ "step": 124
753
  },
754
  {
755
+ "epoch": 0.42,
756
  "learning_rate": 0.000248,
757
+ "loss": 4.3815,
758
+ "step": 125
759
  },
760
  {
761
+ "epoch": 0.42,
762
  "learning_rate": 0.00025,
763
+ "loss": 4.318,
764
+ "step": 126
765
  },
766
  {
767
+ "epoch": 0.43,
768
  "learning_rate": 0.000252,
769
+ "loss": 4.336,
770
+ "step": 127
771
  },
772
  {
773
+ "epoch": 0.43,
774
  "learning_rate": 0.000254,
775
+ "loss": 4.2494,
776
+ "step": 128
777
  },
778
  {
779
+ "epoch": 0.43,
780
  "learning_rate": 0.000256,
781
+ "loss": 4.3796,
782
+ "step": 129
783
  },
784
  {
785
+ "epoch": 0.44,
786
  "learning_rate": 0.00025800000000000004,
787
+ "loss": 4.4553,
788
+ "step": 130
789
  },
790
  {
791
+ "epoch": 0.44,
792
  "learning_rate": 0.00026000000000000003,
793
+ "loss": 4.3515,
794
+ "step": 131
795
  },
796
  {
797
+ "epoch": 0.44,
798
  "learning_rate": 0.000262,
799
+ "loss": 4.1923,
800
+ "step": 132
801
  },
802
  {
803
+ "epoch": 0.45,
804
  "learning_rate": 0.000264,
805
+ "loss": 4.5483,
806
+ "step": 133
807
  },
808
  {
809
+ "epoch": 0.45,
810
  "learning_rate": 0.000266,
811
+ "loss": 4.3097,
812
+ "step": 134
813
  },
814
  {
815
+ "epoch": 0.45,
816
  "learning_rate": 0.000268,
817
+ "loss": 4.411,
818
+ "step": 135
819
  },
820
  {
821
+ "epoch": 0.46,
822
  "learning_rate": 0.00027,
823
+ "loss": 4.3845,
824
+ "step": 136
825
  },
826
  {
827
+ "epoch": 0.46,
828
  "learning_rate": 0.00027200000000000005,
829
+ "loss": 4.261,
830
+ "step": 137
831
  },
832
  {
833
+ "epoch": 0.46,
834
  "learning_rate": 0.00027400000000000005,
835
+ "loss": 4.2935,
836
+ "step": 138
837
  },
838
  {
839
+ "epoch": 0.47,
840
  "learning_rate": 0.00027600000000000004,
841
+ "loss": 4.3502,
842
+ "step": 139
843
  },
844
  {
845
+ "epoch": 0.47,
846
  "learning_rate": 0.00027800000000000004,
847
+ "loss": 4.2937,
848
+ "step": 140
849
  },
850
  {
851
+ "epoch": 0.47,
852
  "learning_rate": 0.00028000000000000003,
853
+ "loss": 4.5061,
854
+ "step": 141
855
  },
856
  {
857
+ "epoch": 0.48,
858
  "learning_rate": 0.00028199999999999997,
859
+ "loss": 4.5488,
860
+ "step": 142
861
  },
862
  {
863
+ "epoch": 0.48,
864
  "learning_rate": 0.00028399999999999996,
865
+ "loss": 4.3519,
866
+ "step": 143
867
  },
868
  {
869
+ "epoch": 0.48,
870
  "learning_rate": 0.00028599999999999996,
871
+ "loss": 4.4849,
872
+ "step": 144
873
  },
874
  {
875
+ "epoch": 0.49,
876
  "learning_rate": 0.000288,
877
+ "loss": 4.6754,
878
+ "step": 145
879
  },
880
  {
881
+ "epoch": 0.49,
882
  "learning_rate": 0.00029,
883
+ "loss": 4.3389,
884
+ "step": 146
885
  },
886
  {
887
+ "epoch": 0.49,
888
  "learning_rate": 0.000292,
889
+ "loss": 5.3417,
890
+ "step": 147
891
  },
892
  {
893
+ "epoch": 0.5,
894
  "learning_rate": 0.000294,
895
+ "loss": 4.739,
896
+ "step": 148
897
  },
898
  {
899
+ "epoch": 0.5,
900
  "learning_rate": 0.000296,
901
+ "loss": 4.8816,
902
+ "step": 149
903
  },
904
  {
905
+ "epoch": 0.5,
906
  "learning_rate": 0.000298,
907
+ "loss": 5.0976,
908
+ "step": 150
909
  },
910
  {
911
+ "epoch": 0.51,
912
  "learning_rate": 0.0003,
913
+ "loss": 4.6187,
914
+ "step": 151
915
  },
916
  {
917
+ "epoch": 0.51,
918
  "learning_rate": 0.000302,
919
+ "loss": 4.5796,
920
+ "step": 152
921
  },
922
  {
923
+ "epoch": 0.51,
924
  "learning_rate": 0.000304,
925
+ "loss": 4.4965,
926
+ "step": 153
927
  },
928
  {
929
+ "epoch": 0.52,
930
  "learning_rate": 0.000306,
931
+ "loss": 4.4367,
932
+ "step": 154
933
  },
934
  {
935
+ "epoch": 0.52,
936
  "learning_rate": 0.000308,
937
+ "loss": 4.3843,
938
+ "step": 155
939
  },
940
  {
941
+ "epoch": 0.52,
942
  "learning_rate": 0.00031,
943
+ "loss": 4.4425,
944
+ "step": 156
945
  },
946
  {
947
+ "epoch": 0.53,
948
  "learning_rate": 0.000312,
949
+ "loss": 4.3224,
950
+ "step": 157
951
  },
952
  {
953
+ "epoch": 0.53,
954
  "learning_rate": 0.000314,
955
+ "loss": 4.3704,
956
+ "step": 158
957
  },
958
  {
959
+ "epoch": 0.53,
960
  "learning_rate": 0.000316,
961
+ "loss": 4.3799,
962
+ "step": 159
963
  },
964
  {
965
+ "epoch": 0.54,
966
  "learning_rate": 0.00031800000000000003,
967
+ "loss": 4.4944,
968
+ "step": 160
969
  },
970
  {
971
+ "epoch": 0.54,
972
  "learning_rate": 0.00032,
973
+ "loss": 4.4363,
974
+ "step": 161
975
  },
976
  {
977
+ "epoch": 0.54,
978
  "learning_rate": 0.000322,
979
+ "loss": 4.2812,
980
+ "step": 162
981
  },
982
  {
983
+ "epoch": 0.55,
984
  "learning_rate": 0.000324,
985
+ "loss": 4.4024,
986
+ "step": 163
987
  },
988
  {
989
+ "epoch": 0.55,
990
  "learning_rate": 0.000326,
991
+ "loss": 4.2637,
992
+ "step": 164
993
  },
994
  {
995
+ "epoch": 0.55,
996
  "learning_rate": 0.000328,
997
+ "loss": 4.3806,
998
+ "step": 165
999
  },
1000
  {
1001
+ "epoch": 0.56,
1002
  "learning_rate": 0.00033,
1003
+ "loss": 4.3163,
1004
+ "step": 166
1005
  },
1006
  {
1007
+ "epoch": 0.56,
1008
  "learning_rate": 0.00033200000000000005,
1009
+ "loss": 4.2933,
1010
+ "step": 167
1011
  },
1012
  {
1013
+ "epoch": 0.56,
1014
  "learning_rate": 0.00033400000000000004,
1015
+ "loss": 4.4188,
1016
+ "step": 168
1017
  },
1018
  {
1019
+ "epoch": 0.57,
1020
  "learning_rate": 0.00033600000000000004,
1021
+ "loss": 4.4134,
1022
+ "step": 169
1023
  },
1024
  {
1025
+ "epoch": 0.57,
1026
  "learning_rate": 0.00033800000000000003,
1027
+ "loss": 4.4127,
1028
+ "step": 170
1029
  },
1030
  {
1031
+ "epoch": 0.58,
1032
  "learning_rate": 0.00034,
1033
+ "loss": 4.2583,
1034
+ "step": 171
1035
  },
1036
  {
1037
+ "epoch": 0.58,
1038
  "learning_rate": 0.000342,
1039
+ "loss": 4.3188,
1040
+ "step": 172
1041
  },
1042
  {
1043
+ "epoch": 0.58,
1044
  "learning_rate": 0.00034399999999999996,
1045
+ "loss": 4.3439,
1046
+ "step": 173
1047
  },
1048
  {
1049
+ "epoch": 0.59,
1050
  "learning_rate": 0.000346,
1051
+ "loss": 4.4033,
1052
+ "step": 174
1053
  },
1054
  {
1055
+ "epoch": 0.59,
1056
  "learning_rate": 0.000348,
1057
+ "loss": 4.5878,
1058
+ "step": 175
1059
  },
1060
  {
1061
+ "epoch": 0.59,
1062
  "learning_rate": 0.00035,
1063
+ "loss": 4.4151,
1064
+ "step": 176
1065
  },
1066
  {
1067
+ "epoch": 0.6,
1068
  "learning_rate": 0.000352,
1069
+ "loss": 4.4215,
1070
+ "step": 177
1071
  },
1072
  {
1073
+ "epoch": 0.6,
1074
  "learning_rate": 0.000354,
1075
+ "loss": 4.305,
1076
+ "step": 178
1077
  },
1078
  {
1079
+ "epoch": 0.6,
1080
  "learning_rate": 0.000356,
1081
+ "loss": 4.2926,
1082
+ "step": 179
1083
  },
1084
  {
1085
+ "epoch": 0.61,
1086
  "learning_rate": 0.000358,
1087
+ "loss": 4.4736,
1088
+ "step": 180
1089
  },
1090
  {
1091
+ "epoch": 0.61,
1092
  "learning_rate": 0.00035999999999999997,
1093
+ "loss": 4.4711,
1094
+ "step": 181
1095
  },
1096
  {
1097
+ "epoch": 0.61,
1098
  "learning_rate": 0.000362,
1099
+ "loss": 4.4852,
1100
+ "step": 182
1101
  },
1102
  {
1103
+ "epoch": 0.62,
1104
  "learning_rate": 0.000364,
1105
+ "loss": 4.4917,
1106
+ "step": 183
1107
  },
1108
  {
1109
+ "epoch": 0.62,
1110
  "learning_rate": 0.000366,
1111
+ "loss": 4.4806,
1112
+ "step": 184
1113
  },
1114
  {
1115
+ "epoch": 0.62,
1116
  "learning_rate": 0.000368,
1117
+ "loss": 4.4324,
1118
+ "step": 185
1119
  },
1120
  {
1121
+ "epoch": 0.63,
1122
  "learning_rate": 0.00037,
1123
+ "loss": 4.3491,
1124
+ "step": 186
1125
  },
1126
  {
1127
+ "epoch": 0.63,
1128
  "learning_rate": 0.000372,
1129
+ "loss": 4.4705,
1130
+ "step": 187
1131
  },
1132
  {
1133
+ "epoch": 0.63,
1134
  "learning_rate": 0.000374,
1135
+ "loss": 4.3801,
1136
+ "step": 188
1137
  },
1138
  {
1139
+ "epoch": 0.64,
1140
  "learning_rate": 0.00037600000000000003,
1141
+ "loss": 4.3337,
1142
+ "step": 189
1143
  },
1144
  {
1145
+ "epoch": 0.64,
1146
  "learning_rate": 0.000378,
1147
+ "loss": 4.4856,
1148
+ "step": 190
1149
  },
1150
  {
1151
+ "epoch": 0.64,
1152
  "learning_rate": 0.00038,
1153
+ "loss": 4.5416,
1154
+ "step": 191
1155
  },
1156
  {
1157
+ "epoch": 0.65,
1158
  "learning_rate": 0.000382,
1159
+ "loss": 4.6047,
1160
+ "step": 192
1161
  },
1162
  {
1163
+ "epoch": 0.65,
1164
  "learning_rate": 0.000384,
1165
+ "loss": 4.5076,
1166
+ "step": 193
1167
  },
1168
  {
1169
+ "epoch": 0.65,
1170
  "learning_rate": 0.000386,
1171
+ "loss": 4.6097,
1172
+ "step": 194
1173
  },
1174
  {
1175
+ "epoch": 0.66,
1176
  "learning_rate": 0.000388,
1177
+ "loss": 4.4656,
1178
+ "step": 195
1179
  },
1180
  {
1181
+ "epoch": 0.66,
1182
  "learning_rate": 0.00039000000000000005,
1183
+ "loss": 4.5988,
1184
+ "step": 196
1185
  },
1186
  {
1187
+ "epoch": 0.66,
1188
  "learning_rate": 0.00039200000000000004,
1189
+ "loss": 4.4527,
1190
+ "step": 197
1191
  },
1192
  {
1193
+ "epoch": 0.67,
1194
  "learning_rate": 0.00039400000000000004,
1195
+ "loss": 5.1282,
1196
+ "step": 198
1197
  },
1198
  {
1199
+ "epoch": 0.67,
1200
  "learning_rate": 0.00039600000000000003,
1201
+ "loss": 4.5988,
1202
+ "step": 199
1203
  },
1204
  {
1205
+ "epoch": 0.67,
1206
  "learning_rate": 0.000398,
1207
+ "loss": 5.5887,
1208
+ "step": 200
1209
  },
1210
  {
1211
+ "epoch": 0.68,
1212
  "learning_rate": 0.0004,
1213
+ "loss": 6.4326,
1214
+ "step": 201
1215
  },
1216
  {
1217
+ "epoch": 0.68,
1218
  "learning_rate": 0.000402,
1219
+ "loss": 4.8245,
1220
+ "step": 202
1221
  },
1222
  {
1223
+ "epoch": 0.68,
1224
  "learning_rate": 0.000404,
1225
+ "loss": 4.625,
1226
+ "step": 203
1227
  },
1228
  {
1229
+ "epoch": 0.69,
1230
  "learning_rate": 0.00040600000000000006,
1231
+ "loss": 4.7209,
1232
+ "step": 204
1233
  },
1234
  {
1235
+ "epoch": 0.69,
1236
  "learning_rate": 0.000408,
1237
+ "loss": 4.6639,
1238
+ "step": 205
1239
  },
1240
  {
1241
+ "epoch": 0.69,
1242
  "learning_rate": 0.00041,
1243
+ "loss": 4.709,
1244
+ "step": 206
1245
  },
1246
  {
1247
+ "epoch": 0.7,
1248
  "learning_rate": 0.000412,
1249
+ "loss": 5.1396,
1250
+ "step": 207
1251
  },
1252
  {
1253
+ "epoch": 0.7,
1254
  "learning_rate": 0.000414,
1255
+ "loss": 4.5794,
1256
+ "step": 208
1257
  },
1258
  {
1259
+ "epoch": 0.7,
1260
  "learning_rate": 0.000416,
1261
+ "loss": 4.7483,
1262
+ "step": 209
1263
  },
1264
  {
1265
+ "epoch": 0.71,
1266
  "learning_rate": 0.00041799999999999997,
1267
+ "loss": 4.6424,
1268
+ "step": 210
1269
  },
1270
  {
1271
+ "epoch": 0.71,
1272
  "learning_rate": 0.00042,
1273
+ "loss": 4.4329,
1274
+ "step": 211
1275
  },
1276
  {
1277
+ "epoch": 0.71,
1278
  "learning_rate": 0.000422,
1279
+ "loss": 4.5834,
1280
+ "step": 212
1281
  },
1282
  {
1283
+ "epoch": 0.72,
1284
  "learning_rate": 0.000424,
1285
+ "loss": 4.893,
1286
+ "step": 213
1287
  },
1288
  {
1289
+ "epoch": 0.72,
1290
  "learning_rate": 0.000426,
1291
+ "loss": 4.552,
1292
+ "step": 214
1293
  },
1294
  {
1295
+ "epoch": 0.72,
1296
  "learning_rate": 0.000428,
1297
+ "loss": 4.6231,
1298
+ "step": 215
1299
  },
1300
  {
1301
+ "epoch": 0.73,
1302
  "learning_rate": 0.00043,
1303
+ "loss": 4.4696,
1304
+ "step": 216
1305
  },
1306
  {
1307
+ "epoch": 0.73,
1308
  "learning_rate": 0.000432,
1309
+ "loss": 4.5927,
1310
+ "step": 217
1311
  },
1312
  {
1313
+ "epoch": 0.73,
1314
  "learning_rate": 0.00043400000000000003,
1315
+ "loss": 4.3821,
1316
+ "step": 218
1317
  },
1318
  {
1319
+ "epoch": 0.74,
1320
  "learning_rate": 0.000436,
1321
+ "loss": 4.3723,
1322
+ "step": 219
1323
  },
1324
  {
1325
+ "epoch": 0.74,
1326
  "learning_rate": 0.000438,
1327
+ "loss": 4.5752,
1328
+ "step": 220
1329
  },
1330
  {
1331
+ "epoch": 0.74,
1332
  "learning_rate": 0.00044,
1333
+ "loss": 4.4816,
1334
+ "step": 221
1335
  },
1336
  {
1337
+ "epoch": 0.75,
1338
  "learning_rate": 0.000442,
1339
+ "loss": 4.4694,
1340
+ "step": 222
1341
  },
1342
  {
1343
+ "epoch": 0.75,
1344
  "learning_rate": 0.000444,
1345
+ "loss": 4.4168,
1346
+ "step": 223
1347
  },
1348
  {
1349
+ "epoch": 0.75,
1350
  "learning_rate": 0.000446,
1351
+ "loss": 4.5059,
1352
+ "step": 224
1353
  },
1354
  {
1355
+ "epoch": 0.76,
1356
  "learning_rate": 0.000448,
1357
+ "loss": 4.4529,
1358
+ "step": 225
1359
  },
1360
  {
1361
+ "epoch": 0.76,
1362
  "learning_rate": 0.00045000000000000004,
1363
+ "loss": 4.6363,
1364
+ "step": 226
1365
  },
1366
  {
1367
+ "epoch": 0.76,
1368
  "learning_rate": 0.00045200000000000004,
1369
+ "loss": 4.6407,
1370
+ "step": 227
1371
  },
1372
  {
1373
+ "epoch": 0.77,
1374
  "learning_rate": 0.00045400000000000003,
1375
+ "loss": 4.3059,
1376
+ "step": 228
1377
  },
1378
  {
1379
+ "epoch": 0.77,
1380
  "learning_rate": 0.000456,
1381
+ "loss": 4.4393,
1382
+ "step": 229
1383
  },
1384
  {
1385
+ "epoch": 0.77,
1386
  "learning_rate": 0.000458,
1387
+ "loss": 4.4847,
1388
+ "step": 230
1389
  },
1390
  {
1391
+ "epoch": 0.78,
1392
  "learning_rate": 0.00046,
1393
+ "loss": 4.7014,
1394
+ "step": 231
1395
  },
1396
  {
1397
+ "epoch": 0.78,
1398
  "learning_rate": 0.000462,
1399
+ "loss": 4.6399,
1400
+ "step": 232
1401
  },
1402
  {
1403
+ "epoch": 0.78,
1404
  "learning_rate": 0.00046400000000000006,
1405
+ "loss": 4.6032,
1406
+ "step": 233
1407
  },
1408
  {
1409
+ "epoch": 0.79,
1410
  "learning_rate": 0.00046600000000000005,
1411
+ "loss": 4.7998,
1412
+ "step": 234
1413
  },
1414
  {
1415
+ "epoch": 0.79,
1416
  "learning_rate": 0.00046800000000000005,
1417
+ "loss": 4.6568,
1418
+ "step": 235
1419
  },
1420
  {
1421
+ "epoch": 0.79,
1422
  "learning_rate": 0.00047,
1423
+ "loss": 4.5006,
1424
+ "step": 236
1425
  },
1426
  {
1427
+ "epoch": 0.8,
1428
  "learning_rate": 0.000472,
1429
+ "loss": 4.5085,
1430
+ "step": 237
1431
  },
1432
  {
1433
+ "epoch": 0.8,
1434
  "learning_rate": 0.000474,
1435
+ "loss": 4.5638,
1436
+ "step": 238
1437
  },
1438
  {
1439
+ "epoch": 0.8,
1440
  "learning_rate": 0.00047599999999999997,
1441
+ "loss": 4.5744,
1442
+ "step": 239
1443
  },
1444
  {
1445
+ "epoch": 0.81,
1446
  "learning_rate": 0.00047799999999999996,
1447
+ "loss": 4.6586,
1448
+ "step": 240
1449
  },
1450
  {
1451
+ "epoch": 0.81,
1452
  "learning_rate": 0.00048,
1453
+ "loss": 5.0386,
1454
+ "step": 241
1455
  },
1456
  {
1457
+ "epoch": 0.81,
1458
  "learning_rate": 0.000482,
1459
+ "loss": 4.9982,
1460
+ "step": 242
1461
  },
1462
  {
1463
+ "epoch": 0.82,
1464
  "learning_rate": 0.000484,
1465
+ "loss": 4.6487,
1466
+ "step": 243
1467
  },
1468
  {
1469
+ "epoch": 0.82,
1470
  "learning_rate": 0.000486,
1471
+ "loss": 4.4706,
1472
+ "step": 244
1473
  },
1474
  {
1475
+ "epoch": 0.82,
1476
  "learning_rate": 0.000488,
1477
+ "loss": 4.581,
1478
+ "step": 245
1479
  },
1480
  {
1481
+ "epoch": 0.83,
1482
  "learning_rate": 0.00049,
1483
+ "loss": 4.6588,
1484
+ "step": 246
1485
  },
1486
  {
1487
+ "epoch": 0.83,
1488
  "learning_rate": 0.000492,
1489
+ "loss": 4.6926,
1490
+ "step": 247
1491
  },
1492
  {
1493
+ "epoch": 0.83,
1494
  "learning_rate": 0.000494,
1495
+ "loss": 4.712,
1496
+ "step": 248
1497
  },
1498
  {
1499
+ "epoch": 0.84,
1500
  "learning_rate": 0.000496,
1501
+ "loss": 5.1018,
1502
+ "step": 249
1503
  },
1504
  {
1505
+ "epoch": 0.84,
1506
  "learning_rate": 0.000498,
1507
+ "loss": 4.5331,
1508
+ "step": 250
1509
  },
1510
  {
1511
+ "epoch": 0.84,
1512
  "learning_rate": 0.0005,
1513
+ "loss": 4.8435,
1514
+ "step": 251
1515
  },
1516
  {
1517
+ "epoch": 0.85,
1518
  "learning_rate": 0.0005020000000000001,
1519
+ "loss": 4.7466,
1520
+ "step": 252
1521
  },
1522
  {
1523
+ "epoch": 0.85,
1524
  "learning_rate": 0.000504,
1525
+ "loss": 4.7877,
1526
+ "step": 253
1527
  },
1528
  {
1529
+ "epoch": 0.85,
1530
  "learning_rate": 0.000506,
1531
+ "loss": 4.6377,
1532
+ "step": 254
1533
  },
1534
  {
1535
+ "epoch": 0.86,
1536
  "learning_rate": 0.000508,
1537
+ "loss": 4.6048,
1538
+ "step": 255
1539
  },
1540
  {
1541
+ "epoch": 0.86,
1542
  "learning_rate": 0.00051,
1543
+ "loss": 4.6086,
1544
+ "step": 256
1545
  },
1546
  {
1547
+ "epoch": 0.86,
1548
  "learning_rate": 0.000512,
1549
+ "loss": 4.8381,
1550
+ "step": 257
1551
  },
1552
  {
1553
+ "epoch": 0.87,
1554
  "learning_rate": 0.000514,
1555
+ "loss": 4.5629,
1556
+ "step": 258
1557
  },
1558
  {
1559
+ "epoch": 0.87,
1560
  "learning_rate": 0.0005160000000000001,
1561
+ "loss": 4.6615,
1562
+ "step": 259
1563
  },
1564
  {
1565
+ "epoch": 0.87,
1566
  "learning_rate": 0.000518,
1567
+ "loss": 4.5271,
1568
+ "step": 260
1569
  },
1570
  {
1571
+ "epoch": 0.88,
1572
  "learning_rate": 0.0005200000000000001,
1573
+ "loss": 4.6963,
1574
+ "step": 261
1575
  },
1576
  {
1577
+ "epoch": 0.88,
1578
  "learning_rate": 0.000522,
1579
+ "loss": 4.616,
1580
+ "step": 262
1581
  },
1582
  {
1583
+ "epoch": 0.88,
1584
  "learning_rate": 0.000524,
1585
+ "loss": 4.6217,
1586
+ "step": 263
1587
  },
1588
  {
1589
+ "epoch": 0.89,
1590
  "learning_rate": 0.000526,
1591
+ "loss": 4.5987,
1592
+ "step": 264
1593
  },
1594
  {
1595
+ "epoch": 0.89,
1596
  "learning_rate": 0.000528,
1597
+ "loss": 4.6084,
1598
+ "step": 265
1599
  },
1600
  {
1601
+ "epoch": 0.89,
1602
  "learning_rate": 0.0005300000000000001,
1603
+ "loss": 4.7973,
1604
+ "step": 266
1605
  },
1606
  {
1607
+ "epoch": 0.9,
1608
  "learning_rate": 0.000532,
1609
+ "loss": 4.5744,
1610
+ "step": 267
1611
  },
1612
  {
1613
+ "epoch": 0.9,
1614
  "learning_rate": 0.0005340000000000001,
1615
+ "loss": 4.7483,
1616
+ "step": 268
1617
  },
1618
  {
1619
+ "epoch": 0.9,
1620
  "learning_rate": 0.000536,
1621
+ "loss": 4.4083,
1622
+ "step": 269
1623
  },
1624
  {
1625
+ "epoch": 0.91,
1626
  "learning_rate": 0.0005380000000000001,
1627
+ "loss": 4.6659,
1628
+ "step": 270
1629
  },
1630
  {
1631
+ "epoch": 0.91,
1632
  "learning_rate": 0.00054,
1633
+ "loss": 4.5806,
1634
+ "step": 271
1635
  },
1636
  {
1637
+ "epoch": 0.91,
1638
  "learning_rate": 0.0005420000000000001,
1639
+ "loss": 4.4699,
1640
+ "step": 272
1641
  },
1642
  {
1643
+ "epoch": 0.92,
1644
  "learning_rate": 0.0005440000000000001,
1645
+ "loss": 4.6826,
1646
+ "step": 273
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1647
  },
1648
  {
1649
  "epoch": 0.92,
1650
+ "learning_rate": 0.000546,
1651
+ "loss": 4.5738,
1652
+ "step": 274
1653
  },
1654
  {
1655
  "epoch": 0.92,
1656
+ "learning_rate": 0.0005480000000000001,
1657
+ "loss": 4.5848,
1658
+ "step": 275
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1659
  },
1660
  {
1661
  "epoch": 0.93,
1662
+ "learning_rate": 0.00055,
1663
+ "loss": 4.6156,
1664
+ "step": 276
1665
  },
1666
  {
1667
  "epoch": 0.93,
1668
+ "learning_rate": 0.0005520000000000001,
1669
+ "loss": 4.6793,
1670
+ "step": 277
1671
  },
1672
  {
1673
  "epoch": 0.93,
1674
+ "learning_rate": 0.000554,
1675
+ "loss": 4.6557,
1676
+ "step": 278
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1677
  },
1678
  {
1679
  "epoch": 0.94,
1680
+ "learning_rate": 0.0005560000000000001,
1681
+ "loss": 4.6794,
1682
+ "step": 279
1683
  },
1684
  {
1685
  "epoch": 0.94,
1686
+ "learning_rate": 0.000558,
1687
+ "loss": 4.4396,
1688
+ "step": 280
1689
  },
1690
  {
1691
  "epoch": 0.94,
1692
+ "learning_rate": 0.0005600000000000001,
1693
+ "loss": 4.7257,
1694
+ "step": 281
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1695
  },
1696
  {
1697
  "epoch": 0.95,
1698
+ "learning_rate": 0.0005620000000000001,
1699
+ "loss": 4.4927,
1700
+ "step": 282
1701
  },
1702
  {
1703
  "epoch": 0.95,
1704
+ "learning_rate": 0.0005639999999999999,
1705
+ "loss": 4.6712,
1706
+ "step": 283
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1707
  },
1708
  {
1709
  "epoch": 0.96,
1710
+ "learning_rate": 0.000566,
1711
+ "loss": 4.4852,
1712
+ "step": 284
1713
  },
1714
  {
1715
  "epoch": 0.96,
1716
+ "learning_rate": 0.0005679999999999999,
1717
+ "loss": 4.7465,
1718
+ "step": 285
1719
  },
1720
  {
1721
  "epoch": 0.96,
1722
+ "learning_rate": 0.00057,
1723
+ "loss": 4.7233,
1724
+ "step": 286
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1725
  },
1726
  {
1727
  "epoch": 0.97,
1728
+ "learning_rate": 0.0005719999999999999,
1729
+ "loss": 4.6061,
1730
+ "step": 287
1731
  },
1732
  {
1733
  "epoch": 0.97,
1734
+ "learning_rate": 0.000574,
1735
+ "loss": 4.6499,
1736
+ "step": 288
1737
  },
1738
  {
1739
  "epoch": 0.97,
1740
+ "learning_rate": 0.000576,
1741
+ "loss": 4.5556,
1742
+ "step": 289
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1743
  },
1744
  {
1745
  "epoch": 0.98,
1746
+ "learning_rate": 0.000578,
1747
+ "loss": 4.6277,
1748
+ "step": 290
1749
  },
1750
  {
1751
  "epoch": 0.98,
1752
+ "learning_rate": 0.00058,
1753
+ "loss": 4.6354,
1754
+ "step": 291
1755
  },
1756
  {
1757
  "epoch": 0.98,
1758
+ "learning_rate": 0.0005819999999999999,
1759
+ "loss": 4.7174,
1760
+ "step": 292
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1761
  },
1762
  {
1763
  "epoch": 0.99,
1764
+ "learning_rate": 0.000584,
1765
+ "loss": 4.5303,
1766
+ "step": 293
1767
  },
1768
  {
1769
  "epoch": 0.99,
1770
+ "learning_rate": 0.0005859999999999999,
1771
+ "loss": 5.5856,
1772
+ "step": 294
1773
  },
1774
  {
1775
  "epoch": 0.99,
1776
+ "learning_rate": 0.000588,
1777
+ "loss": 4.7103,
1778
+ "step": 295
 
 
 
 
 
 
1779
  },
1780
  {
1781
  "epoch": 1.0,
1782
+ "learning_rate": 0.00059,
1783
+ "loss": 5.2133,
1784
+ "step": 296
1785
  },
1786
  {
1787
  "epoch": 1.0,
1788
+ "learning_rate": 0.000592,
1789
+ "loss": 4.7346,
1790
+ "step": 297
1791
  },
1792
  {
1793
  "epoch": 1.0,
1794
+ "step": 297,
1795
  "total_flos": 0.0,
1796
+ "train_loss": 4.473529295487837,
1797
+ "train_runtime": 4542.6558,
1798
+ "train_samples_per_second": 6.282,
1799
+ "train_steps_per_second": 0.065
1800
  }
1801
  ],
1802
+ "max_steps": 297,
1803
  "num_train_epochs": 1,
1804
  "total_flos": 0.0,
1805
  "trial_name": null,