koya1 commited on
Commit
14e5106
1 Parent(s): 14e11bd

End of training

Browse files
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 3.25,
3
- "eval_accuracy": 0.8114285714285714,
4
- "eval_loss": 0.6237285733222961,
5
- "eval_runtime": 52.2354,
6
- "eval_samples_per_second": 3.35,
7
- "eval_steps_per_second": 3.35
8
  }
 
1
  {
2
+ "epoch": 9.1,
3
+ "eval_accuracy": 0.7885714285714286,
4
+ "eval_loss": 0.6197737455368042,
5
+ "eval_runtime": 54.239,
6
+ "eval_samples_per_second": 3.226,
7
+ "eval_steps_per_second": 3.226
8
  }
runs/Mar30_06-55-30_04b9fd5247a2/events.out.tfevents.1680162526.04b9fd5247a2.349.5 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:88525c4c7bec23f8d1d79193f13542103791655a3fe7967eb7c19496b3bb421f
3
- size 363
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc1095209a46e4c290ff94a0be9e8c4189960fcde7c3b48e724ffc216f1dbc45
3
+ size 686
test_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 3.25,
3
- "eval_accuracy": 0.8114285714285714,
4
- "eval_loss": 0.6237285733222961,
5
- "eval_runtime": 52.2354,
6
- "eval_samples_per_second": 3.35,
7
- "eval_steps_per_second": 3.35
8
  }
 
1
  {
2
+ "epoch": 9.1,
3
+ "eval_accuracy": 0.7885714285714286,
4
+ "eval_loss": 0.6197737455368042,
5
+ "eval_runtime": 54.239,
6
+ "eval_samples_per_second": 3.226,
7
+ "eval_steps_per_second": 3.226
8
  }
trainer_state.json CHANGED
@@ -1,955 +1,2323 @@
1
  {
2
- "best_metric": 0.7422680412371134,
3
- "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-1460",
4
- "epoch": 3.25,
5
- "global_step": 1460,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
- "epoch": 0.01,
12
- "learning_rate": 3.4246575342465754e-06,
13
- "loss": 2.5993,
14
  "step": 10
15
  },
16
  {
17
  "epoch": 0.01,
18
- "learning_rate": 6.849315068493151e-06,
19
- "loss": 2.5853,
20
  "step": 20
21
  },
22
  {
23
- "epoch": 0.02,
24
- "learning_rate": 1.0273972602739726e-05,
25
- "loss": 2.6647,
26
  "step": 30
27
  },
28
  {
29
- "epoch": 0.03,
30
- "learning_rate": 1.3698630136986302e-05,
31
- "loss": 2.5315,
32
  "step": 40
33
  },
34
  {
35
- "epoch": 0.03,
36
- "learning_rate": 1.7123287671232875e-05,
37
- "loss": 2.4656,
38
  "step": 50
39
  },
40
  {
41
- "epoch": 0.04,
42
- "learning_rate": 2.0547945205479453e-05,
43
- "loss": 2.6351,
44
  "step": 60
45
  },
46
  {
47
- "epoch": 0.05,
48
- "learning_rate": 2.3972602739726026e-05,
49
- "loss": 2.662,
50
  "step": 70
51
  },
52
  {
53
- "epoch": 0.05,
54
- "learning_rate": 2.7397260273972603e-05,
55
- "loss": 2.3767,
56
  "step": 80
57
  },
58
  {
59
- "epoch": 0.06,
60
- "learning_rate": 3.082191780821918e-05,
61
- "loss": 2.3325,
62
  "step": 90
63
  },
64
  {
65
- "epoch": 0.07,
66
- "learning_rate": 3.424657534246575e-05,
67
- "loss": 2.7198,
68
  "step": 100
69
  },
70
  {
71
- "epoch": 0.08,
72
- "learning_rate": 3.767123287671233e-05,
73
- "loss": 2.4336,
74
  "step": 110
75
  },
76
  {
77
- "epoch": 0.08,
78
- "learning_rate": 4.1095890410958905e-05,
79
- "loss": 2.4116,
80
  "step": 120
81
  },
82
  {
83
- "epoch": 0.09,
84
- "learning_rate": 4.452054794520548e-05,
85
- "loss": 2.6112,
86
  "step": 130
87
  },
88
  {
89
- "epoch": 0.1,
90
- "learning_rate": 4.794520547945205e-05,
91
- "loss": 2.3558,
92
  "step": 140
93
  },
94
  {
95
- "epoch": 0.1,
96
- "learning_rate": 4.984779299847793e-05,
97
- "loss": 2.1907,
98
  "step": 150
99
  },
100
  {
101
- "epoch": 0.11,
102
- "learning_rate": 4.9467275494672755e-05,
103
- "loss": 2.1687,
104
  "step": 160
105
  },
106
  {
107
- "epoch": 0.12,
108
- "learning_rate": 4.908675799086758e-05,
109
- "loss": 2.2809,
110
  "step": 170
111
  },
112
  {
113
- "epoch": 0.12,
114
- "learning_rate": 4.8706240487062404e-05,
115
- "loss": 2.4247,
116
  "step": 180
117
  },
118
  {
119
- "epoch": 0.13,
120
- "learning_rate": 4.8325722983257235e-05,
121
- "loss": 2.2747,
122
  "step": 190
123
  },
124
  {
125
- "epoch": 0.14,
126
- "learning_rate": 4.794520547945205e-05,
127
- "loss": 2.2174,
128
  "step": 200
129
  },
130
  {
131
- "epoch": 0.14,
132
- "learning_rate": 4.7564687975646884e-05,
133
- "loss": 2.4167,
134
  "step": 210
135
  },
136
  {
137
- "epoch": 0.15,
138
- "learning_rate": 4.71841704718417e-05,
139
- "loss": 2.2652,
140
  "step": 220
141
  },
142
  {
143
- "epoch": 0.16,
144
- "learning_rate": 4.680365296803653e-05,
145
- "loss": 2.2426,
146
  "step": 230
147
  },
148
  {
149
- "epoch": 0.16,
150
- "learning_rate": 4.642313546423136e-05,
151
- "loss": 2.3514,
152
  "step": 240
153
  },
154
  {
155
- "epoch": 0.17,
156
- "learning_rate": 4.604261796042618e-05,
157
- "loss": 1.8138,
158
  "step": 250
159
  },
160
  {
161
- "epoch": 0.18,
162
- "learning_rate": 4.5662100456621006e-05,
163
- "loss": 2.6617,
164
  "step": 260
165
  },
166
  {
167
- "epoch": 0.18,
168
- "learning_rate": 4.528158295281583e-05,
169
- "loss": 1.6823,
170
  "step": 270
171
  },
172
  {
173
- "epoch": 0.19,
174
- "learning_rate": 4.4901065449010655e-05,
175
- "loss": 2.382,
176
  "step": 280
177
  },
178
  {
179
- "epoch": 0.2,
180
- "learning_rate": 4.452054794520548e-05,
181
- "loss": 2.1086,
182
  "step": 290
183
  },
184
  {
185
- "epoch": 0.21,
186
- "learning_rate": 4.41400304414003e-05,
187
- "loss": 1.8853,
188
  "step": 300
189
  },
190
  {
191
- "epoch": 0.21,
192
- "learning_rate": 4.3759512937595135e-05,
193
- "loss": 1.8514,
194
  "step": 310
195
  },
196
  {
197
- "epoch": 0.22,
198
- "learning_rate": 4.337899543378995e-05,
199
- "loss": 2.6215,
200
  "step": 320
201
  },
202
  {
203
- "epoch": 0.23,
204
- "learning_rate": 4.299847792998478e-05,
205
- "loss": 2.2046,
206
  "step": 330
207
  },
208
  {
209
- "epoch": 0.23,
210
- "learning_rate": 4.26179604261796e-05,
211
- "loss": 2.602,
212
  "step": 340
213
  },
214
  {
215
- "epoch": 0.24,
216
- "learning_rate": 4.223744292237443e-05,
217
- "loss": 2.3634,
218
  "step": 350
219
  },
220
  {
221
- "epoch": 0.25,
222
- "learning_rate": 4.1856925418569256e-05,
223
- "loss": 2.0135,
224
  "step": 360
225
  },
226
  {
227
- "epoch": 0.25,
228
- "eval_accuracy": 0.14432989690721648,
229
- "eval_loss": 2.1087806224823,
230
- "eval_runtime": 27.8992,
231
- "eval_samples_per_second": 3.477,
232
- "eval_steps_per_second": 3.477,
233
  "step": 365
234
  },
235
  {
236
  "epoch": 1.0,
237
- "learning_rate": 4.147640791476408e-05,
238
- "loss": 2.381,
239
  "step": 370
240
  },
241
  {
242
- "epoch": 1.01,
243
- "learning_rate": 4.1095890410958905e-05,
244
- "loss": 1.991,
245
  "step": 380
246
  },
247
  {
248
- "epoch": 1.02,
249
- "learning_rate": 4.071537290715373e-05,
250
- "loss": 1.4985,
251
  "step": 390
252
  },
253
  {
254
- "epoch": 1.02,
255
- "learning_rate": 4.0334855403348554e-05,
256
- "loss": 1.761,
257
  "step": 400
258
  },
259
  {
260
- "epoch": 1.03,
261
- "learning_rate": 3.995433789954338e-05,
262
- "loss": 2.366,
263
  "step": 410
264
  },
265
  {
266
- "epoch": 1.04,
267
- "learning_rate": 3.95738203957382e-05,
268
- "loss": 2.3994,
269
  "step": 420
270
  },
271
  {
272
- "epoch": 1.04,
273
- "learning_rate": 3.9193302891933034e-05,
274
- "loss": 2.032,
275
  "step": 430
276
  },
277
  {
278
- "epoch": 1.05,
279
- "learning_rate": 3.881278538812785e-05,
280
- "loss": 1.9747,
281
  "step": 440
282
  },
283
  {
284
- "epoch": 1.06,
285
- "learning_rate": 3.843226788432268e-05,
286
- "loss": 1.925,
287
  "step": 450
288
  },
289
  {
290
- "epoch": 1.07,
291
- "learning_rate": 3.80517503805175e-05,
292
- "loss": 1.7874,
293
  "step": 460
294
  },
295
  {
296
- "epoch": 1.07,
297
- "learning_rate": 3.767123287671233e-05,
298
- "loss": 1.9562,
299
  "step": 470
300
  },
301
  {
302
- "epoch": 1.08,
303
- "learning_rate": 3.7290715372907156e-05,
304
- "loss": 2.988,
305
  "step": 480
306
  },
307
  {
308
- "epoch": 1.09,
309
- "learning_rate": 3.691019786910198e-05,
310
- "loss": 1.6423,
311
  "step": 490
312
  },
313
  {
314
- "epoch": 1.09,
315
- "learning_rate": 3.6529680365296805e-05,
316
- "loss": 2.0119,
317
  "step": 500
318
  },
319
  {
320
- "epoch": 1.1,
321
- "learning_rate": 3.614916286149163e-05,
322
- "loss": 1.6883,
323
  "step": 510
324
  },
325
  {
326
- "epoch": 1.11,
327
- "learning_rate": 3.5768645357686453e-05,
328
- "loss": 1.7275,
329
  "step": 520
330
  },
331
  {
332
- "epoch": 1.11,
333
- "learning_rate": 3.538812785388128e-05,
334
- "loss": 1.0309,
335
  "step": 530
336
  },
337
  {
338
- "epoch": 1.12,
339
- "learning_rate": 3.50076103500761e-05,
340
- "loss": 2.0449,
341
  "step": 540
342
  },
343
  {
344
- "epoch": 1.13,
345
- "learning_rate": 3.462709284627093e-05,
346
- "loss": 1.1414,
347
  "step": 550
348
  },
349
  {
350
- "epoch": 1.13,
351
- "learning_rate": 3.424657534246575e-05,
352
- "loss": 1.0559,
353
  "step": 560
354
  },
355
  {
356
- "epoch": 1.14,
357
- "learning_rate": 3.386605783866058e-05,
358
- "loss": 1.4151,
359
  "step": 570
360
  },
361
  {
362
- "epoch": 1.15,
363
- "learning_rate": 3.34855403348554e-05,
364
- "loss": 1.5515,
365
  "step": 580
366
  },
367
  {
368
- "epoch": 1.15,
369
- "learning_rate": 3.310502283105023e-05,
370
- "loss": 1.2185,
371
  "step": 590
372
  },
373
  {
374
- "epoch": 1.16,
375
- "learning_rate": 3.2724505327245055e-05,
376
- "loss": 1.3974,
377
  "step": 600
378
  },
379
  {
380
- "epoch": 1.17,
381
- "learning_rate": 3.234398782343988e-05,
382
- "loss": 1.3453,
383
  "step": 610
384
  },
385
  {
386
- "epoch": 1.17,
387
- "learning_rate": 3.1963470319634704e-05,
388
- "loss": 1.8057,
389
  "step": 620
390
  },
391
  {
392
- "epoch": 1.18,
393
- "learning_rate": 3.158295281582953e-05,
394
- "loss": 0.9315,
395
  "step": 630
396
  },
397
  {
398
- "epoch": 1.19,
399
- "learning_rate": 3.120243531202435e-05,
400
- "loss": 1.2106,
401
  "step": 640
402
  },
403
  {
404
- "epoch": 1.2,
405
- "learning_rate": 3.082191780821918e-05,
406
- "loss": 0.7568,
407
  "step": 650
408
  },
409
  {
410
- "epoch": 1.2,
411
- "learning_rate": 3.0441400304414e-05,
412
- "loss": 1.0881,
413
  "step": 660
414
  },
415
  {
416
- "epoch": 1.21,
417
- "learning_rate": 3.006088280060883e-05,
418
- "loss": 1.1191,
419
  "step": 670
420
  },
421
  {
422
- "epoch": 1.22,
423
- "learning_rate": 2.9680365296803654e-05,
424
- "loss": 0.3567,
425
  "step": 680
426
  },
427
  {
428
- "epoch": 1.22,
429
- "learning_rate": 2.929984779299848e-05,
430
- "loss": 0.828,
431
  "step": 690
432
  },
433
  {
434
- "epoch": 1.23,
435
- "learning_rate": 2.8919330289193303e-05,
436
- "loss": 0.6539,
437
  "step": 700
438
  },
439
  {
440
- "epoch": 1.24,
441
- "learning_rate": 2.853881278538813e-05,
442
- "loss": 1.2757,
443
  "step": 710
444
  },
445
  {
446
- "epoch": 1.24,
447
- "learning_rate": 2.815829528158295e-05,
448
- "loss": 1.9999,
449
  "step": 720
450
  },
451
  {
452
- "epoch": 1.25,
453
- "learning_rate": 2.777777777777778e-05,
454
- "loss": 1.6817,
455
  "step": 730
456
  },
457
  {
458
- "epoch": 1.25,
459
- "eval_accuracy": 0.3917525773195876,
460
- "eval_loss": 2.0997157096862793,
461
- "eval_runtime": 25.2952,
462
- "eval_samples_per_second": 3.835,
463
- "eval_steps_per_second": 3.835,
464
  "step": 730
465
  },
466
  {
467
- "epoch": 2.01,
468
- "learning_rate": 2.7397260273972603e-05,
469
- "loss": 0.9235,
470
  "step": 740
471
  },
472
  {
473
  "epoch": 2.01,
474
- "learning_rate": 2.701674277016743e-05,
475
- "loss": 1.7029,
476
  "step": 750
477
  },
478
  {
479
- "epoch": 2.02,
480
- "learning_rate": 2.6636225266362252e-05,
481
- "loss": 1.0662,
482
  "step": 760
483
  },
484
  {
485
- "epoch": 2.03,
486
- "learning_rate": 2.625570776255708e-05,
487
- "loss": 0.9232,
488
  "step": 770
489
  },
490
  {
491
- "epoch": 2.03,
492
- "learning_rate": 2.58751902587519e-05,
493
- "loss": 0.9683,
494
  "step": 780
495
  },
496
  {
497
- "epoch": 2.04,
498
- "learning_rate": 2.549467275494673e-05,
499
- "loss": 1.005,
500
  "step": 790
501
  },
502
  {
503
- "epoch": 2.05,
504
- "learning_rate": 2.5114155251141553e-05,
505
- "loss": 1.0024,
506
  "step": 800
507
  },
508
  {
509
- "epoch": 2.05,
510
- "learning_rate": 2.4733637747336378e-05,
511
- "loss": 1.0185,
512
  "step": 810
513
  },
514
  {
515
- "epoch": 2.06,
516
- "learning_rate": 2.4353120243531202e-05,
517
- "loss": 0.289,
518
  "step": 820
519
  },
520
  {
521
- "epoch": 2.07,
522
- "learning_rate": 2.3972602739726026e-05,
523
- "loss": 1.3332,
524
  "step": 830
525
  },
526
  {
527
- "epoch": 2.08,
528
- "learning_rate": 2.359208523592085e-05,
529
- "loss": 1.2137,
530
  "step": 840
531
  },
532
  {
533
- "epoch": 2.08,
534
- "learning_rate": 2.321156773211568e-05,
535
- "loss": 0.8686,
536
  "step": 850
537
  },
538
  {
539
- "epoch": 2.09,
540
- "learning_rate": 2.2831050228310503e-05,
541
- "loss": 1.8639,
542
  "step": 860
543
  },
544
  {
545
- "epoch": 2.1,
546
- "learning_rate": 2.2450532724505327e-05,
547
- "loss": 0.7075,
548
  "step": 870
549
  },
550
  {
551
- "epoch": 2.1,
552
- "learning_rate": 2.207001522070015e-05,
553
- "loss": 0.4791,
554
  "step": 880
555
  },
556
  {
557
- "epoch": 2.11,
558
- "learning_rate": 2.1689497716894976e-05,
559
- "loss": 0.8184,
560
  "step": 890
561
  },
562
  {
563
- "epoch": 2.12,
564
- "learning_rate": 2.13089802130898e-05,
565
- "loss": 1.0992,
566
  "step": 900
567
  },
568
  {
569
- "epoch": 2.12,
570
- "learning_rate": 2.0928462709284628e-05,
571
- "loss": 0.7498,
572
  "step": 910
573
  },
574
  {
575
- "epoch": 2.13,
576
- "learning_rate": 2.0547945205479453e-05,
577
- "loss": 0.2501,
578
  "step": 920
579
  },
580
  {
581
- "epoch": 2.14,
582
- "learning_rate": 2.0167427701674277e-05,
583
- "loss": 0.2121,
584
  "step": 930
585
  },
586
  {
587
- "epoch": 2.14,
588
- "learning_rate": 1.97869101978691e-05,
589
- "loss": 0.3343,
590
  "step": 940
591
  },
592
  {
593
- "epoch": 2.15,
594
- "learning_rate": 1.9406392694063926e-05,
595
- "loss": 0.7544,
596
  "step": 950
597
  },
598
  {
599
- "epoch": 2.16,
600
- "learning_rate": 1.902587519025875e-05,
601
- "loss": 1.3621,
602
  "step": 960
603
  },
604
  {
605
- "epoch": 2.16,
606
- "learning_rate": 1.8645357686453578e-05,
607
- "loss": 0.5267,
608
  "step": 970
609
  },
610
  {
611
- "epoch": 2.17,
612
- "learning_rate": 1.8264840182648402e-05,
613
- "loss": 0.8902,
614
  "step": 980
615
  },
616
  {
617
- "epoch": 2.18,
618
- "learning_rate": 1.7884322678843227e-05,
619
- "loss": 0.5535,
620
  "step": 990
621
  },
622
  {
623
- "epoch": 2.18,
624
- "learning_rate": 1.750380517503805e-05,
625
- "loss": 0.8907,
626
  "step": 1000
627
  },
628
  {
629
- "epoch": 2.19,
630
- "learning_rate": 1.7123287671232875e-05,
631
- "loss": 0.1521,
632
  "step": 1010
633
  },
634
  {
635
- "epoch": 2.2,
636
- "learning_rate": 1.67427701674277e-05,
637
- "loss": 0.2744,
638
  "step": 1020
639
  },
640
  {
641
- "epoch": 2.21,
642
- "learning_rate": 1.6362252663622528e-05,
643
- "loss": 0.17,
644
  "step": 1030
645
  },
646
  {
647
- "epoch": 2.21,
648
- "learning_rate": 1.5981735159817352e-05,
649
- "loss": 0.1956,
650
  "step": 1040
651
  },
652
  {
653
- "epoch": 2.22,
654
- "learning_rate": 1.5601217656012176e-05,
655
- "loss": 0.9093,
656
  "step": 1050
657
  },
658
  {
659
- "epoch": 2.23,
660
- "learning_rate": 1.5220700152207e-05,
661
- "loss": 1.2878,
662
  "step": 1060
663
  },
664
  {
665
- "epoch": 2.23,
666
- "learning_rate": 1.4840182648401827e-05,
667
- "loss": 0.6323,
668
  "step": 1070
669
  },
670
  {
671
- "epoch": 2.24,
672
- "learning_rate": 1.4459665144596651e-05,
673
- "loss": 0.4712,
674
  "step": 1080
675
  },
676
  {
677
- "epoch": 2.25,
678
- "learning_rate": 1.4079147640791476e-05,
679
- "loss": 0.1918,
680
  "step": 1090
681
  },
682
  {
683
- "epoch": 2.25,
684
- "eval_accuracy": 0.6185567010309279,
685
- "eval_loss": 1.2655441761016846,
686
- "eval_runtime": 25.5435,
687
- "eval_samples_per_second": 3.797,
688
- "eval_steps_per_second": 3.797,
689
  "step": 1095
690
  },
691
  {
692
  "epoch": 3.0,
693
- "learning_rate": 1.3698630136986302e-05,
694
- "loss": 0.629,
695
  "step": 1100
696
  },
697
  {
698
- "epoch": 3.01,
699
- "learning_rate": 1.3318112633181126e-05,
700
- "loss": 0.4214,
701
  "step": 1110
702
  },
703
  {
704
- "epoch": 3.02,
705
- "learning_rate": 1.293759512937595e-05,
706
- "loss": 0.7807,
707
  "step": 1120
708
  },
709
  {
710
- "epoch": 3.02,
711
- "learning_rate": 1.2557077625570777e-05,
712
- "loss": 0.2658,
713
  "step": 1130
714
  },
715
  {
716
- "epoch": 3.03,
717
- "learning_rate": 1.2176560121765601e-05,
718
- "loss": 0.8271,
719
  "step": 1140
720
  },
721
  {
722
- "epoch": 3.04,
723
- "learning_rate": 1.1796042617960425e-05,
724
- "loss": 0.3046,
725
  "step": 1150
726
  },
727
  {
728
- "epoch": 3.04,
729
- "learning_rate": 1.1415525114155251e-05,
730
- "loss": 0.3587,
731
  "step": 1160
732
  },
733
  {
734
- "epoch": 3.05,
735
- "learning_rate": 1.1035007610350076e-05,
736
- "loss": 0.1176,
737
  "step": 1170
738
  },
739
  {
740
- "epoch": 3.06,
741
- "learning_rate": 1.06544901065449e-05,
742
- "loss": 0.1104,
743
  "step": 1180
744
  },
745
  {
746
- "epoch": 3.07,
747
- "learning_rate": 1.0273972602739726e-05,
748
- "loss": 0.8839,
749
  "step": 1190
750
  },
751
  {
752
- "epoch": 3.07,
753
- "learning_rate": 9.89345509893455e-06,
754
- "loss": 0.1449,
755
  "step": 1200
756
  },
757
  {
758
- "epoch": 3.08,
759
- "learning_rate": 9.512937595129375e-06,
760
- "loss": 0.1052,
761
  "step": 1210
762
  },
763
  {
764
- "epoch": 3.09,
765
- "learning_rate": 9.132420091324201e-06,
766
- "loss": 0.0785,
767
  "step": 1220
768
  },
769
  {
770
- "epoch": 3.09,
771
- "learning_rate": 8.751902587519026e-06,
772
- "loss": 0.3151,
773
  "step": 1230
774
  },
775
  {
776
- "epoch": 3.1,
777
- "learning_rate": 8.37138508371385e-06,
778
- "loss": 1.218,
779
  "step": 1240
780
  },
781
  {
782
- "epoch": 3.11,
783
- "learning_rate": 7.990867579908676e-06,
784
- "loss": 0.1809,
785
  "step": 1250
786
  },
787
  {
788
- "epoch": 3.11,
789
- "learning_rate": 7.6103500761035e-06,
790
- "loss": 0.1806,
791
  "step": 1260
792
  },
793
  {
794
- "epoch": 3.12,
795
- "learning_rate": 7.229832572298326e-06,
796
- "loss": 0.1936,
797
  "step": 1270
798
  },
799
  {
800
- "epoch": 3.13,
801
- "learning_rate": 6.849315068493151e-06,
802
- "loss": 0.0504,
803
  "step": 1280
804
  },
805
  {
806
- "epoch": 3.13,
807
- "learning_rate": 6.468797564687975e-06,
808
- "loss": 0.9438,
809
  "step": 1290
810
  },
811
  {
812
- "epoch": 3.14,
813
- "learning_rate": 6.0882800608828005e-06,
814
- "loss": 0.0356,
815
  "step": 1300
816
  },
817
  {
818
- "epoch": 3.15,
819
- "learning_rate": 5.707762557077626e-06,
820
- "loss": 0.9749,
821
  "step": 1310
822
  },
823
  {
824
- "epoch": 3.15,
825
- "learning_rate": 5.32724505327245e-06,
826
- "loss": 0.2376,
827
  "step": 1320
828
  },
829
  {
830
- "epoch": 3.16,
831
- "learning_rate": 4.946727549467275e-06,
832
- "loss": 0.3259,
833
  "step": 1330
834
  },
835
  {
836
- "epoch": 3.17,
837
- "learning_rate": 4.566210045662101e-06,
838
- "loss": 0.2959,
839
  "step": 1340
840
  },
841
  {
842
- "epoch": 3.17,
843
- "learning_rate": 4.185692541856925e-06,
844
- "loss": 0.0705,
845
  "step": 1350
846
  },
847
  {
848
- "epoch": 3.18,
849
- "learning_rate": 3.80517503805175e-06,
850
- "loss": 0.5899,
851
  "step": 1360
852
  },
853
  {
854
- "epoch": 3.19,
855
- "learning_rate": 3.4246575342465754e-06,
856
- "loss": 0.2143,
857
  "step": 1370
858
  },
859
  {
860
- "epoch": 3.2,
861
- "learning_rate": 3.0441400304414002e-06,
862
- "loss": 0.265,
863
  "step": 1380
864
  },
865
  {
866
- "epoch": 3.2,
867
- "learning_rate": 2.663622526636225e-06,
868
- "loss": 0.0809,
869
  "step": 1390
870
  },
871
  {
872
- "epoch": 3.21,
873
- "learning_rate": 2.2831050228310503e-06,
874
- "loss": 0.0135,
875
  "step": 1400
876
  },
877
  {
878
- "epoch": 3.22,
879
- "learning_rate": 1.902587519025875e-06,
880
- "loss": 0.0089,
881
  "step": 1410
882
  },
883
  {
884
- "epoch": 3.22,
885
- "learning_rate": 1.5220700152207001e-06,
886
- "loss": 0.0718,
887
  "step": 1420
888
  },
889
  {
890
- "epoch": 3.23,
891
- "learning_rate": 1.1415525114155251e-06,
892
- "loss": 0.5197,
893
  "step": 1430
894
  },
895
  {
896
- "epoch": 3.24,
897
- "learning_rate": 7.610350076103501e-07,
898
- "loss": 0.3288,
899
  "step": 1440
900
  },
901
  {
902
- "epoch": 3.24,
903
- "learning_rate": 3.8051750380517503e-07,
904
- "loss": 0.1716,
905
  "step": 1450
906
  },
907
  {
908
- "epoch": 3.25,
909
- "learning_rate": 0.0,
910
- "loss": 0.4802,
911
  "step": 1460
912
  },
913
  {
914
- "epoch": 3.25,
915
- "eval_accuracy": 0.7422680412371134,
916
- "eval_loss": 1.1125394105911255,
917
- "eval_runtime": 26.9468,
918
- "eval_samples_per_second": 3.6,
919
- "eval_steps_per_second": 3.6,
920
  "step": 1460
921
  },
922
  {
923
- "epoch": 3.25,
924
- "step": 1460,
925
- "total_flos": 1.8194165442831974e+18,
926
- "train_loss": 1.2557036288133632,
927
- "train_runtime": 1150.1678,
928
- "train_samples_per_second": 1.269,
929
- "train_steps_per_second": 1.269
930
  },
931
  {
932
- "epoch": 3.25,
933
- "eval_accuracy": 0.8114285714285714,
934
- "eval_loss": 0.6237284541130066,
935
- "eval_runtime": 57.0411,
936
- "eval_samples_per_second": 3.068,
937
- "eval_steps_per_second": 3.068,
938
- "step": 1460
939
  },
940
  {
941
- "epoch": 3.25,
942
- "eval_accuracy": 0.8114285714285714,
943
- "eval_loss": 0.6237285733222961,
944
- "eval_runtime": 52.2354,
945
- "eval_samples_per_second": 3.35,
946
- "eval_steps_per_second": 3.35,
947
- "step": 1460
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
948
  }
949
  ],
950
- "max_steps": 1460,
951
  "num_train_epochs": 9223372036854775807,
952
- "total_flos": 1.8194165442831974e+18,
953
  "trial_name": null,
954
  "trial_params": null
955
  }
 
1
  {
2
+ "best_metric": 0.8041237113402062,
3
+ "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-730",
4
+ "epoch": 9.1,
5
+ "global_step": 3650,
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": 1.3698630136986302e-06,
13
+ "loss": 0.0423,
14
  "step": 10
15
  },
16
  {
17
  "epoch": 0.01,
18
+ "learning_rate": 2.7397260273972604e-06,
19
+ "loss": 0.4626,
20
  "step": 20
21
  },
22
  {
23
+ "epoch": 0.01,
24
+ "learning_rate": 4.10958904109589e-06,
25
+ "loss": 0.009,
26
  "step": 30
27
  },
28
  {
29
+ "epoch": 0.01,
30
+ "learning_rate": 5.479452054794521e-06,
31
+ "loss": 0.1166,
32
  "step": 40
33
  },
34
  {
35
+ "epoch": 0.01,
36
+ "learning_rate": 6.849315068493151e-06,
37
+ "loss": 0.8124,
38
  "step": 50
39
  },
40
  {
41
+ "epoch": 0.02,
42
+ "learning_rate": 8.21917808219178e-06,
43
+ "loss": 0.0274,
44
  "step": 60
45
  },
46
  {
47
+ "epoch": 0.02,
48
+ "learning_rate": 9.589041095890411e-06,
49
+ "loss": 0.1435,
50
  "step": 70
51
  },
52
  {
53
+ "epoch": 0.02,
54
+ "learning_rate": 1.0958904109589042e-05,
55
+ "loss": 0.159,
56
  "step": 80
57
  },
58
  {
59
+ "epoch": 0.02,
60
+ "learning_rate": 1.2328767123287671e-05,
61
+ "loss": 0.2835,
62
  "step": 90
63
  },
64
  {
65
+ "epoch": 0.03,
66
+ "learning_rate": 1.3698630136986302e-05,
67
+ "loss": 0.0095,
68
  "step": 100
69
  },
70
  {
71
+ "epoch": 0.03,
72
+ "learning_rate": 1.5068493150684931e-05,
73
+ "loss": 0.1436,
74
  "step": 110
75
  },
76
  {
77
+ "epoch": 0.03,
78
+ "learning_rate": 1.643835616438356e-05,
79
+ "loss": 0.0331,
80
  "step": 120
81
  },
82
  {
83
+ "epoch": 0.04,
84
+ "learning_rate": 1.780821917808219e-05,
85
+ "loss": 0.4701,
86
  "step": 130
87
  },
88
  {
89
+ "epoch": 0.04,
90
+ "learning_rate": 1.9178082191780822e-05,
91
+ "loss": 0.1941,
92
  "step": 140
93
  },
94
  {
95
+ "epoch": 0.04,
96
+ "learning_rate": 2.0547945205479453e-05,
97
+ "loss": 0.0103,
98
  "step": 150
99
  },
100
  {
101
+ "epoch": 0.04,
102
+ "learning_rate": 2.1917808219178083e-05,
103
+ "loss": 0.4455,
104
  "step": 160
105
  },
106
  {
107
+ "epoch": 0.05,
108
+ "learning_rate": 2.328767123287671e-05,
109
+ "loss": 0.0121,
110
  "step": 170
111
  },
112
  {
113
+ "epoch": 0.05,
114
+ "learning_rate": 2.4657534246575342e-05,
115
+ "loss": 0.3688,
116
  "step": 180
117
  },
118
  {
119
+ "epoch": 0.05,
120
+ "learning_rate": 2.6027397260273973e-05,
121
+ "loss": 0.1885,
122
  "step": 190
123
  },
124
  {
125
+ "epoch": 0.05,
126
+ "learning_rate": 2.7397260273972603e-05,
127
+ "loss": 0.6733,
128
  "step": 200
129
  },
130
  {
131
+ "epoch": 0.06,
132
+ "learning_rate": 2.8767123287671234e-05,
133
+ "loss": 0.1224,
134
  "step": 210
135
  },
136
  {
137
+ "epoch": 0.06,
138
+ "learning_rate": 3.0136986301369862e-05,
139
+ "loss": 0.9795,
140
  "step": 220
141
  },
142
  {
143
+ "epoch": 0.06,
144
+ "learning_rate": 3.1506849315068496e-05,
145
+ "loss": 0.3799,
146
  "step": 230
147
  },
148
  {
149
+ "epoch": 0.07,
150
+ "learning_rate": 3.287671232876712e-05,
151
+ "loss": 0.7247,
152
  "step": 240
153
  },
154
  {
155
+ "epoch": 0.07,
156
+ "learning_rate": 3.424657534246575e-05,
157
+ "loss": 1.259,
158
  "step": 250
159
  },
160
  {
161
+ "epoch": 0.07,
162
+ "learning_rate": 3.561643835616438e-05,
163
+ "loss": 0.1054,
164
  "step": 260
165
  },
166
  {
167
+ "epoch": 0.07,
168
+ "learning_rate": 3.698630136986301e-05,
169
+ "loss": 0.0743,
170
  "step": 270
171
  },
172
  {
173
+ "epoch": 0.08,
174
+ "learning_rate": 3.8356164383561644e-05,
175
+ "loss": 1.3547,
176
  "step": 280
177
  },
178
  {
179
+ "epoch": 0.08,
180
+ "learning_rate": 3.9726027397260274e-05,
181
+ "loss": 0.7676,
182
  "step": 290
183
  },
184
  {
185
+ "epoch": 0.08,
186
+ "learning_rate": 4.1095890410958905e-05,
187
+ "loss": 0.8231,
188
  "step": 300
189
  },
190
  {
191
+ "epoch": 0.08,
192
+ "learning_rate": 4.2465753424657536e-05,
193
+ "loss": 1.247,
194
  "step": 310
195
  },
196
  {
197
+ "epoch": 0.09,
198
+ "learning_rate": 4.383561643835617e-05,
199
+ "loss": 0.5459,
200
  "step": 320
201
  },
202
  {
203
+ "epoch": 0.09,
204
+ "learning_rate": 4.520547945205479e-05,
205
+ "loss": 0.8819,
206
  "step": 330
207
  },
208
  {
209
+ "epoch": 0.09,
210
+ "learning_rate": 4.657534246575342e-05,
211
+ "loss": 0.664,
212
  "step": 340
213
  },
214
  {
215
+ "epoch": 0.1,
216
+ "learning_rate": 4.794520547945205e-05,
217
+ "loss": 1.572,
218
  "step": 350
219
  },
220
  {
221
+ "epoch": 0.1,
222
+ "learning_rate": 4.9315068493150684e-05,
223
+ "loss": 1.1996,
224
  "step": 360
225
  },
226
  {
227
+ "epoch": 0.1,
228
+ "eval_accuracy": 0.5567010309278351,
229
+ "eval_loss": 2.4460103511810303,
230
+ "eval_runtime": 26.1399,
231
+ "eval_samples_per_second": 3.711,
232
+ "eval_steps_per_second": 3.711,
233
  "step": 365
234
  },
235
  {
236
  "epoch": 1.0,
237
+ "learning_rate": 4.992389649923896e-05,
238
+ "loss": 1.3704,
239
  "step": 370
240
  },
241
  {
242
+ "epoch": 1.0,
243
+ "learning_rate": 4.977168949771689e-05,
244
+ "loss": 0.3425,
245
  "step": 380
246
  },
247
  {
248
+ "epoch": 1.01,
249
+ "learning_rate": 4.961948249619483e-05,
250
+ "loss": 0.9157,
251
  "step": 390
252
  },
253
  {
254
+ "epoch": 1.01,
255
+ "learning_rate": 4.9467275494672755e-05,
256
+ "loss": 1.1445,
257
  "step": 400
258
  },
259
  {
260
+ "epoch": 1.01,
261
+ "learning_rate": 4.9315068493150684e-05,
262
+ "loss": 1.7488,
263
  "step": 410
264
  },
265
  {
266
+ "epoch": 1.02,
267
+ "learning_rate": 4.916286149162861e-05,
268
+ "loss": 0.684,
269
  "step": 420
270
  },
271
  {
272
+ "epoch": 1.02,
273
+ "learning_rate": 4.901065449010655e-05,
274
+ "loss": 0.6746,
275
  "step": 430
276
  },
277
  {
278
+ "epoch": 1.02,
279
+ "learning_rate": 4.8858447488584476e-05,
280
+ "loss": 1.6721,
281
  "step": 440
282
  },
283
  {
284
+ "epoch": 1.02,
285
+ "learning_rate": 4.8706240487062404e-05,
286
+ "loss": 1.8393,
287
  "step": 450
288
  },
289
  {
290
+ "epoch": 1.03,
291
+ "learning_rate": 4.855403348554033e-05,
292
+ "loss": 1.2233,
293
  "step": 460
294
  },
295
  {
296
+ "epoch": 1.03,
297
+ "learning_rate": 4.840182648401827e-05,
298
+ "loss": 1.2178,
299
  "step": 470
300
  },
301
  {
302
+ "epoch": 1.03,
303
+ "learning_rate": 4.8249619482496196e-05,
304
+ "loss": 1.0698,
305
  "step": 480
306
  },
307
  {
308
+ "epoch": 1.03,
309
+ "learning_rate": 4.8097412480974124e-05,
310
+ "loss": 1.1187,
311
  "step": 490
312
  },
313
  {
314
+ "epoch": 1.04,
315
+ "learning_rate": 4.794520547945205e-05,
316
+ "loss": 1.1796,
317
  "step": 500
318
  },
319
  {
320
+ "epoch": 1.04,
321
+ "learning_rate": 4.779299847792998e-05,
322
+ "loss": 1.4509,
323
  "step": 510
324
  },
325
  {
326
+ "epoch": 1.04,
327
+ "learning_rate": 4.7640791476407916e-05,
328
+ "loss": 0.5875,
329
  "step": 520
330
  },
331
  {
332
+ "epoch": 1.05,
333
+ "learning_rate": 4.7488584474885845e-05,
334
+ "loss": 0.7167,
335
  "step": 530
336
  },
337
  {
338
+ "epoch": 1.05,
339
+ "learning_rate": 4.733637747336377e-05,
340
+ "loss": 0.5177,
341
  "step": 540
342
  },
343
  {
344
+ "epoch": 1.05,
345
+ "learning_rate": 4.71841704718417e-05,
346
+ "loss": 1.7653,
347
  "step": 550
348
  },
349
  {
350
+ "epoch": 1.05,
351
+ "learning_rate": 4.703196347031964e-05,
352
+ "loss": 0.693,
353
  "step": 560
354
  },
355
  {
356
+ "epoch": 1.06,
357
+ "learning_rate": 4.6879756468797565e-05,
358
+ "loss": 0.3745,
359
  "step": 570
360
  },
361
  {
362
+ "epoch": 1.06,
363
+ "learning_rate": 4.6727549467275493e-05,
364
+ "loss": 1.3281,
365
  "step": 580
366
  },
367
  {
368
+ "epoch": 1.06,
369
+ "learning_rate": 4.657534246575342e-05,
370
+ "loss": 1.4898,
371
  "step": 590
372
  },
373
  {
374
+ "epoch": 1.06,
375
+ "learning_rate": 4.642313546423136e-05,
376
+ "loss": 1.3792,
377
  "step": 600
378
  },
379
  {
380
+ "epoch": 1.07,
381
+ "learning_rate": 4.6270928462709285e-05,
382
+ "loss": 0.8248,
383
  "step": 610
384
  },
385
  {
386
+ "epoch": 1.07,
387
+ "learning_rate": 4.6118721461187214e-05,
388
+ "loss": 2.1622,
389
  "step": 620
390
  },
391
  {
392
+ "epoch": 1.07,
393
+ "learning_rate": 4.596651445966514e-05,
394
+ "loss": 0.8459,
395
  "step": 630
396
  },
397
  {
398
+ "epoch": 1.08,
399
+ "learning_rate": 4.581430745814308e-05,
400
+ "loss": 0.9529,
401
  "step": 640
402
  },
403
  {
404
+ "epoch": 1.08,
405
+ "learning_rate": 4.5662100456621006e-05,
406
+ "loss": 1.2896,
407
  "step": 650
408
  },
409
  {
410
+ "epoch": 1.08,
411
+ "learning_rate": 4.5509893455098934e-05,
412
+ "loss": 0.8538,
413
  "step": 660
414
  },
415
  {
416
+ "epoch": 1.08,
417
+ "learning_rate": 4.535768645357686e-05,
418
+ "loss": 1.972,
419
  "step": 670
420
  },
421
  {
422
+ "epoch": 1.09,
423
+ "learning_rate": 4.520547945205479e-05,
424
+ "loss": 2.0179,
425
  "step": 680
426
  },
427
  {
428
+ "epoch": 1.09,
429
+ "learning_rate": 4.5053272450532726e-05,
430
+ "loss": 0.9832,
431
  "step": 690
432
  },
433
  {
434
+ "epoch": 1.09,
435
+ "learning_rate": 4.4901065449010655e-05,
436
+ "loss": 0.5654,
437
  "step": 700
438
  },
439
  {
440
+ "epoch": 1.09,
441
+ "learning_rate": 4.474885844748858e-05,
442
+ "loss": 0.8281,
443
  "step": 710
444
  },
445
  {
446
+ "epoch": 1.1,
447
+ "learning_rate": 4.459665144596651e-05,
448
+ "loss": 0.2361,
449
  "step": 720
450
  },
451
  {
452
+ "epoch": 1.1,
453
+ "learning_rate": 4.4444444444444447e-05,
454
+ "loss": 1.4174,
455
  "step": 730
456
  },
457
  {
458
+ "epoch": 1.1,
459
+ "eval_accuracy": 0.8041237113402062,
460
+ "eval_loss": 0.6252075433731079,
461
+ "eval_runtime": 26.3776,
462
+ "eval_samples_per_second": 3.677,
463
+ "eval_steps_per_second": 3.677,
464
  "step": 730
465
  },
466
  {
467
+ "epoch": 2.0,
468
+ "learning_rate": 4.4292237442922375e-05,
469
+ "loss": 0.9256,
470
  "step": 740
471
  },
472
  {
473
  "epoch": 2.01,
474
+ "learning_rate": 4.41400304414003e-05,
475
+ "loss": 0.4378,
476
  "step": 750
477
  },
478
  {
479
+ "epoch": 2.01,
480
+ "learning_rate": 4.398782343987823e-05,
481
+ "loss": 0.4734,
482
  "step": 760
483
  },
484
  {
485
+ "epoch": 2.01,
486
+ "learning_rate": 4.383561643835617e-05,
487
+ "loss": 0.451,
488
  "step": 770
489
  },
490
  {
491
+ "epoch": 2.01,
492
+ "learning_rate": 4.3683409436834095e-05,
493
+ "loss": 0.6936,
494
  "step": 780
495
  },
496
  {
497
+ "epoch": 2.02,
498
+ "learning_rate": 4.3531202435312024e-05,
499
+ "loss": 0.5293,
500
  "step": 790
501
  },
502
  {
503
+ "epoch": 2.02,
504
+ "learning_rate": 4.337899543378995e-05,
505
+ "loss": 0.0223,
506
  "step": 800
507
  },
508
  {
509
+ "epoch": 2.02,
510
+ "learning_rate": 4.322678843226789e-05,
511
+ "loss": 0.1403,
512
  "step": 810
513
  },
514
  {
515
+ "epoch": 2.02,
516
+ "learning_rate": 4.3074581430745816e-05,
517
+ "loss": 1.3285,
518
  "step": 820
519
  },
520
  {
521
+ "epoch": 2.03,
522
+ "learning_rate": 4.2922374429223744e-05,
523
+ "loss": 0.0107,
524
  "step": 830
525
  },
526
  {
527
+ "epoch": 2.03,
528
+ "learning_rate": 4.277016742770167e-05,
529
+ "loss": 0.9825,
530
  "step": 840
531
  },
532
  {
533
+ "epoch": 2.03,
534
+ "learning_rate": 4.26179604261796e-05,
535
+ "loss": 0.4971,
536
  "step": 850
537
  },
538
  {
539
+ "epoch": 2.04,
540
+ "learning_rate": 4.2465753424657536e-05,
541
+ "loss": 0.0056,
542
  "step": 860
543
  },
544
  {
545
+ "epoch": 2.04,
546
+ "learning_rate": 4.2313546423135464e-05,
547
+ "loss": 2.5121,
548
  "step": 870
549
  },
550
  {
551
+ "epoch": 2.04,
552
+ "learning_rate": 4.216133942161339e-05,
553
+ "loss": 1.3604,
554
  "step": 880
555
  },
556
  {
557
+ "epoch": 2.04,
558
+ "learning_rate": 4.200913242009132e-05,
559
+ "loss": 0.1579,
560
  "step": 890
561
  },
562
  {
563
+ "epoch": 2.05,
564
+ "learning_rate": 4.1856925418569256e-05,
565
+ "loss": 0.4522,
566
  "step": 900
567
  },
568
  {
569
+ "epoch": 2.05,
570
+ "learning_rate": 4.1704718417047185e-05,
571
+ "loss": 0.1113,
572
  "step": 910
573
  },
574
  {
575
+ "epoch": 2.05,
576
+ "learning_rate": 4.155251141552511e-05,
577
+ "loss": 0.8716,
578
  "step": 920
579
  },
580
  {
581
+ "epoch": 2.05,
582
+ "learning_rate": 4.140030441400304e-05,
583
+ "loss": 0.5942,
584
  "step": 930
585
  },
586
  {
587
+ "epoch": 2.06,
588
+ "learning_rate": 4.124809741248098e-05,
589
+ "loss": 0.1398,
590
  "step": 940
591
  },
592
  {
593
+ "epoch": 2.06,
594
+ "learning_rate": 4.1095890410958905e-05,
595
+ "loss": 1.299,
596
  "step": 950
597
  },
598
  {
599
+ "epoch": 2.06,
600
+ "learning_rate": 4.0943683409436834e-05,
601
+ "loss": 0.7001,
602
  "step": 960
603
  },
604
  {
605
+ "epoch": 2.07,
606
+ "learning_rate": 4.079147640791476e-05,
607
+ "loss": 1.363,
608
  "step": 970
609
  },
610
  {
611
+ "epoch": 2.07,
612
+ "learning_rate": 4.063926940639269e-05,
613
+ "loss": 0.7667,
614
  "step": 980
615
  },
616
  {
617
+ "epoch": 2.07,
618
+ "learning_rate": 4.0487062404870626e-05,
619
+ "loss": 0.3073,
620
  "step": 990
621
  },
622
  {
623
+ "epoch": 2.07,
624
+ "learning_rate": 4.0334855403348554e-05,
625
+ "loss": 0.069,
626
  "step": 1000
627
  },
628
  {
629
+ "epoch": 2.08,
630
+ "learning_rate": 4.018264840182648e-05,
631
+ "loss": 0.7712,
632
  "step": 1010
633
  },
634
  {
635
+ "epoch": 2.08,
636
+ "learning_rate": 4.003044140030441e-05,
637
+ "loss": 0.8272,
638
  "step": 1020
639
  },
640
  {
641
+ "epoch": 2.08,
642
+ "learning_rate": 3.9878234398782346e-05,
643
+ "loss": 0.6164,
644
  "step": 1030
645
  },
646
  {
647
+ "epoch": 2.08,
648
+ "learning_rate": 3.9726027397260274e-05,
649
+ "loss": 0.0011,
650
  "step": 1040
651
  },
652
  {
653
+ "epoch": 2.09,
654
+ "learning_rate": 3.95738203957382e-05,
655
+ "loss": 0.7969,
656
  "step": 1050
657
  },
658
  {
659
+ "epoch": 2.09,
660
+ "learning_rate": 3.942161339421613e-05,
661
+ "loss": 0.3793,
662
  "step": 1060
663
  },
664
  {
665
+ "epoch": 2.09,
666
+ "learning_rate": 3.9269406392694066e-05,
667
+ "loss": 1.3395,
668
  "step": 1070
669
  },
670
  {
671
+ "epoch": 2.1,
672
+ "learning_rate": 3.9117199391171995e-05,
673
+ "loss": 0.0117,
674
  "step": 1080
675
  },
676
  {
677
+ "epoch": 2.1,
678
+ "learning_rate": 3.896499238964992e-05,
679
+ "loss": 0.0092,
680
  "step": 1090
681
  },
682
  {
683
+ "epoch": 2.1,
684
+ "eval_accuracy": 0.7628865979381443,
685
+ "eval_loss": 1.2255264520645142,
686
+ "eval_runtime": 26.1764,
687
+ "eval_samples_per_second": 3.706,
688
+ "eval_steps_per_second": 3.706,
689
  "step": 1095
690
  },
691
  {
692
  "epoch": 3.0,
693
+ "learning_rate": 3.881278538812785e-05,
694
+ "loss": 0.3452,
695
  "step": 1100
696
  },
697
  {
698
+ "epoch": 3.0,
699
+ "learning_rate": 3.866057838660579e-05,
700
+ "loss": 0.0055,
701
  "step": 1110
702
  },
703
  {
704
+ "epoch": 3.01,
705
+ "learning_rate": 3.8508371385083715e-05,
706
+ "loss": 0.0049,
707
  "step": 1120
708
  },
709
  {
710
+ "epoch": 3.01,
711
+ "learning_rate": 3.8356164383561644e-05,
712
+ "loss": 0.0045,
713
  "step": 1130
714
  },
715
  {
716
+ "epoch": 3.01,
717
+ "learning_rate": 3.820395738203957e-05,
718
+ "loss": 0.266,
719
  "step": 1140
720
  },
721
  {
722
+ "epoch": 3.02,
723
+ "learning_rate": 3.80517503805175e-05,
724
+ "loss": 0.709,
725
  "step": 1150
726
  },
727
  {
728
+ "epoch": 3.02,
729
+ "learning_rate": 3.7899543378995436e-05,
730
+ "loss": 0.0077,
731
  "step": 1160
732
  },
733
  {
734
+ "epoch": 3.02,
735
+ "learning_rate": 3.7747336377473364e-05,
736
+ "loss": 0.0024,
737
  "step": 1170
738
  },
739
  {
740
+ "epoch": 3.02,
741
+ "learning_rate": 3.759512937595129e-05,
742
+ "loss": 0.0016,
743
  "step": 1180
744
  },
745
  {
746
+ "epoch": 3.03,
747
+ "learning_rate": 3.744292237442922e-05,
748
+ "loss": 0.0326,
749
  "step": 1190
750
  },
751
  {
752
+ "epoch": 3.03,
753
+ "learning_rate": 3.7290715372907156e-05,
754
+ "loss": 0.0517,
755
  "step": 1200
756
  },
757
  {
758
+ "epoch": 3.03,
759
+ "learning_rate": 3.7138508371385084e-05,
760
+ "loss": 0.0405,
761
  "step": 1210
762
  },
763
  {
764
+ "epoch": 3.03,
765
+ "learning_rate": 3.698630136986301e-05,
766
+ "loss": 0.7458,
767
  "step": 1220
768
  },
769
  {
770
+ "epoch": 3.04,
771
+ "learning_rate": 3.683409436834094e-05,
772
+ "loss": 0.002,
773
  "step": 1230
774
  },
775
  {
776
+ "epoch": 3.04,
777
+ "learning_rate": 3.6681887366818876e-05,
778
+ "loss": 0.0044,
779
  "step": 1240
780
  },
781
  {
782
+ "epoch": 3.04,
783
+ "learning_rate": 3.6529680365296805e-05,
784
+ "loss": 0.6527,
785
  "step": 1250
786
  },
787
  {
788
+ "epoch": 3.05,
789
+ "learning_rate": 3.637747336377473e-05,
790
+ "loss": 0.1117,
791
  "step": 1260
792
  },
793
  {
794
+ "epoch": 3.05,
795
+ "learning_rate": 3.622526636225266e-05,
796
+ "loss": 1.9262,
797
  "step": 1270
798
  },
799
  {
800
+ "epoch": 3.05,
801
+ "learning_rate": 3.60730593607306e-05,
802
+ "loss": 0.0107,
803
  "step": 1280
804
  },
805
  {
806
+ "epoch": 3.05,
807
+ "learning_rate": 3.5920852359208525e-05,
808
+ "loss": 0.7209,
809
  "step": 1290
810
  },
811
  {
812
+ "epoch": 3.06,
813
+ "learning_rate": 3.5768645357686453e-05,
814
+ "loss": 0.0092,
815
  "step": 1300
816
  },
817
  {
818
+ "epoch": 3.06,
819
+ "learning_rate": 3.561643835616438e-05,
820
+ "loss": 1.4381,
821
  "step": 1310
822
  },
823
  {
824
+ "epoch": 3.06,
825
+ "learning_rate": 3.546423135464231e-05,
826
+ "loss": 0.3603,
827
  "step": 1320
828
  },
829
  {
830
+ "epoch": 3.06,
831
+ "learning_rate": 3.5312024353120245e-05,
832
+ "loss": 0.5943,
833
  "step": 1330
834
  },
835
  {
836
+ "epoch": 3.07,
837
+ "learning_rate": 3.5159817351598174e-05,
838
+ "loss": 0.0023,
839
  "step": 1340
840
  },
841
  {
842
+ "epoch": 3.07,
843
+ "learning_rate": 3.50076103500761e-05,
844
+ "loss": 0.3005,
845
  "step": 1350
846
  },
847
  {
848
+ "epoch": 3.07,
849
+ "learning_rate": 3.485540334855403e-05,
850
+ "loss": 0.1295,
851
  "step": 1360
852
  },
853
  {
854
+ "epoch": 3.08,
855
+ "learning_rate": 3.4703196347031966e-05,
856
+ "loss": 0.0016,
857
  "step": 1370
858
  },
859
  {
860
+ "epoch": 3.08,
861
+ "learning_rate": 3.4550989345509894e-05,
862
+ "loss": 0.002,
863
  "step": 1380
864
  },
865
  {
866
+ "epoch": 3.08,
867
+ "learning_rate": 3.439878234398782e-05,
868
+ "loss": 0.1203,
869
  "step": 1390
870
  },
871
  {
872
+ "epoch": 3.08,
873
+ "learning_rate": 3.424657534246575e-05,
874
+ "loss": 0.1421,
875
  "step": 1400
876
  },
877
  {
878
+ "epoch": 3.09,
879
+ "learning_rate": 3.4094368340943686e-05,
880
+ "loss": 0.0017,
881
  "step": 1410
882
  },
883
  {
884
+ "epoch": 3.09,
885
+ "learning_rate": 3.3942161339421615e-05,
886
+ "loss": 0.0011,
887
  "step": 1420
888
  },
889
  {
890
+ "epoch": 3.09,
891
+ "learning_rate": 3.378995433789954e-05,
892
+ "loss": 1.7634,
893
  "step": 1430
894
  },
895
  {
896
+ "epoch": 3.09,
897
+ "learning_rate": 3.363774733637747e-05,
898
+ "loss": 0.265,
899
  "step": 1440
900
  },
901
  {
902
+ "epoch": 3.1,
903
+ "learning_rate": 3.34855403348554e-05,
904
+ "loss": 0.0013,
905
  "step": 1450
906
  },
907
  {
908
+ "epoch": 3.1,
909
+ "learning_rate": 3.3333333333333335e-05,
910
+ "loss": 0.765,
911
  "step": 1460
912
  },
913
  {
914
+ "epoch": 3.1,
915
+ "eval_accuracy": 0.7731958762886598,
916
+ "eval_loss": 0.9237529039382935,
917
+ "eval_runtime": 26.0231,
918
+ "eval_samples_per_second": 3.727,
919
+ "eval_steps_per_second": 3.727,
920
  "step": 1460
921
  },
922
  {
923
+ "epoch": 4.0,
924
+ "learning_rate": 3.318112633181126e-05,
925
+ "loss": 0.0017,
926
+ "step": 1470
 
 
 
927
  },
928
  {
929
+ "epoch": 4.01,
930
+ "learning_rate": 3.302891933028919e-05,
931
+ "loss": 0.0014,
932
+ "step": 1480
 
 
 
933
  },
934
  {
935
+ "epoch": 4.01,
936
+ "learning_rate": 3.287671232876712e-05,
937
+ "loss": 1.1364,
938
+ "step": 1490
939
+ },
940
+ {
941
+ "epoch": 4.01,
942
+ "learning_rate": 3.2724505327245055e-05,
943
+ "loss": 0.087,
944
+ "step": 1500
945
+ },
946
+ {
947
+ "epoch": 4.01,
948
+ "learning_rate": 3.2572298325722984e-05,
949
+ "loss": 0.5396,
950
+ "step": 1510
951
+ },
952
+ {
953
+ "epoch": 4.02,
954
+ "learning_rate": 3.242009132420091e-05,
955
+ "loss": 0.0016,
956
+ "step": 1520
957
+ },
958
+ {
959
+ "epoch": 4.02,
960
+ "learning_rate": 3.226788432267884e-05,
961
+ "loss": 0.6971,
962
+ "step": 1530
963
+ },
964
+ {
965
+ "epoch": 4.02,
966
+ "learning_rate": 3.2115677321156776e-05,
967
+ "loss": 0.936,
968
+ "step": 1540
969
+ },
970
+ {
971
+ "epoch": 4.02,
972
+ "learning_rate": 3.1963470319634704e-05,
973
+ "loss": 0.0014,
974
+ "step": 1550
975
+ },
976
+ {
977
+ "epoch": 4.03,
978
+ "learning_rate": 3.181126331811263e-05,
979
+ "loss": 0.2137,
980
+ "step": 1560
981
+ },
982
+ {
983
+ "epoch": 4.03,
984
+ "learning_rate": 3.165905631659056e-05,
985
+ "loss": 0.0012,
986
+ "step": 1570
987
+ },
988
+ {
989
+ "epoch": 4.03,
990
+ "learning_rate": 3.1506849315068496e-05,
991
+ "loss": 0.0047,
992
+ "step": 1580
993
+ },
994
+ {
995
+ "epoch": 4.04,
996
+ "learning_rate": 3.1354642313546424e-05,
997
+ "loss": 0.0012,
998
+ "step": 1590
999
+ },
1000
+ {
1001
+ "epoch": 4.04,
1002
+ "learning_rate": 3.120243531202435e-05,
1003
+ "loss": 0.0018,
1004
+ "step": 1600
1005
+ },
1006
+ {
1007
+ "epoch": 4.04,
1008
+ "learning_rate": 3.105022831050228e-05,
1009
+ "loss": 0.0288,
1010
+ "step": 1610
1011
+ },
1012
+ {
1013
+ "epoch": 4.04,
1014
+ "learning_rate": 3.089802130898021e-05,
1015
+ "loss": 0.0017,
1016
+ "step": 1620
1017
+ },
1018
+ {
1019
+ "epoch": 4.05,
1020
+ "learning_rate": 3.0745814307458145e-05,
1021
+ "loss": 0.0012,
1022
+ "step": 1630
1023
+ },
1024
+ {
1025
+ "epoch": 4.05,
1026
+ "learning_rate": 3.059360730593607e-05,
1027
+ "loss": 0.0142,
1028
+ "step": 1640
1029
+ },
1030
+ {
1031
+ "epoch": 4.05,
1032
+ "learning_rate": 3.0441400304414e-05,
1033
+ "loss": 0.1674,
1034
+ "step": 1650
1035
+ },
1036
+ {
1037
+ "epoch": 4.05,
1038
+ "learning_rate": 3.0289193302891933e-05,
1039
+ "loss": 0.4027,
1040
+ "step": 1660
1041
+ },
1042
+ {
1043
+ "epoch": 4.06,
1044
+ "learning_rate": 3.0136986301369862e-05,
1045
+ "loss": 0.0012,
1046
+ "step": 1670
1047
+ },
1048
+ {
1049
+ "epoch": 4.06,
1050
+ "learning_rate": 2.9984779299847794e-05,
1051
+ "loss": 0.0027,
1052
+ "step": 1680
1053
+ },
1054
+ {
1055
+ "epoch": 4.06,
1056
+ "learning_rate": 2.9832572298325722e-05,
1057
+ "loss": 0.0011,
1058
+ "step": 1690
1059
+ },
1060
+ {
1061
+ "epoch": 4.07,
1062
+ "learning_rate": 2.9680365296803654e-05,
1063
+ "loss": 0.001,
1064
+ "step": 1700
1065
+ },
1066
+ {
1067
+ "epoch": 4.07,
1068
+ "learning_rate": 2.9528158295281582e-05,
1069
+ "loss": 0.1668,
1070
+ "step": 1710
1071
+ },
1072
+ {
1073
+ "epoch": 4.07,
1074
+ "learning_rate": 2.9375951293759514e-05,
1075
+ "loss": 0.5638,
1076
+ "step": 1720
1077
+ },
1078
+ {
1079
+ "epoch": 4.07,
1080
+ "learning_rate": 2.9223744292237442e-05,
1081
+ "loss": 0.0009,
1082
+ "step": 1730
1083
+ },
1084
+ {
1085
+ "epoch": 4.08,
1086
+ "learning_rate": 2.9071537290715374e-05,
1087
+ "loss": 1.0001,
1088
+ "step": 1740
1089
+ },
1090
+ {
1091
+ "epoch": 4.08,
1092
+ "learning_rate": 2.8919330289193303e-05,
1093
+ "loss": 0.7653,
1094
+ "step": 1750
1095
+ },
1096
+ {
1097
+ "epoch": 4.08,
1098
+ "learning_rate": 2.8767123287671234e-05,
1099
+ "loss": 0.0047,
1100
+ "step": 1760
1101
+ },
1102
+ {
1103
+ "epoch": 4.08,
1104
+ "learning_rate": 2.8614916286149163e-05,
1105
+ "loss": 0.0018,
1106
+ "step": 1770
1107
+ },
1108
+ {
1109
+ "epoch": 4.09,
1110
+ "learning_rate": 2.846270928462709e-05,
1111
+ "loss": 0.1766,
1112
+ "step": 1780
1113
+ },
1114
+ {
1115
+ "epoch": 4.09,
1116
+ "learning_rate": 2.8310502283105023e-05,
1117
+ "loss": 0.652,
1118
+ "step": 1790
1119
+ },
1120
+ {
1121
+ "epoch": 4.09,
1122
+ "learning_rate": 2.815829528158295e-05,
1123
+ "loss": 0.8817,
1124
+ "step": 1800
1125
+ },
1126
+ {
1127
+ "epoch": 4.1,
1128
+ "learning_rate": 2.8006088280060883e-05,
1129
+ "loss": 0.001,
1130
+ "step": 1810
1131
+ },
1132
+ {
1133
+ "epoch": 4.1,
1134
+ "learning_rate": 2.785388127853881e-05,
1135
+ "loss": 0.1714,
1136
+ "step": 1820
1137
+ },
1138
+ {
1139
+ "epoch": 4.1,
1140
+ "eval_accuracy": 0.7938144329896907,
1141
+ "eval_loss": 1.044911503791809,
1142
+ "eval_runtime": 26.3617,
1143
+ "eval_samples_per_second": 3.68,
1144
+ "eval_steps_per_second": 3.68,
1145
+ "step": 1825
1146
+ },
1147
+ {
1148
+ "epoch": 5.0,
1149
+ "learning_rate": 2.7701674277016743e-05,
1150
+ "loss": 1.0092,
1151
+ "step": 1830
1152
+ },
1153
+ {
1154
+ "epoch": 5.0,
1155
+ "learning_rate": 2.7549467275494672e-05,
1156
+ "loss": 0.2696,
1157
+ "step": 1840
1158
+ },
1159
+ {
1160
+ "epoch": 5.01,
1161
+ "learning_rate": 2.7397260273972603e-05,
1162
+ "loss": 0.0089,
1163
+ "step": 1850
1164
+ },
1165
+ {
1166
+ "epoch": 5.01,
1167
+ "learning_rate": 2.7245053272450532e-05,
1168
+ "loss": 0.8207,
1169
+ "step": 1860
1170
+ },
1171
+ {
1172
+ "epoch": 5.01,
1173
+ "learning_rate": 2.7092846270928464e-05,
1174
+ "loss": 0.0021,
1175
+ "step": 1870
1176
+ },
1177
+ {
1178
+ "epoch": 5.02,
1179
+ "learning_rate": 2.6940639269406392e-05,
1180
+ "loss": 0.5146,
1181
+ "step": 1880
1182
+ },
1183
+ {
1184
+ "epoch": 5.02,
1185
+ "learning_rate": 2.6788432267884324e-05,
1186
+ "loss": 0.0009,
1187
+ "step": 1890
1188
+ },
1189
+ {
1190
+ "epoch": 5.02,
1191
+ "learning_rate": 2.6636225266362252e-05,
1192
+ "loss": 0.1109,
1193
+ "step": 1900
1194
+ },
1195
+ {
1196
+ "epoch": 5.02,
1197
+ "learning_rate": 2.6484018264840184e-05,
1198
+ "loss": 0.0012,
1199
+ "step": 1910
1200
+ },
1201
+ {
1202
+ "epoch": 5.03,
1203
+ "learning_rate": 2.6331811263318112e-05,
1204
+ "loss": 0.0016,
1205
+ "step": 1920
1206
+ },
1207
+ {
1208
+ "epoch": 5.03,
1209
+ "learning_rate": 2.617960426179604e-05,
1210
+ "loss": 0.8104,
1211
+ "step": 1930
1212
+ },
1213
+ {
1214
+ "epoch": 5.03,
1215
+ "learning_rate": 2.6027397260273973e-05,
1216
+ "loss": 1.1465,
1217
+ "step": 1940
1218
+ },
1219
+ {
1220
+ "epoch": 5.03,
1221
+ "learning_rate": 2.58751902587519e-05,
1222
+ "loss": 0.0055,
1223
+ "step": 1950
1224
+ },
1225
+ {
1226
+ "epoch": 5.04,
1227
+ "learning_rate": 2.5722983257229833e-05,
1228
+ "loss": 0.028,
1229
+ "step": 1960
1230
+ },
1231
+ {
1232
+ "epoch": 5.04,
1233
+ "learning_rate": 2.557077625570776e-05,
1234
+ "loss": 0.0615,
1235
+ "step": 1970
1236
+ },
1237
+ {
1238
+ "epoch": 5.04,
1239
+ "learning_rate": 2.5418569254185693e-05,
1240
+ "loss": 0.2375,
1241
+ "step": 1980
1242
+ },
1243
+ {
1244
+ "epoch": 5.05,
1245
+ "learning_rate": 2.526636225266362e-05,
1246
+ "loss": 0.7314,
1247
+ "step": 1990
1248
+ },
1249
+ {
1250
+ "epoch": 5.05,
1251
+ "learning_rate": 2.5114155251141553e-05,
1252
+ "loss": 0.0128,
1253
+ "step": 2000
1254
+ },
1255
+ {
1256
+ "epoch": 5.05,
1257
+ "learning_rate": 2.496194824961948e-05,
1258
+ "loss": 0.7367,
1259
+ "step": 2010
1260
+ },
1261
+ {
1262
+ "epoch": 5.05,
1263
+ "learning_rate": 2.4809741248097413e-05,
1264
+ "loss": 0.001,
1265
+ "step": 2020
1266
+ },
1267
+ {
1268
+ "epoch": 5.06,
1269
+ "learning_rate": 2.4657534246575342e-05,
1270
+ "loss": 0.0012,
1271
+ "step": 2030
1272
+ },
1273
+ {
1274
+ "epoch": 5.06,
1275
+ "learning_rate": 2.4505327245053274e-05,
1276
+ "loss": 0.0032,
1277
+ "step": 2040
1278
+ },
1279
+ {
1280
+ "epoch": 5.06,
1281
+ "learning_rate": 2.4353120243531202e-05,
1282
+ "loss": 0.5217,
1283
+ "step": 2050
1284
+ },
1285
+ {
1286
+ "epoch": 5.06,
1287
+ "learning_rate": 2.4200913242009134e-05,
1288
+ "loss": 0.0011,
1289
+ "step": 2060
1290
+ },
1291
+ {
1292
+ "epoch": 5.07,
1293
+ "learning_rate": 2.4048706240487062e-05,
1294
+ "loss": 0.0039,
1295
+ "step": 2070
1296
+ },
1297
+ {
1298
+ "epoch": 5.07,
1299
+ "learning_rate": 2.389649923896499e-05,
1300
+ "loss": 0.0009,
1301
+ "step": 2080
1302
+ },
1303
+ {
1304
+ "epoch": 5.07,
1305
+ "learning_rate": 2.3744292237442922e-05,
1306
+ "loss": 0.0013,
1307
+ "step": 2090
1308
+ },
1309
+ {
1310
+ "epoch": 5.08,
1311
+ "learning_rate": 2.359208523592085e-05,
1312
+ "loss": 0.0018,
1313
+ "step": 2100
1314
+ },
1315
+ {
1316
+ "epoch": 5.08,
1317
+ "learning_rate": 2.3439878234398783e-05,
1318
+ "loss": 0.0011,
1319
+ "step": 2110
1320
+ },
1321
+ {
1322
+ "epoch": 5.08,
1323
+ "learning_rate": 2.328767123287671e-05,
1324
+ "loss": 0.0262,
1325
+ "step": 2120
1326
+ },
1327
+ {
1328
+ "epoch": 5.08,
1329
+ "learning_rate": 2.3135464231354643e-05,
1330
+ "loss": 0.001,
1331
+ "step": 2130
1332
+ },
1333
+ {
1334
+ "epoch": 5.09,
1335
+ "learning_rate": 2.298325722983257e-05,
1336
+ "loss": 0.0012,
1337
+ "step": 2140
1338
+ },
1339
+ {
1340
+ "epoch": 5.09,
1341
+ "learning_rate": 2.2831050228310503e-05,
1342
+ "loss": 0.0012,
1343
+ "step": 2150
1344
+ },
1345
+ {
1346
+ "epoch": 5.09,
1347
+ "learning_rate": 2.267884322678843e-05,
1348
+ "loss": 0.0112,
1349
+ "step": 2160
1350
+ },
1351
+ {
1352
+ "epoch": 5.09,
1353
+ "learning_rate": 2.2526636225266363e-05,
1354
+ "loss": 0.0593,
1355
+ "step": 2170
1356
+ },
1357
+ {
1358
+ "epoch": 5.1,
1359
+ "learning_rate": 2.237442922374429e-05,
1360
+ "loss": 0.0009,
1361
+ "step": 2180
1362
+ },
1363
+ {
1364
+ "epoch": 5.1,
1365
+ "learning_rate": 2.2222222222222223e-05,
1366
+ "loss": 0.09,
1367
+ "step": 2190
1368
+ },
1369
+ {
1370
+ "epoch": 5.1,
1371
+ "eval_accuracy": 0.7628865979381443,
1372
+ "eval_loss": 1.474695086479187,
1373
+ "eval_runtime": 27.4624,
1374
+ "eval_samples_per_second": 3.532,
1375
+ "eval_steps_per_second": 3.532,
1376
+ "step": 2190
1377
+ },
1378
+ {
1379
+ "epoch": 6.0,
1380
+ "learning_rate": 2.207001522070015e-05,
1381
+ "loss": 0.0282,
1382
+ "step": 2200
1383
+ },
1384
+ {
1385
+ "epoch": 6.01,
1386
+ "learning_rate": 2.1917808219178083e-05,
1387
+ "loss": 0.001,
1388
+ "step": 2210
1389
+ },
1390
+ {
1391
+ "epoch": 6.01,
1392
+ "learning_rate": 2.1765601217656012e-05,
1393
+ "loss": 0.1037,
1394
+ "step": 2220
1395
+ },
1396
+ {
1397
+ "epoch": 6.01,
1398
+ "learning_rate": 2.1613394216133944e-05,
1399
+ "loss": 0.0009,
1400
+ "step": 2230
1401
+ },
1402
+ {
1403
+ "epoch": 6.01,
1404
+ "learning_rate": 2.1461187214611872e-05,
1405
+ "loss": 0.0006,
1406
+ "step": 2240
1407
+ },
1408
+ {
1409
+ "epoch": 6.02,
1410
+ "learning_rate": 2.13089802130898e-05,
1411
+ "loss": 0.0007,
1412
+ "step": 2250
1413
+ },
1414
+ {
1415
+ "epoch": 6.02,
1416
+ "learning_rate": 2.1156773211567732e-05,
1417
+ "loss": 0.0007,
1418
+ "step": 2260
1419
+ },
1420
+ {
1421
+ "epoch": 6.02,
1422
+ "learning_rate": 2.100456621004566e-05,
1423
+ "loss": 0.0007,
1424
+ "step": 2270
1425
+ },
1426
+ {
1427
+ "epoch": 6.02,
1428
+ "learning_rate": 2.0852359208523592e-05,
1429
+ "loss": 0.0007,
1430
+ "step": 2280
1431
+ },
1432
+ {
1433
+ "epoch": 6.03,
1434
+ "learning_rate": 2.070015220700152e-05,
1435
+ "loss": 0.0007,
1436
+ "step": 2290
1437
+ },
1438
+ {
1439
+ "epoch": 6.03,
1440
+ "learning_rate": 2.0547945205479453e-05,
1441
+ "loss": 0.0007,
1442
+ "step": 2300
1443
+ },
1444
+ {
1445
+ "epoch": 6.03,
1446
+ "learning_rate": 2.039573820395738e-05,
1447
+ "loss": 0.0031,
1448
+ "step": 2310
1449
+ },
1450
+ {
1451
+ "epoch": 6.04,
1452
+ "learning_rate": 2.0243531202435313e-05,
1453
+ "loss": 0.2555,
1454
+ "step": 2320
1455
+ },
1456
+ {
1457
+ "epoch": 6.04,
1458
+ "learning_rate": 2.009132420091324e-05,
1459
+ "loss": 0.0008,
1460
+ "step": 2330
1461
+ },
1462
+ {
1463
+ "epoch": 6.04,
1464
+ "learning_rate": 1.9939117199391173e-05,
1465
+ "loss": 0.0008,
1466
+ "step": 2340
1467
+ },
1468
+ {
1469
+ "epoch": 6.04,
1470
+ "learning_rate": 1.97869101978691e-05,
1471
+ "loss": 0.0006,
1472
+ "step": 2350
1473
+ },
1474
+ {
1475
+ "epoch": 6.05,
1476
+ "learning_rate": 1.9634703196347033e-05,
1477
+ "loss": 0.0009,
1478
+ "step": 2360
1479
+ },
1480
+ {
1481
+ "epoch": 6.05,
1482
+ "learning_rate": 1.948249619482496e-05,
1483
+ "loss": 0.0385,
1484
+ "step": 2370
1485
+ },
1486
+ {
1487
+ "epoch": 6.05,
1488
+ "learning_rate": 1.9330289193302893e-05,
1489
+ "loss": 0.0008,
1490
+ "step": 2380
1491
+ },
1492
+ {
1493
+ "epoch": 6.05,
1494
+ "learning_rate": 1.9178082191780822e-05,
1495
+ "loss": 0.0006,
1496
+ "step": 2390
1497
+ },
1498
+ {
1499
+ "epoch": 6.06,
1500
+ "learning_rate": 1.902587519025875e-05,
1501
+ "loss": 0.0009,
1502
+ "step": 2400
1503
+ },
1504
+ {
1505
+ "epoch": 6.06,
1506
+ "learning_rate": 1.8873668188736682e-05,
1507
+ "loss": 0.0006,
1508
+ "step": 2410
1509
+ },
1510
+ {
1511
+ "epoch": 6.06,
1512
+ "learning_rate": 1.872146118721461e-05,
1513
+ "loss": 0.0008,
1514
+ "step": 2420
1515
+ },
1516
+ {
1517
+ "epoch": 6.07,
1518
+ "learning_rate": 1.8569254185692542e-05,
1519
+ "loss": 0.0008,
1520
+ "step": 2430
1521
+ },
1522
+ {
1523
+ "epoch": 6.07,
1524
+ "learning_rate": 1.841704718417047e-05,
1525
+ "loss": 0.0007,
1526
+ "step": 2440
1527
+ },
1528
+ {
1529
+ "epoch": 6.07,
1530
+ "learning_rate": 1.8264840182648402e-05,
1531
+ "loss": 0.0008,
1532
+ "step": 2450
1533
+ },
1534
+ {
1535
+ "epoch": 6.07,
1536
+ "learning_rate": 1.811263318112633e-05,
1537
+ "loss": 0.0008,
1538
+ "step": 2460
1539
+ },
1540
+ {
1541
+ "epoch": 6.08,
1542
+ "learning_rate": 1.7960426179604263e-05,
1543
+ "loss": 0.0006,
1544
+ "step": 2470
1545
+ },
1546
+ {
1547
+ "epoch": 6.08,
1548
+ "learning_rate": 1.780821917808219e-05,
1549
+ "loss": 0.0006,
1550
+ "step": 2480
1551
+ },
1552
+ {
1553
+ "epoch": 6.08,
1554
+ "learning_rate": 1.7656012176560123e-05,
1555
+ "loss": 0.1391,
1556
+ "step": 2490
1557
+ },
1558
+ {
1559
+ "epoch": 6.08,
1560
+ "learning_rate": 1.750380517503805e-05,
1561
+ "loss": 0.0006,
1562
+ "step": 2500
1563
+ },
1564
+ {
1565
+ "epoch": 6.09,
1566
+ "learning_rate": 1.7351598173515983e-05,
1567
+ "loss": 0.0005,
1568
+ "step": 2510
1569
+ },
1570
+ {
1571
+ "epoch": 6.09,
1572
+ "learning_rate": 1.719939117199391e-05,
1573
+ "loss": 0.0005,
1574
+ "step": 2520
1575
+ },
1576
+ {
1577
+ "epoch": 6.09,
1578
+ "learning_rate": 1.7047184170471843e-05,
1579
+ "loss": 0.5689,
1580
+ "step": 2530
1581
+ },
1582
+ {
1583
+ "epoch": 6.1,
1584
+ "learning_rate": 1.689497716894977e-05,
1585
+ "loss": 0.0007,
1586
+ "step": 2540
1587
+ },
1588
+ {
1589
+ "epoch": 6.1,
1590
+ "learning_rate": 1.67427701674277e-05,
1591
+ "loss": 0.0006,
1592
+ "step": 2550
1593
+ },
1594
+ {
1595
+ "epoch": 6.1,
1596
+ "eval_accuracy": 0.7835051546391752,
1597
+ "eval_loss": 1.3157318830490112,
1598
+ "eval_runtime": 27.9871,
1599
+ "eval_samples_per_second": 3.466,
1600
+ "eval_steps_per_second": 3.466,
1601
+ "step": 2555
1602
+ },
1603
+ {
1604
+ "epoch": 7.0,
1605
+ "learning_rate": 1.659056316590563e-05,
1606
+ "loss": 0.0005,
1607
+ "step": 2560
1608
+ },
1609
+ {
1610
+ "epoch": 7.0,
1611
+ "learning_rate": 1.643835616438356e-05,
1612
+ "loss": 0.0006,
1613
+ "step": 2570
1614
+ },
1615
+ {
1616
+ "epoch": 7.01,
1617
+ "learning_rate": 1.6286149162861492e-05,
1618
+ "loss": 0.0006,
1619
+ "step": 2580
1620
+ },
1621
+ {
1622
+ "epoch": 7.01,
1623
+ "learning_rate": 1.613394216133942e-05,
1624
+ "loss": 0.0055,
1625
+ "step": 2590
1626
+ },
1627
+ {
1628
+ "epoch": 7.01,
1629
+ "learning_rate": 1.5981735159817352e-05,
1630
+ "loss": 0.0007,
1631
+ "step": 2600
1632
+ },
1633
+ {
1634
+ "epoch": 7.02,
1635
+ "learning_rate": 1.582952815829528e-05,
1636
+ "loss": 0.0007,
1637
+ "step": 2610
1638
+ },
1639
+ {
1640
+ "epoch": 7.02,
1641
+ "learning_rate": 1.5677321156773212e-05,
1642
+ "loss": 0.0008,
1643
+ "step": 2620
1644
+ },
1645
+ {
1646
+ "epoch": 7.02,
1647
+ "learning_rate": 1.552511415525114e-05,
1648
+ "loss": 0.0008,
1649
+ "step": 2630
1650
+ },
1651
+ {
1652
+ "epoch": 7.02,
1653
+ "learning_rate": 1.5372907153729072e-05,
1654
+ "loss": 0.0007,
1655
+ "step": 2640
1656
+ },
1657
+ {
1658
+ "epoch": 7.03,
1659
+ "learning_rate": 1.5220700152207e-05,
1660
+ "loss": 0.0006,
1661
+ "step": 2650
1662
+ },
1663
+ {
1664
+ "epoch": 7.03,
1665
+ "learning_rate": 1.5068493150684931e-05,
1666
+ "loss": 0.0008,
1667
+ "step": 2660
1668
+ },
1669
+ {
1670
+ "epoch": 7.03,
1671
+ "learning_rate": 1.4916286149162861e-05,
1672
+ "loss": 0.0007,
1673
+ "step": 2670
1674
+ },
1675
+ {
1676
+ "epoch": 7.03,
1677
+ "learning_rate": 1.4764079147640791e-05,
1678
+ "loss": 0.0132,
1679
+ "step": 2680
1680
+ },
1681
+ {
1682
+ "epoch": 7.04,
1683
+ "learning_rate": 1.4611872146118721e-05,
1684
+ "loss": 0.0017,
1685
+ "step": 2690
1686
+ },
1687
+ {
1688
+ "epoch": 7.04,
1689
+ "learning_rate": 1.4459665144596651e-05,
1690
+ "loss": 0.0006,
1691
+ "step": 2700
1692
+ },
1693
+ {
1694
+ "epoch": 7.04,
1695
+ "learning_rate": 1.4307458143074581e-05,
1696
+ "loss": 0.0007,
1697
+ "step": 2710
1698
+ },
1699
+ {
1700
+ "epoch": 7.05,
1701
+ "learning_rate": 1.4155251141552511e-05,
1702
+ "loss": 0.0005,
1703
+ "step": 2720
1704
+ },
1705
+ {
1706
+ "epoch": 7.05,
1707
+ "learning_rate": 1.4003044140030442e-05,
1708
+ "loss": 0.8613,
1709
+ "step": 2730
1710
+ },
1711
+ {
1712
+ "epoch": 7.05,
1713
+ "learning_rate": 1.3850837138508372e-05,
1714
+ "loss": 0.0008,
1715
+ "step": 2740
1716
+ },
1717
+ {
1718
+ "epoch": 7.05,
1719
+ "learning_rate": 1.3698630136986302e-05,
1720
+ "loss": 0.0008,
1721
+ "step": 2750
1722
+ },
1723
+ {
1724
+ "epoch": 7.06,
1725
+ "learning_rate": 1.3546423135464232e-05,
1726
+ "loss": 0.0007,
1727
+ "step": 2760
1728
+ },
1729
+ {
1730
+ "epoch": 7.06,
1731
+ "learning_rate": 1.3394216133942162e-05,
1732
+ "loss": 0.0004,
1733
+ "step": 2770
1734
+ },
1735
+ {
1736
+ "epoch": 7.06,
1737
+ "learning_rate": 1.3242009132420092e-05,
1738
+ "loss": 0.0005,
1739
+ "step": 2780
1740
+ },
1741
+ {
1742
+ "epoch": 7.06,
1743
+ "learning_rate": 1.308980213089802e-05,
1744
+ "loss": 0.0006,
1745
+ "step": 2790
1746
+ },
1747
+ {
1748
+ "epoch": 7.07,
1749
+ "learning_rate": 1.293759512937595e-05,
1750
+ "loss": 0.0007,
1751
+ "step": 2800
1752
+ },
1753
+ {
1754
+ "epoch": 7.07,
1755
+ "learning_rate": 1.278538812785388e-05,
1756
+ "loss": 0.008,
1757
+ "step": 2810
1758
+ },
1759
+ {
1760
+ "epoch": 7.07,
1761
+ "learning_rate": 1.263318112633181e-05,
1762
+ "loss": 0.0005,
1763
+ "step": 2820
1764
+ },
1765
+ {
1766
+ "epoch": 7.08,
1767
+ "learning_rate": 1.248097412480974e-05,
1768
+ "loss": 0.7276,
1769
+ "step": 2830
1770
+ },
1771
+ {
1772
+ "epoch": 7.08,
1773
+ "learning_rate": 1.2328767123287671e-05,
1774
+ "loss": 0.0005,
1775
+ "step": 2840
1776
+ },
1777
+ {
1778
+ "epoch": 7.08,
1779
+ "learning_rate": 1.2176560121765601e-05,
1780
+ "loss": 0.0009,
1781
+ "step": 2850
1782
+ },
1783
+ {
1784
+ "epoch": 7.08,
1785
+ "learning_rate": 1.2024353120243531e-05,
1786
+ "loss": 0.0005,
1787
+ "step": 2860
1788
+ },
1789
+ {
1790
+ "epoch": 7.09,
1791
+ "learning_rate": 1.1872146118721461e-05,
1792
+ "loss": 0.0006,
1793
+ "step": 2870
1794
+ },
1795
+ {
1796
+ "epoch": 7.09,
1797
+ "learning_rate": 1.1719939117199391e-05,
1798
+ "loss": 0.009,
1799
+ "step": 2880
1800
+ },
1801
+ {
1802
+ "epoch": 7.09,
1803
+ "learning_rate": 1.1567732115677321e-05,
1804
+ "loss": 0.0008,
1805
+ "step": 2890
1806
+ },
1807
+ {
1808
+ "epoch": 7.09,
1809
+ "learning_rate": 1.1415525114155251e-05,
1810
+ "loss": 0.0587,
1811
+ "step": 2900
1812
+ },
1813
+ {
1814
+ "epoch": 7.1,
1815
+ "learning_rate": 1.1263318112633182e-05,
1816
+ "loss": 0.0006,
1817
+ "step": 2910
1818
+ },
1819
+ {
1820
+ "epoch": 7.1,
1821
+ "learning_rate": 1.1111111111111112e-05,
1822
+ "loss": 0.0004,
1823
+ "step": 2920
1824
+ },
1825
+ {
1826
+ "epoch": 7.1,
1827
+ "eval_accuracy": 0.7216494845360825,
1828
+ "eval_loss": 1.6183393001556396,
1829
+ "eval_runtime": 27.6439,
1830
+ "eval_samples_per_second": 3.509,
1831
+ "eval_steps_per_second": 3.509,
1832
+ "step": 2920
1833
+ },
1834
+ {
1835
+ "epoch": 8.0,
1836
+ "learning_rate": 1.0958904109589042e-05,
1837
+ "loss": 0.0006,
1838
+ "step": 2930
1839
+ },
1840
+ {
1841
+ "epoch": 8.01,
1842
+ "learning_rate": 1.0806697108066972e-05,
1843
+ "loss": 0.0016,
1844
+ "step": 2940
1845
+ },
1846
+ {
1847
+ "epoch": 8.01,
1848
+ "learning_rate": 1.06544901065449e-05,
1849
+ "loss": 0.0007,
1850
+ "step": 2950
1851
+ },
1852
+ {
1853
+ "epoch": 8.01,
1854
+ "learning_rate": 1.050228310502283e-05,
1855
+ "loss": 0.8142,
1856
+ "step": 2960
1857
+ },
1858
+ {
1859
+ "epoch": 8.01,
1860
+ "learning_rate": 1.035007610350076e-05,
1861
+ "loss": 0.0007,
1862
+ "step": 2970
1863
+ },
1864
+ {
1865
+ "epoch": 8.02,
1866
+ "learning_rate": 1.019786910197869e-05,
1867
+ "loss": 0.0007,
1868
+ "step": 2980
1869
+ },
1870
+ {
1871
+ "epoch": 8.02,
1872
+ "learning_rate": 1.004566210045662e-05,
1873
+ "loss": 0.0006,
1874
+ "step": 2990
1875
+ },
1876
+ {
1877
+ "epoch": 8.02,
1878
+ "learning_rate": 9.89345509893455e-06,
1879
+ "loss": 0.0007,
1880
+ "step": 3000
1881
+ },
1882
+ {
1883
+ "epoch": 8.02,
1884
+ "learning_rate": 9.74124809741248e-06,
1885
+ "loss": 0.0006,
1886
+ "step": 3010
1887
+ },
1888
+ {
1889
+ "epoch": 8.03,
1890
+ "learning_rate": 9.589041095890411e-06,
1891
+ "loss": 0.0006,
1892
+ "step": 3020
1893
+ },
1894
+ {
1895
+ "epoch": 8.03,
1896
+ "learning_rate": 9.436834094368341e-06,
1897
+ "loss": 0.0005,
1898
+ "step": 3030
1899
+ },
1900
+ {
1901
+ "epoch": 8.03,
1902
+ "learning_rate": 9.284627092846271e-06,
1903
+ "loss": 0.0424,
1904
+ "step": 3040
1905
+ },
1906
+ {
1907
+ "epoch": 8.04,
1908
+ "learning_rate": 9.132420091324201e-06,
1909
+ "loss": 0.0006,
1910
+ "step": 3050
1911
+ },
1912
+ {
1913
+ "epoch": 8.04,
1914
+ "learning_rate": 8.980213089802131e-06,
1915
+ "loss": 0.001,
1916
+ "step": 3060
1917
+ },
1918
+ {
1919
+ "epoch": 8.04,
1920
+ "learning_rate": 8.828006088280061e-06,
1921
+ "loss": 0.0006,
1922
+ "step": 3070
1923
+ },
1924
+ {
1925
+ "epoch": 8.04,
1926
+ "learning_rate": 8.675799086757991e-06,
1927
+ "loss": 0.0014,
1928
+ "step": 3080
1929
+ },
1930
+ {
1931
+ "epoch": 8.05,
1932
+ "learning_rate": 8.523592085235922e-06,
1933
+ "loss": 0.0005,
1934
+ "step": 3090
1935
+ },
1936
+ {
1937
+ "epoch": 8.05,
1938
+ "learning_rate": 8.37138508371385e-06,
1939
+ "loss": 0.0006,
1940
+ "step": 3100
1941
+ },
1942
+ {
1943
+ "epoch": 8.05,
1944
+ "learning_rate": 8.21917808219178e-06,
1945
+ "loss": 0.0006,
1946
+ "step": 3110
1947
+ },
1948
+ {
1949
+ "epoch": 8.05,
1950
+ "learning_rate": 8.06697108066971e-06,
1951
+ "loss": 0.0017,
1952
+ "step": 3120
1953
+ },
1954
+ {
1955
+ "epoch": 8.06,
1956
+ "learning_rate": 7.91476407914764e-06,
1957
+ "loss": 0.0005,
1958
+ "step": 3130
1959
+ },
1960
+ {
1961
+ "epoch": 8.06,
1962
+ "learning_rate": 7.76255707762557e-06,
1963
+ "loss": 0.0007,
1964
+ "step": 3140
1965
+ },
1966
+ {
1967
+ "epoch": 8.06,
1968
+ "learning_rate": 7.6103500761035e-06,
1969
+ "loss": 0.0005,
1970
+ "step": 3150
1971
+ },
1972
+ {
1973
+ "epoch": 8.07,
1974
+ "learning_rate": 7.4581430745814305e-06,
1975
+ "loss": 0.0005,
1976
+ "step": 3160
1977
+ },
1978
+ {
1979
+ "epoch": 8.07,
1980
+ "learning_rate": 7.305936073059361e-06,
1981
+ "loss": 0.0005,
1982
+ "step": 3170
1983
+ },
1984
+ {
1985
+ "epoch": 8.07,
1986
+ "learning_rate": 7.153729071537291e-06,
1987
+ "loss": 0.0006,
1988
+ "step": 3180
1989
+ },
1990
+ {
1991
+ "epoch": 8.07,
1992
+ "learning_rate": 7.001522070015221e-06,
1993
+ "loss": 0.0005,
1994
+ "step": 3190
1995
+ },
1996
+ {
1997
+ "epoch": 8.08,
1998
+ "learning_rate": 6.849315068493151e-06,
1999
+ "loss": 0.0005,
2000
+ "step": 3200
2001
+ },
2002
+ {
2003
+ "epoch": 8.08,
2004
+ "learning_rate": 6.697108066971081e-06,
2005
+ "loss": 0.0005,
2006
+ "step": 3210
2007
+ },
2008
+ {
2009
+ "epoch": 8.08,
2010
+ "learning_rate": 6.54490106544901e-06,
2011
+ "loss": 0.004,
2012
+ "step": 3220
2013
+ },
2014
+ {
2015
+ "epoch": 8.08,
2016
+ "learning_rate": 6.39269406392694e-06,
2017
+ "loss": 0.0006,
2018
+ "step": 3230
2019
+ },
2020
+ {
2021
+ "epoch": 8.09,
2022
+ "learning_rate": 6.24048706240487e-06,
2023
+ "loss": 0.0007,
2024
+ "step": 3240
2025
+ },
2026
+ {
2027
+ "epoch": 8.09,
2028
+ "learning_rate": 6.0882800608828005e-06,
2029
+ "loss": 0.0016,
2030
+ "step": 3250
2031
+ },
2032
+ {
2033
+ "epoch": 8.09,
2034
+ "learning_rate": 5.936073059360731e-06,
2035
+ "loss": 0.0005,
2036
+ "step": 3260
2037
+ },
2038
+ {
2039
+ "epoch": 8.1,
2040
+ "learning_rate": 5.783866057838661e-06,
2041
+ "loss": 0.0006,
2042
+ "step": 3270
2043
+ },
2044
+ {
2045
+ "epoch": 8.1,
2046
+ "learning_rate": 5.631659056316591e-06,
2047
+ "loss": 0.0012,
2048
+ "step": 3280
2049
+ },
2050
+ {
2051
+ "epoch": 8.1,
2052
+ "eval_accuracy": 0.7731958762886598,
2053
+ "eval_loss": 1.5373260974884033,
2054
+ "eval_runtime": 28.4296,
2055
+ "eval_samples_per_second": 3.412,
2056
+ "eval_steps_per_second": 3.412,
2057
+ "step": 3285
2058
+ },
2059
+ {
2060
+ "epoch": 9.0,
2061
+ "learning_rate": 5.479452054794521e-06,
2062
+ "loss": 0.0005,
2063
+ "step": 3290
2064
+ },
2065
+ {
2066
+ "epoch": 9.0,
2067
+ "learning_rate": 5.32724505327245e-06,
2068
+ "loss": 0.0005,
2069
+ "step": 3300
2070
+ },
2071
+ {
2072
+ "epoch": 9.01,
2073
+ "learning_rate": 5.17503805175038e-06,
2074
+ "loss": 0.0013,
2075
+ "step": 3310
2076
+ },
2077
+ {
2078
+ "epoch": 9.01,
2079
+ "learning_rate": 5.02283105022831e-06,
2080
+ "loss": 0.0005,
2081
+ "step": 3320
2082
+ },
2083
+ {
2084
+ "epoch": 9.01,
2085
+ "learning_rate": 4.87062404870624e-06,
2086
+ "loss": 0.0006,
2087
+ "step": 3330
2088
+ },
2089
+ {
2090
+ "epoch": 9.02,
2091
+ "learning_rate": 4.7184170471841705e-06,
2092
+ "loss": 0.0005,
2093
+ "step": 3340
2094
+ },
2095
+ {
2096
+ "epoch": 9.02,
2097
+ "learning_rate": 4.566210045662101e-06,
2098
+ "loss": 0.0005,
2099
+ "step": 3350
2100
+ },
2101
+ {
2102
+ "epoch": 9.02,
2103
+ "learning_rate": 4.414003044140031e-06,
2104
+ "loss": 0.0004,
2105
+ "step": 3360
2106
+ },
2107
+ {
2108
+ "epoch": 9.02,
2109
+ "learning_rate": 4.261796042617961e-06,
2110
+ "loss": 0.0008,
2111
+ "step": 3370
2112
+ },
2113
+ {
2114
+ "epoch": 9.03,
2115
+ "learning_rate": 4.10958904109589e-06,
2116
+ "loss": 0.0005,
2117
+ "step": 3380
2118
+ },
2119
+ {
2120
+ "epoch": 9.03,
2121
+ "learning_rate": 3.95738203957382e-06,
2122
+ "loss": 0.0005,
2123
+ "step": 3390
2124
+ },
2125
+ {
2126
+ "epoch": 9.03,
2127
+ "learning_rate": 3.80517503805175e-06,
2128
+ "loss": 0.0007,
2129
+ "step": 3400
2130
+ },
2131
+ {
2132
+ "epoch": 9.03,
2133
+ "learning_rate": 3.6529680365296803e-06,
2134
+ "loss": 0.0008,
2135
+ "step": 3410
2136
+ },
2137
+ {
2138
+ "epoch": 9.04,
2139
+ "learning_rate": 3.5007610350076104e-06,
2140
+ "loss": 0.0005,
2141
+ "step": 3420
2142
+ },
2143
+ {
2144
+ "epoch": 9.04,
2145
+ "learning_rate": 3.3485540334855405e-06,
2146
+ "loss": 0.0005,
2147
+ "step": 3430
2148
+ },
2149
+ {
2150
+ "epoch": 9.04,
2151
+ "learning_rate": 3.19634703196347e-06,
2152
+ "loss": 0.0007,
2153
+ "step": 3440
2154
+ },
2155
+ {
2156
+ "epoch": 9.05,
2157
+ "learning_rate": 3.0441400304414002e-06,
2158
+ "loss": 0.0006,
2159
+ "step": 3450
2160
+ },
2161
+ {
2162
+ "epoch": 9.05,
2163
+ "learning_rate": 2.8919330289193303e-06,
2164
+ "loss": 0.0004,
2165
+ "step": 3460
2166
+ },
2167
+ {
2168
+ "epoch": 9.05,
2169
+ "learning_rate": 2.7397260273972604e-06,
2170
+ "loss": 0.0005,
2171
+ "step": 3470
2172
+ },
2173
+ {
2174
+ "epoch": 9.05,
2175
+ "learning_rate": 2.58751902587519e-06,
2176
+ "loss": 0.0004,
2177
+ "step": 3480
2178
+ },
2179
+ {
2180
+ "epoch": 9.06,
2181
+ "learning_rate": 2.43531202435312e-06,
2182
+ "loss": 0.0005,
2183
+ "step": 3490
2184
+ },
2185
+ {
2186
+ "epoch": 9.06,
2187
+ "learning_rate": 2.2831050228310503e-06,
2188
+ "loss": 0.0006,
2189
+ "step": 3500
2190
+ },
2191
+ {
2192
+ "epoch": 9.06,
2193
+ "learning_rate": 2.1308980213089804e-06,
2194
+ "loss": 0.0006,
2195
+ "step": 3510
2196
+ },
2197
+ {
2198
+ "epoch": 9.06,
2199
+ "learning_rate": 1.97869101978691e-06,
2200
+ "loss": 0.0682,
2201
+ "step": 3520
2202
+ },
2203
+ {
2204
+ "epoch": 9.07,
2205
+ "learning_rate": 1.8264840182648401e-06,
2206
+ "loss": 0.0004,
2207
+ "step": 3530
2208
+ },
2209
+ {
2210
+ "epoch": 9.07,
2211
+ "learning_rate": 1.6742770167427702e-06,
2212
+ "loss": 0.0005,
2213
+ "step": 3540
2214
+ },
2215
+ {
2216
+ "epoch": 9.07,
2217
+ "learning_rate": 1.5220700152207001e-06,
2218
+ "loss": 0.0005,
2219
+ "step": 3550
2220
+ },
2221
+ {
2222
+ "epoch": 9.08,
2223
+ "learning_rate": 1.3698630136986302e-06,
2224
+ "loss": 0.0006,
2225
+ "step": 3560
2226
+ },
2227
+ {
2228
+ "epoch": 9.08,
2229
+ "learning_rate": 1.21765601217656e-06,
2230
+ "loss": 0.0007,
2231
+ "step": 3570
2232
+ },
2233
+ {
2234
+ "epoch": 9.08,
2235
+ "learning_rate": 1.0654490106544902e-06,
2236
+ "loss": 0.0005,
2237
+ "step": 3580
2238
+ },
2239
+ {
2240
+ "epoch": 9.08,
2241
+ "learning_rate": 9.132420091324201e-07,
2242
+ "loss": 0.0005,
2243
+ "step": 3590
2244
+ },
2245
+ {
2246
+ "epoch": 9.09,
2247
+ "learning_rate": 7.610350076103501e-07,
2248
+ "loss": 0.7116,
2249
+ "step": 3600
2250
+ },
2251
+ {
2252
+ "epoch": 9.09,
2253
+ "learning_rate": 6.0882800608828e-07,
2254
+ "loss": 0.0694,
2255
+ "step": 3610
2256
+ },
2257
+ {
2258
+ "epoch": 9.09,
2259
+ "learning_rate": 4.5662100456621004e-07,
2260
+ "loss": 0.0006,
2261
+ "step": 3620
2262
+ },
2263
+ {
2264
+ "epoch": 9.09,
2265
+ "learning_rate": 3.0441400304414e-07,
2266
+ "loss": 0.0006,
2267
+ "step": 3630
2268
+ },
2269
+ {
2270
+ "epoch": 9.1,
2271
+ "learning_rate": 1.5220700152207e-07,
2272
+ "loss": 0.0004,
2273
+ "step": 3640
2274
+ },
2275
+ {
2276
+ "epoch": 9.1,
2277
+ "learning_rate": 0.0,
2278
+ "loss": 0.0005,
2279
+ "step": 3650
2280
+ },
2281
+ {
2282
+ "epoch": 9.1,
2283
+ "eval_accuracy": 0.7628865979381443,
2284
+ "eval_loss": 1.67752206325531,
2285
+ "eval_runtime": 28.4688,
2286
+ "eval_samples_per_second": 3.407,
2287
+ "eval_steps_per_second": 3.407,
2288
+ "step": 3650
2289
+ },
2290
+ {
2291
+ "epoch": 9.1,
2292
+ "step": 3650,
2293
+ "total_flos": 4.5485413607079936e+18,
2294
+ "train_loss": 0.30888694136824507,
2295
+ "train_runtime": 2930.7636,
2296
+ "train_samples_per_second": 1.245,
2297
+ "train_steps_per_second": 1.245
2298
+ },
2299
+ {
2300
+ "epoch": 9.1,
2301
+ "eval_accuracy": 0.7897727272727273,
2302
+ "eval_loss": 0.6162788271903992,
2303
+ "eval_runtime": 56.0111,
2304
+ "eval_samples_per_second": 3.142,
2305
+ "eval_steps_per_second": 3.142,
2306
+ "step": 3650
2307
+ },
2308
+ {
2309
+ "epoch": 9.1,
2310
+ "eval_accuracy": 0.7885714285714286,
2311
+ "eval_loss": 0.6197737455368042,
2312
+ "eval_runtime": 54.239,
2313
+ "eval_samples_per_second": 3.226,
2314
+ "eval_steps_per_second": 3.226,
2315
+ "step": 3650
2316
  }
2317
  ],
2318
+ "max_steps": 3650,
2319
  "num_train_epochs": 9223372036854775807,
2320
+ "total_flos": 4.5485413607079936e+18,
2321
  "trial_name": null,
2322
  "trial_params": null
2323
  }