AkshatSurolia commited on
Commit
71aaa02
1 Parent(s): 07c3608

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +646 -0
trainer_state.json ADDED
@@ -0,0 +1,646 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.027106858789920807,
3
+ "best_model_checkpoint": "./deit-base-distilled-mask-finetuned/checkpoint-600",
4
+ "epoch": 1.9955654101995566,
5
+ "global_step": 900,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.02,
12
+ "learning_rate": 0.00019889135254988916,
13
+ "loss": 0.0451,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.04,
18
+ "learning_rate": 0.00019778270509977829,
19
+ "loss": 0.0465,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.07,
24
+ "learning_rate": 0.0001966740576496674,
25
+ "loss": 0.2191,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.09,
30
+ "learning_rate": 0.00019556541019955653,
31
+ "loss": 0.0804,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.11,
36
+ "learning_rate": 0.00019445676274944569,
37
+ "loss": 0.0719,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.13,
42
+ "learning_rate": 0.00019334811529933484,
43
+ "loss": 0.0518,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.16,
48
+ "learning_rate": 0.00019223946784922396,
49
+ "loss": 0.0433,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.18,
54
+ "learning_rate": 0.00019124168514412417,
55
+ "loss": 0.2892,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.2,
60
+ "learning_rate": 0.00019013303769401332,
61
+ "loss": 0.0692,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.22,
66
+ "learning_rate": 0.00018902439024390244,
67
+ "loss": 0.0765,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.22,
72
+ "eval_accuracy": 0.9888765294771968,
73
+ "eval_loss": 0.05117267370223999,
74
+ "eval_runtime": 195.5169,
75
+ "eval_samples_per_second": 18.392,
76
+ "eval_steps_per_second": 2.302,
77
+ "step": 100
78
+ },
79
+ {
80
+ "epoch": 0.24,
81
+ "learning_rate": 0.0001879157427937916,
82
+ "loss": 0.1107,
83
+ "step": 110
84
+ },
85
+ {
86
+ "epoch": 0.27,
87
+ "learning_rate": 0.00018680709534368072,
88
+ "loss": 0.0967,
89
+ "step": 120
90
+ },
91
+ {
92
+ "epoch": 0.29,
93
+ "learning_rate": 0.00018569844789356984,
94
+ "loss": 0.0852,
95
+ "step": 130
96
+ },
97
+ {
98
+ "epoch": 0.31,
99
+ "learning_rate": 0.000184589800443459,
100
+ "loss": 0.082,
101
+ "step": 140
102
+ },
103
+ {
104
+ "epoch": 0.33,
105
+ "learning_rate": 0.00018348115299334812,
106
+ "loss": 0.0172,
107
+ "step": 150
108
+ },
109
+ {
110
+ "epoch": 0.35,
111
+ "learning_rate": 0.00018237250554323727,
112
+ "loss": 0.0379,
113
+ "step": 160
114
+ },
115
+ {
116
+ "epoch": 0.38,
117
+ "learning_rate": 0.0001812638580931264,
118
+ "loss": 0.0773,
119
+ "step": 170
120
+ },
121
+ {
122
+ "epoch": 0.4,
123
+ "learning_rate": 0.00018015521064301552,
124
+ "loss": 0.0329,
125
+ "step": 180
126
+ },
127
+ {
128
+ "epoch": 0.42,
129
+ "learning_rate": 0.00017904656319290467,
130
+ "loss": 0.0688,
131
+ "step": 190
132
+ },
133
+ {
134
+ "epoch": 0.44,
135
+ "learning_rate": 0.0001779379157427938,
136
+ "loss": 0.0533,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.44,
141
+ "eval_accuracy": 0.9908231368186874,
142
+ "eval_loss": 0.037399593740701675,
143
+ "eval_runtime": 194.4149,
144
+ "eval_samples_per_second": 18.497,
145
+ "eval_steps_per_second": 2.315,
146
+ "step": 200
147
+ },
148
+ {
149
+ "epoch": 0.47,
150
+ "learning_rate": 0.00017682926829268295,
151
+ "loss": 0.0536,
152
+ "step": 210
153
+ },
154
+ {
155
+ "epoch": 0.49,
156
+ "learning_rate": 0.00017572062084257207,
157
+ "loss": 0.0144,
158
+ "step": 220
159
+ },
160
+ {
161
+ "epoch": 0.51,
162
+ "learning_rate": 0.0001746119733924612,
163
+ "loss": 0.0535,
164
+ "step": 230
165
+ },
166
+ {
167
+ "epoch": 0.53,
168
+ "learning_rate": 0.00017350332594235035,
169
+ "loss": 0.0569,
170
+ "step": 240
171
+ },
172
+ {
173
+ "epoch": 0.55,
174
+ "learning_rate": 0.00017239467849223947,
175
+ "loss": 0.0408,
176
+ "step": 250
177
+ },
178
+ {
179
+ "epoch": 0.58,
180
+ "learning_rate": 0.00017128603104212862,
181
+ "loss": 0.0529,
182
+ "step": 260
183
+ },
184
+ {
185
+ "epoch": 0.6,
186
+ "learning_rate": 0.00017017738359201775,
187
+ "loss": 0.088,
188
+ "step": 270
189
+ },
190
+ {
191
+ "epoch": 0.62,
192
+ "learning_rate": 0.00016906873614190687,
193
+ "loss": 0.0147,
194
+ "step": 280
195
+ },
196
+ {
197
+ "epoch": 0.64,
198
+ "learning_rate": 0.00016796008869179602,
199
+ "loss": 0.0369,
200
+ "step": 290
201
+ },
202
+ {
203
+ "epoch": 0.67,
204
+ "learning_rate": 0.00016685144124168515,
205
+ "loss": 0.0442,
206
+ "step": 300
207
+ },
208
+ {
209
+ "epoch": 0.67,
210
+ "eval_accuracy": 0.9885984427141268,
211
+ "eval_loss": 0.03961510211229324,
212
+ "eval_runtime": 194.5523,
213
+ "eval_samples_per_second": 18.483,
214
+ "eval_steps_per_second": 2.313,
215
+ "step": 300
216
+ },
217
+ {
218
+ "epoch": 0.69,
219
+ "learning_rate": 0.0001657427937915743,
220
+ "loss": 0.0191,
221
+ "step": 310
222
+ },
223
+ {
224
+ "epoch": 0.71,
225
+ "learning_rate": 0.00016463414634146343,
226
+ "loss": 0.044,
227
+ "step": 320
228
+ },
229
+ {
230
+ "epoch": 0.73,
231
+ "learning_rate": 0.00016352549889135255,
232
+ "loss": 0.0274,
233
+ "step": 330
234
+ },
235
+ {
236
+ "epoch": 0.75,
237
+ "learning_rate": 0.0001624168514412417,
238
+ "loss": 0.064,
239
+ "step": 340
240
+ },
241
+ {
242
+ "epoch": 0.78,
243
+ "learning_rate": 0.00016130820399113083,
244
+ "loss": 0.0478,
245
+ "step": 350
246
+ },
247
+ {
248
+ "epoch": 0.8,
249
+ "learning_rate": 0.00016019955654101998,
250
+ "loss": 0.0532,
251
+ "step": 360
252
+ },
253
+ {
254
+ "epoch": 0.82,
255
+ "learning_rate": 0.0001590909090909091,
256
+ "loss": 0.05,
257
+ "step": 370
258
+ },
259
+ {
260
+ "epoch": 0.84,
261
+ "learning_rate": 0.00015798226164079823,
262
+ "loss": 0.0197,
263
+ "step": 380
264
+ },
265
+ {
266
+ "epoch": 0.86,
267
+ "learning_rate": 0.00015687361419068738,
268
+ "loss": 0.0492,
269
+ "step": 390
270
+ },
271
+ {
272
+ "epoch": 0.89,
273
+ "learning_rate": 0.0001557649667405765,
274
+ "loss": 0.0359,
275
+ "step": 400
276
+ },
277
+ {
278
+ "epoch": 0.89,
279
+ "eval_accuracy": 0.9885984427141268,
280
+ "eval_loss": 0.034940723329782486,
281
+ "eval_runtime": 193.9815,
282
+ "eval_samples_per_second": 18.538,
283
+ "eval_steps_per_second": 2.32,
284
+ "step": 400
285
+ },
286
+ {
287
+ "epoch": 0.91,
288
+ "learning_rate": 0.00015465631929046565,
289
+ "loss": 0.0405,
290
+ "step": 410
291
+ },
292
+ {
293
+ "epoch": 0.93,
294
+ "learning_rate": 0.00015354767184035478,
295
+ "loss": 0.03,
296
+ "step": 420
297
+ },
298
+ {
299
+ "epoch": 0.95,
300
+ "learning_rate": 0.0001524390243902439,
301
+ "loss": 0.0272,
302
+ "step": 430
303
+ },
304
+ {
305
+ "epoch": 0.98,
306
+ "learning_rate": 0.00015133037694013303,
307
+ "loss": 0.0366,
308
+ "step": 440
309
+ },
310
+ {
311
+ "epoch": 1.0,
312
+ "learning_rate": 0.00015022172949002218,
313
+ "loss": 0.0676,
314
+ "step": 450
315
+ },
316
+ {
317
+ "epoch": 1.02,
318
+ "learning_rate": 0.00014911308203991133,
319
+ "loss": 0.0185,
320
+ "step": 460
321
+ },
322
+ {
323
+ "epoch": 1.04,
324
+ "learning_rate": 0.00014800443458980045,
325
+ "loss": 0.0053,
326
+ "step": 470
327
+ },
328
+ {
329
+ "epoch": 1.06,
330
+ "learning_rate": 0.00014689578713968958,
331
+ "loss": 0.0309,
332
+ "step": 480
333
+ },
334
+ {
335
+ "epoch": 1.09,
336
+ "learning_rate": 0.0001457871396895787,
337
+ "loss": 0.0183,
338
+ "step": 490
339
+ },
340
+ {
341
+ "epoch": 1.11,
342
+ "learning_rate": 0.00014467849223946785,
343
+ "loss": 0.0777,
344
+ "step": 500
345
+ },
346
+ {
347
+ "epoch": 1.11,
348
+ "eval_accuracy": 0.9869299221357063,
349
+ "eval_loss": 0.042109716683626175,
350
+ "eval_runtime": 193.8529,
351
+ "eval_samples_per_second": 18.55,
352
+ "eval_steps_per_second": 2.321,
353
+ "step": 500
354
+ },
355
+ {
356
+ "epoch": 1.13,
357
+ "learning_rate": 0.000143569844789357,
358
+ "loss": 0.0159,
359
+ "step": 510
360
+ },
361
+ {
362
+ "epoch": 1.15,
363
+ "learning_rate": 0.00014246119733924613,
364
+ "loss": 0.0645,
365
+ "step": 520
366
+ },
367
+ {
368
+ "epoch": 1.18,
369
+ "learning_rate": 0.00014135254988913525,
370
+ "loss": 0.0632,
371
+ "step": 530
372
+ },
373
+ {
374
+ "epoch": 1.2,
375
+ "learning_rate": 0.00014024390243902438,
376
+ "loss": 0.0251,
377
+ "step": 540
378
+ },
379
+ {
380
+ "epoch": 1.22,
381
+ "learning_rate": 0.00013913525498891353,
382
+ "loss": 0.036,
383
+ "step": 550
384
+ },
385
+ {
386
+ "epoch": 1.24,
387
+ "learning_rate": 0.00013802660753880268,
388
+ "loss": 0.0389,
389
+ "step": 560
390
+ },
391
+ {
392
+ "epoch": 1.26,
393
+ "learning_rate": 0.0001369179600886918,
394
+ "loss": 0.0353,
395
+ "step": 570
396
+ },
397
+ {
398
+ "epoch": 1.29,
399
+ "learning_rate": 0.00013580931263858093,
400
+ "loss": 0.0376,
401
+ "step": 580
402
+ },
403
+ {
404
+ "epoch": 1.31,
405
+ "learning_rate": 0.00013470066518847006,
406
+ "loss": 0.0416,
407
+ "step": 590
408
+ },
409
+ {
410
+ "epoch": 1.33,
411
+ "learning_rate": 0.0001335920177383592,
412
+ "loss": 0.0051,
413
+ "step": 600
414
+ },
415
+ {
416
+ "epoch": 1.33,
417
+ "eval_accuracy": 0.9922135706340378,
418
+ "eval_loss": 0.027106858789920807,
419
+ "eval_runtime": 193.7443,
420
+ "eval_samples_per_second": 18.561,
421
+ "eval_steps_per_second": 2.323,
422
+ "step": 600
423
+ },
424
+ {
425
+ "epoch": 1.35,
426
+ "learning_rate": 0.00013248337028824836,
427
+ "loss": 0.0277,
428
+ "step": 610
429
+ },
430
+ {
431
+ "epoch": 1.37,
432
+ "learning_rate": 0.00013137472283813748,
433
+ "loss": 0.0466,
434
+ "step": 620
435
+ },
436
+ {
437
+ "epoch": 1.4,
438
+ "learning_rate": 0.0001302660753880266,
439
+ "loss": 0.0365,
440
+ "step": 630
441
+ },
442
+ {
443
+ "epoch": 1.42,
444
+ "learning_rate": 0.00012915742793791573,
445
+ "loss": 0.0079,
446
+ "step": 640
447
+ },
448
+ {
449
+ "epoch": 1.44,
450
+ "learning_rate": 0.00012804878048780488,
451
+ "loss": 0.0291,
452
+ "step": 650
453
+ },
454
+ {
455
+ "epoch": 1.46,
456
+ "learning_rate": 0.000126940133037694,
457
+ "loss": 0.0694,
458
+ "step": 660
459
+ },
460
+ {
461
+ "epoch": 1.49,
462
+ "learning_rate": 0.00012583148558758316,
463
+ "loss": 0.0027,
464
+ "step": 670
465
+ },
466
+ {
467
+ "epoch": 1.51,
468
+ "learning_rate": 0.0001247228381374723,
469
+ "loss": 0.0179,
470
+ "step": 680
471
+ },
472
+ {
473
+ "epoch": 1.53,
474
+ "learning_rate": 0.0001236141906873614,
475
+ "loss": 0.0206,
476
+ "step": 690
477
+ },
478
+ {
479
+ "epoch": 1.55,
480
+ "learning_rate": 0.00012250554323725056,
481
+ "loss": 0.0112,
482
+ "step": 700
483
+ },
484
+ {
485
+ "epoch": 1.55,
486
+ "eval_accuracy": 0.9849833147942157,
487
+ "eval_loss": 0.051631152629852295,
488
+ "eval_runtime": 194.3203,
489
+ "eval_samples_per_second": 18.506,
490
+ "eval_steps_per_second": 2.316,
491
+ "step": 700
492
+ },
493
+ {
494
+ "epoch": 1.57,
495
+ "learning_rate": 0.00012139689578713968,
496
+ "loss": 0.0037,
497
+ "step": 710
498
+ },
499
+ {
500
+ "epoch": 1.6,
501
+ "learning_rate": 0.00012028824833702884,
502
+ "loss": 0.0051,
503
+ "step": 720
504
+ },
505
+ {
506
+ "epoch": 1.62,
507
+ "learning_rate": 0.00011917960088691797,
508
+ "loss": 0.0014,
509
+ "step": 730
510
+ },
511
+ {
512
+ "epoch": 1.64,
513
+ "learning_rate": 0.0001180709534368071,
514
+ "loss": 0.0018,
515
+ "step": 740
516
+ },
517
+ {
518
+ "epoch": 1.66,
519
+ "learning_rate": 0.00011696230598669624,
520
+ "loss": 0.0445,
521
+ "step": 750
522
+ },
523
+ {
524
+ "epoch": 1.69,
525
+ "learning_rate": 0.00011585365853658536,
526
+ "loss": 0.0294,
527
+ "step": 760
528
+ },
529
+ {
530
+ "epoch": 1.71,
531
+ "learning_rate": 0.00011474501108647451,
532
+ "loss": 0.0247,
533
+ "step": 770
534
+ },
535
+ {
536
+ "epoch": 1.73,
537
+ "learning_rate": 0.00011363636363636365,
538
+ "loss": 0.0069,
539
+ "step": 780
540
+ },
541
+ {
542
+ "epoch": 1.75,
543
+ "learning_rate": 0.00011252771618625277,
544
+ "loss": 0.0612,
545
+ "step": 790
546
+ },
547
+ {
548
+ "epoch": 1.77,
549
+ "learning_rate": 0.00011141906873614191,
550
+ "loss": 0.0152,
551
+ "step": 800
552
+ },
553
+ {
554
+ "epoch": 1.77,
555
+ "eval_accuracy": 0.9824805339265851,
556
+ "eval_loss": 0.05840621143579483,
557
+ "eval_runtime": 194.8587,
558
+ "eval_samples_per_second": 18.454,
559
+ "eval_steps_per_second": 2.309,
560
+ "step": 800
561
+ },
562
+ {
563
+ "epoch": 1.8,
564
+ "learning_rate": 0.00011031042128603104,
565
+ "loss": 0.0137,
566
+ "step": 810
567
+ },
568
+ {
569
+ "epoch": 1.82,
570
+ "learning_rate": 0.00010920177383592019,
571
+ "loss": 0.0118,
572
+ "step": 820
573
+ },
574
+ {
575
+ "epoch": 1.84,
576
+ "learning_rate": 0.00010809312638580931,
577
+ "loss": 0.0246,
578
+ "step": 830
579
+ },
580
+ {
581
+ "epoch": 1.86,
582
+ "learning_rate": 0.00010698447893569845,
583
+ "loss": 0.0246,
584
+ "step": 840
585
+ },
586
+ {
587
+ "epoch": 1.88,
588
+ "learning_rate": 0.0001058758314855876,
589
+ "loss": 0.036,
590
+ "step": 850
591
+ },
592
+ {
593
+ "epoch": 1.91,
594
+ "learning_rate": 0.00010476718403547671,
595
+ "loss": 0.0248,
596
+ "step": 860
597
+ },
598
+ {
599
+ "epoch": 1.93,
600
+ "learning_rate": 0.00010365853658536586,
601
+ "loss": 0.019,
602
+ "step": 870
603
+ },
604
+ {
605
+ "epoch": 1.95,
606
+ "learning_rate": 0.00010254988913525499,
607
+ "loss": 0.0387,
608
+ "step": 880
609
+ },
610
+ {
611
+ "epoch": 1.97,
612
+ "learning_rate": 0.00010144124168514413,
613
+ "loss": 0.0127,
614
+ "step": 890
615
+ },
616
+ {
617
+ "epoch": 2.0,
618
+ "learning_rate": 0.00010033259423503328,
619
+ "loss": 0.0339,
620
+ "step": 900
621
+ },
622
+ {
623
+ "epoch": 2.0,
624
+ "eval_accuracy": 0.9905450500556173,
625
+ "eval_loss": 0.027991166338324547,
626
+ "eval_runtime": 194.6989,
627
+ "eval_samples_per_second": 18.47,
628
+ "eval_steps_per_second": 2.311,
629
+ "step": 900
630
+ },
631
+ {
632
+ "epoch": 2.0,
633
+ "step": 900,
634
+ "total_flos": 2.2314992806549094e+18,
635
+ "train_loss": 0.04378239723129405,
636
+ "train_runtime": 5836.8754,
637
+ "train_samples_per_second": 9.887,
638
+ "train_steps_per_second": 0.309
639
+ }
640
+ ],
641
+ "max_steps": 1804,
642
+ "num_train_epochs": 4,
643
+ "total_flos": 2.2314992806549094e+18,
644
+ "trial_name": null,
645
+ "trial_params": null
646
+ }