Polly1231 commited on
Commit
8c3bb24
·
verified ·
1 Parent(s): 07488b4

Upload trainer_state.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. trainer_state.json +606 -0
trainer_state.json ADDED
@@ -0,0 +1,606 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 96,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.03,
13
+ "learning_rate": 6.666666666666667e-05,
14
+ "loss": 1.7624,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.06,
19
+ "learning_rate": 0.00013333333333333334,
20
+ "loss": 1.7445,
21
+ "step": 2
22
+ },
23
+ {
24
+ "epoch": 0.09,
25
+ "learning_rate": 0.0002,
26
+ "loss": 1.6128,
27
+ "step": 3
28
+ },
29
+ {
30
+ "epoch": 0.12,
31
+ "learning_rate": 0.0001999429490929718,
32
+ "loss": 1.4354,
33
+ "step": 4
34
+ },
35
+ {
36
+ "epoch": 0.16,
37
+ "learning_rate": 0.00019977186146800707,
38
+ "loss": 1.3906,
39
+ "step": 5
40
+ },
41
+ {
42
+ "epoch": 0.19,
43
+ "learning_rate": 0.00019948693233918952,
44
+ "loss": 1.297,
45
+ "step": 6
46
+ },
47
+ {
48
+ "epoch": 0.22,
49
+ "learning_rate": 0.00019908848681582391,
50
+ "loss": 1.2876,
51
+ "step": 7
52
+ },
53
+ {
54
+ "epoch": 0.25,
55
+ "learning_rate": 0.00019857697953148037,
56
+ "loss": 1.2704,
57
+ "step": 8
58
+ },
59
+ {
60
+ "epoch": 0.28,
61
+ "learning_rate": 0.00019795299412524945,
62
+ "loss": 1.2342,
63
+ "step": 9
64
+ },
65
+ {
66
+ "epoch": 0.31,
67
+ "learning_rate": 0.00019721724257579907,
68
+ "loss": 1.1611,
69
+ "step": 10
70
+ },
71
+ {
72
+ "epoch": 0.34,
73
+ "learning_rate": 0.0001963705643889941,
74
+ "loss": 1.2132,
75
+ "step": 11
76
+ },
77
+ {
78
+ "epoch": 0.38,
79
+ "learning_rate": 0.00019541392564000488,
80
+ "loss": 1.23,
81
+ "step": 12
82
+ },
83
+ {
84
+ "epoch": 0.41,
85
+ "learning_rate": 0.00019434841787099803,
86
+ "loss": 1.181,
87
+ "step": 13
88
+ },
89
+ {
90
+ "epoch": 0.44,
91
+ "learning_rate": 0.00019317525684566685,
92
+ "loss": 1.1662,
93
+ "step": 14
94
+ },
95
+ {
96
+ "epoch": 0.47,
97
+ "learning_rate": 0.00019189578116202307,
98
+ "loss": 1.1026,
99
+ "step": 15
100
+ },
101
+ {
102
+ "epoch": 0.5,
103
+ "learning_rate": 0.00019051145072503215,
104
+ "loss": 1.1725,
105
+ "step": 16
106
+ },
107
+ {
108
+ "epoch": 0.53,
109
+ "learning_rate": 0.00018902384508083517,
110
+ "loss": 1.1523,
111
+ "step": 17
112
+ },
113
+ {
114
+ "epoch": 0.56,
115
+ "learning_rate": 0.00018743466161445823,
116
+ "loss": 1.166,
117
+ "step": 18
118
+ },
119
+ {
120
+ "epoch": 0.59,
121
+ "learning_rate": 0.0001857457136130651,
122
+ "loss": 1.1499,
123
+ "step": 19
124
+ },
125
+ {
126
+ "epoch": 0.62,
127
+ "learning_rate": 0.00018395892819696389,
128
+ "loss": 1.1366,
129
+ "step": 20
130
+ },
131
+ {
132
+ "epoch": 0.66,
133
+ "learning_rate": 0.00018207634412072764,
134
+ "loss": 1.1057,
135
+ "step": 21
136
+ },
137
+ {
138
+ "epoch": 0.69,
139
+ "learning_rate": 0.00018010010944693848,
140
+ "loss": 1.0965,
141
+ "step": 22
142
+ },
143
+ {
144
+ "epoch": 0.72,
145
+ "learning_rate": 0.0001780324790952092,
146
+ "loss": 1.1557,
147
+ "step": 23
148
+ },
149
+ {
150
+ "epoch": 0.75,
151
+ "learning_rate": 0.0001758758122692791,
152
+ "loss": 1.1568,
153
+ "step": 24
154
+ },
155
+ {
156
+ "epoch": 0.78,
157
+ "learning_rate": 0.00017363256976511972,
158
+ "loss": 1.1355,
159
+ "step": 25
160
+ },
161
+ {
162
+ "epoch": 0.81,
163
+ "learning_rate": 0.00017130531116312203,
164
+ "loss": 1.0729,
165
+ "step": 26
166
+ },
167
+ {
168
+ "epoch": 0.84,
169
+ "learning_rate": 0.00016889669190756868,
170
+ "loss": 1.0658,
171
+ "step": 27
172
+ },
173
+ {
174
+ "epoch": 0.88,
175
+ "learning_rate": 0.00016640946027672392,
176
+ "loss": 1.1382,
177
+ "step": 28
178
+ },
179
+ {
180
+ "epoch": 0.91,
181
+ "learning_rate": 0.00016384645424699835,
182
+ "loss": 1.0855,
183
+ "step": 29
184
+ },
185
+ {
186
+ "epoch": 0.94,
187
+ "learning_rate": 0.0001612105982547663,
188
+ "loss": 1.1532,
189
+ "step": 30
190
+ },
191
+ {
192
+ "epoch": 0.97,
193
+ "learning_rate": 0.00015850489985953076,
194
+ "loss": 1.1607,
195
+ "step": 31
196
+ },
197
+ {
198
+ "epoch": 1.0,
199
+ "learning_rate": 0.00015573244631224365,
200
+ "loss": 1.0461,
201
+ "step": 32
202
+ },
203
+ {
204
+ "epoch": 1.03,
205
+ "learning_rate": 0.00015289640103269625,
206
+ "loss": 0.9625,
207
+ "step": 33
208
+ },
209
+ {
210
+ "epoch": 1.06,
211
+ "learning_rate": 0.00015000000000000001,
212
+ "loss": 0.9087,
213
+ "step": 34
214
+ },
215
+ {
216
+ "epoch": 1.09,
217
+ "learning_rate": 0.0001470465480602756,
218
+ "loss": 0.8878,
219
+ "step": 35
220
+ },
221
+ {
222
+ "epoch": 1.12,
223
+ "learning_rate": 0.00014403941515576344,
224
+ "loss": 0.8778,
225
+ "step": 36
226
+ },
227
+ {
228
+ "epoch": 1.16,
229
+ "learning_rate": 0.00014098203247965875,
230
+ "loss": 0.9648,
231
+ "step": 37
232
+ },
233
+ {
234
+ "epoch": 1.19,
235
+ "learning_rate": 0.0001378778885610576,
236
+ "loss": 0.8955,
237
+ "step": 38
238
+ },
239
+ {
240
+ "epoch": 1.22,
241
+ "learning_rate": 0.00013473052528448201,
242
+ "loss": 0.8645,
243
+ "step": 39
244
+ },
245
+ {
246
+ "epoch": 1.25,
247
+ "learning_rate": 0.00013154353384852558,
248
+ "loss": 0.8806,
249
+ "step": 40
250
+ },
251
+ {
252
+ "epoch": 1.28,
253
+ "learning_rate": 0.00012832055066823038,
254
+ "loss": 0.8299,
255
+ "step": 41
256
+ },
257
+ {
258
+ "epoch": 1.31,
259
+ "learning_rate": 0.00012506525322587207,
260
+ "loss": 0.839,
261
+ "step": 42
262
+ },
263
+ {
264
+ "epoch": 1.34,
265
+ "learning_rate": 0.00012178135587488515,
266
+ "loss": 0.8633,
267
+ "step": 43
268
+ },
269
+ {
270
+ "epoch": 1.38,
271
+ "learning_rate": 0.00011847260560171896,
272
+ "loss": 0.8526,
273
+ "step": 44
274
+ },
275
+ {
276
+ "epoch": 1.41,
277
+ "learning_rate": 0.00011514277775045768,
278
+ "loss": 0.9084,
279
+ "step": 45
280
+ },
281
+ {
282
+ "epoch": 1.44,
283
+ "learning_rate": 0.00011179567171508463,
284
+ "loss": 0.8485,
285
+ "step": 46
286
+ },
287
+ {
288
+ "epoch": 1.47,
289
+ "learning_rate": 0.00010843510660430447,
290
+ "loss": 0.8431,
291
+ "step": 47
292
+ },
293
+ {
294
+ "epoch": 1.5,
295
+ "learning_rate": 0.00010506491688387127,
296
+ "loss": 0.8836,
297
+ "step": 48
298
+ },
299
+ {
300
+ "epoch": 1.53,
301
+ "learning_rate": 0.0001016889480013931,
302
+ "loss": 0.8329,
303
+ "step": 49
304
+ },
305
+ {
306
+ "epoch": 1.56,
307
+ "learning_rate": 9.83110519986069e-05,
308
+ "loss": 0.8403,
309
+ "step": 50
310
+ },
311
+ {
312
+ "epoch": 1.59,
313
+ "learning_rate": 9.493508311612874e-05,
314
+ "loss": 0.7999,
315
+ "step": 51
316
+ },
317
+ {
318
+ "epoch": 1.62,
319
+ "learning_rate": 9.156489339569554e-05,
320
+ "loss": 0.8522,
321
+ "step": 52
322
+ },
323
+ {
324
+ "epoch": 1.66,
325
+ "learning_rate": 8.820432828491542e-05,
326
+ "loss": 0.8663,
327
+ "step": 53
328
+ },
329
+ {
330
+ "epoch": 1.69,
331
+ "learning_rate": 8.485722224954237e-05,
332
+ "loss": 0.8942,
333
+ "step": 54
334
+ },
335
+ {
336
+ "epoch": 1.72,
337
+ "learning_rate": 8.15273943982811e-05,
338
+ "loss": 0.8413,
339
+ "step": 55
340
+ },
341
+ {
342
+ "epoch": 1.75,
343
+ "learning_rate": 7.821864412511485e-05,
344
+ "loss": 0.8353,
345
+ "step": 56
346
+ },
347
+ {
348
+ "epoch": 1.78,
349
+ "learning_rate": 7.493474677412794e-05,
350
+ "loss": 0.8274,
351
+ "step": 57
352
+ },
353
+ {
354
+ "epoch": 1.81,
355
+ "learning_rate": 7.16794493317696e-05,
356
+ "loss": 0.8657,
357
+ "step": 58
358
+ },
359
+ {
360
+ "epoch": 1.84,
361
+ "learning_rate": 6.845646615147445e-05,
362
+ "loss": 0.857,
363
+ "step": 59
364
+ },
365
+ {
366
+ "epoch": 1.88,
367
+ "learning_rate": 6.526947471551798e-05,
368
+ "loss": 0.82,
369
+ "step": 60
370
+ },
371
+ {
372
+ "epoch": 1.91,
373
+ "learning_rate": 6.21221114389424e-05,
374
+ "loss": 0.8453,
375
+ "step": 61
376
+ },
377
+ {
378
+ "epoch": 1.94,
379
+ "learning_rate": 5.901796752034128e-05,
380
+ "loss": 0.828,
381
+ "step": 62
382
+ },
383
+ {
384
+ "epoch": 1.97,
385
+ "learning_rate": 5.596058484423656e-05,
386
+ "loss": 0.8287,
387
+ "step": 63
388
+ },
389
+ {
390
+ "epoch": 2.0,
391
+ "learning_rate": 5.2953451939724454e-05,
392
+ "loss": 0.7682,
393
+ "step": 64
394
+ },
395
+ {
396
+ "epoch": 2.03,
397
+ "learning_rate": 5.000000000000002e-05,
398
+ "loss": 0.669,
399
+ "step": 65
400
+ },
401
+ {
402
+ "epoch": 2.06,
403
+ "learning_rate": 4.710359896730379e-05,
404
+ "loss": 0.6887,
405
+ "step": 66
406
+ },
407
+ {
408
+ "epoch": 2.09,
409
+ "learning_rate": 4.426755368775637e-05,
410
+ "loss": 0.6653,
411
+ "step": 67
412
+ },
413
+ {
414
+ "epoch": 2.12,
415
+ "learning_rate": 4.149510014046922e-05,
416
+ "loss": 0.6591,
417
+ "step": 68
418
+ },
419
+ {
420
+ "epoch": 2.16,
421
+ "learning_rate": 3.878940174523371e-05,
422
+ "loss": 0.6193,
423
+ "step": 69
424
+ },
425
+ {
426
+ "epoch": 2.19,
427
+ "learning_rate": 3.615354575300166e-05,
428
+ "loss": 0.6349,
429
+ "step": 70
430
+ },
431
+ {
432
+ "epoch": 2.22,
433
+ "learning_rate": 3.3590539723276083e-05,
434
+ "loss": 0.6766,
435
+ "step": 71
436
+ },
437
+ {
438
+ "epoch": 2.25,
439
+ "learning_rate": 3.110330809243134e-05,
440
+ "loss": 0.6399,
441
+ "step": 72
442
+ },
443
+ {
444
+ "epoch": 2.28,
445
+ "learning_rate": 2.869468883687798e-05,
446
+ "loss": 0.668,
447
+ "step": 73
448
+ },
449
+ {
450
+ "epoch": 2.31,
451
+ "learning_rate": 2.6367430234880284e-05,
452
+ "loss": 0.6316,
453
+ "step": 74
454
+ },
455
+ {
456
+ "epoch": 2.34,
457
+ "learning_rate": 2.4124187730720917e-05,
458
+ "loss": 0.6236,
459
+ "step": 75
460
+ },
461
+ {
462
+ "epoch": 2.38,
463
+ "learning_rate": 2.1967520904790827e-05,
464
+ "loss": 0.6256,
465
+ "step": 76
466
+ },
467
+ {
468
+ "epoch": 2.41,
469
+ "learning_rate": 1.9899890553061562e-05,
470
+ "loss": 0.6268,
471
+ "step": 77
472
+ },
473
+ {
474
+ "epoch": 2.44,
475
+ "learning_rate": 1.7923655879272393e-05,
476
+ "loss": 0.6326,
477
+ "step": 78
478
+ },
479
+ {
480
+ "epoch": 2.47,
481
+ "learning_rate": 1.60410718030361e-05,
482
+ "loss": 0.6234,
483
+ "step": 79
484
+ },
485
+ {
486
+ "epoch": 2.5,
487
+ "learning_rate": 1.425428638693489e-05,
488
+ "loss": 0.6475,
489
+ "step": 80
490
+ },
491
+ {
492
+ "epoch": 2.53,
493
+ "learning_rate": 1.2565338385541792e-05,
494
+ "loss": 0.6525,
495
+ "step": 81
496
+ },
497
+ {
498
+ "epoch": 2.56,
499
+ "learning_rate": 1.097615491916485e-05,
500
+ "loss": 0.6506,
501
+ "step": 82
502
+ },
503
+ {
504
+ "epoch": 2.59,
505
+ "learning_rate": 9.488549274967872e-06,
506
+ "loss": 0.6068,
507
+ "step": 83
508
+ },
509
+ {
510
+ "epoch": 2.62,
511
+ "learning_rate": 8.10421883797694e-06,
512
+ "loss": 0.6521,
513
+ "step": 84
514
+ },
515
+ {
516
+ "epoch": 2.66,
517
+ "learning_rate": 6.824743154333157e-06,
518
+ "loss": 0.6288,
519
+ "step": 85
520
+ },
521
+ {
522
+ "epoch": 2.69,
523
+ "learning_rate": 5.651582129001986e-06,
524
+ "loss": 0.6158,
525
+ "step": 86
526
+ },
527
+ {
528
+ "epoch": 2.72,
529
+ "learning_rate": 4.586074359995119e-06,
530
+ "loss": 0.6562,
531
+ "step": 87
532
+ },
533
+ {
534
+ "epoch": 2.75,
535
+ "learning_rate": 3.6294356110059157e-06,
536
+ "loss": 0.6376,
537
+ "step": 88
538
+ },
539
+ {
540
+ "epoch": 2.78,
541
+ "learning_rate": 2.7827574242009437e-06,
542
+ "loss": 0.649,
543
+ "step": 89
544
+ },
545
+ {
546
+ "epoch": 2.81,
547
+ "learning_rate": 2.0470058747505516e-06,
548
+ "loss": 0.6226,
549
+ "step": 90
550
+ },
551
+ {
552
+ "epoch": 2.84,
553
+ "learning_rate": 1.4230204685196203e-06,
554
+ "loss": 0.6499,
555
+ "step": 91
556
+ },
557
+ {
558
+ "epoch": 2.88,
559
+ "learning_rate": 9.11513184176116e-07,
560
+ "loss": 0.6173,
561
+ "step": 92
562
+ },
563
+ {
564
+ "epoch": 2.91,
565
+ "learning_rate": 5.130676608104845e-07,
566
+ "loss": 0.6381,
567
+ "step": 93
568
+ },
569
+ {
570
+ "epoch": 2.94,
571
+ "learning_rate": 2.2813853199292746e-07,
572
+ "loss": 0.5941,
573
+ "step": 94
574
+ },
575
+ {
576
+ "epoch": 2.97,
577
+ "learning_rate": 5.705090702819993e-08,
578
+ "loss": 0.6088,
579
+ "step": 95
580
+ },
581
+ {
582
+ "epoch": 3.0,
583
+ "learning_rate": 0.0,
584
+ "loss": 0.5529,
585
+ "step": 96
586
+ },
587
+ {
588
+ "epoch": 3.0,
589
+ "step": 96,
590
+ "total_flos": 6492865167360.0,
591
+ "train_loss": 0.9074693458775679,
592
+ "train_runtime": 1320.5347,
593
+ "train_samples_per_second": 4.544,
594
+ "train_steps_per_second": 0.073
595
+ }
596
+ ],
597
+ "logging_steps": 1.0,
598
+ "max_steps": 96,
599
+ "num_input_tokens_seen": 0,
600
+ "num_train_epochs": 3,
601
+ "save_steps": 50000,
602
+ "total_flos": 6492865167360.0,
603
+ "train_batch_size": 32,
604
+ "trial_name": null,
605
+ "trial_params": null
606
+ }