srcocotero commited on
Commit
cc2b1bd
1 Parent(s): e64cd9b

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +655 -0
trainer_state.json ADDED
@@ -0,0 +1,655 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 52629,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 2.9714986034315684e-05,
13
+ "loss": 3.9696,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.06,
18
+ "learning_rate": 2.942997206863136e-05,
19
+ "loss": 3.1419,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.09,
24
+ "learning_rate": 2.9144958102947044e-05,
25
+ "loss": 2.8913,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.11,
30
+ "learning_rate": 2.8859944137262728e-05,
31
+ "loss": 2.7378,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.14,
36
+ "learning_rate": 2.857493017157841e-05,
37
+ "loss": 2.6655,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.17,
42
+ "learning_rate": 2.8289916205894088e-05,
43
+ "loss": 2.5075,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.2,
48
+ "learning_rate": 2.800490224020977e-05,
49
+ "loss": 2.5241,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.23,
54
+ "learning_rate": 2.7719888274525454e-05,
55
+ "loss": 2.3847,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.26,
60
+ "learning_rate": 2.743487430884113e-05,
61
+ "loss": 2.3737,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.29,
66
+ "learning_rate": 2.7149860343156814e-05,
67
+ "loss": 2.295,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.31,
72
+ "learning_rate": 2.6864846377472498e-05,
73
+ "loss": 2.2504,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.34,
78
+ "learning_rate": 2.657983241178818e-05,
79
+ "loss": 2.1942,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.37,
84
+ "learning_rate": 2.6294818446103858e-05,
85
+ "loss": 2.1828,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.4,
90
+ "learning_rate": 2.600980448041954e-05,
91
+ "loss": 2.1573,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.43,
96
+ "learning_rate": 2.5724790514735224e-05,
97
+ "loss": 2.1698,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.46,
102
+ "learning_rate": 2.54397765490509e-05,
103
+ "loss": 2.0895,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.48,
108
+ "learning_rate": 2.5154762583366584e-05,
109
+ "loss": 2.0382,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.51,
114
+ "learning_rate": 2.4869748617682268e-05,
115
+ "loss": 1.987,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.54,
120
+ "learning_rate": 2.458473465199795e-05,
121
+ "loss": 1.9637,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.57,
126
+ "learning_rate": 2.4299720686313628e-05,
127
+ "loss": 1.9302,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.6,
132
+ "learning_rate": 2.401470672062931e-05,
133
+ "loss": 1.9569,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.63,
138
+ "learning_rate": 2.3729692754944995e-05,
139
+ "loss": 1.9173,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.66,
144
+ "learning_rate": 2.344467878926067e-05,
145
+ "loss": 1.8948,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.68,
150
+ "learning_rate": 2.3159664823576355e-05,
151
+ "loss": 1.8932,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.71,
156
+ "learning_rate": 2.2874650857892038e-05,
157
+ "loss": 1.8795,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.74,
162
+ "learning_rate": 2.258963689220772e-05,
163
+ "loss": 1.8534,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.77,
168
+ "learning_rate": 2.2304622926523398e-05,
169
+ "loss": 1.8206,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.8,
174
+ "learning_rate": 2.201960896083908e-05,
175
+ "loss": 1.8155,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.83,
180
+ "learning_rate": 2.1734594995154765e-05,
181
+ "loss": 1.8103,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.86,
186
+ "learning_rate": 2.1449581029470445e-05,
187
+ "loss": 1.7853,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.88,
192
+ "learning_rate": 2.1164567063786125e-05,
193
+ "loss": 1.7908,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.91,
198
+ "learning_rate": 2.0879553098101808e-05,
199
+ "loss": 1.8047,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.94,
204
+ "learning_rate": 2.059453913241749e-05,
205
+ "loss": 1.7753,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.97,
210
+ "learning_rate": 2.0309525166733168e-05,
211
+ "loss": 1.7658,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 1.0,
216
+ "learning_rate": 2.002451120104885e-05,
217
+ "loss": 1.7315,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 1.03,
222
+ "learning_rate": 1.9739497235364535e-05,
223
+ "loss": 1.5751,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 1.05,
228
+ "learning_rate": 1.9454483269680215e-05,
229
+ "loss": 1.5829,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 1.08,
234
+ "learning_rate": 1.9169469303995895e-05,
235
+ "loss": 1.6288,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 1.11,
240
+ "learning_rate": 1.8884455338311578e-05,
241
+ "loss": 1.6086,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 1.14,
246
+ "learning_rate": 1.859944137262726e-05,
247
+ "loss": 1.5986,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 1.17,
252
+ "learning_rate": 1.8314427406942938e-05,
253
+ "loss": 1.5133,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 1.2,
258
+ "learning_rate": 1.802941344125862e-05,
259
+ "loss": 1.5322,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 1.23,
264
+ "learning_rate": 1.7744399475574305e-05,
265
+ "loss": 1.5677,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 1.25,
270
+ "learning_rate": 1.7459385509889985e-05,
271
+ "loss": 1.5486,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 1.28,
276
+ "learning_rate": 1.7174371544205665e-05,
277
+ "loss": 1.5557,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 1.31,
282
+ "learning_rate": 1.688935757852135e-05,
283
+ "loss": 1.5739,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 1.34,
288
+ "learning_rate": 1.6604343612837032e-05,
289
+ "loss": 1.5718,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 1.37,
294
+ "learning_rate": 1.631932964715271e-05,
295
+ "loss": 1.5901,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 1.4,
300
+ "learning_rate": 1.6034315681468392e-05,
301
+ "loss": 1.5141,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 1.43,
306
+ "learning_rate": 1.5749301715784075e-05,
307
+ "loss": 1.5126,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 1.45,
312
+ "learning_rate": 1.5464287750099755e-05,
313
+ "loss": 1.5208,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 1.48,
318
+ "learning_rate": 1.5179273784415437e-05,
319
+ "loss": 1.5288,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 1.51,
324
+ "learning_rate": 1.4894259818731119e-05,
325
+ "loss": 1.5958,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 1.54,
330
+ "learning_rate": 1.4609245853046799e-05,
331
+ "loss": 1.5581,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 1.57,
336
+ "learning_rate": 1.4324231887362482e-05,
337
+ "loss": 1.4951,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 1.6,
342
+ "learning_rate": 1.4039217921678162e-05,
343
+ "loss": 1.53,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 1.62,
348
+ "learning_rate": 1.3754203955993845e-05,
349
+ "loss": 1.5718,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 1.65,
354
+ "learning_rate": 1.3469189990309525e-05,
355
+ "loss": 1.5892,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 1.68,
360
+ "learning_rate": 1.3184176024625207e-05,
361
+ "loss": 1.4618,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 1.71,
366
+ "learning_rate": 1.2899162058940889e-05,
367
+ "loss": 1.5026,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 1.74,
372
+ "learning_rate": 1.2614148093256569e-05,
373
+ "loss": 1.5375,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 1.77,
378
+ "learning_rate": 1.2329134127572252e-05,
379
+ "loss": 1.5061,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 1.8,
384
+ "learning_rate": 1.2044120161887932e-05,
385
+ "loss": 1.4563,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 1.82,
390
+ "learning_rate": 1.1759106196203616e-05,
391
+ "loss": 1.5999,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 1.85,
396
+ "learning_rate": 1.1474092230519296e-05,
397
+ "loss": 1.4604,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 1.88,
402
+ "learning_rate": 1.1189078264834977e-05,
403
+ "loss": 1.4915,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 1.91,
408
+ "learning_rate": 1.0904064299150659e-05,
409
+ "loss": 1.5452,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 1.94,
414
+ "learning_rate": 1.0619050333466339e-05,
415
+ "loss": 1.4683,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 1.97,
420
+ "learning_rate": 1.0334036367782022e-05,
421
+ "loss": 1.5058,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 2.0,
426
+ "learning_rate": 1.0049022402097702e-05,
427
+ "loss": 1.4581,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 2.02,
432
+ "learning_rate": 9.764008436413386e-06,
433
+ "loss": 1.3859,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 2.05,
438
+ "learning_rate": 9.478994470729066e-06,
439
+ "loss": 1.356,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 2.08,
444
+ "learning_rate": 9.193980505044747e-06,
445
+ "loss": 1.3701,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 2.11,
450
+ "learning_rate": 8.908966539360429e-06,
451
+ "loss": 1.341,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 2.14,
456
+ "learning_rate": 8.623952573676109e-06,
457
+ "loss": 1.3811,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 2.17,
462
+ "learning_rate": 8.338938607991792e-06,
463
+ "loss": 1.3036,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 2.19,
468
+ "learning_rate": 8.053924642307472e-06,
469
+ "loss": 1.3512,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 2.22,
474
+ "learning_rate": 7.768910676623156e-06,
475
+ "loss": 1.3579,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 2.25,
480
+ "learning_rate": 7.483896710938836e-06,
481
+ "loss": 1.3188,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 2.28,
486
+ "learning_rate": 7.1988827452545175e-06,
487
+ "loss": 1.3191,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 2.31,
492
+ "learning_rate": 6.913868779570199e-06,
493
+ "loss": 1.3198,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 2.34,
498
+ "learning_rate": 6.628854813885881e-06,
499
+ "loss": 1.4049,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 2.37,
504
+ "learning_rate": 6.343840848201562e-06,
505
+ "loss": 1.3345,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 2.39,
510
+ "learning_rate": 6.0588268825172434e-06,
511
+ "loss": 1.3574,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 2.42,
516
+ "learning_rate": 5.773812916832925e-06,
517
+ "loss": 1.3654,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 2.45,
522
+ "learning_rate": 5.488798951148607e-06,
523
+ "loss": 1.3169,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 2.48,
528
+ "learning_rate": 5.203784985464288e-06,
529
+ "loss": 1.3223,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 2.51,
534
+ "learning_rate": 4.918771019779969e-06,
535
+ "loss": 1.345,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 2.54,
540
+ "learning_rate": 4.633757054095651e-06,
541
+ "loss": 1.3325,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 2.57,
546
+ "learning_rate": 4.348743088411332e-06,
547
+ "loss": 1.2813,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 2.59,
552
+ "learning_rate": 4.063729122727014e-06,
553
+ "loss": 1.3331,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 2.62,
558
+ "learning_rate": 3.778715157042695e-06,
559
+ "loss": 1.343,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 2.65,
564
+ "learning_rate": 3.4937011913583766e-06,
565
+ "loss": 1.3073,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 2.68,
570
+ "learning_rate": 3.2086872256740582e-06,
571
+ "loss": 1.3007,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 2.71,
576
+ "learning_rate": 2.92367325998974e-06,
577
+ "loss": 1.3505,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 2.74,
582
+ "learning_rate": 2.6386592943054208e-06,
583
+ "loss": 1.352,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 2.76,
588
+ "learning_rate": 2.3536453286211025e-06,
589
+ "loss": 1.3208,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 2.79,
594
+ "learning_rate": 2.068631362936784e-06,
595
+ "loss": 1.3556,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 2.82,
600
+ "learning_rate": 1.7836173972524654e-06,
601
+ "loss": 1.3009,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 2.85,
606
+ "learning_rate": 1.498603431568147e-06,
607
+ "loss": 1.3798,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 2.88,
612
+ "learning_rate": 1.2135894658838284e-06,
613
+ "loss": 1.2971,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 2.91,
618
+ "learning_rate": 9.285755001995098e-07,
619
+ "loss": 1.3536,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 2.94,
624
+ "learning_rate": 6.435615345151913e-07,
625
+ "loss": 1.3207,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 2.96,
630
+ "learning_rate": 3.5854756883087273e-07,
631
+ "loss": 1.2801,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 2.99,
636
+ "learning_rate": 7.353360314655419e-08,
637
+ "loss": 1.3456,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 3.0,
642
+ "step": 52629,
643
+ "total_flos": 2554509580406784.0,
644
+ "train_loss": 1.6810464263743796,
645
+ "train_runtime": 1840.7807,
646
+ "train_samples_per_second": 142.951,
647
+ "train_steps_per_second": 28.591
648
+ }
649
+ ],
650
+ "max_steps": 52629,
651
+ "num_train_epochs": 3,
652
+ "total_flos": 2554509580406784.0,
653
+ "trial_name": null,
654
+ "trial_params": null
655
+ }