tuanio commited on
Commit
e6fbce7
1 Parent(s): b7f0732

End of training

Browse files
Files changed (1) hide show
  1. trainer_state.json +664 -0
trainer_state.json ADDED
@@ -0,0 +1,664 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 100.0,
5
+ "eval_steps": 100,
6
+ "global_step": 10600,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.94,
13
+ "learning_rate": 1.3773584905660377e-06,
14
+ "loss": 126.7325,
15
+ "step": 100
16
+ },
17
+ {
18
+ "epoch": 1.89,
19
+ "learning_rate": 3.2641509433962264e-06,
20
+ "loss": 92.4996,
21
+ "step": 200
22
+ },
23
+ {
24
+ "epoch": 2.83,
25
+ "learning_rate": 5.150943396226416e-06,
26
+ "loss": 46.8471,
27
+ "step": 300
28
+ },
29
+ {
30
+ "epoch": 3.77,
31
+ "learning_rate": 7.037735849056604e-06,
32
+ "loss": 29.0405,
33
+ "step": 400
34
+ },
35
+ {
36
+ "epoch": 4.72,
37
+ "learning_rate": 8.924528301886793e-06,
38
+ "loss": 21.3856,
39
+ "step": 500
40
+ },
41
+ {
42
+ "epoch": 5.66,
43
+ "learning_rate": 1.0811320754716984e-05,
44
+ "loss": 16.0048,
45
+ "step": 600
46
+ },
47
+ {
48
+ "epoch": 6.6,
49
+ "learning_rate": 1.2698113207547172e-05,
50
+ "loss": 11.667,
51
+ "step": 700
52
+ },
53
+ {
54
+ "epoch": 7.55,
55
+ "learning_rate": 1.458490566037736e-05,
56
+ "loss": 8.5327,
57
+ "step": 800
58
+ },
59
+ {
60
+ "epoch": 8.49,
61
+ "learning_rate": 1.6471698113207547e-05,
62
+ "loss": 6.5511,
63
+ "step": 900
64
+ },
65
+ {
66
+ "epoch": 9.43,
67
+ "learning_rate": 1.8358490566037736e-05,
68
+ "loss": 5.4833,
69
+ "step": 1000
70
+ },
71
+ {
72
+ "epoch": 10.38,
73
+ "learning_rate": 1.9999908365494266e-05,
74
+ "loss": 4.8445,
75
+ "step": 1100
76
+ },
77
+ {
78
+ "epoch": 11.32,
79
+ "learning_rate": 1.999307723208382e-05,
80
+ "loss": 4.5783,
81
+ "step": 1200
82
+ },
83
+ {
84
+ "epoch": 12.26,
85
+ "learning_rate": 1.9975410248472147e-05,
86
+ "loss": 4.458,
87
+ "step": 1300
88
+ },
89
+ {
90
+ "epoch": 13.21,
91
+ "learning_rate": 1.9946926571599945e-05,
92
+ "loss": 4.4055,
93
+ "step": 1400
94
+ },
95
+ {
96
+ "epoch": 14.15,
97
+ "learning_rate": 1.9907657087334364e-05,
98
+ "loss": 4.3862,
99
+ "step": 1500
100
+ },
101
+ {
102
+ "epoch": 15.09,
103
+ "learning_rate": 1.9857644376978347e-05,
104
+ "loss": 4.3751,
105
+ "step": 1600
106
+ },
107
+ {
108
+ "epoch": 16.04,
109
+ "learning_rate": 1.9796942671098212e-05,
110
+ "loss": 4.3656,
111
+ "step": 1700
112
+ },
113
+ {
114
+ "epoch": 16.98,
115
+ "learning_rate": 1.9725617790719523e-05,
116
+ "loss": 4.3571,
117
+ "step": 1800
118
+ },
119
+ {
120
+ "epoch": 17.92,
121
+ "learning_rate": 1.9643747075954974e-05,
122
+ "loss": 4.3475,
123
+ "step": 1900
124
+ },
125
+ {
126
+ "epoch": 18.87,
127
+ "learning_rate": 1.9551419302141784e-05,
128
+ "loss": 4.3313,
129
+ "step": 2000
130
+ },
131
+ {
132
+ "epoch": 19.81,
133
+ "learning_rate": 1.9448734583579404e-05,
134
+ "loss": 4.3083,
135
+ "step": 2100
136
+ },
137
+ {
138
+ "epoch": 20.75,
139
+ "learning_rate": 1.9335804264972018e-05,
140
+ "loss": 4.2959,
141
+ "step": 2200
142
+ },
143
+ {
144
+ "epoch": 21.7,
145
+ "learning_rate": 1.9212750800693506e-05,
146
+ "loss": 4.2644,
147
+ "step": 2300
148
+ },
149
+ {
150
+ "epoch": 22.64,
151
+ "learning_rate": 1.90797076220058e-05,
152
+ "loss": 4.2342,
153
+ "step": 2400
154
+ },
155
+ {
156
+ "epoch": 23.58,
157
+ "learning_rate": 1.8936818992374606e-05,
158
+ "loss": 4.1978,
159
+ "step": 2500
160
+ },
161
+ {
162
+ "epoch": 24.53,
163
+ "learning_rate": 1.8784239851039375e-05,
164
+ "loss": 4.1374,
165
+ "step": 2600
166
+ },
167
+ {
168
+ "epoch": 25.47,
169
+ "learning_rate": 1.8622135645007168e-05,
170
+ "loss": 3.9802,
171
+ "step": 2700
172
+ },
173
+ {
174
+ "epoch": 26.42,
175
+ "learning_rate": 1.845068214965255e-05,
176
+ "loss": 3.6945,
177
+ "step": 2800
178
+ },
179
+ {
180
+ "epoch": 27.36,
181
+ "learning_rate": 1.8270065278118092e-05,
182
+ "loss": 3.3409,
183
+ "step": 2900
184
+ },
185
+ {
186
+ "epoch": 28.3,
187
+ "learning_rate": 1.8080480879722103e-05,
188
+ "loss": 2.8779,
189
+ "step": 3000
190
+ },
191
+ {
192
+ "epoch": 29.25,
193
+ "learning_rate": 1.7882134527592244e-05,
194
+ "loss": 2.4274,
195
+ "step": 3100
196
+ },
197
+ {
198
+ "epoch": 30.19,
199
+ "learning_rate": 1.7679461483820617e-05,
200
+ "loss": 2.1192,
201
+ "step": 3200
202
+ },
203
+ {
204
+ "epoch": 31.13,
205
+ "learning_rate": 1.7464409907419976e-05,
206
+ "loss": 1.8713,
207
+ "step": 3300
208
+ },
209
+ {
210
+ "epoch": 32.08,
211
+ "learning_rate": 1.7241264405022477e-05,
212
+ "loss": 1.6762,
213
+ "step": 3400
214
+ },
215
+ {
216
+ "epoch": 33.02,
217
+ "learning_rate": 1.701026694125833e-05,
218
+ "loss": 1.5089,
219
+ "step": 3500
220
+ },
221
+ {
222
+ "epoch": 33.96,
223
+ "learning_rate": 1.6774090768906323e-05,
224
+ "loss": 1.3725,
225
+ "step": 3600
226
+ },
227
+ {
228
+ "epoch": 34.91,
229
+ "learning_rate": 1.6528221181905217e-05,
230
+ "loss": 1.2504,
231
+ "step": 3700
232
+ },
233
+ {
234
+ "epoch": 35.85,
235
+ "learning_rate": 1.6275272811746063e-05,
236
+ "loss": 1.1491,
237
+ "step": 3800
238
+ },
239
+ {
240
+ "epoch": 36.79,
241
+ "learning_rate": 1.6018150231520486e-05,
242
+ "loss": 1.0699,
243
+ "step": 3900
244
+ },
245
+ {
246
+ "epoch": 37.74,
247
+ "learning_rate": 1.575193832697346e-05,
248
+ "loss": 0.9869,
249
+ "step": 4000
250
+ },
251
+ {
252
+ "epoch": 38.68,
253
+ "learning_rate": 1.5479489390466186e-05,
254
+ "loss": 0.9305,
255
+ "step": 4100
256
+ },
257
+ {
258
+ "epoch": 39.62,
259
+ "learning_rate": 1.520109884810164e-05,
260
+ "loss": 0.8711,
261
+ "step": 4200
262
+ },
263
+ {
264
+ "epoch": 40.57,
265
+ "learning_rate": 1.4917068568678037e-05,
266
+ "loss": 0.8215,
267
+ "step": 4300
268
+ },
269
+ {
270
+ "epoch": 41.51,
271
+ "learning_rate": 1.4627706536361717e-05,
272
+ "loss": 0.7728,
273
+ "step": 4400
274
+ },
275
+ {
276
+ "epoch": 42.45,
277
+ "learning_rate": 1.4333326516728909e-05,
278
+ "loss": 0.7473,
279
+ "step": 4500
280
+ },
281
+ {
282
+ "epoch": 43.4,
283
+ "learning_rate": 1.4034247716538545e-05,
284
+ "loss": 0.7033,
285
+ "step": 4600
286
+ },
287
+ {
288
+ "epoch": 44.34,
289
+ "learning_rate": 1.373079443760497e-05,
290
+ "loss": 0.6874,
291
+ "step": 4700
292
+ },
293
+ {
294
+ "epoch": 45.28,
295
+ "learning_rate": 1.3423295725145908e-05,
296
+ "loss": 0.6774,
297
+ "step": 4800
298
+ },
299
+ {
300
+ "epoch": 46.23,
301
+ "learning_rate": 1.3112085010987002e-05,
302
+ "loss": 0.6289,
303
+ "step": 4900
304
+ },
305
+ {
306
+ "epoch": 47.17,
307
+ "learning_rate": 1.279749975200978e-05,
308
+ "loss": 0.6088,
309
+ "step": 5000
310
+ },
311
+ {
312
+ "epoch": 48.11,
313
+ "learning_rate": 1.247988106423513e-05,
314
+ "loss": 0.5807,
315
+ "step": 5100
316
+ },
317
+ {
318
+ "epoch": 49.06,
319
+ "learning_rate": 1.2159573352939014e-05,
320
+ "loss": 0.5688,
321
+ "step": 5200
322
+ },
323
+ {
324
+ "epoch": 50.0,
325
+ "learning_rate": 1.1836923939201543e-05,
326
+ "loss": 0.5522,
327
+ "step": 5300
328
+ },
329
+ {
330
+ "epoch": 50.94,
331
+ "learning_rate": 1.1512282683294342e-05,
332
+ "loss": 0.5322,
333
+ "step": 5400
334
+ },
335
+ {
336
+ "epoch": 51.89,
337
+ "learning_rate": 1.1186001605314571e-05,
338
+ "loss": 0.5294,
339
+ "step": 5500
340
+ },
341
+ {
342
+ "epoch": 52.83,
343
+ "learning_rate": 1.0861715375003623e-05,
344
+ "loss": 0.5092,
345
+ "step": 5600
346
+ },
347
+ {
348
+ "epoch": 53.77,
349
+ "learning_rate": 1.0533224988478176e-05,
350
+ "loss": 0.5154,
351
+ "step": 5700
352
+ },
353
+ {
354
+ "epoch": 54.72,
355
+ "learning_rate": 1.0204156407071964e-05,
356
+ "loss": 0.5496,
357
+ "step": 5800
358
+ },
359
+ {
360
+ "epoch": 55.66,
361
+ "learning_rate": 9.874866451589151e-06,
362
+ "loss": 0.501,
363
+ "step": 5900
364
+ },
365
+ {
366
+ "epoch": 56.6,
367
+ "learning_rate": 9.545712182877714e-06,
368
+ "loss": 0.489,
369
+ "step": 6000
370
+ },
371
+ {
372
+ "epoch": 57.55,
373
+ "learning_rate": 9.21705051465562e-06,
374
+ "loss": 0.4805,
375
+ "step": 6100
376
+ },
377
+ {
378
+ "epoch": 58.49,
379
+ "learning_rate": 8.889237826496551e-06,
380
+ "loss": 0.4786,
381
+ "step": 6200
382
+ },
383
+ {
384
+ "epoch": 59.43,
385
+ "learning_rate": 8.562629577394817e-06,
386
+ "loss": 0.4576,
387
+ "step": 6300
388
+ },
389
+ {
390
+ "epoch": 60.38,
391
+ "learning_rate": 8.237579920328478e-06,
392
+ "loss": 0.4606,
393
+ "step": 6400
394
+ },
395
+ {
396
+ "epoch": 61.32,
397
+ "learning_rate": 7.91444131823864e-06,
398
+ "loss": 0.438,
399
+ "step": 6500
400
+ },
401
+ {
402
+ "epoch": 62.26,
403
+ "learning_rate": 7.593564161841318e-06,
404
+ "loss": 0.4471,
405
+ "step": 6600
406
+ },
407
+ {
408
+ "epoch": 63.21,
409
+ "learning_rate": 7.275296389686258e-06,
410
+ "loss": 0.4348,
411
+ "step": 6700
412
+ },
413
+ {
414
+ "epoch": 64.15,
415
+ "learning_rate": 6.959983110874782e-06,
416
+ "loss": 0.429,
417
+ "step": 6800
418
+ },
419
+ {
420
+ "epoch": 65.09,
421
+ "learning_rate": 6.647966230845655e-06,
422
+ "loss": 0.4101,
423
+ "step": 6900
424
+ },
425
+ {
426
+ "epoch": 66.04,
427
+ "learning_rate": 6.339584080634824e-06,
428
+ "loss": 0.4191,
429
+ "step": 7000
430
+ },
431
+ {
432
+ "epoch": 66.98,
433
+ "learning_rate": 6.03517105001101e-06,
434
+ "loss": 0.422,
435
+ "step": 7100
436
+ },
437
+ {
438
+ "epoch": 67.92,
439
+ "learning_rate": 5.735057224884939e-06,
440
+ "loss": 0.418,
441
+ "step": 7200
442
+ },
443
+ {
444
+ "epoch": 68.87,
445
+ "learning_rate": 5.439568029385422e-06,
446
+ "loss": 0.3971,
447
+ "step": 7300
448
+ },
449
+ {
450
+ "epoch": 69.81,
451
+ "learning_rate": 5.14902387299036e-06,
452
+ "loss": 0.3938,
453
+ "step": 7400
454
+ },
455
+ {
456
+ "epoch": 70.75,
457
+ "learning_rate": 4.863739803095299e-06,
458
+ "loss": 0.387,
459
+ "step": 7500
460
+ },
461
+ {
462
+ "epoch": 71.7,
463
+ "learning_rate": 4.584025163396323e-06,
464
+ "loss": 0.3915,
465
+ "step": 7600
466
+ },
467
+ {
468
+ "epoch": 72.64,
469
+ "learning_rate": 4.310183258457632e-06,
470
+ "loss": 0.3865,
471
+ "step": 7700
472
+ },
473
+ {
474
+ "epoch": 73.58,
475
+ "learning_rate": 4.042511024827573e-06,
476
+ "loss": 0.3771,
477
+ "step": 7800
478
+ },
479
+ {
480
+ "epoch": 74.53,
481
+ "learning_rate": 3.7838779196384036e-06,
482
+ "loss": 0.3776,
483
+ "step": 7900
484
+ },
485
+ {
486
+ "epoch": 75.47,
487
+ "learning_rate": 3.5293399536806903e-06,
488
+ "loss": 0.3716,
489
+ "step": 8000
490
+ },
491
+ {
492
+ "epoch": 76.42,
493
+ "learning_rate": 3.2818183552879e-06,
494
+ "loss": 0.3717,
495
+ "step": 8100
496
+ },
497
+ {
498
+ "epoch": 77.36,
499
+ "learning_rate": 3.041581520960881e-06,
500
+ "loss": 0.3701,
501
+ "step": 8200
502
+ },
503
+ {
504
+ "epoch": 78.3,
505
+ "learning_rate": 2.8088899480708677e-06,
506
+ "loss": 0.369,
507
+ "step": 8300
508
+ },
509
+ {
510
+ "epoch": 79.25,
511
+ "learning_rate": 2.5839959523929126e-06,
512
+ "loss": 0.3556,
513
+ "step": 8400
514
+ },
515
+ {
516
+ "epoch": 80.19,
517
+ "learning_rate": 2.3671433945108712e-06,
518
+ "loss": 0.3618,
519
+ "step": 8500
520
+ },
521
+ {
522
+ "epoch": 81.13,
523
+ "learning_rate": 2.15856741539067e-06,
524
+ "loss": 0.367,
525
+ "step": 8600
526
+ },
527
+ {
528
+ "epoch": 82.08,
529
+ "learning_rate": 1.9604521050536506e-06,
530
+ "loss": 0.3557,
531
+ "step": 8700
532
+ },
533
+ {
534
+ "epoch": 83.02,
535
+ "learning_rate": 1.7690103228094845e-06,
536
+ "loss": 0.3637,
537
+ "step": 8800
538
+ },
539
+ {
540
+ "epoch": 83.96,
541
+ "learning_rate": 1.5864936963585164e-06,
542
+ "loss": 0.3524,
543
+ "step": 8900
544
+ },
545
+ {
546
+ "epoch": 84.91,
547
+ "learning_rate": 1.413100134991111e-06,
548
+ "loss": 0.3581,
549
+ "step": 9000
550
+ },
551
+ {
552
+ "epoch": 85.85,
553
+ "learning_rate": 1.2490176555326306e-06,
554
+ "loss": 0.3546,
555
+ "step": 9100
556
+ },
557
+ {
558
+ "epoch": 86.79,
559
+ "learning_rate": 1.0944241784701025e-06,
560
+ "loss": 0.352,
561
+ "step": 9200
562
+ },
563
+ {
564
+ "epoch": 87.74,
565
+ "learning_rate": 9.49487335026692e-07,
566
+ "loss": 0.3451,
567
+ "step": 9300
568
+ },
569
+ {
570
+ "epoch": 88.68,
571
+ "learning_rate": 8.143642853931688e-07,
572
+ "loss": 0.3632,
573
+ "step": 9400
574
+ },
575
+ {
576
+ "epoch": 89.62,
577
+ "learning_rate": 6.892015483135095e-07,
578
+ "loss": 0.3469,
579
+ "step": 9500
580
+ },
581
+ {
582
+ "epoch": 90.57,
583
+ "learning_rate": 5.741348422094128e-07,
584
+ "loss": 0.339,
585
+ "step": 9600
586
+ },
587
+ {
588
+ "epoch": 91.51,
589
+ "learning_rate": 4.69288938015946e-07,
590
+ "loss": 0.3671,
591
+ "step": 9700
592
+ },
593
+ {
594
+ "epoch": 92.45,
595
+ "learning_rate": 3.747775238879847e-07,
596
+ "loss": 0.3441,
597
+ "step": 9800
598
+ },
599
+ {
600
+ "epoch": 93.4,
601
+ "learning_rate": 2.907030819240908e-07,
602
+ "loss": 0.3582,
603
+ "step": 9900
604
+ },
605
+ {
606
+ "epoch": 94.34,
607
+ "learning_rate": 2.1715677704150395e-07,
608
+ "loss": 0.3402,
609
+ "step": 10000
610
+ },
611
+ {
612
+ "epoch": 95.28,
613
+ "learning_rate": 1.5421835812278408e-07,
614
+ "loss": 0.3499,
615
+ "step": 10100
616
+ },
617
+ {
618
+ "epoch": 96.23,
619
+ "learning_rate": 1.0195607154123932e-07,
620
+ "loss": 0.3519,
621
+ "step": 10200
622
+ },
623
+ {
624
+ "epoch": 97.17,
625
+ "learning_rate": 6.042658715895932e-08,
626
+ "loss": 0.348,
627
+ "step": 10300
628
+ },
629
+ {
630
+ "epoch": 98.11,
631
+ "learning_rate": 2.967493687765277e-08,
632
+ "loss": 0.3553,
633
+ "step": 10400
634
+ },
635
+ {
636
+ "epoch": 99.06,
637
+ "learning_rate": 9.734465808957095e-09,
638
+ "loss": 0.3431,
639
+ "step": 10500
640
+ },
641
+ {
642
+ "epoch": 100.0,
643
+ "learning_rate": 6.267961171368963e-10,
644
+ "loss": 0.3459,
645
+ "step": 10600
646
+ },
647
+ {
648
+ "epoch": 100.0,
649
+ "step": 10600,
650
+ "total_flos": 6.514875776263782e+16,
651
+ "train_loss": 4.663586199778431,
652
+ "train_runtime": 6099.7809,
653
+ "train_samples_per_second": 55.477,
654
+ "train_steps_per_second": 1.738
655
+ }
656
+ ],
657
+ "logging_steps": 100,
658
+ "max_steps": 10600,
659
+ "num_train_epochs": 100,
660
+ "save_steps": 100,
661
+ "total_flos": 6.514875776263782e+16,
662
+ "trial_name": null,
663
+ "trial_params": null
664
+ }