Davlan commited on
Commit
0fae682
1 Parent(s): 024aa2a

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +799 -0
trainer_state.json ADDED
@@ -0,0 +1,799 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.9999768663104076,
5
+ "global_step": 64839,
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": 4.9614429587131204e-05,
13
+ "loss": 1.9292,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.05,
18
+ "learning_rate": 4.9228859174262406e-05,
19
+ "loss": 1.8232,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.07,
24
+ "learning_rate": 4.884328876139361e-05,
25
+ "loss": 1.7474,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.09,
30
+ "learning_rate": 4.845771834852481e-05,
31
+ "loss": 1.6787,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.12,
36
+ "learning_rate": 4.807214793565602e-05,
37
+ "loss": 1.6332,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.14,
42
+ "learning_rate": 4.768657752278721e-05,
43
+ "loss": 1.584,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.16,
48
+ "learning_rate": 4.7301007109918414e-05,
49
+ "loss": 1.5654,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.19,
54
+ "learning_rate": 4.691543669704962e-05,
55
+ "loss": 1.5113,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.21,
60
+ "learning_rate": 4.652986628418082e-05,
61
+ "loss": 1.4909,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.23,
66
+ "learning_rate": 4.614429587131202e-05,
67
+ "loss": 1.4654,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.25,
72
+ "learning_rate": 4.575872545844323e-05,
73
+ "loss": 1.4466,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.28,
78
+ "learning_rate": 4.537315504557442e-05,
79
+ "loss": 1.4305,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.3,
84
+ "learning_rate": 4.4987584632705625e-05,
85
+ "loss": 1.3963,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.32,
90
+ "learning_rate": 4.4602014219836826e-05,
91
+ "loss": 1.388,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.35,
96
+ "learning_rate": 4.4216443806968035e-05,
97
+ "loss": 1.3647,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.37,
102
+ "learning_rate": 4.383087339409923e-05,
103
+ "loss": 1.3549,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.39,
108
+ "learning_rate": 4.344530298123043e-05,
109
+ "loss": 1.336,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.42,
114
+ "learning_rate": 4.305973256836164e-05,
115
+ "loss": 1.3306,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.44,
120
+ "learning_rate": 4.2674162155492835e-05,
121
+ "loss": 1.3242,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.46,
126
+ "learning_rate": 4.2288591742624036e-05,
127
+ "loss": 1.3018,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.49,
132
+ "learning_rate": 4.1903021329755245e-05,
133
+ "loss": 1.2957,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.51,
138
+ "learning_rate": 4.151745091688644e-05,
139
+ "loss": 1.2851,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.53,
144
+ "learning_rate": 4.113188050401764e-05,
145
+ "loss": 1.2768,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.56,
150
+ "learning_rate": 4.074631009114885e-05,
151
+ "loss": 1.2619,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.58,
156
+ "learning_rate": 4.036073967828005e-05,
157
+ "loss": 1.2513,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.6,
162
+ "learning_rate": 3.997516926541125e-05,
163
+ "loss": 1.2438,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.62,
168
+ "learning_rate": 3.9589598852542455e-05,
169
+ "loss": 1.2194,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.65,
174
+ "learning_rate": 3.920402843967366e-05,
175
+ "loss": 1.2121,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.67,
180
+ "learning_rate": 3.881845802680485e-05,
181
+ "loss": 1.2242,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.69,
186
+ "learning_rate": 3.843288761393606e-05,
187
+ "loss": 1.2131,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.72,
192
+ "learning_rate": 3.804731720106726e-05,
193
+ "loss": 1.2157,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.74,
198
+ "learning_rate": 3.7661746788198464e-05,
199
+ "loss": 1.1866,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.76,
204
+ "learning_rate": 3.7276176375329665e-05,
205
+ "loss": 1.1897,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.79,
210
+ "learning_rate": 3.689060596246087e-05,
211
+ "loss": 1.181,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.81,
216
+ "learning_rate": 3.650503554959207e-05,
217
+ "loss": 1.1739,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.83,
222
+ "learning_rate": 3.611946513672327e-05,
223
+ "loss": 1.1757,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.86,
228
+ "learning_rate": 3.573389472385447e-05,
229
+ "loss": 1.1519,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.88,
234
+ "learning_rate": 3.5348324310985674e-05,
235
+ "loss": 1.1544,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.9,
240
+ "learning_rate": 3.4962753898116876e-05,
241
+ "loss": 1.1543,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.93,
246
+ "learning_rate": 3.457718348524808e-05,
247
+ "loss": 1.1473,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.95,
252
+ "learning_rate": 3.419161307237928e-05,
253
+ "loss": 1.1385,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.97,
258
+ "learning_rate": 3.380604265951048e-05,
259
+ "loss": 1.1219,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.99,
264
+ "learning_rate": 3.342047224664168e-05,
265
+ "loss": 1.1335,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 1.02,
270
+ "learning_rate": 3.3034901833772884e-05,
271
+ "loss": 1.1096,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 1.04,
276
+ "learning_rate": 3.2649331420904086e-05,
277
+ "loss": 1.1093,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 1.06,
282
+ "learning_rate": 3.226376100803529e-05,
283
+ "loss": 1.1033,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 1.09,
288
+ "learning_rate": 3.187819059516649e-05,
289
+ "loss": 1.0932,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 1.11,
294
+ "learning_rate": 3.149262018229769e-05,
295
+ "loss": 1.0927,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 1.13,
300
+ "learning_rate": 3.110704976942889e-05,
301
+ "loss": 1.0957,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 1.16,
306
+ "learning_rate": 3.0721479356560094e-05,
307
+ "loss": 1.0855,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 1.18,
312
+ "learning_rate": 3.03359089436913e-05,
313
+ "loss": 1.089,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 1.2,
318
+ "learning_rate": 2.99503385308225e-05,
319
+ "loss": 1.0788,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 1.23,
324
+ "learning_rate": 2.95647681179537e-05,
325
+ "loss": 1.0798,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 1.25,
330
+ "learning_rate": 2.9179197705084905e-05,
331
+ "loss": 1.07,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 1.27,
336
+ "learning_rate": 2.8793627292216106e-05,
337
+ "loss": 1.0736,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 1.3,
342
+ "learning_rate": 2.8408056879347305e-05,
343
+ "loss": 1.0503,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 1.32,
348
+ "learning_rate": 2.802248646647851e-05,
349
+ "loss": 1.0581,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 1.34,
354
+ "learning_rate": 2.763691605360971e-05,
355
+ "loss": 1.048,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 1.36,
360
+ "learning_rate": 2.725134564074091e-05,
361
+ "loss": 1.044,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 1.39,
366
+ "learning_rate": 2.6865775227872115e-05,
367
+ "loss": 1.0392,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 1.41,
372
+ "learning_rate": 2.6480204815003317e-05,
373
+ "loss": 1.0381,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 1.43,
378
+ "learning_rate": 2.609463440213452e-05,
379
+ "loss": 1.0404,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 1.46,
384
+ "learning_rate": 2.570906398926572e-05,
385
+ "loss": 1.0296,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 1.48,
390
+ "learning_rate": 2.532349357639692e-05,
391
+ "loss": 1.0272,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 1.5,
396
+ "learning_rate": 2.4937923163528123e-05,
397
+ "loss": 1.0349,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 1.53,
402
+ "learning_rate": 2.455235275065933e-05,
403
+ "loss": 1.0093,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 1.55,
408
+ "learning_rate": 2.4166782337790527e-05,
409
+ "loss": 1.0055,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 1.57,
414
+ "learning_rate": 2.3781211924921732e-05,
415
+ "loss": 1.0172,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 1.6,
420
+ "learning_rate": 2.3395641512052934e-05,
421
+ "loss": 1.007,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 1.62,
426
+ "learning_rate": 2.3010071099184132e-05,
427
+ "loss": 0.9954,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 1.64,
432
+ "learning_rate": 2.2624500686315337e-05,
433
+ "loss": 1.0058,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 1.67,
438
+ "learning_rate": 2.223893027344654e-05,
439
+ "loss": 1.0001,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 1.69,
444
+ "learning_rate": 2.185335986057774e-05,
445
+ "loss": 0.9934,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 1.71,
450
+ "learning_rate": 2.1467789447708942e-05,
451
+ "loss": 0.9829,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 1.74,
456
+ "learning_rate": 2.1082219034840144e-05,
457
+ "loss": 0.9994,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 1.76,
462
+ "learning_rate": 2.0696648621971345e-05,
463
+ "loss": 0.9782,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 1.78,
468
+ "learning_rate": 2.0311078209102547e-05,
469
+ "loss": 0.9843,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 1.8,
474
+ "learning_rate": 1.992550779623375e-05,
475
+ "loss": 0.981,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 1.83,
480
+ "learning_rate": 1.953993738336495e-05,
481
+ "loss": 0.98,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 1.85,
486
+ "learning_rate": 1.9154366970496156e-05,
487
+ "loss": 0.9663,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 1.87,
492
+ "learning_rate": 1.8768796557627354e-05,
493
+ "loss": 0.9705,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 1.9,
498
+ "learning_rate": 1.8383226144758556e-05,
499
+ "loss": 0.9871,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 1.92,
504
+ "learning_rate": 1.799765573188976e-05,
505
+ "loss": 0.9679,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 1.94,
510
+ "learning_rate": 1.761208531902096e-05,
511
+ "loss": 0.9672,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 1.97,
516
+ "learning_rate": 1.7226514906152164e-05,
517
+ "loss": 0.9618,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 1.99,
522
+ "learning_rate": 1.6840944493283366e-05,
523
+ "loss": 0.9621,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 2.01,
528
+ "learning_rate": 1.6455374080414564e-05,
529
+ "loss": 0.9445,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 2.04,
534
+ "learning_rate": 1.606980366754577e-05,
535
+ "loss": 0.9448,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 2.06,
540
+ "learning_rate": 1.568423325467697e-05,
541
+ "loss": 0.9405,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 2.08,
546
+ "learning_rate": 1.5298662841808173e-05,
547
+ "loss": 0.9352,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 2.11,
552
+ "learning_rate": 1.4913092428939374e-05,
553
+ "loss": 0.9308,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 2.13,
558
+ "learning_rate": 1.4527522016070574e-05,
559
+ "loss": 0.9281,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 2.15,
564
+ "learning_rate": 1.4141951603201778e-05,
565
+ "loss": 0.9344,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 2.17,
570
+ "learning_rate": 1.3756381190332978e-05,
571
+ "loss": 0.9337,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 2.2,
576
+ "learning_rate": 1.3370810777464183e-05,
577
+ "loss": 0.9272,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 2.22,
582
+ "learning_rate": 1.2985240364595383e-05,
583
+ "loss": 0.9302,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 2.24,
588
+ "learning_rate": 1.2599669951726583e-05,
589
+ "loss": 0.9297,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 2.27,
594
+ "learning_rate": 1.2214099538857788e-05,
595
+ "loss": 0.9238,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 2.29,
600
+ "learning_rate": 1.1828529125988988e-05,
601
+ "loss": 0.9181,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 2.31,
606
+ "learning_rate": 1.144295871312019e-05,
607
+ "loss": 0.9133,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 2.34,
612
+ "learning_rate": 1.1057388300251391e-05,
613
+ "loss": 0.9244,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 2.36,
618
+ "learning_rate": 1.0671817887382595e-05,
619
+ "loss": 0.9086,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 2.38,
624
+ "learning_rate": 1.0286247474513797e-05,
625
+ "loss": 0.9055,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 2.41,
630
+ "learning_rate": 9.900677061644998e-06,
631
+ "loss": 0.9084,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 2.43,
636
+ "learning_rate": 9.5151066487762e-06,
637
+ "loss": 0.9082,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 2.45,
642
+ "learning_rate": 9.129536235907402e-06,
643
+ "loss": 0.9062,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 2.48,
648
+ "learning_rate": 8.743965823038603e-06,
649
+ "loss": 0.8988,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 2.5,
654
+ "learning_rate": 8.358395410169805e-06,
655
+ "loss": 0.9025,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 2.52,
660
+ "learning_rate": 7.972824997301008e-06,
661
+ "loss": 0.8846,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 2.54,
666
+ "learning_rate": 7.5872545844322085e-06,
667
+ "loss": 0.9002,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 2.57,
672
+ "learning_rate": 7.201684171563411e-06,
673
+ "loss": 0.8875,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 2.59,
678
+ "learning_rate": 6.816113758694613e-06,
679
+ "loss": 0.8902,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 2.61,
684
+ "learning_rate": 6.430543345825815e-06,
685
+ "loss": 0.894,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 2.64,
690
+ "learning_rate": 6.044972932957017e-06,
691
+ "loss": 0.8845,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 2.66,
696
+ "learning_rate": 5.659402520088219e-06,
697
+ "loss": 0.8866,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 2.68,
702
+ "learning_rate": 5.273832107219421e-06,
703
+ "loss": 0.8833,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 2.71,
708
+ "learning_rate": 4.888261694350622e-06,
709
+ "loss": 0.8906,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 2.73,
714
+ "learning_rate": 4.502691281481825e-06,
715
+ "loss": 0.8869,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 2.75,
720
+ "learning_rate": 4.117120868613026e-06,
721
+ "loss": 0.8848,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 2.78,
726
+ "learning_rate": 3.731550455744228e-06,
727
+ "loss": 0.8744,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 2.8,
732
+ "learning_rate": 3.34598004287543e-06,
733
+ "loss": 0.8698,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 2.82,
738
+ "learning_rate": 2.960409630006632e-06,
739
+ "loss": 0.876,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 2.85,
744
+ "learning_rate": 2.574839217137834e-06,
745
+ "loss": 0.8739,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 2.87,
750
+ "learning_rate": 2.1892688042690357e-06,
751
+ "loss": 0.8827,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 2.89,
756
+ "learning_rate": 1.8036983914002379e-06,
757
+ "loss": 0.8773,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 2.91,
762
+ "learning_rate": 1.4181279785314396e-06,
763
+ "loss": 0.8599,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 2.94,
768
+ "learning_rate": 1.0325575656626413e-06,
769
+ "loss": 0.8711,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 2.96,
774
+ "learning_rate": 6.469871527938432e-07,
775
+ "loss": 0.8778,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 2.98,
780
+ "learning_rate": 2.614167399250451e-07,
781
+ "loss": 0.8713,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 3.0,
786
+ "step": 64839,
787
+ "total_flos": 6.047604206657864e+17,
788
+ "train_loss": 1.089792190311953,
789
+ "train_runtime": 32594.2653,
790
+ "train_samples_per_second": 19.893,
791
+ "train_steps_per_second": 1.989
792
+ }
793
+ ],
794
+ "max_steps": 64839,
795
+ "num_train_epochs": 3,
796
+ "total_flos": 6.047604206657864e+17,
797
+ "trial_name": null,
798
+ "trial_params": null
799
+ }