SushantGautam commited on
Commit
b8235f7
·
1 Parent(s): 4486ebb

End of training

Browse files
Files changed (4) hide show
  1. all_results.json +13 -11
  2. eval_results.json +8 -11
  3. train_results.json +6 -6
  4. trainer_state.json +74 -1538
all_results.json CHANGED
@@ -1,15 +1,17 @@
1
  {
2
- "epoch": 5.0,
 
3
  "eval_gen_len": 1.0,
4
- "eval_loss": 5.282332897186279,
5
  "eval_rouge1": 0.0,
6
  "eval_rouge2": 0.0,
7
  "eval_rougeL": 0.0,
8
  "eval_rougeLsum": 0.0,
9
- "eval_runtime": 0.8186,
10
- "eval_samples": 176,
11
- "eval_samples_per_second": 214.997,
12
- "eval_steps_per_second": 13.437,
 
13
  "predict_gen_len": 1.0,
14
  "predict_loss": 5.266044616699219,
15
  "predict_rouge1": 0.0,
@@ -20,9 +22,9 @@
20
  "predict_samples": 3527,
21
  "predict_samples_per_second": 237.858,
22
  "predict_steps_per_second": 14.904,
23
- "train_loss": 6.612753483034529,
24
- "train_runtime": 33.2807,
25
- "train_samples": 3351,
26
- "train_samples_per_second": 503.444,
27
- "train_steps_per_second": 7.963
28
  }
 
1
  {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.5793418051482567,
4
  "eval_gen_len": 1.0,
5
+ "eval_loss": 2.115241050720215,
6
  "eval_rouge1": 0.0,
7
  "eval_rouge2": 0.0,
8
  "eval_rougeL": 0.0,
9
  "eval_rougeLsum": 0.0,
10
+ "eval_runtime": 0.2024,
11
+ "eval_samples": 3,
12
+ "eval_samples_per_second": 14.82,
13
+ "eval_steps_per_second": 4.94,
14
+ "perplexity": 8.291584217592838,
15
  "predict_gen_len": 1.0,
16
  "predict_loss": 5.266044616699219,
17
  "predict_rouge1": 0.0,
 
22
  "predict_samples": 3527,
23
  "predict_samples_per_second": 237.858,
24
  "predict_steps_per_second": 14.904,
25
+ "train_loss": 2.7257756051563082,
26
+ "train_runtime": 7.0042,
27
+ "train_samples": 49,
28
+ "train_samples_per_second": 20.987,
29
+ "train_steps_per_second": 2.998
30
  }
eval_results.json CHANGED
@@ -1,13 +1,10 @@
1
  {
2
- "epoch": 5.0,
3
- "eval_gen_len": 1.0,
4
- "eval_loss": 5.282332897186279,
5
- "eval_rouge1": 0.0,
6
- "eval_rouge2": 0.0,
7
- "eval_rougeL": 0.0,
8
- "eval_rougeLsum": 0.0,
9
- "eval_runtime": 0.8186,
10
- "eval_samples": 176,
11
- "eval_samples_per_second": 214.997,
12
- "eval_steps_per_second": 13.437
13
  }
 
1
  {
2
+ "epoch": 3.0,
3
+ "eval_accuracy": 0.5793418051482567,
4
+ "eval_loss": 2.115241050720215,
5
+ "eval_runtime": 0.2024,
6
+ "eval_samples": 3,
7
+ "eval_samples_per_second": 14.82,
8
+ "eval_steps_per_second": 4.94,
9
+ "perplexity": 8.291584217592838
 
 
 
10
  }
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 5.0,
3
- "train_loss": 6.612753483034529,
4
- "train_runtime": 33.2807,
5
- "train_samples": 3351,
6
- "train_samples_per_second": 503.444,
7
- "train_steps_per_second": 7.963
8
  }
 
1
  {
2
+ "epoch": 3.0,
3
+ "train_loss": 2.7257756051563082,
4
+ "train_runtime": 7.0042,
5
+ "train_samples": 49,
6
+ "train_samples_per_second": 20.987,
7
+ "train_steps_per_second": 2.998
8
  }
trainer_state.json CHANGED
@@ -1,1615 +1,151 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 5.0,
5
- "global_step": 265,
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": 4.9811320754716985e-05,
13
- "loss": 18.4602,
14
  "step": 1
15
  },
16
  {
17
- "epoch": 0.04,
18
- "learning_rate": 4.962264150943397e-05,
19
- "loss": 16.3339,
20
  "step": 2
21
  },
22
  {
23
- "epoch": 0.06,
24
- "learning_rate": 4.943396226415095e-05,
25
- "loss": 14.7128,
26
  "step": 3
27
  },
28
  {
29
- "epoch": 0.08,
30
- "learning_rate": 4.9245283018867924e-05,
31
- "loss": 13.1405,
32
  "step": 4
33
  },
34
  {
35
- "epoch": 0.09,
36
- "learning_rate": 4.9056603773584906e-05,
37
- "loss": 12.334,
38
  "step": 5
39
  },
40
  {
41
- "epoch": 0.11,
42
- "learning_rate": 4.886792452830189e-05,
43
- "loss": 11.5351,
44
  "step": 6
45
  },
46
  {
47
- "epoch": 0.13,
48
- "learning_rate": 4.867924528301887e-05,
49
- "loss": 11.2585,
50
  "step": 7
51
  },
52
  {
53
- "epoch": 0.15,
54
- "learning_rate": 4.849056603773585e-05,
55
- "loss": 10.6678,
56
  "step": 8
57
  },
58
  {
59
- "epoch": 0.17,
60
- "learning_rate": 4.8301886792452835e-05,
61
- "loss": 10.3116,
62
  "step": 9
63
  },
64
  {
65
- "epoch": 0.19,
66
- "learning_rate": 4.811320754716982e-05,
67
- "loss": 10.1386,
68
  "step": 10
69
  },
70
  {
71
- "epoch": 0.21,
72
- "learning_rate": 4.792452830188679e-05,
73
- "loss": 9.7791,
74
  "step": 11
75
  },
76
  {
77
- "epoch": 0.23,
78
- "learning_rate": 4.7735849056603775e-05,
79
- "loss": 9.4858,
80
  "step": 12
81
  },
82
  {
83
- "epoch": 0.25,
84
- "learning_rate": 4.754716981132076e-05,
85
- "loss": 9.4354,
86
  "step": 13
87
  },
88
  {
89
- "epoch": 0.26,
90
- "learning_rate": 4.735849056603774e-05,
91
- "loss": 9.1703,
92
  "step": 14
93
  },
94
  {
95
- "epoch": 0.28,
96
- "learning_rate": 4.716981132075472e-05,
97
- "loss": 9.0127,
98
  "step": 15
99
  },
100
  {
101
- "epoch": 0.3,
102
- "learning_rate": 4.6981132075471704e-05,
103
- "loss": 8.978,
104
  "step": 16
105
  },
106
  {
107
- "epoch": 0.32,
108
- "learning_rate": 4.679245283018868e-05,
109
- "loss": 8.8674,
110
  "step": 17
111
  },
112
  {
113
- "epoch": 0.34,
114
- "learning_rate": 4.660377358490566e-05,
115
- "loss": 8.7663,
116
  "step": 18
117
  },
118
  {
119
- "epoch": 0.36,
120
- "learning_rate": 4.641509433962264e-05,
121
- "loss": 8.8133,
122
  "step": 19
123
  },
124
  {
125
- "epoch": 0.38,
126
- "learning_rate": 4.6226415094339625e-05,
127
- "loss": 8.5456,
128
  "step": 20
129
  },
130
- {
131
- "epoch": 0.4,
132
- "learning_rate": 4.603773584905661e-05,
133
- "loss": 8.4904,
134
- "step": 21
135
- },
136
- {
137
- "epoch": 0.42,
138
- "learning_rate": 4.584905660377359e-05,
139
- "loss": 8.4489,
140
- "step": 22
141
- },
142
- {
143
- "epoch": 0.43,
144
- "learning_rate": 4.566037735849057e-05,
145
- "loss": 8.4346,
146
- "step": 23
147
- },
148
- {
149
- "epoch": 0.45,
150
- "learning_rate": 4.547169811320755e-05,
151
- "loss": 8.1324,
152
- "step": 24
153
- },
154
- {
155
- "epoch": 0.47,
156
- "learning_rate": 4.528301886792453e-05,
157
- "loss": 8.1567,
158
- "step": 25
159
- },
160
- {
161
- "epoch": 0.49,
162
- "learning_rate": 4.509433962264151e-05,
163
- "loss": 8.1832,
164
- "step": 26
165
- },
166
- {
167
- "epoch": 0.51,
168
- "learning_rate": 4.4905660377358494e-05,
169
- "loss": 8.1429,
170
- "step": 27
171
- },
172
- {
173
- "epoch": 0.53,
174
- "learning_rate": 4.4716981132075476e-05,
175
- "loss": 7.9504,
176
- "step": 28
177
- },
178
- {
179
- "epoch": 0.55,
180
- "learning_rate": 4.452830188679246e-05,
181
- "loss": 7.9522,
182
- "step": 29
183
- },
184
- {
185
- "epoch": 0.57,
186
- "learning_rate": 4.433962264150944e-05,
187
- "loss": 7.7705,
188
- "step": 30
189
- },
190
- {
191
- "epoch": 0.58,
192
- "learning_rate": 4.4150943396226416e-05,
193
- "loss": 7.7457,
194
- "step": 31
195
- },
196
- {
197
- "epoch": 0.6,
198
- "learning_rate": 4.39622641509434e-05,
199
- "loss": 7.8053,
200
- "step": 32
201
- },
202
- {
203
- "epoch": 0.62,
204
- "learning_rate": 4.377358490566038e-05,
205
- "loss": 7.7519,
206
- "step": 33
207
- },
208
- {
209
- "epoch": 0.64,
210
- "learning_rate": 4.358490566037736e-05,
211
- "loss": 7.8011,
212
- "step": 34
213
- },
214
- {
215
- "epoch": 0.66,
216
- "learning_rate": 4.3396226415094345e-05,
217
- "loss": 7.7119,
218
- "step": 35
219
- },
220
- {
221
- "epoch": 0.68,
222
- "learning_rate": 4.320754716981133e-05,
223
- "loss": 7.4472,
224
- "step": 36
225
- },
226
- {
227
- "epoch": 0.7,
228
- "learning_rate": 4.301886792452831e-05,
229
- "loss": 7.5688,
230
- "step": 37
231
- },
232
- {
233
- "epoch": 0.72,
234
- "learning_rate": 4.2830188679245284e-05,
235
- "loss": 7.3592,
236
- "step": 38
237
- },
238
- {
239
- "epoch": 0.74,
240
- "learning_rate": 4.2641509433962266e-05,
241
- "loss": 7.4869,
242
- "step": 39
243
- },
244
- {
245
- "epoch": 0.75,
246
- "learning_rate": 4.245283018867925e-05,
247
- "loss": 7.3489,
248
- "step": 40
249
- },
250
- {
251
- "epoch": 0.77,
252
- "learning_rate": 4.226415094339623e-05,
253
- "loss": 7.412,
254
- "step": 41
255
- },
256
- {
257
- "epoch": 0.79,
258
- "learning_rate": 4.207547169811321e-05,
259
- "loss": 7.2706,
260
- "step": 42
261
- },
262
- {
263
- "epoch": 0.81,
264
- "learning_rate": 4.1886792452830195e-05,
265
- "loss": 7.2465,
266
- "step": 43
267
- },
268
- {
269
- "epoch": 0.83,
270
- "learning_rate": 4.169811320754717e-05,
271
- "loss": 7.337,
272
- "step": 44
273
- },
274
- {
275
- "epoch": 0.85,
276
- "learning_rate": 4.150943396226415e-05,
277
- "loss": 7.2097,
278
- "step": 45
279
- },
280
- {
281
- "epoch": 0.87,
282
- "learning_rate": 4.1320754716981135e-05,
283
- "loss": 7.2128,
284
- "step": 46
285
- },
286
- {
287
- "epoch": 0.89,
288
- "learning_rate": 4.113207547169812e-05,
289
- "loss": 7.2417,
290
- "step": 47
291
- },
292
- {
293
- "epoch": 0.91,
294
- "learning_rate": 4.09433962264151e-05,
295
- "loss": 7.2044,
296
- "step": 48
297
- },
298
- {
299
- "epoch": 0.92,
300
- "learning_rate": 4.075471698113208e-05,
301
- "loss": 7.1435,
302
- "step": 49
303
- },
304
- {
305
- "epoch": 0.94,
306
- "learning_rate": 4.0566037735849064e-05,
307
- "loss": 7.075,
308
- "step": 50
309
- },
310
- {
311
- "epoch": 0.96,
312
- "learning_rate": 4.037735849056604e-05,
313
- "loss": 6.9571,
314
- "step": 51
315
- },
316
- {
317
- "epoch": 0.98,
318
- "learning_rate": 4.018867924528302e-05,
319
- "loss": 6.7984,
320
- "step": 52
321
- },
322
- {
323
- "epoch": 1.0,
324
- "learning_rate": 4e-05,
325
- "loss": 6.7736,
326
- "step": 53
327
- },
328
- {
329
- "epoch": 1.02,
330
- "learning_rate": 3.9811320754716985e-05,
331
- "loss": 6.8625,
332
- "step": 54
333
- },
334
- {
335
- "epoch": 1.04,
336
- "learning_rate": 3.962264150943397e-05,
337
- "loss": 7.0221,
338
- "step": 55
339
- },
340
- {
341
- "epoch": 1.06,
342
- "learning_rate": 3.943396226415095e-05,
343
- "loss": 6.969,
344
- "step": 56
345
- },
346
- {
347
- "epoch": 1.08,
348
- "learning_rate": 3.924528301886793e-05,
349
- "loss": 6.9754,
350
- "step": 57
351
- },
352
- {
353
- "epoch": 1.09,
354
- "learning_rate": 3.905660377358491e-05,
355
- "loss": 6.9933,
356
- "step": 58
357
- },
358
- {
359
- "epoch": 1.11,
360
- "learning_rate": 3.886792452830189e-05,
361
- "loss": 6.9458,
362
- "step": 59
363
- },
364
- {
365
- "epoch": 1.13,
366
- "learning_rate": 3.867924528301887e-05,
367
- "loss": 6.9104,
368
- "step": 60
369
- },
370
- {
371
- "epoch": 1.15,
372
- "learning_rate": 3.8490566037735854e-05,
373
- "loss": 6.809,
374
- "step": 61
375
- },
376
- {
377
- "epoch": 1.17,
378
- "learning_rate": 3.8301886792452836e-05,
379
- "loss": 6.7712,
380
- "step": 62
381
- },
382
- {
383
- "epoch": 1.19,
384
- "learning_rate": 3.811320754716982e-05,
385
- "loss": 6.671,
386
- "step": 63
387
- },
388
- {
389
- "epoch": 1.21,
390
- "learning_rate": 3.7924528301886794e-05,
391
- "loss": 6.7017,
392
- "step": 64
393
- },
394
- {
395
- "epoch": 1.23,
396
- "learning_rate": 3.7735849056603776e-05,
397
- "loss": 6.7152,
398
- "step": 65
399
- },
400
- {
401
- "epoch": 1.25,
402
- "learning_rate": 3.754716981132076e-05,
403
- "loss": 6.7993,
404
- "step": 66
405
- },
406
- {
407
- "epoch": 1.26,
408
- "learning_rate": 3.735849056603773e-05,
409
- "loss": 6.6646,
410
- "step": 67
411
- },
412
- {
413
- "epoch": 1.28,
414
- "learning_rate": 3.7169811320754716e-05,
415
- "loss": 6.6046,
416
- "step": 68
417
- },
418
- {
419
- "epoch": 1.3,
420
- "learning_rate": 3.69811320754717e-05,
421
- "loss": 6.6799,
422
- "step": 69
423
- },
424
- {
425
- "epoch": 1.32,
426
- "learning_rate": 3.679245283018868e-05,
427
- "loss": 6.5709,
428
- "step": 70
429
- },
430
- {
431
- "epoch": 1.34,
432
- "learning_rate": 3.660377358490566e-05,
433
- "loss": 6.5575,
434
- "step": 71
435
- },
436
- {
437
- "epoch": 1.36,
438
- "learning_rate": 3.641509433962264e-05,
439
- "loss": 6.8022,
440
- "step": 72
441
- },
442
- {
443
- "epoch": 1.38,
444
- "learning_rate": 3.622641509433962e-05,
445
- "loss": 6.6274,
446
- "step": 73
447
- },
448
- {
449
- "epoch": 1.4,
450
- "learning_rate": 3.60377358490566e-05,
451
- "loss": 6.519,
452
- "step": 74
453
- },
454
- {
455
- "epoch": 1.42,
456
- "learning_rate": 3.5849056603773584e-05,
457
- "loss": 6.4789,
458
- "step": 75
459
- },
460
- {
461
- "epoch": 1.43,
462
- "learning_rate": 3.5660377358490566e-05,
463
- "loss": 6.5077,
464
- "step": 76
465
- },
466
- {
467
- "epoch": 1.45,
468
- "learning_rate": 3.547169811320755e-05,
469
- "loss": 6.4625,
470
- "step": 77
471
- },
472
- {
473
- "epoch": 1.47,
474
- "learning_rate": 3.528301886792453e-05,
475
- "loss": 6.6545,
476
- "step": 78
477
- },
478
- {
479
- "epoch": 1.49,
480
- "learning_rate": 3.5094339622641506e-05,
481
- "loss": 6.58,
482
- "step": 79
483
- },
484
- {
485
- "epoch": 1.51,
486
- "learning_rate": 3.490566037735849e-05,
487
- "loss": 6.7074,
488
- "step": 80
489
- },
490
- {
491
- "epoch": 1.53,
492
- "learning_rate": 3.471698113207547e-05,
493
- "loss": 6.5293,
494
- "step": 81
495
- },
496
- {
497
- "epoch": 1.55,
498
- "learning_rate": 3.452830188679245e-05,
499
- "loss": 6.4572,
500
- "step": 82
501
- },
502
- {
503
- "epoch": 1.57,
504
- "learning_rate": 3.4339622641509435e-05,
505
- "loss": 6.4919,
506
- "step": 83
507
- },
508
- {
509
- "epoch": 1.58,
510
- "learning_rate": 3.415094339622642e-05,
511
- "loss": 6.4721,
512
- "step": 84
513
- },
514
- {
515
- "epoch": 1.6,
516
- "learning_rate": 3.39622641509434e-05,
517
- "loss": 6.4396,
518
- "step": 85
519
- },
520
- {
521
- "epoch": 1.62,
522
- "learning_rate": 3.3773584905660374e-05,
523
- "loss": 6.4422,
524
- "step": 86
525
- },
526
- {
527
- "epoch": 1.64,
528
- "learning_rate": 3.3584905660377356e-05,
529
- "loss": 6.6386,
530
- "step": 87
531
- },
532
- {
533
- "epoch": 1.66,
534
- "learning_rate": 3.339622641509434e-05,
535
- "loss": 6.3216,
536
- "step": 88
537
- },
538
- {
539
- "epoch": 1.68,
540
- "learning_rate": 3.320754716981132e-05,
541
- "loss": 6.5511,
542
- "step": 89
543
- },
544
- {
545
- "epoch": 1.7,
546
- "learning_rate": 3.30188679245283e-05,
547
- "loss": 6.3281,
548
- "step": 90
549
- },
550
- {
551
- "epoch": 1.72,
552
- "learning_rate": 3.2830188679245285e-05,
553
- "loss": 6.3153,
554
- "step": 91
555
- },
556
- {
557
- "epoch": 1.74,
558
- "learning_rate": 3.264150943396227e-05,
559
- "loss": 6.3444,
560
- "step": 92
561
- },
562
- {
563
- "epoch": 1.75,
564
- "learning_rate": 3.245283018867924e-05,
565
- "loss": 6.4761,
566
- "step": 93
567
- },
568
- {
569
- "epoch": 1.77,
570
- "learning_rate": 3.2264150943396225e-05,
571
- "loss": 6.4824,
572
- "step": 94
573
- },
574
- {
575
- "epoch": 1.79,
576
- "learning_rate": 3.207547169811321e-05,
577
- "loss": 6.4817,
578
- "step": 95
579
- },
580
- {
581
- "epoch": 1.81,
582
- "learning_rate": 3.188679245283019e-05,
583
- "loss": 6.4784,
584
- "step": 96
585
- },
586
- {
587
- "epoch": 1.83,
588
- "learning_rate": 3.169811320754717e-05,
589
- "loss": 6.3177,
590
- "step": 97
591
- },
592
- {
593
- "epoch": 1.85,
594
- "learning_rate": 3.1509433962264154e-05,
595
- "loss": 6.304,
596
- "step": 98
597
- },
598
- {
599
- "epoch": 1.87,
600
- "learning_rate": 3.132075471698113e-05,
601
- "loss": 6.176,
602
- "step": 99
603
- },
604
- {
605
- "epoch": 1.89,
606
- "learning_rate": 3.113207547169811e-05,
607
- "loss": 6.3501,
608
- "step": 100
609
- },
610
- {
611
- "epoch": 1.91,
612
- "learning_rate": 3.094339622641509e-05,
613
- "loss": 6.1981,
614
- "step": 101
615
- },
616
- {
617
- "epoch": 1.92,
618
- "learning_rate": 3.0754716981132075e-05,
619
- "loss": 6.2568,
620
- "step": 102
621
- },
622
- {
623
- "epoch": 1.94,
624
- "learning_rate": 3.056603773584906e-05,
625
- "loss": 6.3171,
626
- "step": 103
627
- },
628
- {
629
- "epoch": 1.96,
630
- "learning_rate": 3.0377358490566036e-05,
631
- "loss": 6.2815,
632
- "step": 104
633
- },
634
- {
635
- "epoch": 1.98,
636
- "learning_rate": 3.018867924528302e-05,
637
- "loss": 6.2513,
638
- "step": 105
639
- },
640
- {
641
- "epoch": 2.0,
642
- "learning_rate": 3e-05,
643
- "loss": 6.0677,
644
- "step": 106
645
- },
646
- {
647
- "epoch": 2.02,
648
- "learning_rate": 2.9811320754716983e-05,
649
- "loss": 6.3859,
650
- "step": 107
651
- },
652
- {
653
- "epoch": 2.04,
654
- "learning_rate": 2.9622641509433962e-05,
655
- "loss": 6.2803,
656
- "step": 108
657
- },
658
- {
659
- "epoch": 2.06,
660
- "learning_rate": 2.9433962264150944e-05,
661
- "loss": 6.4167,
662
- "step": 109
663
- },
664
- {
665
- "epoch": 2.08,
666
- "learning_rate": 2.9245283018867926e-05,
667
- "loss": 6.234,
668
- "step": 110
669
- },
670
- {
671
- "epoch": 2.09,
672
- "learning_rate": 2.9056603773584905e-05,
673
- "loss": 6.3275,
674
- "step": 111
675
- },
676
- {
677
- "epoch": 2.11,
678
- "learning_rate": 2.8867924528301887e-05,
679
- "loss": 6.2766,
680
- "step": 112
681
- },
682
- {
683
- "epoch": 2.13,
684
- "learning_rate": 2.867924528301887e-05,
685
- "loss": 6.3146,
686
- "step": 113
687
- },
688
- {
689
- "epoch": 2.15,
690
- "learning_rate": 2.8490566037735848e-05,
691
- "loss": 6.1047,
692
- "step": 114
693
- },
694
- {
695
- "epoch": 2.17,
696
- "learning_rate": 2.830188679245283e-05,
697
- "loss": 6.1752,
698
- "step": 115
699
- },
700
- {
701
- "epoch": 2.19,
702
- "learning_rate": 2.8113207547169812e-05,
703
- "loss": 6.2028,
704
- "step": 116
705
- },
706
- {
707
- "epoch": 2.21,
708
- "learning_rate": 2.7924528301886794e-05,
709
- "loss": 6.0047,
710
- "step": 117
711
- },
712
- {
713
- "epoch": 2.23,
714
- "learning_rate": 2.7735849056603773e-05,
715
- "loss": 6.0241,
716
- "step": 118
717
- },
718
- {
719
- "epoch": 2.25,
720
- "learning_rate": 2.7547169811320755e-05,
721
- "loss": 6.25,
722
- "step": 119
723
- },
724
- {
725
- "epoch": 2.26,
726
- "learning_rate": 2.7358490566037738e-05,
727
- "loss": 6.2054,
728
- "step": 120
729
- },
730
- {
731
- "epoch": 2.28,
732
- "learning_rate": 2.7169811320754716e-05,
733
- "loss": 6.1963,
734
- "step": 121
735
- },
736
- {
737
- "epoch": 2.3,
738
- "learning_rate": 2.69811320754717e-05,
739
- "loss": 6.1233,
740
- "step": 122
741
- },
742
- {
743
- "epoch": 2.32,
744
- "learning_rate": 2.679245283018868e-05,
745
- "loss": 6.1879,
746
- "step": 123
747
- },
748
- {
749
- "epoch": 2.34,
750
- "learning_rate": 2.6603773584905663e-05,
751
- "loss": 6.1532,
752
- "step": 124
753
- },
754
- {
755
- "epoch": 2.36,
756
- "learning_rate": 2.641509433962264e-05,
757
- "loss": 5.8332,
758
- "step": 125
759
- },
760
- {
761
- "epoch": 2.38,
762
- "learning_rate": 2.6226415094339624e-05,
763
- "loss": 5.9921,
764
- "step": 126
765
- },
766
- {
767
- "epoch": 2.4,
768
- "learning_rate": 2.6037735849056606e-05,
769
- "loss": 6.08,
770
- "step": 127
771
- },
772
- {
773
- "epoch": 2.42,
774
- "learning_rate": 2.5849056603773585e-05,
775
- "loss": 5.9691,
776
- "step": 128
777
- },
778
- {
779
- "epoch": 2.43,
780
- "learning_rate": 2.5660377358490567e-05,
781
- "loss": 5.8631,
782
- "step": 129
783
- },
784
- {
785
- "epoch": 2.45,
786
- "learning_rate": 2.547169811320755e-05,
787
- "loss": 5.9265,
788
- "step": 130
789
- },
790
- {
791
- "epoch": 2.47,
792
- "learning_rate": 2.5283018867924528e-05,
793
- "loss": 6.0456,
794
- "step": 131
795
- },
796
- {
797
- "epoch": 2.49,
798
- "learning_rate": 2.509433962264151e-05,
799
- "loss": 6.0374,
800
- "step": 132
801
- },
802
- {
803
- "epoch": 2.51,
804
- "learning_rate": 2.4905660377358492e-05,
805
- "loss": 6.0419,
806
- "step": 133
807
- },
808
- {
809
- "epoch": 2.53,
810
- "learning_rate": 2.4716981132075474e-05,
811
- "loss": 5.9543,
812
- "step": 134
813
- },
814
- {
815
- "epoch": 2.55,
816
- "learning_rate": 2.4528301886792453e-05,
817
- "loss": 6.0268,
818
- "step": 135
819
- },
820
- {
821
- "epoch": 2.57,
822
- "learning_rate": 2.4339622641509435e-05,
823
- "loss": 5.9925,
824
- "step": 136
825
- },
826
- {
827
- "epoch": 2.58,
828
- "learning_rate": 2.4150943396226418e-05,
829
- "loss": 6.16,
830
- "step": 137
831
- },
832
- {
833
- "epoch": 2.6,
834
- "learning_rate": 2.3962264150943396e-05,
835
- "loss": 5.9924,
836
- "step": 138
837
- },
838
- {
839
- "epoch": 2.62,
840
- "learning_rate": 2.377358490566038e-05,
841
- "loss": 5.8427,
842
- "step": 139
843
- },
844
- {
845
- "epoch": 2.64,
846
- "learning_rate": 2.358490566037736e-05,
847
- "loss": 5.9679,
848
- "step": 140
849
- },
850
- {
851
- "epoch": 2.66,
852
- "learning_rate": 2.339622641509434e-05,
853
- "loss": 6.1985,
854
- "step": 141
855
- },
856
- {
857
- "epoch": 2.68,
858
- "learning_rate": 2.320754716981132e-05,
859
- "loss": 5.9773,
860
- "step": 142
861
- },
862
- {
863
- "epoch": 2.7,
864
- "learning_rate": 2.3018867924528304e-05,
865
- "loss": 5.9543,
866
- "step": 143
867
- },
868
- {
869
- "epoch": 2.72,
870
- "learning_rate": 2.2830188679245286e-05,
871
- "loss": 6.0227,
872
- "step": 144
873
- },
874
- {
875
- "epoch": 2.74,
876
- "learning_rate": 2.2641509433962265e-05,
877
- "loss": 5.9294,
878
- "step": 145
879
- },
880
- {
881
- "epoch": 2.75,
882
- "learning_rate": 2.2452830188679247e-05,
883
- "loss": 6.1774,
884
- "step": 146
885
- },
886
- {
887
- "epoch": 2.77,
888
- "learning_rate": 2.226415094339623e-05,
889
- "loss": 6.0009,
890
- "step": 147
891
- },
892
- {
893
- "epoch": 2.79,
894
- "learning_rate": 2.2075471698113208e-05,
895
- "loss": 5.9207,
896
- "step": 148
897
- },
898
- {
899
- "epoch": 2.81,
900
- "learning_rate": 2.188679245283019e-05,
901
- "loss": 5.9735,
902
- "step": 149
903
- },
904
- {
905
- "epoch": 2.83,
906
- "learning_rate": 2.1698113207547172e-05,
907
- "loss": 5.9751,
908
- "step": 150
909
- },
910
- {
911
- "epoch": 2.85,
912
- "learning_rate": 2.1509433962264154e-05,
913
- "loss": 5.8208,
914
- "step": 151
915
- },
916
- {
917
- "epoch": 2.87,
918
- "learning_rate": 2.1320754716981133e-05,
919
- "loss": 5.8645,
920
- "step": 152
921
- },
922
- {
923
- "epoch": 2.89,
924
- "learning_rate": 2.1132075471698115e-05,
925
- "loss": 5.9207,
926
- "step": 153
927
- },
928
- {
929
- "epoch": 2.91,
930
- "learning_rate": 2.0943396226415098e-05,
931
- "loss": 5.972,
932
- "step": 154
933
- },
934
- {
935
- "epoch": 2.92,
936
- "learning_rate": 2.0754716981132076e-05,
937
- "loss": 5.8784,
938
- "step": 155
939
- },
940
- {
941
- "epoch": 2.94,
942
- "learning_rate": 2.056603773584906e-05,
943
- "loss": 6.0364,
944
- "step": 156
945
- },
946
- {
947
- "epoch": 2.96,
948
- "learning_rate": 2.037735849056604e-05,
949
- "loss": 6.0571,
950
- "step": 157
951
- },
952
- {
953
- "epoch": 2.98,
954
- "learning_rate": 2.018867924528302e-05,
955
- "loss": 5.8623,
956
- "step": 158
957
- },
958
  {
959
  "epoch": 3.0,
960
- "learning_rate": 2e-05,
961
- "loss": 5.6061,
962
- "step": 159
963
- },
964
- {
965
- "epoch": 3.02,
966
- "learning_rate": 1.9811320754716984e-05,
967
- "loss": 5.8306,
968
- "step": 160
969
- },
970
- {
971
- "epoch": 3.04,
972
- "learning_rate": 1.9622641509433966e-05,
973
- "loss": 6.0213,
974
- "step": 161
975
- },
976
- {
977
- "epoch": 3.06,
978
- "learning_rate": 1.9433962264150945e-05,
979
- "loss": 6.022,
980
- "step": 162
981
- },
982
- {
983
- "epoch": 3.08,
984
- "learning_rate": 1.9245283018867927e-05,
985
- "loss": 5.9523,
986
- "step": 163
987
- },
988
- {
989
- "epoch": 3.09,
990
- "learning_rate": 1.905660377358491e-05,
991
- "loss": 5.8842,
992
- "step": 164
993
- },
994
- {
995
- "epoch": 3.11,
996
- "learning_rate": 1.8867924528301888e-05,
997
- "loss": 5.9051,
998
- "step": 165
999
- },
1000
- {
1001
- "epoch": 3.13,
1002
- "learning_rate": 1.8679245283018867e-05,
1003
- "loss": 5.91,
1004
- "step": 166
1005
- },
1006
- {
1007
- "epoch": 3.15,
1008
- "learning_rate": 1.849056603773585e-05,
1009
- "loss": 5.7862,
1010
- "step": 167
1011
- },
1012
- {
1013
- "epoch": 3.17,
1014
- "learning_rate": 1.830188679245283e-05,
1015
- "loss": 5.8941,
1016
- "step": 168
1017
- },
1018
- {
1019
- "epoch": 3.19,
1020
- "learning_rate": 1.811320754716981e-05,
1021
- "loss": 6.0136,
1022
- "step": 169
1023
- },
1024
- {
1025
- "epoch": 3.21,
1026
- "learning_rate": 1.7924528301886792e-05,
1027
- "loss": 5.7909,
1028
- "step": 170
1029
- },
1030
- {
1031
- "epoch": 3.23,
1032
- "learning_rate": 1.7735849056603774e-05,
1033
- "loss": 5.8893,
1034
- "step": 171
1035
- },
1036
- {
1037
- "epoch": 3.25,
1038
- "learning_rate": 1.7547169811320753e-05,
1039
- "loss": 5.8309,
1040
- "step": 172
1041
- },
1042
- {
1043
- "epoch": 3.26,
1044
- "learning_rate": 1.7358490566037735e-05,
1045
- "loss": 5.6179,
1046
- "step": 173
1047
- },
1048
- {
1049
- "epoch": 3.28,
1050
- "learning_rate": 1.7169811320754717e-05,
1051
- "loss": 5.8054,
1052
- "step": 174
1053
- },
1054
- {
1055
- "epoch": 3.3,
1056
- "learning_rate": 1.69811320754717e-05,
1057
- "loss": 5.8038,
1058
- "step": 175
1059
- },
1060
- {
1061
- "epoch": 3.32,
1062
- "learning_rate": 1.6792452830188678e-05,
1063
- "loss": 5.8717,
1064
- "step": 176
1065
- },
1066
- {
1067
- "epoch": 3.34,
1068
- "learning_rate": 1.660377358490566e-05,
1069
- "loss": 5.7068,
1070
- "step": 177
1071
- },
1072
- {
1073
- "epoch": 3.36,
1074
- "learning_rate": 1.6415094339622643e-05,
1075
- "loss": 5.8516,
1076
- "step": 178
1077
- },
1078
- {
1079
- "epoch": 3.38,
1080
- "learning_rate": 1.622641509433962e-05,
1081
- "loss": 5.7989,
1082
- "step": 179
1083
- },
1084
- {
1085
- "epoch": 3.4,
1086
- "learning_rate": 1.6037735849056604e-05,
1087
- "loss": 5.9828,
1088
- "step": 180
1089
- },
1090
- {
1091
- "epoch": 3.42,
1092
- "learning_rate": 1.5849056603773586e-05,
1093
- "loss": 5.8526,
1094
- "step": 181
1095
- },
1096
- {
1097
- "epoch": 3.43,
1098
- "learning_rate": 1.5660377358490564e-05,
1099
- "loss": 5.9012,
1100
- "step": 182
1101
- },
1102
- {
1103
- "epoch": 3.45,
1104
- "learning_rate": 1.5471698113207547e-05,
1105
- "loss": 5.9608,
1106
- "step": 183
1107
- },
1108
- {
1109
- "epoch": 3.47,
1110
- "learning_rate": 1.528301886792453e-05,
1111
- "loss": 5.8062,
1112
- "step": 184
1113
- },
1114
- {
1115
- "epoch": 3.49,
1116
- "learning_rate": 1.509433962264151e-05,
1117
- "loss": 5.7624,
1118
- "step": 185
1119
- },
1120
- {
1121
- "epoch": 3.51,
1122
- "learning_rate": 1.4905660377358491e-05,
1123
- "loss": 5.8388,
1124
- "step": 186
1125
- },
1126
- {
1127
- "epoch": 3.53,
1128
- "learning_rate": 1.4716981132075472e-05,
1129
- "loss": 5.8675,
1130
- "step": 187
1131
- },
1132
- {
1133
- "epoch": 3.55,
1134
- "learning_rate": 1.4528301886792452e-05,
1135
- "loss": 5.971,
1136
- "step": 188
1137
- },
1138
- {
1139
- "epoch": 3.57,
1140
- "learning_rate": 1.4339622641509435e-05,
1141
- "loss": 5.8333,
1142
- "step": 189
1143
- },
1144
- {
1145
- "epoch": 3.58,
1146
- "learning_rate": 1.4150943396226415e-05,
1147
- "loss": 5.9516,
1148
- "step": 190
1149
- },
1150
- {
1151
- "epoch": 3.6,
1152
- "learning_rate": 1.3962264150943397e-05,
1153
- "loss": 5.8128,
1154
- "step": 191
1155
- },
1156
- {
1157
- "epoch": 3.62,
1158
- "learning_rate": 1.3773584905660378e-05,
1159
- "loss": 5.6509,
1160
- "step": 192
1161
- },
1162
- {
1163
- "epoch": 3.64,
1164
- "learning_rate": 1.3584905660377358e-05,
1165
- "loss": 5.7938,
1166
- "step": 193
1167
- },
1168
- {
1169
- "epoch": 3.66,
1170
- "learning_rate": 1.339622641509434e-05,
1171
- "loss": 5.736,
1172
- "step": 194
1173
- },
1174
- {
1175
- "epoch": 3.68,
1176
- "learning_rate": 1.320754716981132e-05,
1177
- "loss": 5.6844,
1178
- "step": 195
1179
- },
1180
- {
1181
- "epoch": 3.7,
1182
- "learning_rate": 1.3018867924528303e-05,
1183
- "loss": 5.7359,
1184
- "step": 196
1185
- },
1186
- {
1187
- "epoch": 3.72,
1188
- "learning_rate": 1.2830188679245283e-05,
1189
- "loss": 5.8941,
1190
- "step": 197
1191
- },
1192
- {
1193
- "epoch": 3.74,
1194
- "learning_rate": 1.2641509433962264e-05,
1195
- "loss": 5.8553,
1196
- "step": 198
1197
- },
1198
- {
1199
- "epoch": 3.75,
1200
- "learning_rate": 1.2452830188679246e-05,
1201
- "loss": 5.6521,
1202
- "step": 199
1203
- },
1204
- {
1205
- "epoch": 3.77,
1206
- "learning_rate": 1.2264150943396227e-05,
1207
- "loss": 5.921,
1208
- "step": 200
1209
- },
1210
- {
1211
- "epoch": 3.79,
1212
- "learning_rate": 1.2075471698113209e-05,
1213
- "loss": 5.7235,
1214
- "step": 201
1215
- },
1216
- {
1217
- "epoch": 3.81,
1218
- "learning_rate": 1.188679245283019e-05,
1219
- "loss": 5.8019,
1220
- "step": 202
1221
- },
1222
- {
1223
- "epoch": 3.83,
1224
- "learning_rate": 1.169811320754717e-05,
1225
- "loss": 5.7486,
1226
- "step": 203
1227
- },
1228
- {
1229
- "epoch": 3.85,
1230
- "learning_rate": 1.1509433962264152e-05,
1231
- "loss": 5.7429,
1232
- "step": 204
1233
- },
1234
- {
1235
- "epoch": 3.87,
1236
- "learning_rate": 1.1320754716981132e-05,
1237
- "loss": 5.8486,
1238
- "step": 205
1239
- },
1240
- {
1241
- "epoch": 3.89,
1242
- "learning_rate": 1.1132075471698115e-05,
1243
- "loss": 5.7921,
1244
- "step": 206
1245
- },
1246
- {
1247
- "epoch": 3.91,
1248
- "learning_rate": 1.0943396226415095e-05,
1249
- "loss": 5.8232,
1250
- "step": 207
1251
- },
1252
- {
1253
- "epoch": 3.92,
1254
- "learning_rate": 1.0754716981132077e-05,
1255
- "loss": 6.083,
1256
- "step": 208
1257
- },
1258
- {
1259
- "epoch": 3.94,
1260
- "learning_rate": 1.0566037735849058e-05,
1261
- "loss": 5.7263,
1262
- "step": 209
1263
- },
1264
- {
1265
- "epoch": 3.96,
1266
- "learning_rate": 1.0377358490566038e-05,
1267
- "loss": 5.6783,
1268
- "step": 210
1269
- },
1270
- {
1271
- "epoch": 3.98,
1272
- "learning_rate": 1.018867924528302e-05,
1273
- "loss": 5.7754,
1274
- "step": 211
1275
- },
1276
- {
1277
- "epoch": 4.0,
1278
- "learning_rate": 1e-05,
1279
- "loss": 6.1204,
1280
- "step": 212
1281
- },
1282
- {
1283
- "epoch": 4.02,
1284
- "learning_rate": 9.811320754716983e-06,
1285
- "loss": 5.6473,
1286
- "step": 213
1287
- },
1288
- {
1289
- "epoch": 4.04,
1290
- "learning_rate": 9.622641509433963e-06,
1291
- "loss": 5.8676,
1292
- "step": 214
1293
- },
1294
- {
1295
- "epoch": 4.06,
1296
- "learning_rate": 9.433962264150944e-06,
1297
- "loss": 5.7221,
1298
- "step": 215
1299
- },
1300
- {
1301
- "epoch": 4.08,
1302
- "learning_rate": 9.245283018867924e-06,
1303
- "loss": 5.8749,
1304
- "step": 216
1305
- },
1306
- {
1307
- "epoch": 4.09,
1308
- "learning_rate": 9.056603773584905e-06,
1309
- "loss": 5.7737,
1310
- "step": 217
1311
- },
1312
- {
1313
- "epoch": 4.11,
1314
- "learning_rate": 8.867924528301887e-06,
1315
- "loss": 5.7663,
1316
- "step": 218
1317
- },
1318
- {
1319
- "epoch": 4.13,
1320
- "learning_rate": 8.679245283018868e-06,
1321
- "loss": 5.6642,
1322
- "step": 219
1323
- },
1324
- {
1325
- "epoch": 4.15,
1326
- "learning_rate": 8.49056603773585e-06,
1327
- "loss": 5.815,
1328
- "step": 220
1329
- },
1330
- {
1331
- "epoch": 4.17,
1332
- "learning_rate": 8.30188679245283e-06,
1333
- "loss": 5.6594,
1334
- "step": 221
1335
- },
1336
- {
1337
- "epoch": 4.19,
1338
- "learning_rate": 8.11320754716981e-06,
1339
- "loss": 6.0083,
1340
- "step": 222
1341
- },
1342
- {
1343
- "epoch": 4.21,
1344
- "learning_rate": 7.924528301886793e-06,
1345
- "loss": 5.7564,
1346
- "step": 223
1347
- },
1348
- {
1349
- "epoch": 4.23,
1350
- "learning_rate": 7.735849056603773e-06,
1351
- "loss": 5.677,
1352
- "step": 224
1353
- },
1354
- {
1355
- "epoch": 4.25,
1356
- "learning_rate": 7.547169811320755e-06,
1357
- "loss": 5.8026,
1358
- "step": 225
1359
- },
1360
- {
1361
- "epoch": 4.26,
1362
- "learning_rate": 7.358490566037736e-06,
1363
- "loss": 5.9558,
1364
- "step": 226
1365
- },
1366
- {
1367
- "epoch": 4.28,
1368
- "learning_rate": 7.169811320754717e-06,
1369
- "loss": 5.6896,
1370
- "step": 227
1371
- },
1372
- {
1373
- "epoch": 4.3,
1374
- "learning_rate": 6.981132075471699e-06,
1375
- "loss": 5.8383,
1376
- "step": 228
1377
- },
1378
- {
1379
- "epoch": 4.32,
1380
- "learning_rate": 6.792452830188679e-06,
1381
- "loss": 5.8355,
1382
- "step": 229
1383
- },
1384
- {
1385
- "epoch": 4.34,
1386
- "learning_rate": 6.60377358490566e-06,
1387
- "loss": 5.6567,
1388
- "step": 230
1389
- },
1390
- {
1391
- "epoch": 4.36,
1392
- "learning_rate": 6.415094339622642e-06,
1393
- "loss": 5.7913,
1394
- "step": 231
1395
- },
1396
- {
1397
- "epoch": 4.38,
1398
- "learning_rate": 6.226415094339623e-06,
1399
- "loss": 5.9586,
1400
- "step": 232
1401
- },
1402
- {
1403
- "epoch": 4.4,
1404
- "learning_rate": 6.037735849056604e-06,
1405
- "loss": 5.7501,
1406
- "step": 233
1407
- },
1408
- {
1409
- "epoch": 4.42,
1410
- "learning_rate": 5.849056603773585e-06,
1411
- "loss": 5.6516,
1412
- "step": 234
1413
- },
1414
- {
1415
- "epoch": 4.43,
1416
- "learning_rate": 5.660377358490566e-06,
1417
- "loss": 5.7582,
1418
- "step": 235
1419
- },
1420
- {
1421
- "epoch": 4.45,
1422
- "learning_rate": 5.4716981132075475e-06,
1423
- "loss": 5.6714,
1424
- "step": 236
1425
- },
1426
- {
1427
- "epoch": 4.47,
1428
- "learning_rate": 5.283018867924529e-06,
1429
- "loss": 5.7091,
1430
- "step": 237
1431
- },
1432
- {
1433
- "epoch": 4.49,
1434
- "learning_rate": 5.09433962264151e-06,
1435
- "loss": 5.6222,
1436
- "step": 238
1437
- },
1438
- {
1439
- "epoch": 4.51,
1440
- "learning_rate": 4.9056603773584915e-06,
1441
- "loss": 5.7866,
1442
- "step": 239
1443
- },
1444
- {
1445
- "epoch": 4.53,
1446
- "learning_rate": 4.716981132075472e-06,
1447
- "loss": 5.6759,
1448
- "step": 240
1449
- },
1450
- {
1451
- "epoch": 4.55,
1452
- "learning_rate": 4.5283018867924524e-06,
1453
- "loss": 5.679,
1454
- "step": 241
1455
- },
1456
- {
1457
- "epoch": 4.57,
1458
- "learning_rate": 4.339622641509434e-06,
1459
- "loss": 5.7489,
1460
- "step": 242
1461
- },
1462
- {
1463
- "epoch": 4.58,
1464
- "learning_rate": 4.150943396226415e-06,
1465
- "loss": 5.6387,
1466
- "step": 243
1467
- },
1468
- {
1469
- "epoch": 4.6,
1470
- "learning_rate": 3.962264150943396e-06,
1471
- "loss": 5.6617,
1472
- "step": 244
1473
- },
1474
- {
1475
- "epoch": 4.62,
1476
- "learning_rate": 3.7735849056603773e-06,
1477
- "loss": 5.6338,
1478
- "step": 245
1479
- },
1480
- {
1481
- "epoch": 4.64,
1482
- "learning_rate": 3.5849056603773586e-06,
1483
- "loss": 5.6264,
1484
- "step": 246
1485
- },
1486
- {
1487
- "epoch": 4.66,
1488
- "learning_rate": 3.3962264150943395e-06,
1489
- "loss": 5.8117,
1490
- "step": 247
1491
- },
1492
- {
1493
- "epoch": 4.68,
1494
- "learning_rate": 3.207547169811321e-06,
1495
- "loss": 5.8644,
1496
- "step": 248
1497
- },
1498
- {
1499
- "epoch": 4.7,
1500
- "learning_rate": 3.018867924528302e-06,
1501
- "loss": 5.7927,
1502
- "step": 249
1503
- },
1504
- {
1505
- "epoch": 4.72,
1506
- "learning_rate": 2.830188679245283e-06,
1507
- "loss": 5.7877,
1508
- "step": 250
1509
- },
1510
- {
1511
- "epoch": 4.74,
1512
- "learning_rate": 2.6415094339622644e-06,
1513
- "loss": 5.7415,
1514
- "step": 251
1515
- },
1516
- {
1517
- "epoch": 4.75,
1518
- "learning_rate": 2.4528301886792457e-06,
1519
- "loss": 5.5958,
1520
- "step": 252
1521
- },
1522
- {
1523
- "epoch": 4.77,
1524
- "learning_rate": 2.2641509433962262e-06,
1525
- "loss": 5.4249,
1526
- "step": 253
1527
- },
1528
- {
1529
- "epoch": 4.79,
1530
- "learning_rate": 2.0754716981132075e-06,
1531
- "loss": 5.7735,
1532
- "step": 254
1533
- },
1534
- {
1535
- "epoch": 4.81,
1536
- "learning_rate": 1.8867924528301887e-06,
1537
- "loss": 5.672,
1538
- "step": 255
1539
- },
1540
- {
1541
- "epoch": 4.83,
1542
- "learning_rate": 1.6981132075471698e-06,
1543
- "loss": 5.5705,
1544
- "step": 256
1545
- },
1546
- {
1547
- "epoch": 4.85,
1548
- "learning_rate": 1.509433962264151e-06,
1549
- "loss": 5.8487,
1550
- "step": 257
1551
- },
1552
- {
1553
- "epoch": 4.87,
1554
- "learning_rate": 1.3207547169811322e-06,
1555
- "loss": 5.8151,
1556
- "step": 258
1557
- },
1558
- {
1559
- "epoch": 4.89,
1560
- "learning_rate": 1.1320754716981131e-06,
1561
- "loss": 5.6904,
1562
- "step": 259
1563
- },
1564
- {
1565
- "epoch": 4.91,
1566
- "learning_rate": 9.433962264150943e-07,
1567
- "loss": 5.7021,
1568
- "step": 260
1569
- },
1570
- {
1571
- "epoch": 4.92,
1572
- "learning_rate": 7.547169811320755e-07,
1573
- "loss": 5.7196,
1574
- "step": 261
1575
- },
1576
- {
1577
- "epoch": 4.94,
1578
- "learning_rate": 5.660377358490566e-07,
1579
- "loss": 5.8667,
1580
- "step": 262
1581
- },
1582
- {
1583
- "epoch": 4.96,
1584
- "learning_rate": 3.773584905660378e-07,
1585
- "loss": 5.8458,
1586
- "step": 263
1587
- },
1588
- {
1589
- "epoch": 4.98,
1590
- "learning_rate": 1.886792452830189e-07,
1591
- "loss": 5.7733,
1592
- "step": 264
1593
- },
1594
- {
1595
- "epoch": 5.0,
1596
  "learning_rate": 0.0,
1597
- "loss": 5.8065,
1598
- "step": 265
1599
  },
1600
  {
1601
- "epoch": 5.0,
1602
- "step": 265,
1603
- "total_flos": 209285213159424.0,
1604
- "train_loss": 6.612753483034529,
1605
- "train_runtime": 33.2807,
1606
- "train_samples_per_second": 503.444,
1607
- "train_steps_per_second": 7.963
1608
  }
1609
  ],
1610
- "max_steps": 265,
1611
- "num_train_epochs": 5,
1612
- "total_flos": 209285213159424.0,
1613
  "trial_name": null,
1614
  "trial_params": null
1615
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 21,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
+ "epoch": 0.14,
12
+ "learning_rate": 4.761904761904762e-05,
13
+ "loss": 3.0215,
14
  "step": 1
15
  },
16
  {
17
+ "epoch": 0.29,
18
+ "learning_rate": 4.523809523809524e-05,
19
+ "loss": 3.396,
20
  "step": 2
21
  },
22
  {
23
+ "epoch": 0.43,
24
+ "learning_rate": 4.2857142857142856e-05,
25
+ "loss": 2.7103,
26
  "step": 3
27
  },
28
  {
29
+ "epoch": 0.57,
30
+ "learning_rate": 4.047619047619048e-05,
31
+ "loss": 2.8102,
32
  "step": 4
33
  },
34
  {
35
+ "epoch": 0.71,
36
+ "learning_rate": 3.809523809523809e-05,
37
+ "loss": 2.8078,
38
  "step": 5
39
  },
40
  {
41
+ "epoch": 0.86,
42
+ "learning_rate": 3.571428571428572e-05,
43
+ "loss": 2.6017,
44
  "step": 6
45
  },
46
  {
47
+ "epoch": 1.0,
48
+ "learning_rate": 3.3333333333333335e-05,
49
+ "loss": 3.1308,
50
  "step": 7
51
  },
52
  {
53
+ "epoch": 1.14,
54
+ "learning_rate": 3.095238095238095e-05,
55
+ "loss": 2.806,
56
  "step": 8
57
  },
58
  {
59
+ "epoch": 1.29,
60
+ "learning_rate": 2.857142857142857e-05,
61
+ "loss": 2.6379,
62
  "step": 9
63
  },
64
  {
65
+ "epoch": 1.43,
66
+ "learning_rate": 2.6190476190476192e-05,
67
+ "loss": 2.6436,
68
  "step": 10
69
  },
70
  {
71
+ "epoch": 1.57,
72
+ "learning_rate": 2.380952380952381e-05,
73
+ "loss": 2.6356,
74
  "step": 11
75
  },
76
  {
77
+ "epoch": 1.71,
78
+ "learning_rate": 2.1428571428571428e-05,
79
+ "loss": 2.7459,
80
  "step": 12
81
  },
82
  {
83
+ "epoch": 1.86,
84
+ "learning_rate": 1.9047619047619046e-05,
85
+ "loss": 2.5413,
86
  "step": 13
87
  },
88
  {
89
+ "epoch": 2.0,
90
+ "learning_rate": 1.6666666666666667e-05,
91
+ "loss": 2.67,
92
  "step": 14
93
  },
94
  {
95
+ "epoch": 2.14,
96
+ "learning_rate": 1.4285714285714285e-05,
97
+ "loss": 2.3287,
98
  "step": 15
99
  },
100
  {
101
+ "epoch": 2.29,
102
+ "learning_rate": 1.1904761904761905e-05,
103
+ "loss": 2.5067,
104
  "step": 16
105
  },
106
  {
107
+ "epoch": 2.43,
108
+ "learning_rate": 9.523809523809523e-06,
109
+ "loss": 2.651,
110
  "step": 17
111
  },
112
  {
113
+ "epoch": 2.57,
114
+ "learning_rate": 7.142857142857143e-06,
115
+ "loss": 2.5026,
116
  "step": 18
117
  },
118
  {
119
+ "epoch": 2.71,
120
+ "learning_rate": 4.7619047619047615e-06,
121
+ "loss": 2.7971,
122
  "step": 19
123
  },
124
  {
125
+ "epoch": 2.86,
126
+ "learning_rate": 2.3809523809523808e-06,
127
+ "loss": 2.6718,
128
  "step": 20
129
  },
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
130
  {
131
  "epoch": 3.0,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
132
  "learning_rate": 0.0,
133
+ "loss": 2.6248,
134
+ "step": 21
135
  },
136
  {
137
+ "epoch": 3.0,
138
+ "step": 21,
139
+ "total_flos": 76819857408000.0,
140
+ "train_loss": 2.7257756051563082,
141
+ "train_runtime": 7.0042,
142
+ "train_samples_per_second": 20.987,
143
+ "train_steps_per_second": 2.998
144
  }
145
  ],
146
+ "max_steps": 21,
147
+ "num_train_epochs": 3,
148
+ "total_flos": 76819857408000.0,
149
  "trial_name": null,
150
  "trial_params": null
151
  }