datastreams commited on
Commit
09430af
1 Parent(s): 64bce72

Upload ./trainer_state.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. trainer_state.json +919 -0
trainer_state.json ADDED
@@ -0,0 +1,919 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 16.483516483516482,
5
+ "eval_steps": 500,
6
+ "global_step": 1500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.11,
13
+ "learning_rate": 4.996861268047709e-05,
14
+ "loss": 2.5145,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.22,
19
+ "learning_rate": 4.9937225360954175e-05,
20
+ "loss": 2.8458,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.33,
25
+ "learning_rate": 4.9905838041431265e-05,
26
+ "loss": 2.7597,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.44,
31
+ "learning_rate": 4.987445072190835e-05,
32
+ "loss": 2.8058,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.55,
37
+ "learning_rate": 4.984306340238544e-05,
38
+ "loss": 2.6902,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.66,
43
+ "learning_rate": 4.981167608286253e-05,
44
+ "loss": 2.7747,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.77,
49
+ "learning_rate": 4.978028876333961e-05,
50
+ "loss": 2.9923,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.88,
55
+ "learning_rate": 4.97489014438167e-05,
56
+ "loss": 2.7536,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.99,
61
+ "learning_rate": 4.971751412429379e-05,
62
+ "loss": 2.8079,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 1.1,
67
+ "learning_rate": 4.968612680477088e-05,
68
+ "loss": 2.8927,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 1.21,
73
+ "learning_rate": 4.965473948524796e-05,
74
+ "loss": 2.815,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 1.32,
79
+ "learning_rate": 4.962335216572505e-05,
80
+ "loss": 2.9312,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 1.43,
85
+ "learning_rate": 4.959196484620214e-05,
86
+ "loss": 2.598,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 1.54,
91
+ "learning_rate": 4.956057752667922e-05,
92
+ "loss": 2.6474,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 1.65,
97
+ "learning_rate": 4.952919020715631e-05,
98
+ "loss": 2.7857,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 1.76,
103
+ "learning_rate": 4.9497802887633396e-05,
104
+ "loss": 2.7879,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 1.87,
109
+ "learning_rate": 4.9466415568110485e-05,
110
+ "loss": 2.6259,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 1.98,
115
+ "learning_rate": 4.9435028248587575e-05,
116
+ "loss": 2.6434,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 2.09,
121
+ "learning_rate": 4.940364092906466e-05,
122
+ "loss": 2.6546,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 2.2,
127
+ "learning_rate": 4.937225360954175e-05,
128
+ "loss": 2.7263,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 2.31,
133
+ "learning_rate": 4.934086629001883e-05,
134
+ "loss": 2.843,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 2.42,
139
+ "learning_rate": 4.930947897049592e-05,
140
+ "loss": 2.6683,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 2.53,
145
+ "learning_rate": 4.927809165097301e-05,
146
+ "loss": 2.441,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 2.64,
151
+ "learning_rate": 4.924670433145009e-05,
152
+ "loss": 2.8475,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 2.75,
157
+ "learning_rate": 4.921531701192718e-05,
158
+ "loss": 2.7734,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 2.86,
163
+ "learning_rate": 4.918392969240427e-05,
164
+ "loss": 2.839,
165
+ "step": 260
166
+ },
167
+ {
168
+ "epoch": 2.97,
169
+ "learning_rate": 4.915254237288136e-05,
170
+ "loss": 2.5971,
171
+ "step": 270
172
+ },
173
+ {
174
+ "epoch": 3.08,
175
+ "learning_rate": 4.9121155053358444e-05,
176
+ "loss": 2.5059,
177
+ "step": 280
178
+ },
179
+ {
180
+ "epoch": 3.19,
181
+ "learning_rate": 4.9089767733835534e-05,
182
+ "loss": 2.6131,
183
+ "step": 290
184
+ },
185
+ {
186
+ "epoch": 3.3,
187
+ "learning_rate": 4.9058380414312623e-05,
188
+ "loss": 2.8308,
189
+ "step": 300
190
+ },
191
+ {
192
+ "epoch": 3.41,
193
+ "learning_rate": 4.9026993094789706e-05,
194
+ "loss": 2.7235,
195
+ "step": 310
196
+ },
197
+ {
198
+ "epoch": 3.52,
199
+ "learning_rate": 4.8995605775266796e-05,
200
+ "loss": 2.7609,
201
+ "step": 320
202
+ },
203
+ {
204
+ "epoch": 3.63,
205
+ "learning_rate": 4.896421845574388e-05,
206
+ "loss": 2.7446,
207
+ "step": 330
208
+ },
209
+ {
210
+ "epoch": 3.74,
211
+ "learning_rate": 4.893283113622097e-05,
212
+ "loss": 2.6262,
213
+ "step": 340
214
+ },
215
+ {
216
+ "epoch": 3.85,
217
+ "learning_rate": 4.890144381669806e-05,
218
+ "loss": 2.4824,
219
+ "step": 350
220
+ },
221
+ {
222
+ "epoch": 3.96,
223
+ "learning_rate": 4.887005649717514e-05,
224
+ "loss": 2.8134,
225
+ "step": 360
226
+ },
227
+ {
228
+ "epoch": 4.07,
229
+ "learning_rate": 4.883866917765223e-05,
230
+ "loss": 2.6734,
231
+ "step": 370
232
+ },
233
+ {
234
+ "epoch": 4.18,
235
+ "learning_rate": 4.8807281858129313e-05,
236
+ "loss": 2.6152,
237
+ "step": 380
238
+ },
239
+ {
240
+ "epoch": 4.29,
241
+ "learning_rate": 4.87758945386064e-05,
242
+ "loss": 2.8046,
243
+ "step": 390
244
+ },
245
+ {
246
+ "epoch": 4.4,
247
+ "learning_rate": 4.874450721908349e-05,
248
+ "loss": 2.6603,
249
+ "step": 400
250
+ },
251
+ {
252
+ "epoch": 4.51,
253
+ "learning_rate": 4.8713119899560576e-05,
254
+ "loss": 2.789,
255
+ "step": 410
256
+ },
257
+ {
258
+ "epoch": 4.62,
259
+ "learning_rate": 4.8681732580037665e-05,
260
+ "loss": 2.5235,
261
+ "step": 420
262
+ },
263
+ {
264
+ "epoch": 4.73,
265
+ "learning_rate": 4.8650345260514755e-05,
266
+ "loss": 2.6632,
267
+ "step": 430
268
+ },
269
+ {
270
+ "epoch": 4.84,
271
+ "learning_rate": 4.8618957940991844e-05,
272
+ "loss": 2.6261,
273
+ "step": 440
274
+ },
275
+ {
276
+ "epoch": 4.95,
277
+ "learning_rate": 4.8587570621468934e-05,
278
+ "loss": 2.6481,
279
+ "step": 450
280
+ },
281
+ {
282
+ "epoch": 5.05,
283
+ "learning_rate": 4.855618330194602e-05,
284
+ "loss": 2.6359,
285
+ "step": 460
286
+ },
287
+ {
288
+ "epoch": 5.16,
289
+ "learning_rate": 4.8524795982423107e-05,
290
+ "loss": 2.6283,
291
+ "step": 470
292
+ },
293
+ {
294
+ "epoch": 5.27,
295
+ "learning_rate": 4.849340866290019e-05,
296
+ "loss": 2.6531,
297
+ "step": 480
298
+ },
299
+ {
300
+ "epoch": 5.38,
301
+ "learning_rate": 4.846202134337728e-05,
302
+ "loss": 2.6473,
303
+ "step": 490
304
+ },
305
+ {
306
+ "epoch": 5.49,
307
+ "learning_rate": 4.843063402385437e-05,
308
+ "loss": 2.7426,
309
+ "step": 500
310
+ },
311
+ {
312
+ "epoch": 5.6,
313
+ "learning_rate": 4.839924670433145e-05,
314
+ "loss": 2.7108,
315
+ "step": 510
316
+ },
317
+ {
318
+ "epoch": 5.71,
319
+ "learning_rate": 4.836785938480854e-05,
320
+ "loss": 2.6723,
321
+ "step": 520
322
+ },
323
+ {
324
+ "epoch": 5.82,
325
+ "learning_rate": 4.8336472065285624e-05,
326
+ "loss": 2.6087,
327
+ "step": 530
328
+ },
329
+ {
330
+ "epoch": 5.93,
331
+ "learning_rate": 4.8305084745762714e-05,
332
+ "loss": 2.5313,
333
+ "step": 540
334
+ },
335
+ {
336
+ "epoch": 6.04,
337
+ "learning_rate": 4.82736974262398e-05,
338
+ "loss": 2.6208,
339
+ "step": 550
340
+ },
341
+ {
342
+ "epoch": 6.15,
343
+ "learning_rate": 4.8242310106716886e-05,
344
+ "loss": 2.613,
345
+ "step": 560
346
+ },
347
+ {
348
+ "epoch": 6.26,
349
+ "learning_rate": 4.8210922787193976e-05,
350
+ "loss": 2.645,
351
+ "step": 570
352
+ },
353
+ {
354
+ "epoch": 6.37,
355
+ "learning_rate": 4.817953546767106e-05,
356
+ "loss": 2.521,
357
+ "step": 580
358
+ },
359
+ {
360
+ "epoch": 6.48,
361
+ "learning_rate": 4.814814814814815e-05,
362
+ "loss": 2.7001,
363
+ "step": 590
364
+ },
365
+ {
366
+ "epoch": 6.59,
367
+ "learning_rate": 4.811676082862524e-05,
368
+ "loss": 2.6073,
369
+ "step": 600
370
+ },
371
+ {
372
+ "epoch": 6.7,
373
+ "learning_rate": 4.808537350910233e-05,
374
+ "loss": 2.6636,
375
+ "step": 610
376
+ },
377
+ {
378
+ "epoch": 6.81,
379
+ "learning_rate": 4.805398618957942e-05,
380
+ "loss": 2.5819,
381
+ "step": 620
382
+ },
383
+ {
384
+ "epoch": 6.92,
385
+ "learning_rate": 4.80225988700565e-05,
386
+ "loss": 2.6064,
387
+ "step": 630
388
+ },
389
+ {
390
+ "epoch": 7.03,
391
+ "learning_rate": 4.799121155053359e-05,
392
+ "loss": 2.7406,
393
+ "step": 640
394
+ },
395
+ {
396
+ "epoch": 7.14,
397
+ "learning_rate": 4.795982423101067e-05,
398
+ "loss": 2.5997,
399
+ "step": 650
400
+ },
401
+ {
402
+ "epoch": 7.25,
403
+ "learning_rate": 4.792843691148776e-05,
404
+ "loss": 2.5727,
405
+ "step": 660
406
+ },
407
+ {
408
+ "epoch": 7.36,
409
+ "learning_rate": 4.789704959196485e-05,
410
+ "loss": 2.4625,
411
+ "step": 670
412
+ },
413
+ {
414
+ "epoch": 7.47,
415
+ "learning_rate": 4.7865662272441935e-05,
416
+ "loss": 2.5773,
417
+ "step": 680
418
+ },
419
+ {
420
+ "epoch": 7.58,
421
+ "learning_rate": 4.7834274952919024e-05,
422
+ "loss": 2.5806,
423
+ "step": 690
424
+ },
425
+ {
426
+ "epoch": 7.69,
427
+ "learning_rate": 4.780288763339611e-05,
428
+ "loss": 2.5901,
429
+ "step": 700
430
+ },
431
+ {
432
+ "epoch": 7.8,
433
+ "learning_rate": 4.77715003138732e-05,
434
+ "loss": 2.4906,
435
+ "step": 710
436
+ },
437
+ {
438
+ "epoch": 7.91,
439
+ "learning_rate": 4.7740112994350286e-05,
440
+ "loss": 2.6069,
441
+ "step": 720
442
+ },
443
+ {
444
+ "epoch": 8.02,
445
+ "learning_rate": 4.770872567482737e-05,
446
+ "loss": 2.5363,
447
+ "step": 730
448
+ },
449
+ {
450
+ "epoch": 8.13,
451
+ "learning_rate": 4.767733835530446e-05,
452
+ "loss": 2.5511,
453
+ "step": 740
454
+ },
455
+ {
456
+ "epoch": 8.24,
457
+ "learning_rate": 4.764595103578154e-05,
458
+ "loss": 2.5839,
459
+ "step": 750
460
+ },
461
+ {
462
+ "epoch": 8.35,
463
+ "learning_rate": 4.761456371625863e-05,
464
+ "loss": 2.462,
465
+ "step": 760
466
+ },
467
+ {
468
+ "epoch": 8.46,
469
+ "learning_rate": 4.758317639673572e-05,
470
+ "loss": 2.8533,
471
+ "step": 770
472
+ },
473
+ {
474
+ "epoch": 8.57,
475
+ "learning_rate": 4.755178907721281e-05,
476
+ "loss": 2.3477,
477
+ "step": 780
478
+ },
479
+ {
480
+ "epoch": 8.68,
481
+ "learning_rate": 4.75204017576899e-05,
482
+ "loss": 2.553,
483
+ "step": 790
484
+ },
485
+ {
486
+ "epoch": 8.79,
487
+ "learning_rate": 4.748901443816698e-05,
488
+ "loss": 2.614,
489
+ "step": 800
490
+ },
491
+ {
492
+ "epoch": 8.9,
493
+ "learning_rate": 4.745762711864407e-05,
494
+ "loss": 2.6441,
495
+ "step": 810
496
+ },
497
+ {
498
+ "epoch": 9.01,
499
+ "learning_rate": 4.7426239799121155e-05,
500
+ "loss": 2.6618,
501
+ "step": 820
502
+ },
503
+ {
504
+ "epoch": 9.12,
505
+ "learning_rate": 4.7394852479598245e-05,
506
+ "loss": 2.4887,
507
+ "step": 830
508
+ },
509
+ {
510
+ "epoch": 9.23,
511
+ "learning_rate": 4.7363465160075335e-05,
512
+ "loss": 2.3639,
513
+ "step": 840
514
+ },
515
+ {
516
+ "epoch": 9.34,
517
+ "learning_rate": 4.733207784055242e-05,
518
+ "loss": 2.7387,
519
+ "step": 850
520
+ },
521
+ {
522
+ "epoch": 9.45,
523
+ "learning_rate": 4.730069052102951e-05,
524
+ "loss": 2.4262,
525
+ "step": 860
526
+ },
527
+ {
528
+ "epoch": 9.56,
529
+ "learning_rate": 4.726930320150659e-05,
530
+ "loss": 2.6172,
531
+ "step": 870
532
+ },
533
+ {
534
+ "epoch": 9.67,
535
+ "learning_rate": 4.723791588198368e-05,
536
+ "loss": 2.7115,
537
+ "step": 880
538
+ },
539
+ {
540
+ "epoch": 9.78,
541
+ "learning_rate": 4.720652856246077e-05,
542
+ "loss": 2.5397,
543
+ "step": 890
544
+ },
545
+ {
546
+ "epoch": 9.89,
547
+ "learning_rate": 4.717514124293785e-05,
548
+ "loss": 2.4934,
549
+ "step": 900
550
+ },
551
+ {
552
+ "epoch": 10.0,
553
+ "learning_rate": 4.714375392341494e-05,
554
+ "loss": 2.5521,
555
+ "step": 910
556
+ },
557
+ {
558
+ "epoch": 10.11,
559
+ "learning_rate": 4.7112366603892025e-05,
560
+ "loss": 2.3767,
561
+ "step": 920
562
+ },
563
+ {
564
+ "epoch": 10.22,
565
+ "learning_rate": 4.7080979284369114e-05,
566
+ "loss": 2.5422,
567
+ "step": 930
568
+ },
569
+ {
570
+ "epoch": 10.33,
571
+ "learning_rate": 4.7049591964846204e-05,
572
+ "loss": 2.7068,
573
+ "step": 940
574
+ },
575
+ {
576
+ "epoch": 10.44,
577
+ "learning_rate": 4.7018204645323294e-05,
578
+ "loss": 2.5518,
579
+ "step": 950
580
+ },
581
+ {
582
+ "epoch": 10.55,
583
+ "learning_rate": 4.698681732580038e-05,
584
+ "loss": 2.4129,
585
+ "step": 960
586
+ },
587
+ {
588
+ "epoch": 10.66,
589
+ "learning_rate": 4.6955430006277466e-05,
590
+ "loss": 2.6916,
591
+ "step": 970
592
+ },
593
+ {
594
+ "epoch": 10.77,
595
+ "learning_rate": 4.6924042686754556e-05,
596
+ "loss": 2.5753,
597
+ "step": 980
598
+ },
599
+ {
600
+ "epoch": 10.88,
601
+ "learning_rate": 4.689265536723164e-05,
602
+ "loss": 2.3783,
603
+ "step": 990
604
+ },
605
+ {
606
+ "epoch": 10.99,
607
+ "learning_rate": 4.686126804770873e-05,
608
+ "loss": 2.5766,
609
+ "step": 1000
610
+ },
611
+ {
612
+ "epoch": 11.1,
613
+ "learning_rate": 4.682988072818582e-05,
614
+ "loss": 2.56,
615
+ "step": 1010
616
+ },
617
+ {
618
+ "epoch": 11.21,
619
+ "learning_rate": 4.67984934086629e-05,
620
+ "loss": 2.59,
621
+ "step": 1020
622
+ },
623
+ {
624
+ "epoch": 11.32,
625
+ "learning_rate": 4.676710608913999e-05,
626
+ "loss": 2.5592,
627
+ "step": 1030
628
+ },
629
+ {
630
+ "epoch": 11.43,
631
+ "learning_rate": 4.673571876961707e-05,
632
+ "loss": 2.553,
633
+ "step": 1040
634
+ },
635
+ {
636
+ "epoch": 11.54,
637
+ "learning_rate": 4.670433145009416e-05,
638
+ "loss": 2.564,
639
+ "step": 1050
640
+ },
641
+ {
642
+ "epoch": 11.65,
643
+ "learning_rate": 4.667294413057125e-05,
644
+ "loss": 2.471,
645
+ "step": 1060
646
+ },
647
+ {
648
+ "epoch": 11.76,
649
+ "learning_rate": 4.6641556811048335e-05,
650
+ "loss": 2.4783,
651
+ "step": 1070
652
+ },
653
+ {
654
+ "epoch": 11.87,
655
+ "learning_rate": 4.6610169491525425e-05,
656
+ "loss": 2.5474,
657
+ "step": 1080
658
+ },
659
+ {
660
+ "epoch": 11.98,
661
+ "learning_rate": 4.657878217200251e-05,
662
+ "loss": 2.5478,
663
+ "step": 1090
664
+ },
665
+ {
666
+ "epoch": 12.09,
667
+ "learning_rate": 4.65473948524796e-05,
668
+ "loss": 2.4591,
669
+ "step": 1100
670
+ },
671
+ {
672
+ "epoch": 12.2,
673
+ "learning_rate": 4.651600753295669e-05,
674
+ "loss": 2.4682,
675
+ "step": 1110
676
+ },
677
+ {
678
+ "epoch": 12.31,
679
+ "learning_rate": 4.6484620213433777e-05,
680
+ "loss": 2.5391,
681
+ "step": 1120
682
+ },
683
+ {
684
+ "epoch": 12.42,
685
+ "learning_rate": 4.6453232893910866e-05,
686
+ "loss": 2.6003,
687
+ "step": 1130
688
+ },
689
+ {
690
+ "epoch": 12.53,
691
+ "learning_rate": 4.642184557438795e-05,
692
+ "loss": 2.4073,
693
+ "step": 1140
694
+ },
695
+ {
696
+ "epoch": 12.64,
697
+ "learning_rate": 4.639045825486504e-05,
698
+ "loss": 2.5084,
699
+ "step": 1150
700
+ },
701
+ {
702
+ "epoch": 12.75,
703
+ "learning_rate": 4.635907093534212e-05,
704
+ "loss": 2.4579,
705
+ "step": 1160
706
+ },
707
+ {
708
+ "epoch": 12.86,
709
+ "learning_rate": 4.632768361581921e-05,
710
+ "loss": 2.512,
711
+ "step": 1170
712
+ },
713
+ {
714
+ "epoch": 12.97,
715
+ "learning_rate": 4.62962962962963e-05,
716
+ "loss": 2.6637,
717
+ "step": 1180
718
+ },
719
+ {
720
+ "epoch": 13.08,
721
+ "learning_rate": 4.6264908976773384e-05,
722
+ "loss": 2.3842,
723
+ "step": 1190
724
+ },
725
+ {
726
+ "epoch": 13.19,
727
+ "learning_rate": 4.623352165725047e-05,
728
+ "loss": 2.3314,
729
+ "step": 1200
730
+ },
731
+ {
732
+ "epoch": 13.3,
733
+ "learning_rate": 4.620213433772756e-05,
734
+ "loss": 2.4636,
735
+ "step": 1210
736
+ },
737
+ {
738
+ "epoch": 13.41,
739
+ "learning_rate": 4.6170747018204646e-05,
740
+ "loss": 2.4403,
741
+ "step": 1220
742
+ },
743
+ {
744
+ "epoch": 13.52,
745
+ "learning_rate": 4.6139359698681735e-05,
746
+ "loss": 2.7347,
747
+ "step": 1230
748
+ },
749
+ {
750
+ "epoch": 13.63,
751
+ "learning_rate": 4.610797237915882e-05,
752
+ "loss": 2.4503,
753
+ "step": 1240
754
+ },
755
+ {
756
+ "epoch": 13.74,
757
+ "learning_rate": 4.607658505963591e-05,
758
+ "loss": 2.566,
759
+ "step": 1250
760
+ },
761
+ {
762
+ "epoch": 13.85,
763
+ "learning_rate": 4.6045197740113e-05,
764
+ "loss": 2.3861,
765
+ "step": 1260
766
+ },
767
+ {
768
+ "epoch": 13.96,
769
+ "learning_rate": 4.601381042059008e-05,
770
+ "loss": 2.4167,
771
+ "step": 1270
772
+ },
773
+ {
774
+ "epoch": 14.07,
775
+ "learning_rate": 4.598242310106718e-05,
776
+ "loss": 2.751,
777
+ "step": 1280
778
+ },
779
+ {
780
+ "epoch": 14.18,
781
+ "learning_rate": 4.595103578154426e-05,
782
+ "loss": 2.4427,
783
+ "step": 1290
784
+ },
785
+ {
786
+ "epoch": 14.29,
787
+ "learning_rate": 4.591964846202135e-05,
788
+ "loss": 2.5363,
789
+ "step": 1300
790
+ },
791
+ {
792
+ "epoch": 14.4,
793
+ "learning_rate": 4.588826114249843e-05,
794
+ "loss": 2.3152,
795
+ "step": 1310
796
+ },
797
+ {
798
+ "epoch": 14.51,
799
+ "learning_rate": 4.585687382297552e-05,
800
+ "loss": 2.6868,
801
+ "step": 1320
802
+ },
803
+ {
804
+ "epoch": 14.62,
805
+ "learning_rate": 4.582548650345261e-05,
806
+ "loss": 2.4857,
807
+ "step": 1330
808
+ },
809
+ {
810
+ "epoch": 14.73,
811
+ "learning_rate": 4.5794099183929694e-05,
812
+ "loss": 2.5417,
813
+ "step": 1340
814
+ },
815
+ {
816
+ "epoch": 14.84,
817
+ "learning_rate": 4.5762711864406784e-05,
818
+ "loss": 2.4462,
819
+ "step": 1350
820
+ },
821
+ {
822
+ "epoch": 14.95,
823
+ "learning_rate": 4.573132454488387e-05,
824
+ "loss": 2.3253,
825
+ "step": 1360
826
+ },
827
+ {
828
+ "epoch": 15.05,
829
+ "learning_rate": 4.5699937225360956e-05,
830
+ "loss": 2.674,
831
+ "step": 1370
832
+ },
833
+ {
834
+ "epoch": 15.16,
835
+ "learning_rate": 4.5668549905838046e-05,
836
+ "loss": 2.607,
837
+ "step": 1380
838
+ },
839
+ {
840
+ "epoch": 15.27,
841
+ "learning_rate": 4.563716258631513e-05,
842
+ "loss": 2.3933,
843
+ "step": 1390
844
+ },
845
+ {
846
+ "epoch": 15.38,
847
+ "learning_rate": 4.560577526679222e-05,
848
+ "loss": 2.4335,
849
+ "step": 1400
850
+ },
851
+ {
852
+ "epoch": 15.49,
853
+ "learning_rate": 4.55743879472693e-05,
854
+ "loss": 2.2217,
855
+ "step": 1410
856
+ },
857
+ {
858
+ "epoch": 15.6,
859
+ "learning_rate": 4.554300062774639e-05,
860
+ "loss": 2.4138,
861
+ "step": 1420
862
+ },
863
+ {
864
+ "epoch": 15.71,
865
+ "learning_rate": 4.551161330822348e-05,
866
+ "loss": 2.5214,
867
+ "step": 1430
868
+ },
869
+ {
870
+ "epoch": 15.82,
871
+ "learning_rate": 4.548022598870056e-05,
872
+ "loss": 2.6247,
873
+ "step": 1440
874
+ },
875
+ {
876
+ "epoch": 15.93,
877
+ "learning_rate": 4.544883866917766e-05,
878
+ "loss": 2.495,
879
+ "step": 1450
880
+ },
881
+ {
882
+ "epoch": 16.04,
883
+ "learning_rate": 4.541745134965474e-05,
884
+ "loss": 2.4676,
885
+ "step": 1460
886
+ },
887
+ {
888
+ "epoch": 16.15,
889
+ "learning_rate": 4.538606403013183e-05,
890
+ "loss": 2.5501,
891
+ "step": 1470
892
+ },
893
+ {
894
+ "epoch": 16.26,
895
+ "learning_rate": 4.5354676710608915e-05,
896
+ "loss": 2.612,
897
+ "step": 1480
898
+ },
899
+ {
900
+ "epoch": 16.37,
901
+ "learning_rate": 4.5323289391086005e-05,
902
+ "loss": 2.6939,
903
+ "step": 1490
904
+ },
905
+ {
906
+ "epoch": 16.48,
907
+ "learning_rate": 4.5291902071563094e-05,
908
+ "loss": 2.3641,
909
+ "step": 1500
910
+ }
911
+ ],
912
+ "logging_steps": 10,
913
+ "max_steps": 15930,
914
+ "num_train_epochs": 176,
915
+ "save_steps": 500,
916
+ "total_flos": 7.73986862813184e+16,
917
+ "trial_name": null,
918
+ "trial_params": null
919
+ }