longcld commited on
Commit
3a9499c
1 Parent(s): b69b321
Files changed (5) hide show
  1. optimizer.pt +1 -1
  2. pytorch_model.bin +1 -1
  3. rng_state.pth +1 -1
  4. scheduler.pt +1 -1
  5. trainer_state.json +813 -3
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:270626bfbed24add5f3d23d044c7d4f4d74ccfd8e37d891e8b02071147ebc368
3
  size 352539155
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51cbdd3ef2f219de9935a5ce9c637d27fee5aff007344e7c841792b16ee2755b
3
  size 352539155
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1b3d1c499136ae9d388a242a2de2d3a249b26b9c6a3b6233428612c2e6451ea
3
  size 688500106
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2032bb6d1867faec4d419dccbf2240eebc8dfab068f9e713b3c55e82e626d6ac
3
  size 688500106
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7c8766af0f7b1618f2a59eafc5d46d0d1937c98a3524f791fe749ee641c9b496
3
  size 14657
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33234e8bf39c862e70fdc68a572da9ce20c0de377c5ec534307e121c07e5177a
3
  size 14657
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c110c53a5ecef842fcafbc06d2cf413115544756e036964a3da796defb9000b9
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:635e1c7c2f2b7589d105218c8ecfe96223daf1a80e41f3c595834e97061e8442
3
  size 623
trainer_state.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 0.9449004901671293,
5
- "global_step": 2000,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
@@ -126,11 +126,821 @@
126
  "learning_rate": 9.054820415879018e-05,
127
  "loss": 2.1412,
128
  "step": 2000
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
129
  }
130
  ],
131
  "max_steps": 21160,
132
  "num_train_epochs": 10,
133
- "total_flos": 1.5105279098578944e+16,
134
  "trial_name": null,
135
  "trial_params": null
136
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 7.325045768617493,
5
+ "global_step": 15500,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
 
126
  "learning_rate": 9.054820415879018e-05,
127
  "loss": 2.1412,
128
  "step": 2000
129
+ },
130
+ {
131
+ "epoch": 0.99,
132
+ "learning_rate": 9.007561436672968e-05,
133
+ "loss": 2.2009,
134
+ "step": 2100
135
+ },
136
+ {
137
+ "epoch": 1.04,
138
+ "learning_rate": 8.960302457466919e-05,
139
+ "loss": 2.205,
140
+ "step": 2200
141
+ },
142
+ {
143
+ "epoch": 1.09,
144
+ "learning_rate": 8.91304347826087e-05,
145
+ "loss": 2.1226,
146
+ "step": 2300
147
+ },
148
+ {
149
+ "epoch": 1.13,
150
+ "learning_rate": 8.865784499054821e-05,
151
+ "loss": 2.1963,
152
+ "step": 2400
153
+ },
154
+ {
155
+ "epoch": 1.18,
156
+ "learning_rate": 8.818525519848772e-05,
157
+ "loss": 2.0325,
158
+ "step": 2500
159
+ },
160
+ {
161
+ "epoch": 1.23,
162
+ "learning_rate": 8.771266540642723e-05,
163
+ "loss": 2.1478,
164
+ "step": 2600
165
+ },
166
+ {
167
+ "epoch": 1.28,
168
+ "learning_rate": 8.724007561436674e-05,
169
+ "loss": 2.1116,
170
+ "step": 2700
171
+ },
172
+ {
173
+ "epoch": 1.32,
174
+ "learning_rate": 8.676748582230624e-05,
175
+ "loss": 2.1805,
176
+ "step": 2800
177
+ },
178
+ {
179
+ "epoch": 1.37,
180
+ "learning_rate": 8.629489603024575e-05,
181
+ "loss": 2.109,
182
+ "step": 2900
183
+ },
184
+ {
185
+ "epoch": 1.42,
186
+ "learning_rate": 8.582230623818526e-05,
187
+ "loss": 2.1156,
188
+ "step": 3000
189
+ },
190
+ {
191
+ "epoch": 1.46,
192
+ "learning_rate": 8.534971644612477e-05,
193
+ "loss": 2.152,
194
+ "step": 3100
195
+ },
196
+ {
197
+ "epoch": 1.51,
198
+ "learning_rate": 8.487712665406428e-05,
199
+ "loss": 2.1512,
200
+ "step": 3200
201
+ },
202
+ {
203
+ "epoch": 1.56,
204
+ "learning_rate": 8.440453686200379e-05,
205
+ "loss": 2.1396,
206
+ "step": 3300
207
+ },
208
+ {
209
+ "epoch": 1.61,
210
+ "learning_rate": 8.39319470699433e-05,
211
+ "loss": 2.0951,
212
+ "step": 3400
213
+ },
214
+ {
215
+ "epoch": 1.65,
216
+ "learning_rate": 8.34593572778828e-05,
217
+ "loss": 2.1533,
218
+ "step": 3500
219
+ },
220
+ {
221
+ "epoch": 1.7,
222
+ "learning_rate": 8.298676748582231e-05,
223
+ "loss": 2.0959,
224
+ "step": 3600
225
+ },
226
+ {
227
+ "epoch": 1.75,
228
+ "learning_rate": 8.251417769376182e-05,
229
+ "loss": 2.1345,
230
+ "step": 3700
231
+ },
232
+ {
233
+ "epoch": 1.8,
234
+ "learning_rate": 8.204158790170132e-05,
235
+ "loss": 2.0643,
236
+ "step": 3800
237
+ },
238
+ {
239
+ "epoch": 1.84,
240
+ "learning_rate": 8.156899810964084e-05,
241
+ "loss": 2.0566,
242
+ "step": 3900
243
+ },
244
+ {
245
+ "epoch": 1.89,
246
+ "learning_rate": 8.109640831758035e-05,
247
+ "loss": 2.0729,
248
+ "step": 4000
249
+ },
250
+ {
251
+ "epoch": 1.94,
252
+ "learning_rate": 8.062381852551986e-05,
253
+ "loss": 2.0768,
254
+ "step": 4100
255
+ },
256
+ {
257
+ "epoch": 1.98,
258
+ "learning_rate": 8.015122873345936e-05,
259
+ "loss": 2.1026,
260
+ "step": 4200
261
+ },
262
+ {
263
+ "epoch": 2.03,
264
+ "learning_rate": 7.967863894139886e-05,
265
+ "loss": 2.0804,
266
+ "step": 4300
267
+ },
268
+ {
269
+ "epoch": 2.08,
270
+ "learning_rate": 7.920604914933838e-05,
271
+ "loss": 2.0875,
272
+ "step": 4400
273
+ },
274
+ {
275
+ "epoch": 2.13,
276
+ "learning_rate": 7.873345935727789e-05,
277
+ "loss": 2.039,
278
+ "step": 4500
279
+ },
280
+ {
281
+ "epoch": 2.17,
282
+ "learning_rate": 7.82608695652174e-05,
283
+ "loss": 2.0605,
284
+ "step": 4600
285
+ },
286
+ {
287
+ "epoch": 2.22,
288
+ "learning_rate": 7.77882797731569e-05,
289
+ "loss": 2.0495,
290
+ "step": 4700
291
+ },
292
+ {
293
+ "epoch": 2.27,
294
+ "learning_rate": 7.731568998109642e-05,
295
+ "loss": 2.0146,
296
+ "step": 4800
297
+ },
298
+ {
299
+ "epoch": 2.32,
300
+ "learning_rate": 7.684310018903592e-05,
301
+ "loss": 2.0378,
302
+ "step": 4900
303
+ },
304
+ {
305
+ "epoch": 2.36,
306
+ "learning_rate": 7.637051039697543e-05,
307
+ "loss": 2.0806,
308
+ "step": 5000
309
+ },
310
+ {
311
+ "epoch": 2.41,
312
+ "learning_rate": 7.589792060491494e-05,
313
+ "loss": 2.0675,
314
+ "step": 5100
315
+ },
316
+ {
317
+ "epoch": 2.46,
318
+ "learning_rate": 7.542533081285445e-05,
319
+ "loss": 2.0247,
320
+ "step": 5200
321
+ },
322
+ {
323
+ "epoch": 2.5,
324
+ "learning_rate": 7.495274102079396e-05,
325
+ "loss": 2.0442,
326
+ "step": 5300
327
+ },
328
+ {
329
+ "epoch": 2.55,
330
+ "learning_rate": 7.448015122873347e-05,
331
+ "loss": 2.0232,
332
+ "step": 5400
333
+ },
334
+ {
335
+ "epoch": 2.6,
336
+ "learning_rate": 7.400756143667296e-05,
337
+ "loss": 2.0648,
338
+ "step": 5500
339
+ },
340
+ {
341
+ "epoch": 2.65,
342
+ "learning_rate": 7.353497164461248e-05,
343
+ "loss": 2.0579,
344
+ "step": 5600
345
+ },
346
+ {
347
+ "epoch": 2.69,
348
+ "learning_rate": 7.306238185255199e-05,
349
+ "loss": 2.0438,
350
+ "step": 5700
351
+ },
352
+ {
353
+ "epoch": 2.74,
354
+ "learning_rate": 7.25897920604915e-05,
355
+ "loss": 2.0849,
356
+ "step": 5800
357
+ },
358
+ {
359
+ "epoch": 2.79,
360
+ "learning_rate": 7.211720226843101e-05,
361
+ "loss": 1.9828,
362
+ "step": 5900
363
+ },
364
+ {
365
+ "epoch": 2.84,
366
+ "learning_rate": 7.16446124763705e-05,
367
+ "loss": 1.9923,
368
+ "step": 6000
369
+ },
370
+ {
371
+ "epoch": 2.88,
372
+ "learning_rate": 7.117202268431003e-05,
373
+ "loss": 1.9898,
374
+ "step": 6100
375
+ },
376
+ {
377
+ "epoch": 2.93,
378
+ "learning_rate": 7.069943289224953e-05,
379
+ "loss": 2.0663,
380
+ "step": 6200
381
+ },
382
+ {
383
+ "epoch": 2.98,
384
+ "learning_rate": 7.022684310018904e-05,
385
+ "loss": 1.9967,
386
+ "step": 6300
387
+ },
388
+ {
389
+ "epoch": 3.02,
390
+ "learning_rate": 6.975425330812855e-05,
391
+ "loss": 1.9941,
392
+ "step": 6400
393
+ },
394
+ {
395
+ "epoch": 3.07,
396
+ "learning_rate": 6.928166351606805e-05,
397
+ "loss": 1.9458,
398
+ "step": 6500
399
+ },
400
+ {
401
+ "epoch": 3.12,
402
+ "learning_rate": 6.880907372400757e-05,
403
+ "loss": 2.0411,
404
+ "step": 6600
405
+ },
406
+ {
407
+ "epoch": 3.17,
408
+ "learning_rate": 6.833648393194706e-05,
409
+ "loss": 1.9634,
410
+ "step": 6700
411
+ },
412
+ {
413
+ "epoch": 3.21,
414
+ "learning_rate": 6.786389413988659e-05,
415
+ "loss": 1.9897,
416
+ "step": 6800
417
+ },
418
+ {
419
+ "epoch": 3.26,
420
+ "learning_rate": 6.73913043478261e-05,
421
+ "loss": 1.9798,
422
+ "step": 6900
423
+ },
424
+ {
425
+ "epoch": 3.31,
426
+ "learning_rate": 6.691871455576559e-05,
427
+ "loss": 2.0127,
428
+ "step": 7000
429
+ },
430
+ {
431
+ "epoch": 3.36,
432
+ "learning_rate": 6.644612476370511e-05,
433
+ "loss": 1.9827,
434
+ "step": 7100
435
+ },
436
+ {
437
+ "epoch": 3.4,
438
+ "learning_rate": 6.59735349716446e-05,
439
+ "loss": 1.9949,
440
+ "step": 7200
441
+ },
442
+ {
443
+ "epoch": 3.45,
444
+ "learning_rate": 6.550094517958413e-05,
445
+ "loss": 2.0152,
446
+ "step": 7300
447
+ },
448
+ {
449
+ "epoch": 3.5,
450
+ "learning_rate": 6.502835538752364e-05,
451
+ "loss": 1.959,
452
+ "step": 7400
453
+ },
454
+ {
455
+ "epoch": 3.54,
456
+ "learning_rate": 6.455576559546313e-05,
457
+ "loss": 1.9876,
458
+ "step": 7500
459
+ },
460
+ {
461
+ "epoch": 3.59,
462
+ "learning_rate": 6.408317580340265e-05,
463
+ "loss": 1.9955,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 3.64,
468
+ "learning_rate": 6.361058601134215e-05,
469
+ "loss": 1.9812,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 3.69,
474
+ "learning_rate": 6.313799621928167e-05,
475
+ "loss": 1.9219,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 3.73,
480
+ "learning_rate": 6.266540642722118e-05,
481
+ "loss": 1.9576,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 3.78,
486
+ "learning_rate": 6.219281663516069e-05,
487
+ "loss": 1.9608,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 3.83,
492
+ "learning_rate": 6.17202268431002e-05,
493
+ "loss": 1.9916,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 3.87,
498
+ "learning_rate": 6.124763705103969e-05,
499
+ "loss": 2.0394,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 3.92,
504
+ "learning_rate": 6.0775047258979214e-05,
505
+ "loss": 1.9416,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 3.97,
510
+ "learning_rate": 6.0302457466918716e-05,
511
+ "loss": 2.0283,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 4.02,
516
+ "learning_rate": 5.982986767485823e-05,
517
+ "loss": 2.0244,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 4.06,
522
+ "learning_rate": 5.935727788279773e-05,
523
+ "loss": 1.9132,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 4.11,
528
+ "learning_rate": 5.888468809073724e-05,
529
+ "loss": 1.9917,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 4.16,
534
+ "learning_rate": 5.841209829867676e-05,
535
+ "loss": 1.9355,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 4.21,
540
+ "learning_rate": 5.793950850661626e-05,
541
+ "loss": 1.9248,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 4.25,
546
+ "learning_rate": 5.7466918714555774e-05,
547
+ "loss": 2.0129,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 4.3,
552
+ "learning_rate": 5.6994328922495276e-05,
553
+ "loss": 1.9497,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 4.35,
558
+ "learning_rate": 5.652173913043478e-05,
559
+ "loss": 1.8806,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 4.39,
564
+ "learning_rate": 5.604914933837429e-05,
565
+ "loss": 1.9262,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 4.44,
570
+ "learning_rate": 5.55765595463138e-05,
571
+ "loss": 2.0024,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 4.49,
576
+ "learning_rate": 5.510396975425332e-05,
577
+ "loss": 1.9407,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 4.54,
582
+ "learning_rate": 5.463137996219282e-05,
583
+ "loss": 1.991,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 4.58,
588
+ "learning_rate": 5.415879017013232e-05,
589
+ "loss": 1.964,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 4.63,
594
+ "learning_rate": 5.3686200378071836e-05,
595
+ "loss": 1.9025,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 4.68,
600
+ "learning_rate": 5.3213610586011344e-05,
601
+ "loss": 1.9489,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 4.73,
606
+ "learning_rate": 5.274102079395086e-05,
607
+ "loss": 1.942,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 4.77,
612
+ "learning_rate": 5.226843100189036e-05,
613
+ "loss": 1.9868,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 4.82,
618
+ "learning_rate": 5.179584120982986e-05,
619
+ "loss": 1.9231,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 4.87,
624
+ "learning_rate": 5.132325141776938e-05,
625
+ "loss": 1.9625,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 4.91,
630
+ "learning_rate": 5.085066162570889e-05,
631
+ "loss": 1.9306,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 4.96,
636
+ "learning_rate": 5.03780718336484e-05,
637
+ "loss": 1.9106,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 5.01,
642
+ "learning_rate": 4.9905482041587904e-05,
643
+ "loss": 1.9209,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 5.06,
648
+ "learning_rate": 4.943289224952741e-05,
649
+ "loss": 1.925,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 5.1,
654
+ "learning_rate": 4.896030245746692e-05,
655
+ "loss": 1.9368,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 5.15,
660
+ "learning_rate": 4.848771266540643e-05,
661
+ "loss": 1.9076,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 5.2,
666
+ "learning_rate": 4.801512287334594e-05,
667
+ "loss": 1.8892,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 5.25,
672
+ "learning_rate": 4.754253308128545e-05,
673
+ "loss": 1.9667,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 5.29,
678
+ "learning_rate": 4.7069943289224955e-05,
679
+ "loss": 1.9371,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 5.34,
684
+ "learning_rate": 4.6597353497164464e-05,
685
+ "loss": 1.8549,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 5.39,
690
+ "learning_rate": 4.612476370510397e-05,
691
+ "loss": 1.9157,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 5.43,
696
+ "learning_rate": 4.565217391304348e-05,
697
+ "loss": 1.9051,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 5.48,
702
+ "learning_rate": 4.517958412098299e-05,
703
+ "loss": 1.9175,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 5.53,
708
+ "learning_rate": 4.47069943289225e-05,
709
+ "loss": 1.9367,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 5.58,
714
+ "learning_rate": 4.423440453686201e-05,
715
+ "loss": 1.8351,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 5.62,
720
+ "learning_rate": 4.3761814744801515e-05,
721
+ "loss": 1.8961,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 5.67,
726
+ "learning_rate": 4.3289224952741024e-05,
727
+ "loss": 1.8583,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 5.72,
732
+ "learning_rate": 4.281663516068053e-05,
733
+ "loss": 1.8909,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 5.77,
738
+ "learning_rate": 4.234404536862004e-05,
739
+ "loss": 1.8642,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 5.81,
744
+ "learning_rate": 4.187145557655955e-05,
745
+ "loss": 1.9229,
746
+ "step": 12300
747
+ },
748
+ {
749
+ "epoch": 5.86,
750
+ "learning_rate": 4.139886578449906e-05,
751
+ "loss": 1.9524,
752
+ "step": 12400
753
+ },
754
+ {
755
+ "epoch": 5.91,
756
+ "learning_rate": 4.0926275992438567e-05,
757
+ "loss": 1.9143,
758
+ "step": 12500
759
+ },
760
+ {
761
+ "epoch": 5.95,
762
+ "learning_rate": 4.045368620037807e-05,
763
+ "loss": 1.9229,
764
+ "step": 12600
765
+ },
766
+ {
767
+ "epoch": 6.0,
768
+ "learning_rate": 3.9981096408317584e-05,
769
+ "loss": 1.9368,
770
+ "step": 12700
771
+ },
772
+ {
773
+ "epoch": 6.05,
774
+ "learning_rate": 3.950850661625709e-05,
775
+ "loss": 1.9183,
776
+ "step": 12800
777
+ },
778
+ {
779
+ "epoch": 6.1,
780
+ "learning_rate": 3.90359168241966e-05,
781
+ "loss": 1.8864,
782
+ "step": 12900
783
+ },
784
+ {
785
+ "epoch": 6.14,
786
+ "learning_rate": 3.856332703213611e-05,
787
+ "loss": 1.8453,
788
+ "step": 13000
789
+ },
790
+ {
791
+ "epoch": 6.19,
792
+ "learning_rate": 3.809073724007561e-05,
793
+ "loss": 1.8785,
794
+ "step": 13100
795
+ },
796
+ {
797
+ "epoch": 6.24,
798
+ "learning_rate": 3.7618147448015126e-05,
799
+ "loss": 1.9145,
800
+ "step": 13200
801
+ },
802
+ {
803
+ "epoch": 6.29,
804
+ "learning_rate": 3.7145557655954635e-05,
805
+ "loss": 1.9477,
806
+ "step": 13300
807
+ },
808
+ {
809
+ "epoch": 6.33,
810
+ "learning_rate": 3.6672967863894143e-05,
811
+ "loss": 1.8378,
812
+ "step": 13400
813
+ },
814
+ {
815
+ "epoch": 6.38,
816
+ "learning_rate": 3.620037807183365e-05,
817
+ "loss": 1.843,
818
+ "step": 13500
819
+ },
820
+ {
821
+ "epoch": 6.43,
822
+ "learning_rate": 3.572778827977316e-05,
823
+ "loss": 1.8386,
824
+ "step": 13600
825
+ },
826
+ {
827
+ "epoch": 6.47,
828
+ "learning_rate": 3.525519848771266e-05,
829
+ "loss": 1.9,
830
+ "step": 13700
831
+ },
832
+ {
833
+ "epoch": 6.52,
834
+ "learning_rate": 3.478260869565218e-05,
835
+ "loss": 1.8437,
836
+ "step": 13800
837
+ },
838
+ {
839
+ "epoch": 6.57,
840
+ "learning_rate": 3.4310018903591686e-05,
841
+ "loss": 1.9301,
842
+ "step": 13900
843
+ },
844
+ {
845
+ "epoch": 6.62,
846
+ "learning_rate": 3.3837429111531195e-05,
847
+ "loss": 1.8447,
848
+ "step": 14000
849
+ },
850
+ {
851
+ "epoch": 6.66,
852
+ "learning_rate": 3.33648393194707e-05,
853
+ "loss": 1.8811,
854
+ "step": 14100
855
+ },
856
+ {
857
+ "epoch": 6.71,
858
+ "learning_rate": 3.2892249527410205e-05,
859
+ "loss": 1.8775,
860
+ "step": 14200
861
+ },
862
+ {
863
+ "epoch": 6.76,
864
+ "learning_rate": 3.2419659735349714e-05,
865
+ "loss": 1.9123,
866
+ "step": 14300
867
+ },
868
+ {
869
+ "epoch": 6.81,
870
+ "learning_rate": 3.194706994328923e-05,
871
+ "loss": 1.8236,
872
+ "step": 14400
873
+ },
874
+ {
875
+ "epoch": 6.85,
876
+ "learning_rate": 3.147448015122874e-05,
877
+ "loss": 1.8434,
878
+ "step": 14500
879
+ },
880
+ {
881
+ "epoch": 6.9,
882
+ "learning_rate": 3.1001890359168246e-05,
883
+ "loss": 1.8975,
884
+ "step": 14600
885
+ },
886
+ {
887
+ "epoch": 6.95,
888
+ "learning_rate": 3.052930056710775e-05,
889
+ "loss": 1.887,
890
+ "step": 14700
891
+ },
892
+ {
893
+ "epoch": 6.99,
894
+ "learning_rate": 3.005671077504726e-05,
895
+ "loss": 1.8665,
896
+ "step": 14800
897
+ },
898
+ {
899
+ "epoch": 7.04,
900
+ "learning_rate": 2.9584120982986768e-05,
901
+ "loss": 1.9142,
902
+ "step": 14900
903
+ },
904
+ {
905
+ "epoch": 7.09,
906
+ "learning_rate": 2.9111531190926277e-05,
907
+ "loss": 1.8892,
908
+ "step": 15000
909
+ },
910
+ {
911
+ "epoch": 7.14,
912
+ "learning_rate": 2.863894139886579e-05,
913
+ "loss": 1.8762,
914
+ "step": 15100
915
+ },
916
+ {
917
+ "epoch": 7.18,
918
+ "learning_rate": 2.8166351606805297e-05,
919
+ "loss": 1.8169,
920
+ "step": 15200
921
+ },
922
+ {
923
+ "epoch": 7.23,
924
+ "learning_rate": 2.7693761814744802e-05,
925
+ "loss": 1.8673,
926
+ "step": 15300
927
+ },
928
+ {
929
+ "epoch": 7.28,
930
+ "learning_rate": 2.722117202268431e-05,
931
+ "loss": 1.8615,
932
+ "step": 15400
933
+ },
934
+ {
935
+ "epoch": 7.33,
936
+ "learning_rate": 2.674858223062382e-05,
937
+ "loss": 1.8072,
938
+ "step": 15500
939
  }
940
  ],
941
  "max_steps": 21160,
942
  "num_train_epochs": 10,
943
+ "total_flos": 1.1710826687824896e+17,
944
  "trial_name": null,
945
  "trial_params": null
946
  }