tmnam20 commited on
Commit
a707120
1 Parent(s): 7811173

Upload trainer_state.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. trainer_state.json +892 -0
trainer_state.json ADDED
@@ -0,0 +1,892 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 6315,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "learning_rate": 1.9841646872525734e-05,
14
+ "loss": 0.6101,
15
+ "step": 50
16
+ },
17
+ {
18
+ "epoch": 0.05,
19
+ "learning_rate": 1.9683293745051467e-05,
20
+ "loss": 0.4973,
21
+ "step": 100
22
+ },
23
+ {
24
+ "epoch": 0.07,
25
+ "learning_rate": 1.95249406175772e-05,
26
+ "loss": 0.4698,
27
+ "step": 150
28
+ },
29
+ {
30
+ "epoch": 0.1,
31
+ "learning_rate": 1.936658749010293e-05,
32
+ "loss": 0.4287,
33
+ "step": 200
34
+ },
35
+ {
36
+ "epoch": 0.12,
37
+ "learning_rate": 1.9208234362628665e-05,
38
+ "loss": 0.4536,
39
+ "step": 250
40
+ },
41
+ {
42
+ "epoch": 0.14,
43
+ "learning_rate": 1.9049881235154395e-05,
44
+ "loss": 0.4286,
45
+ "step": 300
46
+ },
47
+ {
48
+ "epoch": 0.17,
49
+ "learning_rate": 1.8891528107680128e-05,
50
+ "loss": 0.404,
51
+ "step": 350
52
+ },
53
+ {
54
+ "epoch": 0.19,
55
+ "learning_rate": 1.873317498020586e-05,
56
+ "loss": 0.4082,
57
+ "step": 400
58
+ },
59
+ {
60
+ "epoch": 0.21,
61
+ "learning_rate": 1.8574821852731593e-05,
62
+ "loss": 0.3773,
63
+ "step": 450
64
+ },
65
+ {
66
+ "epoch": 0.24,
67
+ "learning_rate": 1.8416468725257326e-05,
68
+ "loss": 0.4066,
69
+ "step": 500
70
+ },
71
+ {
72
+ "epoch": 0.24,
73
+ "eval_accuracy": 0.8291284403669725,
74
+ "eval_loss": 0.3869248926639557,
75
+ "eval_runtime": 0.9123,
76
+ "eval_samples_per_second": 955.831,
77
+ "eval_steps_per_second": 60.287,
78
+ "step": 500
79
+ },
80
+ {
81
+ "epoch": 0.26,
82
+ "learning_rate": 1.825811559778306e-05,
83
+ "loss": 0.3933,
84
+ "step": 550
85
+ },
86
+ {
87
+ "epoch": 0.29,
88
+ "learning_rate": 1.809976247030879e-05,
89
+ "loss": 0.3535,
90
+ "step": 600
91
+ },
92
+ {
93
+ "epoch": 0.31,
94
+ "learning_rate": 1.794140934283452e-05,
95
+ "loss": 0.3736,
96
+ "step": 650
97
+ },
98
+ {
99
+ "epoch": 0.33,
100
+ "learning_rate": 1.7783056215360253e-05,
101
+ "loss": 0.3384,
102
+ "step": 700
103
+ },
104
+ {
105
+ "epoch": 0.36,
106
+ "learning_rate": 1.762470308788599e-05,
107
+ "loss": 0.3733,
108
+ "step": 750
109
+ },
110
+ {
111
+ "epoch": 0.38,
112
+ "learning_rate": 1.746634996041172e-05,
113
+ "loss": 0.349,
114
+ "step": 800
115
+ },
116
+ {
117
+ "epoch": 0.4,
118
+ "learning_rate": 1.730799683293745e-05,
119
+ "loss": 0.3517,
120
+ "step": 850
121
+ },
122
+ {
123
+ "epoch": 0.43,
124
+ "learning_rate": 1.7149643705463184e-05,
125
+ "loss": 0.3728,
126
+ "step": 900
127
+ },
128
+ {
129
+ "epoch": 0.45,
130
+ "learning_rate": 1.6991290577988917e-05,
131
+ "loss": 0.3452,
132
+ "step": 950
133
+ },
134
+ {
135
+ "epoch": 0.48,
136
+ "learning_rate": 1.6832937450514647e-05,
137
+ "loss": 0.3414,
138
+ "step": 1000
139
+ },
140
+ {
141
+ "epoch": 0.48,
142
+ "eval_accuracy": 0.8486238532110092,
143
+ "eval_loss": 0.34994572401046753,
144
+ "eval_runtime": 0.9133,
145
+ "eval_samples_per_second": 954.828,
146
+ "eval_steps_per_second": 60.224,
147
+ "step": 1000
148
+ },
149
+ {
150
+ "epoch": 0.5,
151
+ "learning_rate": 1.6674584323040383e-05,
152
+ "loss": 0.329,
153
+ "step": 1050
154
+ },
155
+ {
156
+ "epoch": 0.52,
157
+ "learning_rate": 1.6516231195566115e-05,
158
+ "loss": 0.3401,
159
+ "step": 1100
160
+ },
161
+ {
162
+ "epoch": 0.55,
163
+ "learning_rate": 1.6357878068091845e-05,
164
+ "loss": 0.3335,
165
+ "step": 1150
166
+ },
167
+ {
168
+ "epoch": 0.57,
169
+ "learning_rate": 1.6199524940617578e-05,
170
+ "loss": 0.3066,
171
+ "step": 1200
172
+ },
173
+ {
174
+ "epoch": 0.59,
175
+ "learning_rate": 1.604117181314331e-05,
176
+ "loss": 0.3144,
177
+ "step": 1250
178
+ },
179
+ {
180
+ "epoch": 0.62,
181
+ "learning_rate": 1.5882818685669043e-05,
182
+ "loss": 0.3292,
183
+ "step": 1300
184
+ },
185
+ {
186
+ "epoch": 0.64,
187
+ "learning_rate": 1.5724465558194776e-05,
188
+ "loss": 0.301,
189
+ "step": 1350
190
+ },
191
+ {
192
+ "epoch": 0.67,
193
+ "learning_rate": 1.556611243072051e-05,
194
+ "loss": 0.3254,
195
+ "step": 1400
196
+ },
197
+ {
198
+ "epoch": 0.69,
199
+ "learning_rate": 1.540775930324624e-05,
200
+ "loss": 0.3331,
201
+ "step": 1450
202
+ },
203
+ {
204
+ "epoch": 0.71,
205
+ "learning_rate": 1.5249406175771972e-05,
206
+ "loss": 0.3133,
207
+ "step": 1500
208
+ },
209
+ {
210
+ "epoch": 0.71,
211
+ "eval_accuracy": 0.8509174311926605,
212
+ "eval_loss": 0.37425312399864197,
213
+ "eval_runtime": 0.919,
214
+ "eval_samples_per_second": 948.854,
215
+ "eval_steps_per_second": 59.847,
216
+ "step": 1500
217
+ },
218
+ {
219
+ "epoch": 0.74,
220
+ "learning_rate": 1.5091053048297705e-05,
221
+ "loss": 0.31,
222
+ "step": 1550
223
+ },
224
+ {
225
+ "epoch": 0.76,
226
+ "learning_rate": 1.4932699920823436e-05,
227
+ "loss": 0.3009,
228
+ "step": 1600
229
+ },
230
+ {
231
+ "epoch": 0.78,
232
+ "learning_rate": 1.477434679334917e-05,
233
+ "loss": 0.2757,
234
+ "step": 1650
235
+ },
236
+ {
237
+ "epoch": 0.81,
238
+ "learning_rate": 1.4615993665874902e-05,
239
+ "loss": 0.3219,
240
+ "step": 1700
241
+ },
242
+ {
243
+ "epoch": 0.83,
244
+ "learning_rate": 1.4457640538400634e-05,
245
+ "loss": 0.298,
246
+ "step": 1750
247
+ },
248
+ {
249
+ "epoch": 0.86,
250
+ "learning_rate": 1.4299287410926367e-05,
251
+ "loss": 0.2881,
252
+ "step": 1800
253
+ },
254
+ {
255
+ "epoch": 0.88,
256
+ "learning_rate": 1.4140934283452098e-05,
257
+ "loss": 0.2793,
258
+ "step": 1850
259
+ },
260
+ {
261
+ "epoch": 0.9,
262
+ "learning_rate": 1.3982581155977833e-05,
263
+ "loss": 0.2987,
264
+ "step": 1900
265
+ },
266
+ {
267
+ "epoch": 0.93,
268
+ "learning_rate": 1.3824228028503564e-05,
269
+ "loss": 0.2559,
270
+ "step": 1950
271
+ },
272
+ {
273
+ "epoch": 0.95,
274
+ "learning_rate": 1.3665874901029297e-05,
275
+ "loss": 0.2797,
276
+ "step": 2000
277
+ },
278
+ {
279
+ "epoch": 0.95,
280
+ "eval_accuracy": 0.8474770642201835,
281
+ "eval_loss": 0.41188526153564453,
282
+ "eval_runtime": 0.9208,
283
+ "eval_samples_per_second": 947.047,
284
+ "eval_steps_per_second": 59.733,
285
+ "step": 2000
286
+ },
287
+ {
288
+ "epoch": 0.97,
289
+ "learning_rate": 1.3507521773555028e-05,
290
+ "loss": 0.2811,
291
+ "step": 2050
292
+ },
293
+ {
294
+ "epoch": 1.0,
295
+ "learning_rate": 1.3349168646080762e-05,
296
+ "loss": 0.3067,
297
+ "step": 2100
298
+ },
299
+ {
300
+ "epoch": 1.02,
301
+ "learning_rate": 1.3190815518606495e-05,
302
+ "loss": 0.2245,
303
+ "step": 2150
304
+ },
305
+ {
306
+ "epoch": 1.05,
307
+ "learning_rate": 1.3032462391132226e-05,
308
+ "loss": 0.2384,
309
+ "step": 2200
310
+ },
311
+ {
312
+ "epoch": 1.07,
313
+ "learning_rate": 1.2874109263657959e-05,
314
+ "loss": 0.2247,
315
+ "step": 2250
316
+ },
317
+ {
318
+ "epoch": 1.09,
319
+ "learning_rate": 1.271575613618369e-05,
320
+ "loss": 0.2221,
321
+ "step": 2300
322
+ },
323
+ {
324
+ "epoch": 1.12,
325
+ "learning_rate": 1.2557403008709424e-05,
326
+ "loss": 0.2202,
327
+ "step": 2350
328
+ },
329
+ {
330
+ "epoch": 1.14,
331
+ "learning_rate": 1.2399049881235155e-05,
332
+ "loss": 0.2247,
333
+ "step": 2400
334
+ },
335
+ {
336
+ "epoch": 1.16,
337
+ "learning_rate": 1.2240696753760888e-05,
338
+ "loss": 0.2315,
339
+ "step": 2450
340
+ },
341
+ {
342
+ "epoch": 1.19,
343
+ "learning_rate": 1.208234362628662e-05,
344
+ "loss": 0.236,
345
+ "step": 2500
346
+ },
347
+ {
348
+ "epoch": 1.19,
349
+ "eval_accuracy": 0.8669724770642202,
350
+ "eval_loss": 0.3891030550003052,
351
+ "eval_runtime": 0.9164,
352
+ "eval_samples_per_second": 951.574,
353
+ "eval_steps_per_second": 60.019,
354
+ "step": 2500
355
+ },
356
+ {
357
+ "epoch": 1.21,
358
+ "learning_rate": 1.1923990498812352e-05,
359
+ "loss": 0.2153,
360
+ "step": 2550
361
+ },
362
+ {
363
+ "epoch": 1.24,
364
+ "learning_rate": 1.1765637371338086e-05,
365
+ "loss": 0.2385,
366
+ "step": 2600
367
+ },
368
+ {
369
+ "epoch": 1.26,
370
+ "learning_rate": 1.1607284243863817e-05,
371
+ "loss": 0.236,
372
+ "step": 2650
373
+ },
374
+ {
375
+ "epoch": 1.28,
376
+ "learning_rate": 1.144893111638955e-05,
377
+ "loss": 0.211,
378
+ "step": 2700
379
+ },
380
+ {
381
+ "epoch": 1.31,
382
+ "learning_rate": 1.1290577988915281e-05,
383
+ "loss": 0.2313,
384
+ "step": 2750
385
+ },
386
+ {
387
+ "epoch": 1.33,
388
+ "learning_rate": 1.1132224861441014e-05,
389
+ "loss": 0.2238,
390
+ "step": 2800
391
+ },
392
+ {
393
+ "epoch": 1.35,
394
+ "learning_rate": 1.0973871733966748e-05,
395
+ "loss": 0.2213,
396
+ "step": 2850
397
+ },
398
+ {
399
+ "epoch": 1.38,
400
+ "learning_rate": 1.081551860649248e-05,
401
+ "loss": 0.2171,
402
+ "step": 2900
403
+ },
404
+ {
405
+ "epoch": 1.4,
406
+ "learning_rate": 1.0657165479018212e-05,
407
+ "loss": 0.1877,
408
+ "step": 2950
409
+ },
410
+ {
411
+ "epoch": 1.43,
412
+ "learning_rate": 1.0498812351543943e-05,
413
+ "loss": 0.2202,
414
+ "step": 3000
415
+ },
416
+ {
417
+ "epoch": 1.43,
418
+ "eval_accuracy": 0.8738532110091743,
419
+ "eval_loss": 0.36398938298225403,
420
+ "eval_runtime": 0.9162,
421
+ "eval_samples_per_second": 951.726,
422
+ "eval_steps_per_second": 60.029,
423
+ "step": 3000
424
+ },
425
+ {
426
+ "epoch": 1.45,
427
+ "learning_rate": 1.0340459224069676e-05,
428
+ "loss": 0.2318,
429
+ "step": 3050
430
+ },
431
+ {
432
+ "epoch": 1.47,
433
+ "learning_rate": 1.0182106096595407e-05,
434
+ "loss": 0.2056,
435
+ "step": 3100
436
+ },
437
+ {
438
+ "epoch": 1.5,
439
+ "learning_rate": 1.0023752969121141e-05,
440
+ "loss": 0.2206,
441
+ "step": 3150
442
+ },
443
+ {
444
+ "epoch": 1.52,
445
+ "learning_rate": 9.865399841646872e-06,
446
+ "loss": 0.2048,
447
+ "step": 3200
448
+ },
449
+ {
450
+ "epoch": 1.54,
451
+ "learning_rate": 9.707046714172605e-06,
452
+ "loss": 0.237,
453
+ "step": 3250
454
+ },
455
+ {
456
+ "epoch": 1.57,
457
+ "learning_rate": 9.548693586698338e-06,
458
+ "loss": 0.2514,
459
+ "step": 3300
460
+ },
461
+ {
462
+ "epoch": 1.59,
463
+ "learning_rate": 9.39034045922407e-06,
464
+ "loss": 0.2147,
465
+ "step": 3350
466
+ },
467
+ {
468
+ "epoch": 1.62,
469
+ "learning_rate": 9.231987331749803e-06,
470
+ "loss": 0.2233,
471
+ "step": 3400
472
+ },
473
+ {
474
+ "epoch": 1.64,
475
+ "learning_rate": 9.073634204275536e-06,
476
+ "loss": 0.2236,
477
+ "step": 3450
478
+ },
479
+ {
480
+ "epoch": 1.66,
481
+ "learning_rate": 8.915281076801267e-06,
482
+ "loss": 0.1889,
483
+ "step": 3500
484
+ },
485
+ {
486
+ "epoch": 1.66,
487
+ "eval_accuracy": 0.8681192660550459,
488
+ "eval_loss": 0.38289329409599304,
489
+ "eval_runtime": 0.9155,
490
+ "eval_samples_per_second": 952.465,
491
+ "eval_steps_per_second": 60.075,
492
+ "step": 3500
493
+ },
494
+ {
495
+ "epoch": 1.69,
496
+ "learning_rate": 8.756927949327e-06,
497
+ "loss": 0.2266,
498
+ "step": 3550
499
+ },
500
+ {
501
+ "epoch": 1.71,
502
+ "learning_rate": 8.598574821852733e-06,
503
+ "loss": 0.2293,
504
+ "step": 3600
505
+ },
506
+ {
507
+ "epoch": 1.73,
508
+ "learning_rate": 8.440221694378464e-06,
509
+ "loss": 0.207,
510
+ "step": 3650
511
+ },
512
+ {
513
+ "epoch": 1.76,
514
+ "learning_rate": 8.281868566904197e-06,
515
+ "loss": 0.2282,
516
+ "step": 3700
517
+ },
518
+ {
519
+ "epoch": 1.78,
520
+ "learning_rate": 8.12351543942993e-06,
521
+ "loss": 0.1878,
522
+ "step": 3750
523
+ },
524
+ {
525
+ "epoch": 1.81,
526
+ "learning_rate": 7.965162311955662e-06,
527
+ "loss": 0.2096,
528
+ "step": 3800
529
+ },
530
+ {
531
+ "epoch": 1.83,
532
+ "learning_rate": 7.806809184481395e-06,
533
+ "loss": 0.2033,
534
+ "step": 3850
535
+ },
536
+ {
537
+ "epoch": 1.85,
538
+ "learning_rate": 7.648456057007126e-06,
539
+ "loss": 0.2032,
540
+ "step": 3900
541
+ },
542
+ {
543
+ "epoch": 1.88,
544
+ "learning_rate": 7.490102929532859e-06,
545
+ "loss": 0.1959,
546
+ "step": 3950
547
+ },
548
+ {
549
+ "epoch": 1.9,
550
+ "learning_rate": 7.331749802058591e-06,
551
+ "loss": 0.1847,
552
+ "step": 4000
553
+ },
554
+ {
555
+ "epoch": 1.9,
556
+ "eval_accuracy": 0.8795871559633027,
557
+ "eval_loss": 0.3687029778957367,
558
+ "eval_runtime": 0.9152,
559
+ "eval_samples_per_second": 952.748,
560
+ "eval_steps_per_second": 60.093,
561
+ "step": 4000
562
+ },
563
+ {
564
+ "epoch": 1.92,
565
+ "learning_rate": 7.173396674584323e-06,
566
+ "loss": 0.2019,
567
+ "step": 4050
568
+ },
569
+ {
570
+ "epoch": 1.95,
571
+ "learning_rate": 7.015043547110056e-06,
572
+ "loss": 0.2026,
573
+ "step": 4100
574
+ },
575
+ {
576
+ "epoch": 1.97,
577
+ "learning_rate": 6.856690419635789e-06,
578
+ "loss": 0.2166,
579
+ "step": 4150
580
+ },
581
+ {
582
+ "epoch": 2.0,
583
+ "learning_rate": 6.698337292161521e-06,
584
+ "loss": 0.1776,
585
+ "step": 4200
586
+ },
587
+ {
588
+ "epoch": 2.02,
589
+ "learning_rate": 6.539984164687253e-06,
590
+ "loss": 0.1738,
591
+ "step": 4250
592
+ },
593
+ {
594
+ "epoch": 2.04,
595
+ "learning_rate": 6.3816310372129854e-06,
596
+ "loss": 0.1399,
597
+ "step": 4300
598
+ },
599
+ {
600
+ "epoch": 2.07,
601
+ "learning_rate": 6.223277909738717e-06,
602
+ "loss": 0.1441,
603
+ "step": 4350
604
+ },
605
+ {
606
+ "epoch": 2.09,
607
+ "learning_rate": 6.06492478226445e-06,
608
+ "loss": 0.1578,
609
+ "step": 4400
610
+ },
611
+ {
612
+ "epoch": 2.11,
613
+ "learning_rate": 5.906571654790183e-06,
614
+ "loss": 0.1753,
615
+ "step": 4450
616
+ },
617
+ {
618
+ "epoch": 2.14,
619
+ "learning_rate": 5.748218527315916e-06,
620
+ "loss": 0.1288,
621
+ "step": 4500
622
+ },
623
+ {
624
+ "epoch": 2.14,
625
+ "eval_accuracy": 0.8807339449541285,
626
+ "eval_loss": 0.4524359703063965,
627
+ "eval_runtime": 0.9177,
628
+ "eval_samples_per_second": 950.221,
629
+ "eval_steps_per_second": 59.934,
630
+ "step": 4500
631
+ },
632
+ {
633
+ "epoch": 2.16,
634
+ "learning_rate": 5.5898653998416475e-06,
635
+ "loss": 0.1249,
636
+ "step": 4550
637
+ },
638
+ {
639
+ "epoch": 2.19,
640
+ "learning_rate": 5.431512272367379e-06,
641
+ "loss": 0.1941,
642
+ "step": 4600
643
+ },
644
+ {
645
+ "epoch": 2.21,
646
+ "learning_rate": 5.273159144893112e-06,
647
+ "loss": 0.1798,
648
+ "step": 4650
649
+ },
650
+ {
651
+ "epoch": 2.23,
652
+ "learning_rate": 5.114806017418844e-06,
653
+ "loss": 0.1602,
654
+ "step": 4700
655
+ },
656
+ {
657
+ "epoch": 2.26,
658
+ "learning_rate": 4.956452889944577e-06,
659
+ "loss": 0.1359,
660
+ "step": 4750
661
+ },
662
+ {
663
+ "epoch": 2.28,
664
+ "learning_rate": 4.798099762470309e-06,
665
+ "loss": 0.1581,
666
+ "step": 4800
667
+ },
668
+ {
669
+ "epoch": 2.3,
670
+ "learning_rate": 4.6397466349960415e-06,
671
+ "loss": 0.15,
672
+ "step": 4850
673
+ },
674
+ {
675
+ "epoch": 2.33,
676
+ "learning_rate": 4.481393507521774e-06,
677
+ "loss": 0.1681,
678
+ "step": 4900
679
+ },
680
+ {
681
+ "epoch": 2.35,
682
+ "learning_rate": 4.323040380047506e-06,
683
+ "loss": 0.1527,
684
+ "step": 4950
685
+ },
686
+ {
687
+ "epoch": 2.38,
688
+ "learning_rate": 4.164687252573238e-06,
689
+ "loss": 0.1478,
690
+ "step": 5000
691
+ },
692
+ {
693
+ "epoch": 2.38,
694
+ "eval_accuracy": 0.875,
695
+ "eval_loss": 0.42590492963790894,
696
+ "eval_runtime": 0.9164,
697
+ "eval_samples_per_second": 951.592,
698
+ "eval_steps_per_second": 60.02,
699
+ "step": 5000
700
+ },
701
+ {
702
+ "epoch": 2.4,
703
+ "learning_rate": 4.006334125098972e-06,
704
+ "loss": 0.1698,
705
+ "step": 5050
706
+ },
707
+ {
708
+ "epoch": 2.42,
709
+ "learning_rate": 3.8479809976247036e-06,
710
+ "loss": 0.151,
711
+ "step": 5100
712
+ },
713
+ {
714
+ "epoch": 2.45,
715
+ "learning_rate": 3.6896278701504355e-06,
716
+ "loss": 0.1663,
717
+ "step": 5150
718
+ },
719
+ {
720
+ "epoch": 2.47,
721
+ "learning_rate": 3.531274742676168e-06,
722
+ "loss": 0.162,
723
+ "step": 5200
724
+ },
725
+ {
726
+ "epoch": 2.49,
727
+ "learning_rate": 3.3729216152019006e-06,
728
+ "loss": 0.1691,
729
+ "step": 5250
730
+ },
731
+ {
732
+ "epoch": 2.52,
733
+ "learning_rate": 3.214568487727633e-06,
734
+ "loss": 0.1384,
735
+ "step": 5300
736
+ },
737
+ {
738
+ "epoch": 2.54,
739
+ "learning_rate": 3.0562153602533652e-06,
740
+ "loss": 0.1513,
741
+ "step": 5350
742
+ },
743
+ {
744
+ "epoch": 2.57,
745
+ "learning_rate": 2.897862232779098e-06,
746
+ "loss": 0.1394,
747
+ "step": 5400
748
+ },
749
+ {
750
+ "epoch": 2.59,
751
+ "learning_rate": 2.73950910530483e-06,
752
+ "loss": 0.154,
753
+ "step": 5450
754
+ },
755
+ {
756
+ "epoch": 2.61,
757
+ "learning_rate": 2.581155977830562e-06,
758
+ "loss": 0.1761,
759
+ "step": 5500
760
+ },
761
+ {
762
+ "epoch": 2.61,
763
+ "eval_accuracy": 0.8818807339449541,
764
+ "eval_loss": 0.40600278973579407,
765
+ "eval_runtime": 1.6773,
766
+ "eval_samples_per_second": 519.893,
767
+ "eval_steps_per_second": 32.791,
768
+ "step": 5500
769
+ },
770
+ {
771
+ "epoch": 2.64,
772
+ "learning_rate": 2.422802850356295e-06,
773
+ "loss": 0.163,
774
+ "step": 5550
775
+ },
776
+ {
777
+ "epoch": 2.66,
778
+ "learning_rate": 2.264449722882027e-06,
779
+ "loss": 0.143,
780
+ "step": 5600
781
+ },
782
+ {
783
+ "epoch": 2.68,
784
+ "learning_rate": 2.1060965954077596e-06,
785
+ "loss": 0.1528,
786
+ "step": 5650
787
+ },
788
+ {
789
+ "epoch": 2.71,
790
+ "learning_rate": 1.947743467933492e-06,
791
+ "loss": 0.1403,
792
+ "step": 5700
793
+ },
794
+ {
795
+ "epoch": 2.73,
796
+ "learning_rate": 1.789390340459224e-06,
797
+ "loss": 0.1273,
798
+ "step": 5750
799
+ },
800
+ {
801
+ "epoch": 2.76,
802
+ "learning_rate": 1.6310372129849566e-06,
803
+ "loss": 0.1575,
804
+ "step": 5800
805
+ },
806
+ {
807
+ "epoch": 2.78,
808
+ "learning_rate": 1.472684085510689e-06,
809
+ "loss": 0.1285,
810
+ "step": 5850
811
+ },
812
+ {
813
+ "epoch": 2.8,
814
+ "learning_rate": 1.3143309580364213e-06,
815
+ "loss": 0.1461,
816
+ "step": 5900
817
+ },
818
+ {
819
+ "epoch": 2.83,
820
+ "learning_rate": 1.1559778305621536e-06,
821
+ "loss": 0.154,
822
+ "step": 5950
823
+ },
824
+ {
825
+ "epoch": 2.85,
826
+ "learning_rate": 9.976247030878861e-07,
827
+ "loss": 0.1487,
828
+ "step": 6000
829
+ },
830
+ {
831
+ "epoch": 2.85,
832
+ "eval_accuracy": 0.8807339449541285,
833
+ "eval_loss": 0.4408171772956848,
834
+ "eval_runtime": 0.9153,
835
+ "eval_samples_per_second": 952.703,
836
+ "eval_steps_per_second": 60.09,
837
+ "step": 6000
838
+ },
839
+ {
840
+ "epoch": 2.87,
841
+ "learning_rate": 8.392715756136184e-07,
842
+ "loss": 0.1697,
843
+ "step": 6050
844
+ },
845
+ {
846
+ "epoch": 2.9,
847
+ "learning_rate": 6.809184481393508e-07,
848
+ "loss": 0.1362,
849
+ "step": 6100
850
+ },
851
+ {
852
+ "epoch": 2.92,
853
+ "learning_rate": 5.225653206650831e-07,
854
+ "loss": 0.1442,
855
+ "step": 6150
856
+ },
857
+ {
858
+ "epoch": 2.95,
859
+ "learning_rate": 3.642121931908155e-07,
860
+ "loss": 0.1833,
861
+ "step": 6200
862
+ },
863
+ {
864
+ "epoch": 2.97,
865
+ "learning_rate": 2.0585906571654793e-07,
866
+ "loss": 0.1545,
867
+ "step": 6250
868
+ },
869
+ {
870
+ "epoch": 2.99,
871
+ "learning_rate": 4.750593824228029e-08,
872
+ "loss": 0.1405,
873
+ "step": 6300
874
+ },
875
+ {
876
+ "epoch": 3.0,
877
+ "step": 6315,
878
+ "total_flos": 5774180234831100.0,
879
+ "train_loss": 0.24077867121330254,
880
+ "train_runtime": 816.4155,
881
+ "train_samples_per_second": 247.481,
882
+ "train_steps_per_second": 7.735
883
+ }
884
+ ],
885
+ "logging_steps": 50,
886
+ "max_steps": 6315,
887
+ "num_train_epochs": 3,
888
+ "save_steps": 500,
889
+ "total_flos": 5774180234831100.0,
890
+ "trial_name": null,
891
+ "trial_params": null
892
+ }