pszemraj commited on
Commit
f94e1b2
1 Parent(s): f9cd9c3

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1075 -0
trainer_state.json ADDED
@@ -0,0 +1,1075 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0,
5
+ "global_step": 351,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 5e-05,
13
+ "loss": 0.1291,
14
+ "step": 2
15
+ },
16
+ {
17
+ "epoch": 0.01,
18
+ "learning_rate": 0.0001,
19
+ "loss": 0.1354,
20
+ "step": 4
21
+ },
22
+ {
23
+ "epoch": 0.02,
24
+ "learning_rate": 0.00015000000000000001,
25
+ "loss": 0.1399,
26
+ "step": 6
27
+ },
28
+ {
29
+ "epoch": 0.02,
30
+ "learning_rate": 0.0002,
31
+ "loss": 0.143,
32
+ "step": 8
33
+ },
34
+ {
35
+ "epoch": 0.03,
36
+ "learning_rate": 0.0001999832224185224,
37
+ "loss": 0.1455,
38
+ "step": 10
39
+ },
40
+ {
41
+ "epoch": 0.03,
42
+ "learning_rate": 0.00019993289530383432,
43
+ "loss": 0.1346,
44
+ "step": 12
45
+ },
46
+ {
47
+ "epoch": 0.04,
48
+ "learning_rate": 0.00019984903554328114,
49
+ "loss": 0.1458,
50
+ "step": 14
51
+ },
52
+ {
53
+ "epoch": 0.05,
54
+ "learning_rate": 0.00019973167127614215,
55
+ "loss": 0.1354,
56
+ "step": 16
57
+ },
58
+ {
59
+ "epoch": 0.05,
60
+ "learning_rate": 0.0001995808418841885,
61
+ "loss": 0.1248,
62
+ "step": 18
63
+ },
64
+ {
65
+ "epoch": 0.06,
66
+ "learning_rate": 0.0001993965979784684,
67
+ "loss": 0.1322,
68
+ "step": 20
69
+ },
70
+ {
71
+ "epoch": 0.06,
72
+ "learning_rate": 0.0001991790013823246,
73
+ "loss": 0.1371,
74
+ "step": 22
75
+ },
76
+ {
77
+ "epoch": 0.07,
78
+ "learning_rate": 0.0001989281251106496,
79
+ "loss": 0.1195,
80
+ "step": 24
81
+ },
82
+ {
83
+ "epoch": 0.07,
84
+ "learning_rate": 0.00019864405334538517,
85
+ "loss": 0.1193,
86
+ "step": 26
87
+ },
88
+ {
89
+ "epoch": 0.08,
90
+ "learning_rate": 0.000198326881407275,
91
+ "loss": 0.1313,
92
+ "step": 28
93
+ },
94
+ {
95
+ "epoch": 0.09,
96
+ "learning_rate": 0.00019797671572387984,
97
+ "loss": 0.148,
98
+ "step": 30
99
+ },
100
+ {
101
+ "epoch": 0.09,
102
+ "learning_rate": 0.0001975936737938653,
103
+ "loss": 0.1384,
104
+ "step": 32
105
+ },
106
+ {
107
+ "epoch": 0.1,
108
+ "learning_rate": 0.00019717788414757523,
109
+ "loss": 0.1317,
110
+ "step": 34
111
+ },
112
+ {
113
+ "epoch": 0.1,
114
+ "learning_rate": 0.00019672948630390294,
115
+ "loss": 0.1364,
116
+ "step": 36
117
+ },
118
+ {
119
+ "epoch": 0.11,
120
+ "learning_rate": 0.00019624863072347564,
121
+ "loss": 0.127,
122
+ "step": 38
123
+ },
124
+ {
125
+ "epoch": 0.11,
126
+ "learning_rate": 0.00019573547875816684,
127
+ "loss": 0.1281,
128
+ "step": 40
129
+ },
130
+ {
131
+ "epoch": 0.12,
132
+ "learning_rate": 0.0001951902025969548,
133
+ "loss": 0.1303,
134
+ "step": 42
135
+ },
136
+ {
137
+ "epoch": 0.13,
138
+ "learning_rate": 0.0001946129852081439,
139
+ "loss": 0.1364,
140
+ "step": 44
141
+ },
142
+ {
143
+ "epoch": 0.13,
144
+ "learning_rate": 0.00019400402027796955,
145
+ "loss": 0.1262,
146
+ "step": 46
147
+ },
148
+ {
149
+ "epoch": 0.14,
150
+ "learning_rate": 0.00019336351214560647,
151
+ "loss": 0.1244,
152
+ "step": 48
153
+ },
154
+ {
155
+ "epoch": 0.14,
156
+ "learning_rate": 0.0001926916757346022,
157
+ "loss": 0.1261,
158
+ "step": 50
159
+ },
160
+ {
161
+ "epoch": 0.15,
162
+ "learning_rate": 0.0001919887364807592,
163
+ "loss": 0.1276,
164
+ "step": 52
165
+ },
166
+ {
167
+ "epoch": 0.15,
168
+ "learning_rate": 0.00019125493025648962,
169
+ "loss": 0.1181,
170
+ "step": 54
171
+ },
172
+ {
173
+ "epoch": 0.16,
174
+ "learning_rate": 0.00019049050329166778,
175
+ "loss": 0.1218,
176
+ "step": 56
177
+ },
178
+ {
179
+ "epoch": 0.17,
180
+ "learning_rate": 0.0001896957120910074,
181
+ "loss": 0.1255,
182
+ "step": 58
183
+ },
184
+ {
185
+ "epoch": 0.17,
186
+ "learning_rate": 0.00018887082334799097,
187
+ "loss": 0.1279,
188
+ "step": 60
189
+ },
190
+ {
191
+ "epoch": 0.18,
192
+ "learning_rate": 0.00018801611385538047,
193
+ "loss": 0.1227,
194
+ "step": 62
195
+ },
196
+ {
197
+ "epoch": 0.18,
198
+ "learning_rate": 0.00018713187041233896,
199
+ "loss": 0.132,
200
+ "step": 64
201
+ },
202
+ {
203
+ "epoch": 0.19,
204
+ "learning_rate": 0.00018621838972819458,
205
+ "loss": 0.1249,
206
+ "step": 66
207
+ },
208
+ {
209
+ "epoch": 0.19,
210
+ "learning_rate": 0.00018527597832287954,
211
+ "loss": 0.1182,
212
+ "step": 68
213
+ },
214
+ {
215
+ "epoch": 0.2,
216
+ "learning_rate": 0.00018430495242407658,
217
+ "loss": 0.1325,
218
+ "step": 70
219
+ },
220
+ {
221
+ "epoch": 0.21,
222
+ "learning_rate": 0.00018330563786110834,
223
+ "loss": 0.1209,
224
+ "step": 72
225
+ },
226
+ {
227
+ "epoch": 0.21,
228
+ "learning_rate": 0.0001822783699556049,
229
+ "loss": 0.1322,
230
+ "step": 74
231
+ },
232
+ {
233
+ "epoch": 0.22,
234
+ "learning_rate": 0.00018122349340898595,
235
+ "loss": 0.1225,
236
+ "step": 76
237
+ },
238
+ {
239
+ "epoch": 0.22,
240
+ "learning_rate": 0.00018014136218679567,
241
+ "loss": 0.1209,
242
+ "step": 78
243
+ },
244
+ {
245
+ "epoch": 0.23,
246
+ "learning_rate": 0.00017903233939992906,
247
+ "loss": 0.125,
248
+ "step": 80
249
+ },
250
+ {
251
+ "epoch": 0.23,
252
+ "learning_rate": 0.00017789679718278943,
253
+ "loss": 0.1232,
254
+ "step": 82
255
+ },
256
+ {
257
+ "epoch": 0.24,
258
+ "learning_rate": 0.00017673511656841822,
259
+ "loss": 0.1199,
260
+ "step": 84
261
+ },
262
+ {
263
+ "epoch": 0.25,
264
+ "learning_rate": 0.00017554768736063859,
265
+ "loss": 0.1228,
266
+ "step": 86
267
+ },
268
+ {
269
+ "epoch": 0.25,
270
+ "learning_rate": 0.00017433490800325614,
271
+ "loss": 0.1303,
272
+ "step": 88
273
+ },
274
+ {
275
+ "epoch": 0.26,
276
+ "learning_rate": 0.00017309718544636057,
277
+ "loss": 0.12,
278
+ "step": 90
279
+ },
280
+ {
281
+ "epoch": 0.26,
282
+ "learning_rate": 0.00017183493500977278,
283
+ "loss": 0.1164,
284
+ "step": 92
285
+ },
286
+ {
287
+ "epoch": 0.27,
288
+ "learning_rate": 0.00017054858024368366,
289
+ "loss": 0.123,
290
+ "step": 94
291
+ },
292
+ {
293
+ "epoch": 0.27,
294
+ "learning_rate": 0.00016923855278653114,
295
+ "loss": 0.1156,
296
+ "step": 96
297
+ },
298
+ {
299
+ "epoch": 0.28,
300
+ "learning_rate": 0.00016790529222016328,
301
+ "loss": 0.1199,
302
+ "step": 98
303
+ },
304
+ {
305
+ "epoch": 0.28,
306
+ "learning_rate": 0.00016654924592233568,
307
+ "loss": 0.115,
308
+ "step": 100
309
+ },
310
+ {
311
+ "epoch": 0.29,
312
+ "learning_rate": 0.00016517086891659335,
313
+ "loss": 0.1248,
314
+ "step": 102
315
+ },
316
+ {
317
+ "epoch": 0.3,
318
+ "learning_rate": 0.00016377062371958668,
319
+ "loss": 0.1229,
320
+ "step": 104
321
+ },
322
+ {
323
+ "epoch": 0.3,
324
+ "learning_rate": 0.00016234898018587337,
325
+ "loss": 0.1188,
326
+ "step": 106
327
+ },
328
+ {
329
+ "epoch": 0.31,
330
+ "learning_rate": 0.00016090641535025774,
331
+ "loss": 0.1209,
332
+ "step": 108
333
+ },
334
+ {
335
+ "epoch": 0.31,
336
+ "learning_rate": 0.00015944341326772112,
337
+ "loss": 0.1197,
338
+ "step": 110
339
+ },
340
+ {
341
+ "epoch": 0.32,
342
+ "learning_rate": 0.00015796046485099633,
343
+ "loss": 0.1192,
344
+ "step": 112
345
+ },
346
+ {
347
+ "epoch": 0.32,
348
+ "learning_rate": 0.0001564580677058412,
349
+ "loss": 0.1124,
350
+ "step": 114
351
+ },
352
+ {
353
+ "epoch": 0.33,
354
+ "learning_rate": 0.00015493672596406598,
355
+ "loss": 0.1189,
356
+ "step": 116
357
+ },
358
+ {
359
+ "epoch": 0.34,
360
+ "learning_rate": 0.00015339695011437127,
361
+ "loss": 0.1123,
362
+ "step": 118
363
+ },
364
+ {
365
+ "epoch": 0.34,
366
+ "learning_rate": 0.00015183925683105254,
367
+ "loss": 0.1139,
368
+ "step": 120
369
+ },
370
+ {
371
+ "epoch": 0.35,
372
+ "learning_rate": 0.00015026416880062931,
373
+ "loss": 0.1165,
374
+ "step": 122
375
+ },
376
+ {
377
+ "epoch": 0.35,
378
+ "learning_rate": 0.00014867221454645696,
379
+ "loss": 0.1265,
380
+ "step": 124
381
+ },
382
+ {
383
+ "epoch": 0.36,
384
+ "learning_rate": 0.00014706392825137964,
385
+ "loss": 0.1179,
386
+ "step": 126
387
+ },
388
+ {
389
+ "epoch": 0.36,
390
+ "learning_rate": 0.0001454398495784844,
391
+ "loss": 0.1279,
392
+ "step": 128
393
+ },
394
+ {
395
+ "epoch": 0.37,
396
+ "learning_rate": 0.00014380052349001647,
397
+ "loss": 0.106,
398
+ "step": 130
399
+ },
400
+ {
401
+ "epoch": 0.38,
402
+ "learning_rate": 0.00014214650006451622,
403
+ "loss": 0.118,
404
+ "step": 132
405
+ },
406
+ {
407
+ "epoch": 0.38,
408
+ "learning_rate": 0.00014047833431223938,
409
+ "loss": 0.1276,
410
+ "step": 134
411
+ },
412
+ {
413
+ "epoch": 0.39,
414
+ "learning_rate": 0.00013879658598892254,
415
+ "loss": 0.1255,
416
+ "step": 136
417
+ },
418
+ {
419
+ "epoch": 0.39,
420
+ "learning_rate": 0.000137101819407956,
421
+ "loss": 0.1191,
422
+ "step": 138
423
+ },
424
+ {
425
+ "epoch": 0.4,
426
+ "learning_rate": 0.00013539460325102777,
427
+ "loss": 0.111,
428
+ "step": 140
429
+ },
430
+ {
431
+ "epoch": 0.4,
432
+ "learning_rate": 0.00013367551037730128,
433
+ "loss": 0.1251,
434
+ "step": 142
435
+ },
436
+ {
437
+ "epoch": 0.41,
438
+ "learning_rate": 0.00013194511763119172,
439
+ "loss": 0.1193,
440
+ "step": 144
441
+ },
442
+ {
443
+ "epoch": 0.42,
444
+ "learning_rate": 0.0001302040056488047,
445
+ "loss": 0.11,
446
+ "step": 146
447
+ },
448
+ {
449
+ "epoch": 0.42,
450
+ "learning_rate": 0.00012845275866310324,
451
+ "loss": 0.1033,
452
+ "step": 148
453
+ },
454
+ {
455
+ "epoch": 0.43,
456
+ "learning_rate": 0.00012669196430786713,
457
+ "loss": 0.1149,
458
+ "step": 150
459
+ },
460
+ {
461
+ "epoch": 0.43,
462
+ "learning_rate": 0.00012492221342051154,
463
+ "loss": 0.1231,
464
+ "step": 152
465
+ },
466
+ {
467
+ "epoch": 0.44,
468
+ "learning_rate": 0.00012314409984383066,
469
+ "loss": 0.1244,
470
+ "step": 154
471
+ },
472
+ {
473
+ "epoch": 0.44,
474
+ "learning_rate": 0.00012135822022673263,
475
+ "loss": 0.1125,
476
+ "step": 156
477
+ },
478
+ {
479
+ "epoch": 0.45,
480
+ "learning_rate": 0.00011956517382403321,
481
+ "loss": 0.1102,
482
+ "step": 158
483
+ },
484
+ {
485
+ "epoch": 0.46,
486
+ "learning_rate": 0.0001177655622953746,
487
+ "loss": 0.1175,
488
+ "step": 160
489
+ },
490
+ {
491
+ "epoch": 0.46,
492
+ "learning_rate": 0.00011595998950333793,
493
+ "loss": 0.116,
494
+ "step": 162
495
+ },
496
+ {
497
+ "epoch": 0.47,
498
+ "learning_rate": 0.00011414906131081575,
499
+ "loss": 0.1144,
500
+ "step": 164
501
+ },
502
+ {
503
+ "epoch": 0.47,
504
+ "learning_rate": 0.00011233338537771407,
505
+ "loss": 0.1229,
506
+ "step": 166
507
+ },
508
+ {
509
+ "epoch": 0.48,
510
+ "learning_rate": 0.00011051357095705101,
511
+ "loss": 0.1117,
512
+ "step": 168
513
+ },
514
+ {
515
+ "epoch": 0.48,
516
+ "learning_rate": 0.0001086902286905209,
517
+ "loss": 0.1144,
518
+ "step": 170
519
+ },
520
+ {
521
+ "epoch": 0.49,
522
+ "learning_rate": 0.00010686397040359253,
523
+ "loss": 0.109,
524
+ "step": 172
525
+ },
526
+ {
527
+ "epoch": 0.5,
528
+ "learning_rate": 0.00010503540890020997,
529
+ "loss": 0.1151,
530
+ "step": 174
531
+ },
532
+ {
533
+ "epoch": 0.5,
534
+ "learning_rate": 0.00010320515775716555,
535
+ "loss": 0.1197,
536
+ "step": 176
537
+ },
538
+ {
539
+ "epoch": 0.51,
540
+ "learning_rate": 0.00010137383111821266,
541
+ "loss": 0.1073,
542
+ "step": 178
543
+ },
544
+ {
545
+ "epoch": 0.51,
546
+ "learning_rate": 9.954204348798938e-05,
547
+ "loss": 0.1114,
548
+ "step": 180
549
+ },
550
+ {
551
+ "epoch": 0.52,
552
+ "learning_rate": 9.771040952581998e-05,
553
+ "loss": 0.1158,
554
+ "step": 182
555
+ },
556
+ {
557
+ "epoch": 0.52,
558
+ "learning_rate": 9.587954383946517e-05,
559
+ "loss": 0.1103,
560
+ "step": 184
561
+ },
562
+ {
563
+ "epoch": 0.53,
564
+ "learning_rate": 9.405006077888954e-05,
565
+ "loss": 0.1083,
566
+ "step": 186
567
+ },
568
+ {
569
+ "epoch": 0.54,
570
+ "learning_rate": 9.22225742301153e-05,
571
+ "loss": 0.1137,
572
+ "step": 188
573
+ },
574
+ {
575
+ "epoch": 0.54,
576
+ "learning_rate": 9.039769740923183e-05,
577
+ "loss": 0.1123,
578
+ "step": 190
579
+ },
580
+ {
581
+ "epoch": 0.55,
582
+ "learning_rate": 8.857604265663017e-05,
583
+ "loss": 0.1191,
584
+ "step": 192
585
+ },
586
+ {
587
+ "epoch": 0.55,
588
+ "learning_rate": 8.675822123153103e-05,
589
+ "loss": 0.1119,
590
+ "step": 194
591
+ },
592
+ {
593
+ "epoch": 0.56,
594
+ "learning_rate": 8.494484310687581e-05,
595
+ "loss": 0.1196,
596
+ "step": 196
597
+ },
598
+ {
599
+ "epoch": 0.56,
600
+ "learning_rate": 8.313651676464923e-05,
601
+ "loss": 0.1135,
602
+ "step": 198
603
+ },
604
+ {
605
+ "epoch": 0.57,
606
+ "learning_rate": 8.133384899170225e-05,
607
+ "loss": 0.1092,
608
+ "step": 200
609
+ },
610
+ {
611
+ "epoch": 0.58,
612
+ "learning_rate": 7.953744467614354e-05,
613
+ "loss": 0.112,
614
+ "step": 202
615
+ },
616
+ {
617
+ "epoch": 0.58,
618
+ "learning_rate": 7.774790660436858e-05,
619
+ "loss": 0.1099,
620
+ "step": 204
621
+ },
622
+ {
623
+ "epoch": 0.59,
624
+ "learning_rate": 7.596583525879344e-05,
625
+ "loss": 0.1064,
626
+ "step": 206
627
+ },
628
+ {
629
+ "epoch": 0.59,
630
+ "learning_rate": 7.419182861636218e-05,
631
+ "loss": 0.1165,
632
+ "step": 208
633
+ },
634
+ {
635
+ "epoch": 0.6,
636
+ "learning_rate": 7.242648194789446e-05,
637
+ "loss": 0.1215,
638
+ "step": 210
639
+ },
640
+ {
641
+ "epoch": 0.6,
642
+ "learning_rate": 7.067038761834164e-05,
643
+ "loss": 0.1025,
644
+ "step": 212
645
+ },
646
+ {
647
+ "epoch": 0.61,
648
+ "learning_rate": 6.89241348880176e-05,
649
+ "loss": 0.1109,
650
+ "step": 214
651
+ },
652
+ {
653
+ "epoch": 0.62,
654
+ "learning_rate": 6.718830971487165e-05,
655
+ "loss": 0.1157,
656
+ "step": 216
657
+ },
658
+ {
659
+ "epoch": 0.62,
660
+ "learning_rate": 6.546349455786926e-05,
661
+ "loss": 0.1108,
662
+ "step": 218
663
+ },
664
+ {
665
+ "epoch": 0.63,
666
+ "learning_rate": 6.3750268181547e-05,
667
+ "loss": 0.1043,
668
+ "step": 220
669
+ },
670
+ {
671
+ "epoch": 0.63,
672
+ "learning_rate": 6.204920546180728e-05,
673
+ "loss": 0.108,
674
+ "step": 222
675
+ },
676
+ {
677
+ "epoch": 0.64,
678
+ "learning_rate": 6.036087719301763e-05,
679
+ "loss": 0.1125,
680
+ "step": 224
681
+ },
682
+ {
683
+ "epoch": 0.64,
684
+ "learning_rate": 5.868584989647994e-05,
685
+ "loss": 0.1096,
686
+ "step": 226
687
+ },
688
+ {
689
+ "epoch": 0.65,
690
+ "learning_rate": 5.702468563033306e-05,
691
+ "loss": 0.1114,
692
+ "step": 228
693
+ },
694
+ {
695
+ "epoch": 0.66,
696
+ "learning_rate": 5.5377941800953416e-05,
697
+ "loss": 0.0981,
698
+ "step": 230
699
+ },
700
+ {
701
+ "epoch": 0.66,
702
+ "learning_rate": 5.37461709759165e-05,
703
+ "loss": 0.1147,
704
+ "step": 232
705
+ },
706
+ {
707
+ "epoch": 0.67,
708
+ "learning_rate": 5.2129920698581606e-05,
709
+ "loss": 0.1085,
710
+ "step": 234
711
+ },
712
+ {
713
+ "epoch": 0.67,
714
+ "learning_rate": 5.0529733304363145e-05,
715
+ "loss": 0.1133,
716
+ "step": 236
717
+ },
718
+ {
719
+ "epoch": 0.68,
720
+ "learning_rate": 4.894614573874877e-05,
721
+ "loss": 0.1036,
722
+ "step": 238
723
+ },
724
+ {
725
+ "epoch": 0.68,
726
+ "learning_rate": 4.7379689377126735e-05,
727
+ "loss": 0.1106,
728
+ "step": 240
729
+ },
730
+ {
731
+ "epoch": 0.69,
732
+ "learning_rate": 4.583088984648172e-05,
733
+ "loss": 0.1185,
734
+ "step": 242
735
+ },
736
+ {
737
+ "epoch": 0.7,
738
+ "learning_rate": 4.430026684902017e-05,
739
+ "loss": 0.1127,
740
+ "step": 244
741
+ },
742
+ {
743
+ "epoch": 0.7,
744
+ "learning_rate": 4.278833398778306e-05,
745
+ "loss": 0.1187,
746
+ "step": 246
747
+ },
748
+ {
749
+ "epoch": 0.71,
750
+ "learning_rate": 4.129559859430573e-05,
751
+ "loss": 0.12,
752
+ "step": 248
753
+ },
754
+ {
755
+ "epoch": 0.71,
756
+ "learning_rate": 3.982256155838199e-05,
757
+ "loss": 0.1159,
758
+ "step": 250
759
+ },
760
+ {
761
+ "epoch": 0.72,
762
+ "learning_rate": 3.836971715998967e-05,
763
+ "loss": 0.1278,
764
+ "step": 252
765
+ },
766
+ {
767
+ "epoch": 0.72,
768
+ "learning_rate": 3.693755290343409e-05,
769
+ "loss": 0.1188,
770
+ "step": 254
771
+ },
772
+ {
773
+ "epoch": 0.73,
774
+ "learning_rate": 3.5526549353765296e-05,
775
+ "loss": 0.1091,
776
+ "step": 256
777
+ },
778
+ {
779
+ "epoch": 0.74,
780
+ "learning_rate": 3.413717997552376e-05,
781
+ "loss": 0.1014,
782
+ "step": 258
783
+ },
784
+ {
785
+ "epoch": 0.74,
786
+ "learning_rate": 3.276991097386831e-05,
787
+ "loss": 0.1147,
788
+ "step": 260
789
+ },
790
+ {
791
+ "epoch": 0.75,
792
+ "learning_rate": 3.142520113814059e-05,
793
+ "loss": 0.1144,
794
+ "step": 262
795
+ },
796
+ {
797
+ "epoch": 0.75,
798
+ "learning_rate": 3.010350168791719e-05,
799
+ "loss": 0.1161,
800
+ "step": 264
801
+ },
802
+ {
803
+ "epoch": 0.76,
804
+ "learning_rate": 2.8805256121602398e-05,
805
+ "loss": 0.1145,
806
+ "step": 266
807
+ },
808
+ {
809
+ "epoch": 0.76,
810
+ "learning_rate": 2.7530900067611577e-05,
811
+ "loss": 0.1151,
812
+ "step": 268
813
+ },
814
+ {
815
+ "epoch": 0.77,
816
+ "learning_rate": 2.62808611381953e-05,
817
+ "loss": 0.1149,
818
+ "step": 270
819
+ },
820
+ {
821
+ "epoch": 0.77,
822
+ "learning_rate": 2.5055558785953303e-05,
823
+ "loss": 0.114,
824
+ "step": 272
825
+ },
826
+ {
827
+ "epoch": 0.78,
828
+ "learning_rate": 2.3855404163086558e-05,
829
+ "loss": 0.1187,
830
+ "step": 274
831
+ },
832
+ {
833
+ "epoch": 0.79,
834
+ "learning_rate": 2.268079998343453e-05,
835
+ "loss": 0.1108,
836
+ "step": 276
837
+ },
838
+ {
839
+ "epoch": 0.79,
840
+ "learning_rate": 2.1532140387343735e-05,
841
+ "loss": 0.1112,
842
+ "step": 278
843
+ },
844
+ {
845
+ "epoch": 0.8,
846
+ "learning_rate": 2.0409810809413486e-05,
847
+ "loss": 0.1132,
848
+ "step": 280
849
+ },
850
+ {
851
+ "epoch": 0.8,
852
+ "learning_rate": 1.9314187849162524e-05,
853
+ "loss": 0.1098,
854
+ "step": 282
855
+ },
856
+ {
857
+ "epoch": 0.81,
858
+ "learning_rate": 1.8245639144660532e-05,
859
+ "loss": 0.1073,
860
+ "step": 284
861
+ },
862
+ {
863
+ "epoch": 0.81,
864
+ "learning_rate": 1.720452324916656e-05,
865
+ "loss": 0.1079,
866
+ "step": 286
867
+ },
868
+ {
869
+ "epoch": 0.82,
870
+ "learning_rate": 1.619118951081594e-05,
871
+ "loss": 0.1068,
872
+ "step": 288
873
+ },
874
+ {
875
+ "epoch": 0.83,
876
+ "learning_rate": 1.5205977955395812e-05,
877
+ "loss": 0.1162,
878
+ "step": 290
879
+ },
880
+ {
881
+ "epoch": 0.83,
882
+ "learning_rate": 1.424921917224905e-05,
883
+ "loss": 0.111,
884
+ "step": 292
885
+ },
886
+ {
887
+ "epoch": 0.84,
888
+ "learning_rate": 1.3321234203344435e-05,
889
+ "loss": 0.1145,
890
+ "step": 294
891
+ },
892
+ {
893
+ "epoch": 0.84,
894
+ "learning_rate": 1.2422334435550509e-05,
895
+ "loss": 0.1137,
896
+ "step": 296
897
+ },
898
+ {
899
+ "epoch": 0.85,
900
+ "learning_rate": 1.1552821496149135e-05,
901
+ "loss": 0.1009,
902
+ "step": 298
903
+ },
904
+ {
905
+ "epoch": 0.85,
906
+ "learning_rate": 1.0712987151624054e-05,
907
+ "loss": 0.1055,
908
+ "step": 300
909
+ },
910
+ {
911
+ "epoch": 0.86,
912
+ "learning_rate": 9.903113209758096e-06,
913
+ "loss": 0.1054,
914
+ "step": 302
915
+ },
916
+ {
917
+ "epoch": 0.87,
918
+ "learning_rate": 9.123471425072206e-06,
919
+ "loss": 0.1042,
920
+ "step": 304
921
+ },
922
+ {
923
+ "epoch": 0.87,
924
+ "learning_rate": 8.374323407637742e-06,
925
+ "loss": 0.1032,
926
+ "step": 306
927
+ },
928
+ {
929
+ "epoch": 0.88,
930
+ "learning_rate": 7.655920535292682e-06,
931
+ "loss": 0.104,
932
+ "step": 308
933
+ },
934
+ {
935
+ "epoch": 0.88,
936
+ "learning_rate": 6.968503869291521e-06,
937
+ "loss": 0.1064,
938
+ "step": 310
939
+ },
940
+ {
941
+ "epoch": 0.89,
942
+ "learning_rate": 6.312304073416719e-06,
943
+ "loss": 0.1081,
944
+ "step": 312
945
+ },
946
+ {
947
+ "epoch": 0.89,
948
+ "learning_rate": 5.687541336579128e-06,
949
+ "loss": 0.1151,
950
+ "step": 314
951
+ },
952
+ {
953
+ "epoch": 0.9,
954
+ "learning_rate": 5.094425298933136e-06,
955
+ "loss": 0.1184,
956
+ "step": 316
957
+ },
958
+ {
959
+ "epoch": 0.91,
960
+ "learning_rate": 4.5331549815317174e-06,
961
+ "loss": 0.1061,
962
+ "step": 318
963
+ },
964
+ {
965
+ "epoch": 0.91,
966
+ "learning_rate": 4.003918719544464e-06,
967
+ "loss": 0.1104,
968
+ "step": 320
969
+ },
970
+ {
971
+ "epoch": 0.92,
972
+ "learning_rate": 3.5068940990615197e-06,
973
+ "loss": 0.1022,
974
+ "step": 322
975
+ },
976
+ {
977
+ "epoch": 0.92,
978
+ "learning_rate": 3.0422478975042246e-06,
979
+ "loss": 0.1042,
980
+ "step": 324
981
+ },
982
+ {
983
+ "epoch": 0.93,
984
+ "learning_rate": 2.6101360276626798e-06,
985
+ "loss": 0.1126,
986
+ "step": 326
987
+ },
988
+ {
989
+ "epoch": 0.93,
990
+ "learning_rate": 2.2107034853789288e-06,
991
+ "loss": 0.1084,
992
+ "step": 328
993
+ },
994
+ {
995
+ "epoch": 0.94,
996
+ "learning_rate": 1.8440843008934561e-06,
997
+ "loss": 0.1047,
998
+ "step": 330
999
+ },
1000
+ {
1001
+ "epoch": 0.95,
1002
+ "learning_rate": 1.5104014938710497e-06,
1003
+ "loss": 0.0979,
1004
+ "step": 332
1005
+ },
1006
+ {
1007
+ "epoch": 0.95,
1008
+ "learning_rate": 1.209767032121345e-06,
1009
+ "loss": 0.1147,
1010
+ "step": 334
1011
+ },
1012
+ {
1013
+ "epoch": 0.96,
1014
+ "learning_rate": 9.422817940278772e-07,
1015
+ "loss": 0.1116,
1016
+ "step": 336
1017
+ },
1018
+ {
1019
+ "epoch": 0.96,
1020
+ "learning_rate": 7.080355346981815e-07,
1021
+ "loss": 0.1036,
1022
+ "step": 338
1023
+ },
1024
+ {
1025
+ "epoch": 0.97,
1026
+ "learning_rate": 5.071068558462732e-07,
1027
+ "loss": 0.115,
1028
+ "step": 340
1029
+ },
1030
+ {
1031
+ "epoch": 0.97,
1032
+ "learning_rate": 3.3956317941779005e-07,
1033
+ "loss": 0.108,
1034
+ "step": 342
1035
+ },
1036
+ {
1037
+ "epoch": 0.98,
1038
+ "learning_rate": 2.054607249663665e-07,
1039
+ "loss": 0.1063,
1040
+ "step": 344
1041
+ },
1042
+ {
1043
+ "epoch": 0.99,
1044
+ "learning_rate": 1.0484449078912439e-07,
1045
+ "loss": 0.1028,
1046
+ "step": 346
1047
+ },
1048
+ {
1049
+ "epoch": 0.99,
1050
+ "learning_rate": 3.774823882738421e-08,
1051
+ "loss": 0.1009,
1052
+ "step": 348
1053
+ },
1054
+ {
1055
+ "epoch": 1.0,
1056
+ "learning_rate": 4.194483337860433e-09,
1057
+ "loss": 0.1114,
1058
+ "step": 350
1059
+ },
1060
+ {
1061
+ "epoch": 1.0,
1062
+ "step": 351,
1063
+ "total_flos": 2.4611000956852634e+17,
1064
+ "train_loss": 0.11711755568994756,
1065
+ "train_runtime": 56238.1049,
1066
+ "train_samples_per_second": 3.195,
1067
+ "train_steps_per_second": 0.006
1068
+ }
1069
+ ],
1070
+ "max_steps": 351,
1071
+ "num_train_epochs": 1,
1072
+ "total_flos": 2.4611000956852634e+17,
1073
+ "trial_name": null,
1074
+ "trial_params": null
1075
+ }