Davlan commited on
Commit
b2c511e
1 Parent(s): 18352b3

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1153 -0
trainer_state.json ADDED
@@ -0,0 +1,1153 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 3.0,
5
+ "global_step": 94323,
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": 4.973495329877125e-05,
13
+ "loss": 1.7869,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.03,
18
+ "learning_rate": 4.9469906597542485e-05,
19
+ "loss": 1.7591,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.05,
24
+ "learning_rate": 4.920485989631373e-05,
25
+ "loss": 1.7192,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.06,
30
+ "learning_rate": 4.893981319508498e-05,
31
+ "loss": 1.6754,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.08,
36
+ "learning_rate": 4.867476649385622e-05,
37
+ "loss": 1.6344,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.1,
42
+ "learning_rate": 4.840971979262746e-05,
43
+ "loss": 1.6237,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.11,
48
+ "learning_rate": 4.8144673091398703e-05,
49
+ "loss": 1.5965,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.13,
54
+ "learning_rate": 4.787962639016995e-05,
55
+ "loss": 1.5753,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.14,
60
+ "learning_rate": 4.761457968894119e-05,
61
+ "loss": 1.5598,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.16,
66
+ "learning_rate": 4.734953298771243e-05,
67
+ "loss": 1.5247,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.17,
72
+ "learning_rate": 4.708448628648368e-05,
73
+ "loss": 1.5149,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.19,
78
+ "learning_rate": 4.681943958525493e-05,
79
+ "loss": 1.5089,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.21,
84
+ "learning_rate": 4.655439288402616e-05,
85
+ "loss": 1.4888,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.22,
90
+ "learning_rate": 4.628934618279741e-05,
91
+ "loss": 1.4738,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.24,
96
+ "learning_rate": 4.602429948156866e-05,
97
+ "loss": 1.4535,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.25,
102
+ "learning_rate": 4.57592527803399e-05,
103
+ "loss": 1.474,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.27,
108
+ "learning_rate": 4.549420607911114e-05,
109
+ "loss": 1.4491,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.29,
114
+ "learning_rate": 4.522915937788238e-05,
115
+ "loss": 1.4336,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.3,
120
+ "learning_rate": 4.496411267665363e-05,
121
+ "loss": 1.4323,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.32,
126
+ "learning_rate": 4.469906597542487e-05,
127
+ "loss": 1.4189,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.33,
132
+ "learning_rate": 4.443401927419611e-05,
133
+ "loss": 1.4101,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.35,
138
+ "learning_rate": 4.416897257296736e-05,
139
+ "loss": 1.3992,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.37,
144
+ "learning_rate": 4.390392587173861e-05,
145
+ "loss": 1.3975,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.38,
150
+ "learning_rate": 4.363887917050984e-05,
151
+ "loss": 1.3845,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.4,
156
+ "learning_rate": 4.337383246928109e-05,
157
+ "loss": 1.3795,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.41,
162
+ "learning_rate": 4.310878576805233e-05,
163
+ "loss": 1.3867,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.43,
168
+ "learning_rate": 4.284373906682357e-05,
169
+ "loss": 1.3647,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.45,
174
+ "learning_rate": 4.257869236559482e-05,
175
+ "loss": 1.3638,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.46,
180
+ "learning_rate": 4.231364566436606e-05,
181
+ "loss": 1.3507,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.48,
186
+ "learning_rate": 4.204859896313731e-05,
187
+ "loss": 1.3462,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.49,
192
+ "learning_rate": 4.178355226190855e-05,
193
+ "loss": 1.3511,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.51,
198
+ "learning_rate": 4.151850556067979e-05,
199
+ "loss": 1.3384,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.52,
204
+ "learning_rate": 4.125345885945104e-05,
205
+ "loss": 1.319,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.54,
210
+ "learning_rate": 4.098841215822228e-05,
211
+ "loss": 1.3331,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.56,
216
+ "learning_rate": 4.072336545699352e-05,
217
+ "loss": 1.3119,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.57,
222
+ "learning_rate": 4.045831875576477e-05,
223
+ "loss": 1.32,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.59,
228
+ "learning_rate": 4.019327205453601e-05,
229
+ "loss": 1.3224,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.6,
234
+ "learning_rate": 3.992822535330725e-05,
235
+ "loss": 1.2868,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.62,
240
+ "learning_rate": 3.96631786520785e-05,
241
+ "loss": 1.2955,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.64,
246
+ "learning_rate": 3.939813195084974e-05,
247
+ "loss": 1.2932,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.65,
252
+ "learning_rate": 3.913308524962099e-05,
253
+ "loss": 1.277,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.67,
258
+ "learning_rate": 3.886803854839223e-05,
259
+ "loss": 1.2759,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.68,
264
+ "learning_rate": 3.860299184716347e-05,
265
+ "loss": 1.2793,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.7,
270
+ "learning_rate": 3.833794514593472e-05,
271
+ "loss": 1.2614,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.72,
276
+ "learning_rate": 3.807289844470596e-05,
277
+ "loss": 1.2768,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.73,
282
+ "learning_rate": 3.78078517434772e-05,
283
+ "loss": 1.2615,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.75,
288
+ "learning_rate": 3.754280504224845e-05,
289
+ "loss": 1.2525,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.76,
294
+ "learning_rate": 3.727775834101969e-05,
295
+ "loss": 1.2593,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.78,
300
+ "learning_rate": 3.701271163979093e-05,
301
+ "loss": 1.2435,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.8,
306
+ "learning_rate": 3.674766493856218e-05,
307
+ "loss": 1.241,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.81,
312
+ "learning_rate": 3.648261823733342e-05,
313
+ "loss": 1.2449,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.83,
318
+ "learning_rate": 3.621757153610466e-05,
319
+ "loss": 1.2229,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.84,
324
+ "learning_rate": 3.595252483487591e-05,
325
+ "loss": 1.2296,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.86,
330
+ "learning_rate": 3.568747813364715e-05,
331
+ "loss": 1.2257,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.87,
336
+ "learning_rate": 3.5422431432418397e-05,
337
+ "loss": 1.2292,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.89,
342
+ "learning_rate": 3.515738473118964e-05,
343
+ "loss": 1.2351,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.91,
348
+ "learning_rate": 3.489233802996088e-05,
349
+ "loss": 1.2179,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.92,
354
+ "learning_rate": 3.4627291328732127e-05,
355
+ "loss": 1.2179,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.94,
360
+ "learning_rate": 3.436224462750337e-05,
361
+ "loss": 1.2057,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.95,
366
+ "learning_rate": 3.409719792627461e-05,
367
+ "loss": 1.2081,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.97,
372
+ "learning_rate": 3.3832151225045856e-05,
373
+ "loss": 1.2069,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.99,
378
+ "learning_rate": 3.35671045238171e-05,
379
+ "loss": 1.1903,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 1.0,
384
+ "learning_rate": 3.330205782258834e-05,
385
+ "loss": 1.1958,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 1.02,
390
+ "learning_rate": 3.3037011121359586e-05,
391
+ "loss": 1.1746,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 1.03,
396
+ "learning_rate": 3.277196442013083e-05,
397
+ "loss": 1.1648,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 1.05,
402
+ "learning_rate": 3.2506917718902075e-05,
403
+ "loss": 1.1615,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 1.07,
408
+ "learning_rate": 3.2241871017673316e-05,
409
+ "loss": 1.1732,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 1.08,
414
+ "learning_rate": 3.197682431644456e-05,
415
+ "loss": 1.1664,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 1.1,
420
+ "learning_rate": 3.1711777615215805e-05,
421
+ "loss": 1.1639,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 1.11,
426
+ "learning_rate": 3.1446730913987046e-05,
427
+ "loss": 1.1507,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 1.13,
432
+ "learning_rate": 3.118168421275829e-05,
433
+ "loss": 1.1577,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 1.15,
438
+ "learning_rate": 3.0916637511529535e-05,
439
+ "loss": 1.1563,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 1.16,
444
+ "learning_rate": 3.0651590810300776e-05,
445
+ "loss": 1.1528,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 1.18,
450
+ "learning_rate": 3.0386544109072017e-05,
451
+ "loss": 1.1461,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 1.19,
456
+ "learning_rate": 3.0121497407843262e-05,
457
+ "loss": 1.1574,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 1.21,
462
+ "learning_rate": 2.985645070661451e-05,
463
+ "loss": 1.1503,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 1.22,
468
+ "learning_rate": 2.9591404005385747e-05,
469
+ "loss": 1.1522,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 1.24,
474
+ "learning_rate": 2.9326357304156992e-05,
475
+ "loss": 1.1451,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 1.26,
480
+ "learning_rate": 2.9061310602928236e-05,
481
+ "loss": 1.1335,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 1.27,
486
+ "learning_rate": 2.8796263901699484e-05,
487
+ "loss": 1.1383,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 1.29,
492
+ "learning_rate": 2.8531217200470722e-05,
493
+ "loss": 1.1363,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 1.3,
498
+ "learning_rate": 2.8266170499241966e-05,
499
+ "loss": 1.1277,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 1.32,
504
+ "learning_rate": 2.800112379801321e-05,
505
+ "loss": 1.1332,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 1.34,
510
+ "learning_rate": 2.773607709678446e-05,
511
+ "loss": 1.1249,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 1.35,
516
+ "learning_rate": 2.7471030395555696e-05,
517
+ "loss": 1.1283,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 1.37,
522
+ "learning_rate": 2.720598369432694e-05,
523
+ "loss": 1.1177,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 1.38,
528
+ "learning_rate": 2.694093699309819e-05,
529
+ "loss": 1.1023,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 1.4,
534
+ "learning_rate": 2.6675890291869426e-05,
535
+ "loss": 1.1245,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 1.42,
540
+ "learning_rate": 2.641084359064067e-05,
541
+ "loss": 1.1194,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 1.43,
546
+ "learning_rate": 2.6145796889411915e-05,
547
+ "loss": 1.1132,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 1.45,
552
+ "learning_rate": 2.5880750188183163e-05,
553
+ "loss": 1.1073,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 1.46,
558
+ "learning_rate": 2.56157034869544e-05,
559
+ "loss": 1.0974,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 1.48,
564
+ "learning_rate": 2.5350656785725645e-05,
565
+ "loss": 1.1003,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 1.49,
570
+ "learning_rate": 2.508561008449689e-05,
571
+ "loss": 1.0966,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 1.51,
576
+ "learning_rate": 2.4820563383268134e-05,
577
+ "loss": 1.0928,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 1.53,
582
+ "learning_rate": 2.4555516682039375e-05,
583
+ "loss": 1.0868,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 1.54,
588
+ "learning_rate": 2.429046998081062e-05,
589
+ "loss": 1.0902,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 1.56,
594
+ "learning_rate": 2.4025423279581864e-05,
595
+ "loss": 1.0978,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 1.57,
600
+ "learning_rate": 2.376037657835311e-05,
601
+ "loss": 1.0828,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 1.59,
606
+ "learning_rate": 2.349532987712435e-05,
607
+ "loss": 1.1016,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 1.61,
612
+ "learning_rate": 2.3230283175895594e-05,
613
+ "loss": 1.0892,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 1.62,
618
+ "learning_rate": 2.296523647466684e-05,
619
+ "loss": 1.0832,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 1.64,
624
+ "learning_rate": 2.270018977343808e-05,
625
+ "loss": 1.0668,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 1.65,
630
+ "learning_rate": 2.2435143072209324e-05,
631
+ "loss": 1.0756,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 1.67,
636
+ "learning_rate": 2.217009637098057e-05,
637
+ "loss": 1.0714,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 1.69,
642
+ "learning_rate": 2.1905049669751813e-05,
643
+ "loss": 1.0592,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 1.7,
648
+ "learning_rate": 2.1640002968523054e-05,
649
+ "loss": 1.07,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 1.72,
654
+ "learning_rate": 2.13749562672943e-05,
655
+ "loss": 1.0548,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 1.73,
660
+ "learning_rate": 2.1109909566065543e-05,
661
+ "loss": 1.0589,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 1.75,
666
+ "learning_rate": 2.0844862864836784e-05,
667
+ "loss": 1.0662,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 1.77,
672
+ "learning_rate": 2.057981616360803e-05,
673
+ "loss": 1.0678,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 1.78,
678
+ "learning_rate": 2.0314769462379273e-05,
679
+ "loss": 1.0545,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 1.8,
684
+ "learning_rate": 2.0049722761150517e-05,
685
+ "loss": 1.0486,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 1.81,
690
+ "learning_rate": 1.978467605992176e-05,
691
+ "loss": 1.051,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 1.83,
696
+ "learning_rate": 1.9519629358693003e-05,
697
+ "loss": 1.0389,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 1.84,
702
+ "learning_rate": 1.9254582657464247e-05,
703
+ "loss": 1.0445,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 1.86,
708
+ "learning_rate": 1.8989535956235492e-05,
709
+ "loss": 1.0511,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 1.88,
714
+ "learning_rate": 1.8724489255006733e-05,
715
+ "loss": 1.0384,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 1.89,
720
+ "learning_rate": 1.8459442553777974e-05,
721
+ "loss": 1.0386,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 1.91,
726
+ "learning_rate": 1.819439585254922e-05,
727
+ "loss": 1.0321,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 1.92,
732
+ "learning_rate": 1.7929349151320463e-05,
733
+ "loss": 1.0424,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 1.94,
738
+ "learning_rate": 1.7664302450091707e-05,
739
+ "loss": 1.0421,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 1.96,
744
+ "learning_rate": 1.7399255748862948e-05,
745
+ "loss": 1.0334,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 1.97,
750
+ "learning_rate": 1.7134209047634196e-05,
751
+ "loss": 1.0385,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 1.99,
756
+ "learning_rate": 1.6869162346405437e-05,
757
+ "loss": 1.0265,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 2.0,
762
+ "learning_rate": 1.660411564517668e-05,
763
+ "loss": 1.0245,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 2.02,
768
+ "learning_rate": 1.6339068943947926e-05,
769
+ "loss": 1.0146,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 2.04,
774
+ "learning_rate": 1.6074022242719167e-05,
775
+ "loss": 1.0161,
776
+ "step": 64000
777
+ },
778
+ {
779
+ "epoch": 2.05,
780
+ "learning_rate": 1.580897554149041e-05,
781
+ "loss": 1.0196,
782
+ "step": 64500
783
+ },
784
+ {
785
+ "epoch": 2.07,
786
+ "learning_rate": 1.5543928840261653e-05,
787
+ "loss": 1.0118,
788
+ "step": 65000
789
+ },
790
+ {
791
+ "epoch": 2.08,
792
+ "learning_rate": 1.52788821390329e-05,
793
+ "loss": 0.9996,
794
+ "step": 65500
795
+ },
796
+ {
797
+ "epoch": 2.1,
798
+ "learning_rate": 1.5013835437804142e-05,
799
+ "loss": 1.0026,
800
+ "step": 66000
801
+ },
802
+ {
803
+ "epoch": 2.12,
804
+ "learning_rate": 1.4748788736575386e-05,
805
+ "loss": 1.0156,
806
+ "step": 66500
807
+ },
808
+ {
809
+ "epoch": 2.13,
810
+ "learning_rate": 1.4483742035346629e-05,
811
+ "loss": 1.0017,
812
+ "step": 67000
813
+ },
814
+ {
815
+ "epoch": 2.15,
816
+ "learning_rate": 1.4218695334117873e-05,
817
+ "loss": 1.0129,
818
+ "step": 67500
819
+ },
820
+ {
821
+ "epoch": 2.16,
822
+ "learning_rate": 1.3953648632889116e-05,
823
+ "loss": 1.0093,
824
+ "step": 68000
825
+ },
826
+ {
827
+ "epoch": 2.18,
828
+ "learning_rate": 1.3688601931660359e-05,
829
+ "loss": 0.9939,
830
+ "step": 68500
831
+ },
832
+ {
833
+ "epoch": 2.19,
834
+ "learning_rate": 1.3423555230431603e-05,
835
+ "loss": 0.9891,
836
+ "step": 69000
837
+ },
838
+ {
839
+ "epoch": 2.21,
840
+ "learning_rate": 1.3158508529202846e-05,
841
+ "loss": 0.9978,
842
+ "step": 69500
843
+ },
844
+ {
845
+ "epoch": 2.23,
846
+ "learning_rate": 1.289346182797409e-05,
847
+ "loss": 1.0043,
848
+ "step": 70000
849
+ },
850
+ {
851
+ "epoch": 2.24,
852
+ "learning_rate": 1.2628415126745333e-05,
853
+ "loss": 0.9913,
854
+ "step": 70500
855
+ },
856
+ {
857
+ "epoch": 2.26,
858
+ "learning_rate": 1.2363368425516576e-05,
859
+ "loss": 0.9986,
860
+ "step": 71000
861
+ },
862
+ {
863
+ "epoch": 2.27,
864
+ "learning_rate": 1.209832172428782e-05,
865
+ "loss": 0.9892,
866
+ "step": 71500
867
+ },
868
+ {
869
+ "epoch": 2.29,
870
+ "learning_rate": 1.1833275023059063e-05,
871
+ "loss": 0.9995,
872
+ "step": 72000
873
+ },
874
+ {
875
+ "epoch": 2.31,
876
+ "learning_rate": 1.1568228321830308e-05,
877
+ "loss": 0.9944,
878
+ "step": 72500
879
+ },
880
+ {
881
+ "epoch": 2.32,
882
+ "learning_rate": 1.130318162060155e-05,
883
+ "loss": 0.998,
884
+ "step": 73000
885
+ },
886
+ {
887
+ "epoch": 2.34,
888
+ "learning_rate": 1.1038134919372795e-05,
889
+ "loss": 0.9773,
890
+ "step": 73500
891
+ },
892
+ {
893
+ "epoch": 2.35,
894
+ "learning_rate": 1.0773088218144038e-05,
895
+ "loss": 0.9928,
896
+ "step": 74000
897
+ },
898
+ {
899
+ "epoch": 2.37,
900
+ "learning_rate": 1.050804151691528e-05,
901
+ "loss": 0.9787,
902
+ "step": 74500
903
+ },
904
+ {
905
+ "epoch": 2.39,
906
+ "learning_rate": 1.0242994815686525e-05,
907
+ "loss": 0.978,
908
+ "step": 75000
909
+ },
910
+ {
911
+ "epoch": 2.4,
912
+ "learning_rate": 9.977948114457768e-06,
913
+ "loss": 0.9757,
914
+ "step": 75500
915
+ },
916
+ {
917
+ "epoch": 2.42,
918
+ "learning_rate": 9.712901413229012e-06,
919
+ "loss": 0.9773,
920
+ "step": 76000
921
+ },
922
+ {
923
+ "epoch": 2.43,
924
+ "learning_rate": 9.447854712000255e-06,
925
+ "loss": 0.9747,
926
+ "step": 76500
927
+ },
928
+ {
929
+ "epoch": 2.45,
930
+ "learning_rate": 9.1828080107715e-06,
931
+ "loss": 0.9703,
932
+ "step": 77000
933
+ },
934
+ {
935
+ "epoch": 2.46,
936
+ "learning_rate": 8.917761309542742e-06,
937
+ "loss": 0.9631,
938
+ "step": 77500
939
+ },
940
+ {
941
+ "epoch": 2.48,
942
+ "learning_rate": 8.652714608313986e-06,
943
+ "loss": 0.9754,
944
+ "step": 78000
945
+ },
946
+ {
947
+ "epoch": 2.5,
948
+ "learning_rate": 8.387667907085228e-06,
949
+ "loss": 0.9834,
950
+ "step": 78500
951
+ },
952
+ {
953
+ "epoch": 2.51,
954
+ "learning_rate": 8.122621205856472e-06,
955
+ "loss": 0.9797,
956
+ "step": 79000
957
+ },
958
+ {
959
+ "epoch": 2.53,
960
+ "learning_rate": 7.857574504627715e-06,
961
+ "loss": 0.9761,
962
+ "step": 79500
963
+ },
964
+ {
965
+ "epoch": 2.54,
966
+ "learning_rate": 7.592527803398959e-06,
967
+ "loss": 0.9689,
968
+ "step": 80000
969
+ },
970
+ {
971
+ "epoch": 2.56,
972
+ "learning_rate": 7.327481102170203e-06,
973
+ "loss": 0.9683,
974
+ "step": 80500
975
+ },
976
+ {
977
+ "epoch": 2.58,
978
+ "learning_rate": 7.062434400941446e-06,
979
+ "loss": 0.964,
980
+ "step": 81000
981
+ },
982
+ {
983
+ "epoch": 2.59,
984
+ "learning_rate": 6.79738769971269e-06,
985
+ "loss": 0.968,
986
+ "step": 81500
987
+ },
988
+ {
989
+ "epoch": 2.61,
990
+ "learning_rate": 6.532340998483934e-06,
991
+ "loss": 0.9638,
992
+ "step": 82000
993
+ },
994
+ {
995
+ "epoch": 2.62,
996
+ "learning_rate": 6.267294297255176e-06,
997
+ "loss": 0.9651,
998
+ "step": 82500
999
+ },
1000
+ {
1001
+ "epoch": 2.64,
1002
+ "learning_rate": 6.00224759602642e-06,
1003
+ "loss": 0.9616,
1004
+ "step": 83000
1005
+ },
1006
+ {
1007
+ "epoch": 2.66,
1008
+ "learning_rate": 5.737200894797664e-06,
1009
+ "loss": 0.9586,
1010
+ "step": 83500
1011
+ },
1012
+ {
1013
+ "epoch": 2.67,
1014
+ "learning_rate": 5.472154193568907e-06,
1015
+ "loss": 0.965,
1016
+ "step": 84000
1017
+ },
1018
+ {
1019
+ "epoch": 2.69,
1020
+ "learning_rate": 5.207107492340151e-06,
1021
+ "loss": 0.9616,
1022
+ "step": 84500
1023
+ },
1024
+ {
1025
+ "epoch": 2.7,
1026
+ "learning_rate": 4.942060791111394e-06,
1027
+ "loss": 0.961,
1028
+ "step": 85000
1029
+ },
1030
+ {
1031
+ "epoch": 2.72,
1032
+ "learning_rate": 4.677014089882638e-06,
1033
+ "loss": 0.9471,
1034
+ "step": 85500
1035
+ },
1036
+ {
1037
+ "epoch": 2.74,
1038
+ "learning_rate": 4.411967388653881e-06,
1039
+ "loss": 0.9676,
1040
+ "step": 86000
1041
+ },
1042
+ {
1043
+ "epoch": 2.75,
1044
+ "learning_rate": 4.146920687425124e-06,
1045
+ "loss": 0.9451,
1046
+ "step": 86500
1047
+ },
1048
+ {
1049
+ "epoch": 2.77,
1050
+ "learning_rate": 3.881873986196368e-06,
1051
+ "loss": 0.9501,
1052
+ "step": 87000
1053
+ },
1054
+ {
1055
+ "epoch": 2.78,
1056
+ "learning_rate": 3.6168272849676116e-06,
1057
+ "loss": 0.949,
1058
+ "step": 87500
1059
+ },
1060
+ {
1061
+ "epoch": 2.8,
1062
+ "learning_rate": 3.3517805837388548e-06,
1063
+ "loss": 0.9511,
1064
+ "step": 88000
1065
+ },
1066
+ {
1067
+ "epoch": 2.81,
1068
+ "learning_rate": 3.0867338825100984e-06,
1069
+ "loss": 0.9452,
1070
+ "step": 88500
1071
+ },
1072
+ {
1073
+ "epoch": 2.83,
1074
+ "learning_rate": 2.821687181281342e-06,
1075
+ "loss": 0.9458,
1076
+ "step": 89000
1077
+ },
1078
+ {
1079
+ "epoch": 2.85,
1080
+ "learning_rate": 2.556640480052585e-06,
1081
+ "loss": 0.9519,
1082
+ "step": 89500
1083
+ },
1084
+ {
1085
+ "epoch": 2.86,
1086
+ "learning_rate": 2.2915937788238288e-06,
1087
+ "loss": 0.9391,
1088
+ "step": 90000
1089
+ },
1090
+ {
1091
+ "epoch": 2.88,
1092
+ "learning_rate": 2.0265470775950724e-06,
1093
+ "loss": 0.9483,
1094
+ "step": 90500
1095
+ },
1096
+ {
1097
+ "epoch": 2.89,
1098
+ "learning_rate": 1.7615003763663158e-06,
1099
+ "loss": 0.9483,
1100
+ "step": 91000
1101
+ },
1102
+ {
1103
+ "epoch": 2.91,
1104
+ "learning_rate": 1.4964536751375594e-06,
1105
+ "loss": 0.9454,
1106
+ "step": 91500
1107
+ },
1108
+ {
1109
+ "epoch": 2.93,
1110
+ "learning_rate": 1.2314069739088028e-06,
1111
+ "loss": 0.9371,
1112
+ "step": 92000
1113
+ },
1114
+ {
1115
+ "epoch": 2.94,
1116
+ "learning_rate": 9.663602726800462e-07,
1117
+ "loss": 0.9447,
1118
+ "step": 92500
1119
+ },
1120
+ {
1121
+ "epoch": 2.96,
1122
+ "learning_rate": 7.013135714512897e-07,
1123
+ "loss": 0.9461,
1124
+ "step": 93000
1125
+ },
1126
+ {
1127
+ "epoch": 2.97,
1128
+ "learning_rate": 4.362668702225332e-07,
1129
+ "loss": 0.9397,
1130
+ "step": 93500
1131
+ },
1132
+ {
1133
+ "epoch": 2.99,
1134
+ "learning_rate": 1.7122016899377671e-07,
1135
+ "loss": 0.9397,
1136
+ "step": 94000
1137
+ },
1138
+ {
1139
+ "epoch": 3.0,
1140
+ "step": 94323,
1141
+ "total_flos": 8.797388561511875e+17,
1142
+ "train_loss": 1.148913297653906,
1143
+ "train_runtime": 104744.2409,
1144
+ "train_samples_per_second": 9.005,
1145
+ "train_steps_per_second": 0.901
1146
+ }
1147
+ ],
1148
+ "max_steps": 94323,
1149
+ "num_train_epochs": 3,
1150
+ "total_flos": 8.797388561511875e+17,
1151
+ "trial_name": null,
1152
+ "trial_params": null
1153
+ }