lothritz commited on
Commit
843e91d
1 Parent(s): 00bf4d0

update trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1138 -0
trainer_state.json ADDED
@@ -0,0 +1,1138 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 9.976312925460403,
5
+ "global_step": 1870000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.05,
12
+ "learning_rate": 4.973325366509464e-05,
13
+ "loss": 5.9346,
14
+ "step": 10000
15
+ },
16
+ {
17
+ "epoch": 0.11,
18
+ "learning_rate": 4.946650733018929e-05,
19
+ "loss": 4.8171,
20
+ "step": 20000
21
+ },
22
+ {
23
+ "epoch": 0.16,
24
+ "learning_rate": 4.919976099528393e-05,
25
+ "loss": 4.378,
26
+ "step": 30000
27
+ },
28
+ {
29
+ "epoch": 0.21,
30
+ "learning_rate": 4.893301466037857e-05,
31
+ "loss": 4.0841,
32
+ "step": 40000
33
+ },
34
+ {
35
+ "epoch": 0.27,
36
+ "learning_rate": 4.866626832547321e-05,
37
+ "loss": 3.8807,
38
+ "step": 50000
39
+ },
40
+ {
41
+ "epoch": 0.32,
42
+ "learning_rate": 4.839952199056785e-05,
43
+ "loss": 3.7164,
44
+ "step": 60000
45
+ },
46
+ {
47
+ "epoch": 0.37,
48
+ "learning_rate": 4.8132775655662495e-05,
49
+ "loss": 3.584,
50
+ "step": 70000
51
+ },
52
+ {
53
+ "epoch": 0.43,
54
+ "learning_rate": 4.786602932075713e-05,
55
+ "loss": 3.4784,
56
+ "step": 80000
57
+ },
58
+ {
59
+ "epoch": 0.48,
60
+ "learning_rate": 4.759928298585178e-05,
61
+ "loss": 3.3842,
62
+ "step": 90000
63
+ },
64
+ {
65
+ "epoch": 0.53,
66
+ "learning_rate": 4.733253665094642e-05,
67
+ "loss": 3.3004,
68
+ "step": 100000
69
+ },
70
+ {
71
+ "epoch": 0.59,
72
+ "learning_rate": 4.7065790316041056e-05,
73
+ "loss": 3.2337,
74
+ "step": 110000
75
+ },
76
+ {
77
+ "epoch": 0.64,
78
+ "learning_rate": 4.6799043981135706e-05,
79
+ "loss": 3.1757,
80
+ "step": 120000
81
+ },
82
+ {
83
+ "epoch": 0.69,
84
+ "learning_rate": 4.653229764623034e-05,
85
+ "loss": 3.1198,
86
+ "step": 130000
87
+ },
88
+ {
89
+ "epoch": 0.75,
90
+ "learning_rate": 4.626555131132499e-05,
91
+ "loss": 3.0722,
92
+ "step": 140000
93
+ },
94
+ {
95
+ "epoch": 0.8,
96
+ "learning_rate": 4.5998804976419624e-05,
97
+ "loss": 3.0281,
98
+ "step": 150000
99
+ },
100
+ {
101
+ "epoch": 0.85,
102
+ "learning_rate": 4.573205864151427e-05,
103
+ "loss": 2.9954,
104
+ "step": 160000
105
+ },
106
+ {
107
+ "epoch": 0.91,
108
+ "learning_rate": 4.546531230660891e-05,
109
+ "loss": 2.9605,
110
+ "step": 170000
111
+ },
112
+ {
113
+ "epoch": 0.96,
114
+ "learning_rate": 4.519856597170355e-05,
115
+ "loss": 2.9286,
116
+ "step": 180000
117
+ },
118
+ {
119
+ "epoch": 1.01,
120
+ "learning_rate": 4.493181963679819e-05,
121
+ "loss": 2.8946,
122
+ "step": 190000
123
+ },
124
+ {
125
+ "epoch": 1.07,
126
+ "learning_rate": 4.4665073301892835e-05,
127
+ "loss": 2.8688,
128
+ "step": 200000
129
+ },
130
+ {
131
+ "epoch": 1.12,
132
+ "learning_rate": 4.439832696698747e-05,
133
+ "loss": 2.8424,
134
+ "step": 210000
135
+ },
136
+ {
137
+ "epoch": 1.17,
138
+ "learning_rate": 4.4131580632082116e-05,
139
+ "loss": 2.8178,
140
+ "step": 220000
141
+ },
142
+ {
143
+ "epoch": 1.23,
144
+ "learning_rate": 4.386483429717676e-05,
145
+ "loss": 2.7994,
146
+ "step": 230000
147
+ },
148
+ {
149
+ "epoch": 1.28,
150
+ "learning_rate": 4.35980879622714e-05,
151
+ "loss": 2.784,
152
+ "step": 240000
153
+ },
154
+ {
155
+ "epoch": 1.33,
156
+ "learning_rate": 4.333134162736604e-05,
157
+ "loss": 2.7612,
158
+ "step": 250000
159
+ },
160
+ {
161
+ "epoch": 1.39,
162
+ "learning_rate": 4.3064595292460684e-05,
163
+ "loss": 2.7435,
164
+ "step": 260000
165
+ },
166
+ {
167
+ "epoch": 1.44,
168
+ "learning_rate": 4.279784895755533e-05,
169
+ "loss": 2.7275,
170
+ "step": 270000
171
+ },
172
+ {
173
+ "epoch": 1.49,
174
+ "learning_rate": 4.2531102622649964e-05,
175
+ "loss": 2.7117,
176
+ "step": 280000
177
+ },
178
+ {
179
+ "epoch": 1.55,
180
+ "learning_rate": 4.226435628774461e-05,
181
+ "loss": 2.6908,
182
+ "step": 290000
183
+ },
184
+ {
185
+ "epoch": 1.6,
186
+ "learning_rate": 4.199760995283925e-05,
187
+ "loss": 2.6787,
188
+ "step": 300000
189
+ },
190
+ {
191
+ "epoch": 1.65,
192
+ "learning_rate": 4.173086361793389e-05,
193
+ "loss": 2.6626,
194
+ "step": 310000
195
+ },
196
+ {
197
+ "epoch": 1.71,
198
+ "learning_rate": 4.146411728302853e-05,
199
+ "loss": 2.6431,
200
+ "step": 320000
201
+ },
202
+ {
203
+ "epoch": 1.76,
204
+ "learning_rate": 4.1197370948123176e-05,
205
+ "loss": 2.6329,
206
+ "step": 330000
207
+ },
208
+ {
209
+ "epoch": 1.81,
210
+ "learning_rate": 4.093062461321781e-05,
211
+ "loss": 2.6195,
212
+ "step": 340000
213
+ },
214
+ {
215
+ "epoch": 1.87,
216
+ "learning_rate": 4.066387827831246e-05,
217
+ "loss": 2.6042,
218
+ "step": 350000
219
+ },
220
+ {
221
+ "epoch": 1.92,
222
+ "learning_rate": 4.03971319434071e-05,
223
+ "loss": 2.5967,
224
+ "step": 360000
225
+ },
226
+ {
227
+ "epoch": 1.97,
228
+ "learning_rate": 4.0130385608501744e-05,
229
+ "loss": 2.5843,
230
+ "step": 370000
231
+ },
232
+ {
233
+ "epoch": 2.03,
234
+ "learning_rate": 3.986363927359638e-05,
235
+ "loss": 2.5683,
236
+ "step": 380000
237
+ },
238
+ {
239
+ "epoch": 2.08,
240
+ "learning_rate": 3.9596892938691025e-05,
241
+ "loss": 2.5605,
242
+ "step": 390000
243
+ },
244
+ {
245
+ "epoch": 2.13,
246
+ "learning_rate": 3.933014660378567e-05,
247
+ "loss": 2.5515,
248
+ "step": 400000
249
+ },
250
+ {
251
+ "epoch": 2.19,
252
+ "learning_rate": 3.9063400268880305e-05,
253
+ "loss": 2.5406,
254
+ "step": 410000
255
+ },
256
+ {
257
+ "epoch": 2.24,
258
+ "learning_rate": 3.879665393397495e-05,
259
+ "loss": 2.5315,
260
+ "step": 420000
261
+ },
262
+ {
263
+ "epoch": 2.29,
264
+ "learning_rate": 3.852990759906959e-05,
265
+ "loss": 2.5272,
266
+ "step": 430000
267
+ },
268
+ {
269
+ "epoch": 2.35,
270
+ "learning_rate": 3.826316126416423e-05,
271
+ "loss": 2.5199,
272
+ "step": 440000
273
+ },
274
+ {
275
+ "epoch": 2.4,
276
+ "learning_rate": 3.799641492925887e-05,
277
+ "loss": 2.5061,
278
+ "step": 450000
279
+ },
280
+ {
281
+ "epoch": 2.45,
282
+ "learning_rate": 3.772966859435352e-05,
283
+ "loss": 2.5007,
284
+ "step": 460000
285
+ },
286
+ {
287
+ "epoch": 2.51,
288
+ "learning_rate": 3.746292225944816e-05,
289
+ "loss": 2.4908,
290
+ "step": 470000
291
+ },
292
+ {
293
+ "epoch": 2.56,
294
+ "learning_rate": 3.71961759245428e-05,
295
+ "loss": 2.4817,
296
+ "step": 480000
297
+ },
298
+ {
299
+ "epoch": 2.61,
300
+ "learning_rate": 3.6929429589637434e-05,
301
+ "loss": 2.4731,
302
+ "step": 490000
303
+ },
304
+ {
305
+ "epoch": 2.67,
306
+ "learning_rate": 3.6662683254732085e-05,
307
+ "loss": 2.4659,
308
+ "step": 500000
309
+ },
310
+ {
311
+ "epoch": 2.72,
312
+ "learning_rate": 3.639593691982672e-05,
313
+ "loss": 2.4596,
314
+ "step": 510000
315
+ },
316
+ {
317
+ "epoch": 2.77,
318
+ "learning_rate": 3.6129190584921365e-05,
319
+ "loss": 2.4506,
320
+ "step": 520000
321
+ },
322
+ {
323
+ "epoch": 2.83,
324
+ "learning_rate": 3.586244425001601e-05,
325
+ "loss": 2.4425,
326
+ "step": 530000
327
+ },
328
+ {
329
+ "epoch": 2.88,
330
+ "learning_rate": 3.5595697915110646e-05,
331
+ "loss": 2.4378,
332
+ "step": 540000
333
+ },
334
+ {
335
+ "epoch": 2.93,
336
+ "learning_rate": 3.532895158020529e-05,
337
+ "loss": 2.4301,
338
+ "step": 550000
339
+ },
340
+ {
341
+ "epoch": 2.99,
342
+ "learning_rate": 3.506220524529993e-05,
343
+ "loss": 2.4195,
344
+ "step": 560000
345
+ },
346
+ {
347
+ "epoch": 3.04,
348
+ "learning_rate": 3.479545891039458e-05,
349
+ "loss": 2.4108,
350
+ "step": 570000
351
+ },
352
+ {
353
+ "epoch": 3.09,
354
+ "learning_rate": 3.4528712575489214e-05,
355
+ "loss": 2.4003,
356
+ "step": 580000
357
+ },
358
+ {
359
+ "epoch": 3.15,
360
+ "learning_rate": 3.426196624058385e-05,
361
+ "loss": 2.3948,
362
+ "step": 590000
363
+ },
364
+ {
365
+ "epoch": 3.2,
366
+ "learning_rate": 3.39952199056785e-05,
367
+ "loss": 2.3911,
368
+ "step": 600000
369
+ },
370
+ {
371
+ "epoch": 3.25,
372
+ "learning_rate": 3.372847357077314e-05,
373
+ "loss": 2.3805,
374
+ "step": 610000
375
+ },
376
+ {
377
+ "epoch": 3.31,
378
+ "learning_rate": 3.346172723586778e-05,
379
+ "loss": 2.3719,
380
+ "step": 620000
381
+ },
382
+ {
383
+ "epoch": 3.36,
384
+ "learning_rate": 3.3194980900962426e-05,
385
+ "loss": 2.3685,
386
+ "step": 630000
387
+ },
388
+ {
389
+ "epoch": 3.41,
390
+ "learning_rate": 3.292823456605706e-05,
391
+ "loss": 2.3668,
392
+ "step": 640000
393
+ },
394
+ {
395
+ "epoch": 3.47,
396
+ "learning_rate": 3.2661488231151706e-05,
397
+ "loss": 2.3584,
398
+ "step": 650000
399
+ },
400
+ {
401
+ "epoch": 3.52,
402
+ "learning_rate": 3.239474189624634e-05,
403
+ "loss": 2.3538,
404
+ "step": 660000
405
+ },
406
+ {
407
+ "epoch": 3.57,
408
+ "learning_rate": 3.212799556134099e-05,
409
+ "loss": 2.3481,
410
+ "step": 670000
411
+ },
412
+ {
413
+ "epoch": 3.63,
414
+ "learning_rate": 3.186124922643563e-05,
415
+ "loss": 2.3495,
416
+ "step": 680000
417
+ },
418
+ {
419
+ "epoch": 3.68,
420
+ "learning_rate": 3.159450289153027e-05,
421
+ "loss": 2.3413,
422
+ "step": 690000
423
+ },
424
+ {
425
+ "epoch": 3.73,
426
+ "learning_rate": 3.132775655662492e-05,
427
+ "loss": 2.3325,
428
+ "step": 700000
429
+ },
430
+ {
431
+ "epoch": 3.79,
432
+ "learning_rate": 3.1061010221719555e-05,
433
+ "loss": 2.3269,
434
+ "step": 710000
435
+ },
436
+ {
437
+ "epoch": 3.84,
438
+ "learning_rate": 3.07942638868142e-05,
439
+ "loss": 2.3255,
440
+ "step": 720000
441
+ },
442
+ {
443
+ "epoch": 3.89,
444
+ "learning_rate": 3.052751755190884e-05,
445
+ "loss": 2.3168,
446
+ "step": 730000
447
+ },
448
+ {
449
+ "epoch": 3.95,
450
+ "learning_rate": 3.0260771217003482e-05,
451
+ "loss": 2.3179,
452
+ "step": 740000
453
+ },
454
+ {
455
+ "epoch": 4.0,
456
+ "learning_rate": 2.999402488209812e-05,
457
+ "loss": 2.3098,
458
+ "step": 750000
459
+ },
460
+ {
461
+ "epoch": 4.05,
462
+ "learning_rate": 2.972727854719276e-05,
463
+ "loss": 2.2928,
464
+ "step": 760000
465
+ },
466
+ {
467
+ "epoch": 4.11,
468
+ "learning_rate": 2.9460532212287407e-05,
469
+ "loss": 2.2911,
470
+ "step": 770000
471
+ },
472
+ {
473
+ "epoch": 4.16,
474
+ "learning_rate": 2.9193785877382047e-05,
475
+ "loss": 2.2906,
476
+ "step": 780000
477
+ },
478
+ {
479
+ "epoch": 4.21,
480
+ "learning_rate": 2.8927039542476687e-05,
481
+ "loss": 2.2824,
482
+ "step": 790000
483
+ },
484
+ {
485
+ "epoch": 4.27,
486
+ "learning_rate": 2.866029320757133e-05,
487
+ "loss": 2.2766,
488
+ "step": 800000
489
+ },
490
+ {
491
+ "epoch": 4.32,
492
+ "learning_rate": 2.839354687266597e-05,
493
+ "loss": 2.2722,
494
+ "step": 810000
495
+ },
496
+ {
497
+ "epoch": 4.37,
498
+ "learning_rate": 2.812680053776061e-05,
499
+ "loss": 2.2665,
500
+ "step": 820000
501
+ },
502
+ {
503
+ "epoch": 4.43,
504
+ "learning_rate": 2.786005420285525e-05,
505
+ "loss": 2.259,
506
+ "step": 830000
507
+ },
508
+ {
509
+ "epoch": 4.48,
510
+ "learning_rate": 2.7593307867949895e-05,
511
+ "loss": 2.2584,
512
+ "step": 840000
513
+ },
514
+ {
515
+ "epoch": 4.53,
516
+ "learning_rate": 2.7326561533044536e-05,
517
+ "loss": 2.2524,
518
+ "step": 850000
519
+ },
520
+ {
521
+ "epoch": 4.59,
522
+ "learning_rate": 2.7059815198139176e-05,
523
+ "loss": 2.2536,
524
+ "step": 860000
525
+ },
526
+ {
527
+ "epoch": 4.64,
528
+ "learning_rate": 2.6793068863233823e-05,
529
+ "loss": 2.2446,
530
+ "step": 870000
531
+ },
532
+ {
533
+ "epoch": 4.69,
534
+ "learning_rate": 2.6526322528328463e-05,
535
+ "loss": 2.2439,
536
+ "step": 880000
537
+ },
538
+ {
539
+ "epoch": 4.75,
540
+ "learning_rate": 2.6259576193423104e-05,
541
+ "loss": 2.2389,
542
+ "step": 890000
543
+ },
544
+ {
545
+ "epoch": 4.8,
546
+ "learning_rate": 2.5992829858517747e-05,
547
+ "loss": 2.2362,
548
+ "step": 900000
549
+ },
550
+ {
551
+ "epoch": 4.85,
552
+ "learning_rate": 2.5726083523612388e-05,
553
+ "loss": 2.2313,
554
+ "step": 910000
555
+ },
556
+ {
557
+ "epoch": 4.91,
558
+ "learning_rate": 2.5459337188707028e-05,
559
+ "loss": 2.2283,
560
+ "step": 920000
561
+ },
562
+ {
563
+ "epoch": 4.96,
564
+ "learning_rate": 2.5192590853801668e-05,
565
+ "loss": 2.2224,
566
+ "step": 930000
567
+ },
568
+ {
569
+ "epoch": 5.01,
570
+ "learning_rate": 2.4925844518896312e-05,
571
+ "loss": 2.2175,
572
+ "step": 940000
573
+ },
574
+ {
575
+ "epoch": 5.07,
576
+ "learning_rate": 2.4659098183990952e-05,
577
+ "loss": 2.2067,
578
+ "step": 950000
579
+ },
580
+ {
581
+ "epoch": 5.12,
582
+ "learning_rate": 2.4392351849085596e-05,
583
+ "loss": 2.2029,
584
+ "step": 960000
585
+ },
586
+ {
587
+ "epoch": 5.17,
588
+ "learning_rate": 2.4125605514180236e-05,
589
+ "loss": 2.2014,
590
+ "step": 970000
591
+ },
592
+ {
593
+ "epoch": 5.23,
594
+ "learning_rate": 2.385885917927488e-05,
595
+ "loss": 2.2002,
596
+ "step": 980000
597
+ },
598
+ {
599
+ "epoch": 5.28,
600
+ "learning_rate": 2.3592112844369517e-05,
601
+ "loss": 2.195,
602
+ "step": 990000
603
+ },
604
+ {
605
+ "epoch": 5.33,
606
+ "learning_rate": 2.332536650946416e-05,
607
+ "loss": 2.1898,
608
+ "step": 1000000
609
+ },
610
+ {
611
+ "epoch": 5.39,
612
+ "learning_rate": 2.3058620174558804e-05,
613
+ "loss": 2.1864,
614
+ "step": 1010000
615
+ },
616
+ {
617
+ "epoch": 5.44,
618
+ "learning_rate": 2.2791873839653444e-05,
619
+ "loss": 2.1833,
620
+ "step": 1020000
621
+ },
622
+ {
623
+ "epoch": 5.49,
624
+ "learning_rate": 2.2525127504748088e-05,
625
+ "loss": 2.1799,
626
+ "step": 1030000
627
+ },
628
+ {
629
+ "epoch": 5.55,
630
+ "learning_rate": 2.2258381169842725e-05,
631
+ "loss": 2.1785,
632
+ "step": 1040000
633
+ },
634
+ {
635
+ "epoch": 5.6,
636
+ "learning_rate": 2.199163483493737e-05,
637
+ "loss": 2.1756,
638
+ "step": 1050000
639
+ },
640
+ {
641
+ "epoch": 5.66,
642
+ "learning_rate": 2.1724888500032012e-05,
643
+ "loss": 2.1726,
644
+ "step": 1060000
645
+ },
646
+ {
647
+ "epoch": 5.71,
648
+ "learning_rate": 2.1458142165126653e-05,
649
+ "loss": 2.1664,
650
+ "step": 1070000
651
+ },
652
+ {
653
+ "epoch": 5.76,
654
+ "learning_rate": 2.1191395830221293e-05,
655
+ "loss": 2.1651,
656
+ "step": 1080000
657
+ },
658
+ {
659
+ "epoch": 5.82,
660
+ "learning_rate": 2.0924649495315933e-05,
661
+ "loss": 2.1577,
662
+ "step": 1090000
663
+ },
664
+ {
665
+ "epoch": 5.87,
666
+ "learning_rate": 2.0657903160410577e-05,
667
+ "loss": 2.1472,
668
+ "step": 1100000
669
+ },
670
+ {
671
+ "epoch": 5.92,
672
+ "learning_rate": 2.039115682550522e-05,
673
+ "loss": 2.1458,
674
+ "step": 1110000
675
+ },
676
+ {
677
+ "epoch": 5.98,
678
+ "learning_rate": 2.012441049059986e-05,
679
+ "loss": 2.1432,
680
+ "step": 1120000
681
+ },
682
+ {
683
+ "epoch": 6.03,
684
+ "learning_rate": 1.98576641556945e-05,
685
+ "loss": 2.134,
686
+ "step": 1130000
687
+ },
688
+ {
689
+ "epoch": 6.08,
690
+ "learning_rate": 1.959091782078914e-05,
691
+ "loss": 2.1337,
692
+ "step": 1140000
693
+ },
694
+ {
695
+ "epoch": 6.14,
696
+ "learning_rate": 1.9324171485883785e-05,
697
+ "loss": 2.1337,
698
+ "step": 1150000
699
+ },
700
+ {
701
+ "epoch": 6.19,
702
+ "learning_rate": 1.9057425150978425e-05,
703
+ "loss": 2.128,
704
+ "step": 1160000
705
+ },
706
+ {
707
+ "epoch": 6.24,
708
+ "learning_rate": 1.879067881607307e-05,
709
+ "loss": 2.1262,
710
+ "step": 1170000
711
+ },
712
+ {
713
+ "epoch": 6.3,
714
+ "learning_rate": 1.852393248116771e-05,
715
+ "loss": 2.1209,
716
+ "step": 1180000
717
+ },
718
+ {
719
+ "epoch": 6.35,
720
+ "learning_rate": 1.825718614626235e-05,
721
+ "loss": 2.1178,
722
+ "step": 1190000
723
+ },
724
+ {
725
+ "epoch": 6.4,
726
+ "learning_rate": 1.7990439811356993e-05,
727
+ "loss": 2.1179,
728
+ "step": 1200000
729
+ },
730
+ {
731
+ "epoch": 6.46,
732
+ "learning_rate": 1.7723693476451634e-05,
733
+ "loss": 2.1123,
734
+ "step": 1210000
735
+ },
736
+ {
737
+ "epoch": 6.51,
738
+ "learning_rate": 1.7456947141546277e-05,
739
+ "loss": 2.1098,
740
+ "step": 1220000
741
+ },
742
+ {
743
+ "epoch": 6.56,
744
+ "learning_rate": 1.7190200806640918e-05,
745
+ "loss": 2.1115,
746
+ "step": 1230000
747
+ },
748
+ {
749
+ "epoch": 6.62,
750
+ "learning_rate": 1.6923454471735558e-05,
751
+ "loss": 2.1068,
752
+ "step": 1240000
753
+ },
754
+ {
755
+ "epoch": 6.67,
756
+ "learning_rate": 1.66567081368302e-05,
757
+ "loss": 2.1002,
758
+ "step": 1250000
759
+ },
760
+ {
761
+ "epoch": 6.72,
762
+ "learning_rate": 1.6389961801924842e-05,
763
+ "loss": 2.0962,
764
+ "step": 1260000
765
+ },
766
+ {
767
+ "epoch": 6.78,
768
+ "learning_rate": 1.6123215467019486e-05,
769
+ "loss": 2.0961,
770
+ "step": 1270000
771
+ },
772
+ {
773
+ "epoch": 6.83,
774
+ "learning_rate": 1.5856469132114126e-05,
775
+ "loss": 2.0911,
776
+ "step": 1280000
777
+ },
778
+ {
779
+ "epoch": 6.88,
780
+ "learning_rate": 1.5589722797208766e-05,
781
+ "loss": 2.0916,
782
+ "step": 1290000
783
+ },
784
+ {
785
+ "epoch": 6.94,
786
+ "learning_rate": 1.532297646230341e-05,
787
+ "loss": 2.0854,
788
+ "step": 1300000
789
+ },
790
+ {
791
+ "epoch": 6.99,
792
+ "learning_rate": 1.505623012739805e-05,
793
+ "loss": 2.0846,
794
+ "step": 1310000
795
+ },
796
+ {
797
+ "epoch": 7.04,
798
+ "learning_rate": 1.4789483792492692e-05,
799
+ "loss": 2.072,
800
+ "step": 1320000
801
+ },
802
+ {
803
+ "epoch": 7.1,
804
+ "learning_rate": 1.4522737457587332e-05,
805
+ "loss": 2.0691,
806
+ "step": 1330000
807
+ },
808
+ {
809
+ "epoch": 7.15,
810
+ "learning_rate": 1.4255991122681974e-05,
811
+ "loss": 2.072,
812
+ "step": 1340000
813
+ },
814
+ {
815
+ "epoch": 7.2,
816
+ "learning_rate": 1.3989244787776618e-05,
817
+ "loss": 2.0669,
818
+ "step": 1350000
819
+ },
820
+ {
821
+ "epoch": 7.26,
822
+ "learning_rate": 1.3722498452871257e-05,
823
+ "loss": 2.0646,
824
+ "step": 1360000
825
+ },
826
+ {
827
+ "epoch": 7.31,
828
+ "learning_rate": 1.34557521179659e-05,
829
+ "loss": 2.0573,
830
+ "step": 1370000
831
+ },
832
+ {
833
+ "epoch": 7.36,
834
+ "learning_rate": 1.318900578306054e-05,
835
+ "loss": 2.0487,
836
+ "step": 1380000
837
+ },
838
+ {
839
+ "epoch": 7.42,
840
+ "learning_rate": 1.2922259448155183e-05,
841
+ "loss": 2.0529,
842
+ "step": 1390000
843
+ },
844
+ {
845
+ "epoch": 7.47,
846
+ "learning_rate": 1.2655513113249826e-05,
847
+ "loss": 2.0468,
848
+ "step": 1400000
849
+ },
850
+ {
851
+ "epoch": 7.52,
852
+ "learning_rate": 1.2388766778344465e-05,
853
+ "loss": 2.0455,
854
+ "step": 1410000
855
+ },
856
+ {
857
+ "epoch": 7.58,
858
+ "learning_rate": 1.2122020443439109e-05,
859
+ "loss": 2.046,
860
+ "step": 1420000
861
+ },
862
+ {
863
+ "epoch": 7.63,
864
+ "learning_rate": 1.1855274108533749e-05,
865
+ "loss": 2.0398,
866
+ "step": 1430000
867
+ },
868
+ {
869
+ "epoch": 7.68,
870
+ "learning_rate": 1.158852777362839e-05,
871
+ "loss": 2.0391,
872
+ "step": 1440000
873
+ },
874
+ {
875
+ "epoch": 7.74,
876
+ "learning_rate": 1.1321781438723033e-05,
877
+ "loss": 2.0345,
878
+ "step": 1450000
879
+ },
880
+ {
881
+ "epoch": 7.79,
882
+ "learning_rate": 1.1055035103817673e-05,
883
+ "loss": 2.038,
884
+ "step": 1460000
885
+ },
886
+ {
887
+ "epoch": 7.84,
888
+ "learning_rate": 1.0788288768912317e-05,
889
+ "loss": 2.032,
890
+ "step": 1470000
891
+ },
892
+ {
893
+ "epoch": 7.9,
894
+ "learning_rate": 1.0521542434006957e-05,
895
+ "loss": 2.0271,
896
+ "step": 1480000
897
+ },
898
+ {
899
+ "epoch": 7.95,
900
+ "learning_rate": 1.0254796099101599e-05,
901
+ "loss": 2.0261,
902
+ "step": 1490000
903
+ },
904
+ {
905
+ "epoch": 8.0,
906
+ "learning_rate": 9.988049764196241e-06,
907
+ "loss": 2.0235,
908
+ "step": 1500000
909
+ },
910
+ {
911
+ "epoch": 8.06,
912
+ "learning_rate": 9.721303429290881e-06,
913
+ "loss": 2.0173,
914
+ "step": 1510000
915
+ },
916
+ {
917
+ "epoch": 8.11,
918
+ "learning_rate": 9.454557094385523e-06,
919
+ "loss": 2.0178,
920
+ "step": 1520000
921
+ },
922
+ {
923
+ "epoch": 8.16,
924
+ "learning_rate": 9.187810759480165e-06,
925
+ "loss": 2.0116,
926
+ "step": 1530000
927
+ },
928
+ {
929
+ "epoch": 8.22,
930
+ "learning_rate": 8.921064424574807e-06,
931
+ "loss": 2.0099,
932
+ "step": 1540000
933
+ },
934
+ {
935
+ "epoch": 8.27,
936
+ "learning_rate": 8.654318089669448e-06,
937
+ "loss": 2.0105,
938
+ "step": 1550000
939
+ },
940
+ {
941
+ "epoch": 8.32,
942
+ "learning_rate": 8.38757175476409e-06,
943
+ "loss": 2.0033,
944
+ "step": 1560000
945
+ },
946
+ {
947
+ "epoch": 8.38,
948
+ "learning_rate": 8.120825419858732e-06,
949
+ "loss": 2.0033,
950
+ "step": 1570000
951
+ },
952
+ {
953
+ "epoch": 8.43,
954
+ "learning_rate": 7.854079084953374e-06,
955
+ "loss": 2.0,
956
+ "step": 1580000
957
+ },
958
+ {
959
+ "epoch": 8.48,
960
+ "learning_rate": 7.5873327500480155e-06,
961
+ "loss": 1.9954,
962
+ "step": 1590000
963
+ },
964
+ {
965
+ "epoch": 8.54,
966
+ "learning_rate": 7.320586415142657e-06,
967
+ "loss": 1.9964,
968
+ "step": 1600000
969
+ },
970
+ {
971
+ "epoch": 8.59,
972
+ "learning_rate": 7.053840080237298e-06,
973
+ "loss": 1.9995,
974
+ "step": 1610000
975
+ },
976
+ {
977
+ "epoch": 8.64,
978
+ "learning_rate": 6.787093745331939e-06,
979
+ "loss": 1.9878,
980
+ "step": 1620000
981
+ },
982
+ {
983
+ "epoch": 8.7,
984
+ "learning_rate": 6.52034741042658e-06,
985
+ "loss": 1.988,
986
+ "step": 1630000
987
+ },
988
+ {
989
+ "epoch": 8.75,
990
+ "learning_rate": 6.253601075521223e-06,
991
+ "loss": 1.9861,
992
+ "step": 1640000
993
+ },
994
+ {
995
+ "epoch": 8.8,
996
+ "learning_rate": 5.986854740615865e-06,
997
+ "loss": 1.9828,
998
+ "step": 1650000
999
+ },
1000
+ {
1001
+ "epoch": 8.86,
1002
+ "learning_rate": 5.720108405710506e-06,
1003
+ "loss": 1.9787,
1004
+ "step": 1660000
1005
+ },
1006
+ {
1007
+ "epoch": 8.91,
1008
+ "learning_rate": 5.453362070805147e-06,
1009
+ "loss": 1.9794,
1010
+ "step": 1670000
1011
+ },
1012
+ {
1013
+ "epoch": 8.96,
1014
+ "learning_rate": 5.186615735899789e-06,
1015
+ "loss": 1.9756,
1016
+ "step": 1680000
1017
+ },
1018
+ {
1019
+ "epoch": 9.02,
1020
+ "learning_rate": 4.91986940099443e-06,
1021
+ "loss": 1.9756,
1022
+ "step": 1690000
1023
+ },
1024
+ {
1025
+ "epoch": 9.07,
1026
+ "learning_rate": 4.653123066089072e-06,
1027
+ "loss": 1.9676,
1028
+ "step": 1700000
1029
+ },
1030
+ {
1031
+ "epoch": 9.12,
1032
+ "learning_rate": 4.386376731183714e-06,
1033
+ "loss": 1.967,
1034
+ "step": 1710000
1035
+ },
1036
+ {
1037
+ "epoch": 9.18,
1038
+ "learning_rate": 4.119630396278355e-06,
1039
+ "loss": 1.9656,
1040
+ "step": 1720000
1041
+ },
1042
+ {
1043
+ "epoch": 9.23,
1044
+ "learning_rate": 3.8528840613729966e-06,
1045
+ "loss": 1.9627,
1046
+ "step": 1730000
1047
+ },
1048
+ {
1049
+ "epoch": 9.28,
1050
+ "learning_rate": 3.5861377264676385e-06,
1051
+ "loss": 1.964,
1052
+ "step": 1740000
1053
+ },
1054
+ {
1055
+ "epoch": 9.34,
1056
+ "learning_rate": 3.31939139156228e-06,
1057
+ "loss": 1.9646,
1058
+ "step": 1750000
1059
+ },
1060
+ {
1061
+ "epoch": 9.39,
1062
+ "learning_rate": 3.0526450566569217e-06,
1063
+ "loss": 1.9591,
1064
+ "step": 1760000
1065
+ },
1066
+ {
1067
+ "epoch": 9.44,
1068
+ "learning_rate": 2.7858987217515632e-06,
1069
+ "loss": 1.9623,
1070
+ "step": 1770000
1071
+ },
1072
+ {
1073
+ "epoch": 9.5,
1074
+ "learning_rate": 2.519152386846205e-06,
1075
+ "loss": 1.9544,
1076
+ "step": 1780000
1077
+ },
1078
+ {
1079
+ "epoch": 9.55,
1080
+ "learning_rate": 2.2524060519408464e-06,
1081
+ "loss": 1.9572,
1082
+ "step": 1790000
1083
+ },
1084
+ {
1085
+ "epoch": 9.6,
1086
+ "learning_rate": 1.985659717035488e-06,
1087
+ "loss": 1.9545,
1088
+ "step": 1800000
1089
+ },
1090
+ {
1091
+ "epoch": 9.66,
1092
+ "learning_rate": 1.7189133821301297e-06,
1093
+ "loss": 1.953,
1094
+ "step": 1810000
1095
+ },
1096
+ {
1097
+ "epoch": 9.71,
1098
+ "learning_rate": 1.4521670472247713e-06,
1099
+ "loss": 1.9523,
1100
+ "step": 1820000
1101
+ },
1102
+ {
1103
+ "epoch": 9.76,
1104
+ "learning_rate": 1.1854207123194128e-06,
1105
+ "loss": 1.9528,
1106
+ "step": 1830000
1107
+ },
1108
+ {
1109
+ "epoch": 9.82,
1110
+ "learning_rate": 9.186743774140545e-07,
1111
+ "loss": 1.9482,
1112
+ "step": 1840000
1113
+ },
1114
+ {
1115
+ "epoch": 9.87,
1116
+ "learning_rate": 6.519280425086959e-07,
1117
+ "loss": 1.9511,
1118
+ "step": 1850000
1119
+ },
1120
+ {
1121
+ "epoch": 9.92,
1122
+ "learning_rate": 3.8518170760333756e-07,
1123
+ "loss": 1.947,
1124
+ "step": 1860000
1125
+ },
1126
+ {
1127
+ "epoch": 9.98,
1128
+ "learning_rate": 1.1843537269797913e-07,
1129
+ "loss": 1.9437,
1130
+ "step": 1870000
1131
+ }
1132
+ ],
1133
+ "max_steps": 1874440,
1134
+ "num_train_epochs": 10,
1135
+ "total_flos": 9.135177187546945e+18,
1136
+ "trial_name": null,
1137
+ "trial_params": null
1138
+ }