dannyvas23 commited on
Commit
61600ec
1 Parent(s): 33bea97

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1231 -0
trainer_state.json ADDED
@@ -0,0 +1,1231 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9918692001767565,
3
+ "best_model_checkpoint": "electricidad-small-discriminator-finetuned-clasificacion-comentarios-suicidas/checkpoint-84154",
4
+ "epoch": 15.0,
5
+ "global_step": 90165,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.08,
12
+ "learning_rate": 1.9889092219819223e-05,
13
+ "loss": 0.3744,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.17,
18
+ "learning_rate": 1.977818443963844e-05,
19
+ "loss": 0.2448,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.25,
24
+ "learning_rate": 1.9667276659457665e-05,
25
+ "loss": 0.2103,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.33,
30
+ "learning_rate": 1.9556368879276883e-05,
31
+ "loss": 0.2037,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.42,
36
+ "learning_rate": 1.9445461099096104e-05,
37
+ "loss": 0.1902,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.5,
42
+ "learning_rate": 1.9334553318915322e-05,
43
+ "loss": 0.1874,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.58,
48
+ "learning_rate": 1.9223645538734543e-05,
49
+ "loss": 0.1717,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.67,
54
+ "learning_rate": 1.9112737758553765e-05,
55
+ "loss": 0.1723,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.75,
60
+ "learning_rate": 1.9001829978372986e-05,
61
+ "loss": 0.1697,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.83,
66
+ "learning_rate": 1.8890922198192204e-05,
67
+ "loss": 0.1637,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.91,
72
+ "learning_rate": 1.8780014418011425e-05,
73
+ "loss": 0.1593,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 1.0,
78
+ "learning_rate": 1.8669106637830646e-05,
79
+ "loss": 0.1611,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 1.0,
84
+ "eval_accuracy": 0.9527176314626602,
85
+ "eval_loss": 0.13305681943893433,
86
+ "eval_runtime": 183.6617,
87
+ "eval_samples_per_second": 123.216,
88
+ "eval_steps_per_second": 3.855,
89
+ "step": 6011
90
+ },
91
+ {
92
+ "epoch": 1.08,
93
+ "learning_rate": 1.8558198857649864e-05,
94
+ "loss": 0.1436,
95
+ "step": 6500
96
+ },
97
+ {
98
+ "epoch": 1.16,
99
+ "learning_rate": 1.8447291077469085e-05,
100
+ "loss": 0.1391,
101
+ "step": 7000
102
+ },
103
+ {
104
+ "epoch": 1.25,
105
+ "learning_rate": 1.8336383297288306e-05,
106
+ "loss": 0.1382,
107
+ "step": 7500
108
+ },
109
+ {
110
+ "epoch": 1.33,
111
+ "learning_rate": 1.8225475517107528e-05,
112
+ "loss": 0.1332,
113
+ "step": 8000
114
+ },
115
+ {
116
+ "epoch": 1.41,
117
+ "learning_rate": 1.8114567736926746e-05,
118
+ "loss": 0.1385,
119
+ "step": 8500
120
+ },
121
+ {
122
+ "epoch": 1.5,
123
+ "learning_rate": 1.8003659956745967e-05,
124
+ "loss": 0.136,
125
+ "step": 9000
126
+ },
127
+ {
128
+ "epoch": 1.58,
129
+ "learning_rate": 1.7892752176565188e-05,
130
+ "loss": 0.1426,
131
+ "step": 9500
132
+ },
133
+ {
134
+ "epoch": 1.66,
135
+ "learning_rate": 1.7781844396384406e-05,
136
+ "loss": 0.1308,
137
+ "step": 10000
138
+ },
139
+ {
140
+ "epoch": 1.75,
141
+ "learning_rate": 1.767093661620363e-05,
142
+ "loss": 0.1355,
143
+ "step": 10500
144
+ },
145
+ {
146
+ "epoch": 1.83,
147
+ "learning_rate": 1.7560028836022848e-05,
148
+ "loss": 0.1331,
149
+ "step": 11000
150
+ },
151
+ {
152
+ "epoch": 1.91,
153
+ "learning_rate": 1.744912105584207e-05,
154
+ "loss": 0.1272,
155
+ "step": 11500
156
+ },
157
+ {
158
+ "epoch": 2.0,
159
+ "learning_rate": 1.7338213275661287e-05,
160
+ "loss": 0.1345,
161
+ "step": 12000
162
+ },
163
+ {
164
+ "epoch": 2.0,
165
+ "eval_accuracy": 0.9608042421564296,
166
+ "eval_loss": 0.11096587032079697,
167
+ "eval_runtime": 183.8189,
168
+ "eval_samples_per_second": 123.11,
169
+ "eval_steps_per_second": 3.852,
170
+ "step": 12022
171
+ },
172
+ {
173
+ "epoch": 2.08,
174
+ "learning_rate": 1.722730549548051e-05,
175
+ "loss": 0.1176,
176
+ "step": 12500
177
+ },
178
+ {
179
+ "epoch": 2.16,
180
+ "learning_rate": 1.711639771529973e-05,
181
+ "loss": 0.1138,
182
+ "step": 13000
183
+ },
184
+ {
185
+ "epoch": 2.25,
186
+ "learning_rate": 1.700548993511895e-05,
187
+ "loss": 0.1121,
188
+ "step": 13500
189
+ },
190
+ {
191
+ "epoch": 2.33,
192
+ "learning_rate": 1.689458215493817e-05,
193
+ "loss": 0.1163,
194
+ "step": 14000
195
+ },
196
+ {
197
+ "epoch": 2.41,
198
+ "learning_rate": 1.678367437475739e-05,
199
+ "loss": 0.1115,
200
+ "step": 14500
201
+ },
202
+ {
203
+ "epoch": 2.5,
204
+ "learning_rate": 1.667276659457661e-05,
205
+ "loss": 0.1201,
206
+ "step": 15000
207
+ },
208
+ {
209
+ "epoch": 2.58,
210
+ "learning_rate": 1.656185881439583e-05,
211
+ "loss": 0.1215,
212
+ "step": 15500
213
+ },
214
+ {
215
+ "epoch": 2.66,
216
+ "learning_rate": 1.6450951034215054e-05,
217
+ "loss": 0.112,
218
+ "step": 16000
219
+ },
220
+ {
221
+ "epoch": 2.74,
222
+ "learning_rate": 1.634004325403427e-05,
223
+ "loss": 0.122,
224
+ "step": 16500
225
+ },
226
+ {
227
+ "epoch": 2.83,
228
+ "learning_rate": 1.6229135473853493e-05,
229
+ "loss": 0.1084,
230
+ "step": 17000
231
+ },
232
+ {
233
+ "epoch": 2.91,
234
+ "learning_rate": 1.611822769367271e-05,
235
+ "loss": 0.1208,
236
+ "step": 17500
237
+ },
238
+ {
239
+ "epoch": 2.99,
240
+ "learning_rate": 1.6007319913491932e-05,
241
+ "loss": 0.1085,
242
+ "step": 18000
243
+ },
244
+ {
245
+ "epoch": 3.0,
246
+ "eval_accuracy": 0.9708351745470615,
247
+ "eval_loss": 0.08641747385263443,
248
+ "eval_runtime": 183.9354,
249
+ "eval_samples_per_second": 123.032,
250
+ "eval_steps_per_second": 3.849,
251
+ "step": 18033
252
+ },
253
+ {
254
+ "epoch": 3.08,
255
+ "learning_rate": 1.5896412133311153e-05,
256
+ "loss": 0.1011,
257
+ "step": 18500
258
+ },
259
+ {
260
+ "epoch": 3.16,
261
+ "learning_rate": 1.5785504353130374e-05,
262
+ "loss": 0.1,
263
+ "step": 19000
264
+ },
265
+ {
266
+ "epoch": 3.24,
267
+ "learning_rate": 1.5674596572949596e-05,
268
+ "loss": 0.102,
269
+ "step": 19500
270
+ },
271
+ {
272
+ "epoch": 3.33,
273
+ "learning_rate": 1.5563688792768813e-05,
274
+ "loss": 0.0978,
275
+ "step": 20000
276
+ },
277
+ {
278
+ "epoch": 3.41,
279
+ "learning_rate": 1.5452781012588035e-05,
280
+ "loss": 0.1005,
281
+ "step": 20500
282
+ },
283
+ {
284
+ "epoch": 3.49,
285
+ "learning_rate": 1.5341873232407252e-05,
286
+ "loss": 0.1019,
287
+ "step": 21000
288
+ },
289
+ {
290
+ "epoch": 3.58,
291
+ "learning_rate": 1.5230965452226474e-05,
292
+ "loss": 0.0985,
293
+ "step": 21500
294
+ },
295
+ {
296
+ "epoch": 3.66,
297
+ "learning_rate": 1.5120057672045697e-05,
298
+ "loss": 0.0971,
299
+ "step": 22000
300
+ },
301
+ {
302
+ "epoch": 3.74,
303
+ "learning_rate": 1.5009149891864916e-05,
304
+ "loss": 0.107,
305
+ "step": 22500
306
+ },
307
+ {
308
+ "epoch": 3.83,
309
+ "learning_rate": 1.4898242111684136e-05,
310
+ "loss": 0.1001,
311
+ "step": 23000
312
+ },
313
+ {
314
+ "epoch": 3.91,
315
+ "learning_rate": 1.4787334331503355e-05,
316
+ "loss": 0.0988,
317
+ "step": 23500
318
+ },
319
+ {
320
+ "epoch": 3.99,
321
+ "learning_rate": 1.4676426551322576e-05,
322
+ "loss": 0.0994,
323
+ "step": 24000
324
+ },
325
+ {
326
+ "epoch": 4.0,
327
+ "eval_accuracy": 0.974856385329209,
328
+ "eval_loss": 0.07361753284931183,
329
+ "eval_runtime": 183.7502,
330
+ "eval_samples_per_second": 123.156,
331
+ "eval_steps_per_second": 3.853,
332
+ "step": 24044
333
+ },
334
+ {
335
+ "epoch": 4.08,
336
+ "learning_rate": 1.4565518771141798e-05,
337
+ "loss": 0.0895,
338
+ "step": 24500
339
+ },
340
+ {
341
+ "epoch": 4.16,
342
+ "learning_rate": 1.4454610990961017e-05,
343
+ "loss": 0.0833,
344
+ "step": 25000
345
+ },
346
+ {
347
+ "epoch": 4.24,
348
+ "learning_rate": 1.4343703210780238e-05,
349
+ "loss": 0.0866,
350
+ "step": 25500
351
+ },
352
+ {
353
+ "epoch": 4.33,
354
+ "learning_rate": 1.4232795430599458e-05,
355
+ "loss": 0.0892,
356
+ "step": 26000
357
+ },
358
+ {
359
+ "epoch": 4.41,
360
+ "learning_rate": 1.4121887650418678e-05,
361
+ "loss": 0.0837,
362
+ "step": 26500
363
+ },
364
+ {
365
+ "epoch": 4.49,
366
+ "learning_rate": 1.4010979870237897e-05,
367
+ "loss": 0.0903,
368
+ "step": 27000
369
+ },
370
+ {
371
+ "epoch": 4.57,
372
+ "learning_rate": 1.390007209005712e-05,
373
+ "loss": 0.0825,
374
+ "step": 27500
375
+ },
376
+ {
377
+ "epoch": 4.66,
378
+ "learning_rate": 1.378916430987634e-05,
379
+ "loss": 0.0905,
380
+ "step": 28000
381
+ },
382
+ {
383
+ "epoch": 4.74,
384
+ "learning_rate": 1.3678256529695559e-05,
385
+ "loss": 0.0822,
386
+ "step": 28500
387
+ },
388
+ {
389
+ "epoch": 4.82,
390
+ "learning_rate": 1.3567348749514779e-05,
391
+ "loss": 0.0904,
392
+ "step": 29000
393
+ },
394
+ {
395
+ "epoch": 4.91,
396
+ "learning_rate": 1.3456440969334e-05,
397
+ "loss": 0.0938,
398
+ "step": 29500
399
+ },
400
+ {
401
+ "epoch": 4.99,
402
+ "learning_rate": 1.334553318915322e-05,
403
+ "loss": 0.0905,
404
+ "step": 30000
405
+ },
406
+ {
407
+ "epoch": 5.0,
408
+ "eval_accuracy": 0.9796288113124172,
409
+ "eval_loss": 0.06523095816373825,
410
+ "eval_runtime": 184.074,
411
+ "eval_samples_per_second": 122.94,
412
+ "eval_steps_per_second": 3.846,
413
+ "step": 30055
414
+ },
415
+ {
416
+ "epoch": 5.07,
417
+ "learning_rate": 1.323462540897244e-05,
418
+ "loss": 0.072,
419
+ "step": 30500
420
+ },
421
+ {
422
+ "epoch": 5.16,
423
+ "learning_rate": 1.3123717628791662e-05,
424
+ "loss": 0.0769,
425
+ "step": 31000
426
+ },
427
+ {
428
+ "epoch": 5.24,
429
+ "learning_rate": 1.3012809848610881e-05,
430
+ "loss": 0.0761,
431
+ "step": 31500
432
+ },
433
+ {
434
+ "epoch": 5.32,
435
+ "learning_rate": 1.2901902068430101e-05,
436
+ "loss": 0.0782,
437
+ "step": 32000
438
+ },
439
+ {
440
+ "epoch": 5.41,
441
+ "learning_rate": 1.279099428824932e-05,
442
+ "loss": 0.0682,
443
+ "step": 32500
444
+ },
445
+ {
446
+ "epoch": 5.49,
447
+ "learning_rate": 1.2680086508068543e-05,
448
+ "loss": 0.0799,
449
+ "step": 33000
450
+ },
451
+ {
452
+ "epoch": 5.57,
453
+ "learning_rate": 1.2569178727887763e-05,
454
+ "loss": 0.0766,
455
+ "step": 33500
456
+ },
457
+ {
458
+ "epoch": 5.66,
459
+ "learning_rate": 1.2458270947706982e-05,
460
+ "loss": 0.0762,
461
+ "step": 34000
462
+ },
463
+ {
464
+ "epoch": 5.74,
465
+ "learning_rate": 1.2347363167526204e-05,
466
+ "loss": 0.0775,
467
+ "step": 34500
468
+ },
469
+ {
470
+ "epoch": 5.82,
471
+ "learning_rate": 1.2236455387345423e-05,
472
+ "loss": 0.0756,
473
+ "step": 35000
474
+ },
475
+ {
476
+ "epoch": 5.91,
477
+ "learning_rate": 1.2125547607164643e-05,
478
+ "loss": 0.0784,
479
+ "step": 35500
480
+ },
481
+ {
482
+ "epoch": 5.99,
483
+ "learning_rate": 1.2014639826983866e-05,
484
+ "loss": 0.0807,
485
+ "step": 36000
486
+ },
487
+ {
488
+ "epoch": 6.0,
489
+ "eval_accuracy": 0.9823243482103402,
490
+ "eval_loss": 0.0608493909239769,
491
+ "eval_runtime": 183.9197,
492
+ "eval_samples_per_second": 123.043,
493
+ "eval_steps_per_second": 3.85,
494
+ "step": 36066
495
+ },
496
+ {
497
+ "epoch": 6.07,
498
+ "learning_rate": 1.1903732046803085e-05,
499
+ "loss": 0.0664,
500
+ "step": 36500
501
+ },
502
+ {
503
+ "epoch": 6.16,
504
+ "learning_rate": 1.1792824266622305e-05,
505
+ "loss": 0.0656,
506
+ "step": 37000
507
+ },
508
+ {
509
+ "epoch": 6.24,
510
+ "learning_rate": 1.1681916486441524e-05,
511
+ "loss": 0.0669,
512
+ "step": 37500
513
+ },
514
+ {
515
+ "epoch": 6.32,
516
+ "learning_rate": 1.1571008706260745e-05,
517
+ "loss": 0.0646,
518
+ "step": 38000
519
+ },
520
+ {
521
+ "epoch": 6.4,
522
+ "learning_rate": 1.1460100926079965e-05,
523
+ "loss": 0.067,
524
+ "step": 38500
525
+ },
526
+ {
527
+ "epoch": 6.49,
528
+ "learning_rate": 1.1349193145899186e-05,
529
+ "loss": 0.0708,
530
+ "step": 39000
531
+ },
532
+ {
533
+ "epoch": 6.57,
534
+ "learning_rate": 1.1238285365718407e-05,
535
+ "loss": 0.0729,
536
+ "step": 39500
537
+ },
538
+ {
539
+ "epoch": 6.65,
540
+ "learning_rate": 1.1127377585537627e-05,
541
+ "loss": 0.0682,
542
+ "step": 40000
543
+ },
544
+ {
545
+ "epoch": 6.74,
546
+ "learning_rate": 1.1016469805356846e-05,
547
+ "loss": 0.0747,
548
+ "step": 40500
549
+ },
550
+ {
551
+ "epoch": 6.82,
552
+ "learning_rate": 1.0905562025176066e-05,
553
+ "loss": 0.0697,
554
+ "step": 41000
555
+ },
556
+ {
557
+ "epoch": 6.9,
558
+ "learning_rate": 1.0794654244995286e-05,
559
+ "loss": 0.0695,
560
+ "step": 41500
561
+ },
562
+ {
563
+ "epoch": 6.99,
564
+ "learning_rate": 1.0683746464814508e-05,
565
+ "loss": 0.0692,
566
+ "step": 42000
567
+ },
568
+ {
569
+ "epoch": 7.0,
570
+ "eval_accuracy": 0.9861246133451171,
571
+ "eval_loss": 0.05471781641244888,
572
+ "eval_runtime": 184.0533,
573
+ "eval_samples_per_second": 122.954,
574
+ "eval_steps_per_second": 3.847,
575
+ "step": 42077
576
+ },
577
+ {
578
+ "epoch": 7.07,
579
+ "learning_rate": 1.0572838684633728e-05,
580
+ "loss": 0.062,
581
+ "step": 42500
582
+ },
583
+ {
584
+ "epoch": 7.15,
585
+ "learning_rate": 1.0461930904452948e-05,
586
+ "loss": 0.0573,
587
+ "step": 43000
588
+ },
589
+ {
590
+ "epoch": 7.24,
591
+ "learning_rate": 1.0351023124272169e-05,
592
+ "loss": 0.0599,
593
+ "step": 43500
594
+ },
595
+ {
596
+ "epoch": 7.32,
597
+ "learning_rate": 1.0240115344091388e-05,
598
+ "loss": 0.0577,
599
+ "step": 44000
600
+ },
601
+ {
602
+ "epoch": 7.4,
603
+ "learning_rate": 1.012920756391061e-05,
604
+ "loss": 0.0582,
605
+ "step": 44500
606
+ },
607
+ {
608
+ "epoch": 7.49,
609
+ "learning_rate": 1.001829978372983e-05,
610
+ "loss": 0.0637,
611
+ "step": 45000
612
+ },
613
+ {
614
+ "epoch": 7.57,
615
+ "learning_rate": 9.90739200354905e-06,
616
+ "loss": 0.0658,
617
+ "step": 45500
618
+ },
619
+ {
620
+ "epoch": 7.65,
621
+ "learning_rate": 9.79648422336827e-06,
622
+ "loss": 0.06,
623
+ "step": 46000
624
+ },
625
+ {
626
+ "epoch": 7.74,
627
+ "learning_rate": 9.68557644318749e-06,
628
+ "loss": 0.0575,
629
+ "step": 46500
630
+ },
631
+ {
632
+ "epoch": 7.82,
633
+ "learning_rate": 9.57466866300671e-06,
634
+ "loss": 0.0635,
635
+ "step": 47000
636
+ },
637
+ {
638
+ "epoch": 7.9,
639
+ "learning_rate": 9.46376088282593e-06,
640
+ "loss": 0.0623,
641
+ "step": 47500
642
+ },
643
+ {
644
+ "epoch": 7.99,
645
+ "learning_rate": 9.352853102645151e-06,
646
+ "loss": 0.0604,
647
+ "step": 48000
648
+ },
649
+ {
650
+ "epoch": 8.0,
651
+ "eval_accuracy": 0.9859478568272205,
652
+ "eval_loss": 0.051153190433979034,
653
+ "eval_runtime": 184.082,
654
+ "eval_samples_per_second": 122.934,
655
+ "eval_steps_per_second": 3.846,
656
+ "step": 48088
657
+ },
658
+ {
659
+ "epoch": 8.07,
660
+ "learning_rate": 9.241945322464373e-06,
661
+ "loss": 0.0507,
662
+ "step": 48500
663
+ },
664
+ {
665
+ "epoch": 8.15,
666
+ "learning_rate": 9.131037542283592e-06,
667
+ "loss": 0.0538,
668
+ "step": 49000
669
+ },
670
+ {
671
+ "epoch": 8.23,
672
+ "learning_rate": 9.020129762102813e-06,
673
+ "loss": 0.0506,
674
+ "step": 49500
675
+ },
676
+ {
677
+ "epoch": 8.32,
678
+ "learning_rate": 8.909221981922033e-06,
679
+ "loss": 0.0526,
680
+ "step": 50000
681
+ },
682
+ {
683
+ "epoch": 8.4,
684
+ "learning_rate": 8.798314201741252e-06,
685
+ "loss": 0.0579,
686
+ "step": 50500
687
+ },
688
+ {
689
+ "epoch": 8.48,
690
+ "learning_rate": 8.687406421560474e-06,
691
+ "loss": 0.051,
692
+ "step": 51000
693
+ },
694
+ {
695
+ "epoch": 8.57,
696
+ "learning_rate": 8.576498641379693e-06,
697
+ "loss": 0.0508,
698
+ "step": 51500
699
+ },
700
+ {
701
+ "epoch": 8.65,
702
+ "learning_rate": 8.465590861198913e-06,
703
+ "loss": 0.0566,
704
+ "step": 52000
705
+ },
706
+ {
707
+ "epoch": 8.73,
708
+ "learning_rate": 8.354683081018134e-06,
709
+ "loss": 0.0582,
710
+ "step": 52500
711
+ },
712
+ {
713
+ "epoch": 8.82,
714
+ "learning_rate": 8.243775300837355e-06,
715
+ "loss": 0.0512,
716
+ "step": 53000
717
+ },
718
+ {
719
+ "epoch": 8.9,
720
+ "learning_rate": 8.132867520656575e-06,
721
+ "loss": 0.0544,
722
+ "step": 53500
723
+ },
724
+ {
725
+ "epoch": 8.98,
726
+ "learning_rate": 8.021959740475796e-06,
727
+ "loss": 0.0482,
728
+ "step": 54000
729
+ },
730
+ {
731
+ "epoch": 9.0,
732
+ "eval_accuracy": 0.9897481219619974,
733
+ "eval_loss": 0.04574726149439812,
734
+ "eval_runtime": 184.1495,
735
+ "eval_samples_per_second": 122.889,
736
+ "eval_steps_per_second": 3.845,
737
+ "step": 54099
738
+ },
739
+ {
740
+ "epoch": 9.07,
741
+ "learning_rate": 7.911051960295015e-06,
742
+ "loss": 0.0506,
743
+ "step": 54500
744
+ },
745
+ {
746
+ "epoch": 9.15,
747
+ "learning_rate": 7.800144180114235e-06,
748
+ "loss": 0.0423,
749
+ "step": 55000
750
+ },
751
+ {
752
+ "epoch": 9.23,
753
+ "learning_rate": 7.689236399933456e-06,
754
+ "loss": 0.0458,
755
+ "step": 55500
756
+ },
757
+ {
758
+ "epoch": 9.32,
759
+ "learning_rate": 7.5783286197526765e-06,
760
+ "loss": 0.0454,
761
+ "step": 56000
762
+ },
763
+ {
764
+ "epoch": 9.4,
765
+ "learning_rate": 7.467420839571896e-06,
766
+ "loss": 0.0434,
767
+ "step": 56500
768
+ },
769
+ {
770
+ "epoch": 9.48,
771
+ "learning_rate": 7.356513059391117e-06,
772
+ "loss": 0.0494,
773
+ "step": 57000
774
+ },
775
+ {
776
+ "epoch": 9.57,
777
+ "learning_rate": 7.245605279210337e-06,
778
+ "loss": 0.0477,
779
+ "step": 57500
780
+ },
781
+ {
782
+ "epoch": 9.65,
783
+ "learning_rate": 7.134697499029557e-06,
784
+ "loss": 0.0473,
785
+ "step": 58000
786
+ },
787
+ {
788
+ "epoch": 9.73,
789
+ "learning_rate": 7.0237897188487784e-06,
790
+ "loss": 0.0488,
791
+ "step": 58500
792
+ },
793
+ {
794
+ "epoch": 9.82,
795
+ "learning_rate": 6.912881938667998e-06,
796
+ "loss": 0.0518,
797
+ "step": 59000
798
+ },
799
+ {
800
+ "epoch": 9.9,
801
+ "learning_rate": 6.801974158487219e-06,
802
+ "loss": 0.0482,
803
+ "step": 59500
804
+ },
805
+ {
806
+ "epoch": 9.98,
807
+ "learning_rate": 6.691066378306439e-06,
808
+ "loss": 0.0455,
809
+ "step": 60000
810
+ },
811
+ {
812
+ "epoch": 10.0,
813
+ "eval_accuracy": 0.9880689350419797,
814
+ "eval_loss": 0.049992308020591736,
815
+ "eval_runtime": 183.6678,
816
+ "eval_samples_per_second": 123.212,
817
+ "eval_steps_per_second": 3.855,
818
+ "step": 60110
819
+ },
820
+ {
821
+ "epoch": 10.06,
822
+ "learning_rate": 6.580158598125659e-06,
823
+ "loss": 0.0419,
824
+ "step": 60500
825
+ },
826
+ {
827
+ "epoch": 10.15,
828
+ "learning_rate": 6.4692508179448795e-06,
829
+ "loss": 0.0401,
830
+ "step": 61000
831
+ },
832
+ {
833
+ "epoch": 10.23,
834
+ "learning_rate": 6.3583430377641e-06,
835
+ "loss": 0.0403,
836
+ "step": 61500
837
+ },
838
+ {
839
+ "epoch": 10.31,
840
+ "learning_rate": 6.247435257583319e-06,
841
+ "loss": 0.0375,
842
+ "step": 62000
843
+ },
844
+ {
845
+ "epoch": 10.4,
846
+ "learning_rate": 6.136527477402541e-06,
847
+ "loss": 0.0461,
848
+ "step": 62500
849
+ },
850
+ {
851
+ "epoch": 10.48,
852
+ "learning_rate": 6.025619697221761e-06,
853
+ "loss": 0.0443,
854
+ "step": 63000
855
+ },
856
+ {
857
+ "epoch": 10.56,
858
+ "learning_rate": 5.9147119170409805e-06,
859
+ "loss": 0.0439,
860
+ "step": 63500
861
+ },
862
+ {
863
+ "epoch": 10.65,
864
+ "learning_rate": 5.803804136860202e-06,
865
+ "loss": 0.0374,
866
+ "step": 64000
867
+ },
868
+ {
869
+ "epoch": 10.73,
870
+ "learning_rate": 5.692896356679421e-06,
871
+ "loss": 0.0382,
872
+ "step": 64500
873
+ },
874
+ {
875
+ "epoch": 10.81,
876
+ "learning_rate": 5.581988576498642e-06,
877
+ "loss": 0.0428,
878
+ "step": 65000
879
+ },
880
+ {
881
+ "epoch": 10.9,
882
+ "learning_rate": 5.471080796317863e-06,
883
+ "loss": 0.0401,
884
+ "step": 65500
885
+ },
886
+ {
887
+ "epoch": 10.98,
888
+ "learning_rate": 5.3601730161370824e-06,
889
+ "loss": 0.0434,
890
+ "step": 66000
891
+ },
892
+ {
893
+ "epoch": 11.0,
894
+ "eval_accuracy": 0.990234202386213,
895
+ "eval_loss": 0.046325162053108215,
896
+ "eval_runtime": 183.7954,
897
+ "eval_samples_per_second": 123.126,
898
+ "eval_steps_per_second": 3.852,
899
+ "step": 66121
900
+ },
901
+ {
902
+ "epoch": 11.06,
903
+ "learning_rate": 5.249265235956302e-06,
904
+ "loss": 0.044,
905
+ "step": 66500
906
+ },
907
+ {
908
+ "epoch": 11.15,
909
+ "learning_rate": 5.138357455775523e-06,
910
+ "loss": 0.0348,
911
+ "step": 67000
912
+ },
913
+ {
914
+ "epoch": 11.23,
915
+ "learning_rate": 5.0274496755947436e-06,
916
+ "loss": 0.0358,
917
+ "step": 67500
918
+ },
919
+ {
920
+ "epoch": 11.31,
921
+ "learning_rate": 4.916541895413964e-06,
922
+ "loss": 0.0356,
923
+ "step": 68000
924
+ },
925
+ {
926
+ "epoch": 11.4,
927
+ "learning_rate": 4.805634115233184e-06,
928
+ "loss": 0.0366,
929
+ "step": 68500
930
+ },
931
+ {
932
+ "epoch": 11.48,
933
+ "learning_rate": 4.694726335052404e-06,
934
+ "loss": 0.0361,
935
+ "step": 69000
936
+ },
937
+ {
938
+ "epoch": 11.56,
939
+ "learning_rate": 4.583818554871624e-06,
940
+ "loss": 0.038,
941
+ "step": 69500
942
+ },
943
+ {
944
+ "epoch": 11.65,
945
+ "learning_rate": 4.4729107746908455e-06,
946
+ "loss": 0.0383,
947
+ "step": 70000
948
+ },
949
+ {
950
+ "epoch": 11.73,
951
+ "learning_rate": 4.362002994510065e-06,
952
+ "loss": 0.0429,
953
+ "step": 70500
954
+ },
955
+ {
956
+ "epoch": 11.81,
957
+ "learning_rate": 4.251095214329285e-06,
958
+ "loss": 0.0341,
959
+ "step": 71000
960
+ },
961
+ {
962
+ "epoch": 11.89,
963
+ "learning_rate": 4.140187434148506e-06,
964
+ "loss": 0.0369,
965
+ "step": 71500
966
+ },
967
+ {
968
+ "epoch": 11.98,
969
+ "learning_rate": 4.029279653967726e-06,
970
+ "loss": 0.0343,
971
+ "step": 72000
972
+ },
973
+ {
974
+ "epoch": 12.0,
975
+ "eval_accuracy": 0.9897923110914715,
976
+ "eval_loss": 0.050745535641908646,
977
+ "eval_runtime": 183.7772,
978
+ "eval_samples_per_second": 123.138,
979
+ "eval_steps_per_second": 3.852,
980
+ "step": 72132
981
+ },
982
+ {
983
+ "epoch": 12.06,
984
+ "learning_rate": 3.9183718737869465e-06,
985
+ "loss": 0.0326,
986
+ "step": 72500
987
+ },
988
+ {
989
+ "epoch": 12.14,
990
+ "learning_rate": 3.807464093606167e-06,
991
+ "loss": 0.0315,
992
+ "step": 73000
993
+ },
994
+ {
995
+ "epoch": 12.23,
996
+ "learning_rate": 3.6965563134253873e-06,
997
+ "loss": 0.0338,
998
+ "step": 73500
999
+ },
1000
+ {
1001
+ "epoch": 12.31,
1002
+ "learning_rate": 3.5856485332446072e-06,
1003
+ "loss": 0.0326,
1004
+ "step": 74000
1005
+ },
1006
+ {
1007
+ "epoch": 12.39,
1008
+ "learning_rate": 3.4747407530638276e-06,
1009
+ "loss": 0.0365,
1010
+ "step": 74500
1011
+ },
1012
+ {
1013
+ "epoch": 12.48,
1014
+ "learning_rate": 3.3638329728830484e-06,
1015
+ "loss": 0.0312,
1016
+ "step": 75000
1017
+ },
1018
+ {
1019
+ "epoch": 12.56,
1020
+ "learning_rate": 3.252925192702268e-06,
1021
+ "loss": 0.0321,
1022
+ "step": 75500
1023
+ },
1024
+ {
1025
+ "epoch": 12.64,
1026
+ "learning_rate": 3.1420174125214888e-06,
1027
+ "loss": 0.0281,
1028
+ "step": 76000
1029
+ },
1030
+ {
1031
+ "epoch": 12.73,
1032
+ "learning_rate": 3.031109632340709e-06,
1033
+ "loss": 0.0399,
1034
+ "step": 76500
1035
+ },
1036
+ {
1037
+ "epoch": 12.81,
1038
+ "learning_rate": 2.920201852159929e-06,
1039
+ "loss": 0.0347,
1040
+ "step": 77000
1041
+ },
1042
+ {
1043
+ "epoch": 12.89,
1044
+ "learning_rate": 2.8092940719791495e-06,
1045
+ "loss": 0.0296,
1046
+ "step": 77500
1047
+ },
1048
+ {
1049
+ "epoch": 12.98,
1050
+ "learning_rate": 2.69838629179837e-06,
1051
+ "loss": 0.0329,
1052
+ "step": 78000
1053
+ },
1054
+ {
1055
+ "epoch": 13.0,
1056
+ "eval_accuracy": 0.991736632788334,
1057
+ "eval_loss": 0.04343385621905327,
1058
+ "eval_runtime": 183.857,
1059
+ "eval_samples_per_second": 123.085,
1060
+ "eval_steps_per_second": 3.851,
1061
+ "step": 78143
1062
+ },
1063
+ {
1064
+ "epoch": 13.06,
1065
+ "learning_rate": 2.5874785116175907e-06,
1066
+ "loss": 0.0345,
1067
+ "step": 78500
1068
+ },
1069
+ {
1070
+ "epoch": 13.14,
1071
+ "learning_rate": 2.47657073143681e-06,
1072
+ "loss": 0.0337,
1073
+ "step": 79000
1074
+ },
1075
+ {
1076
+ "epoch": 13.23,
1077
+ "learning_rate": 2.365662951256031e-06,
1078
+ "loss": 0.0269,
1079
+ "step": 79500
1080
+ },
1081
+ {
1082
+ "epoch": 13.31,
1083
+ "learning_rate": 2.254755171075251e-06,
1084
+ "loss": 0.0318,
1085
+ "step": 80000
1086
+ },
1087
+ {
1088
+ "epoch": 13.39,
1089
+ "learning_rate": 2.1438473908944713e-06,
1090
+ "loss": 0.033,
1091
+ "step": 80500
1092
+ },
1093
+ {
1094
+ "epoch": 13.48,
1095
+ "learning_rate": 2.0329396107136917e-06,
1096
+ "loss": 0.0325,
1097
+ "step": 81000
1098
+ },
1099
+ {
1100
+ "epoch": 13.56,
1101
+ "learning_rate": 1.922031830532912e-06,
1102
+ "loss": 0.0241,
1103
+ "step": 81500
1104
+ },
1105
+ {
1106
+ "epoch": 13.64,
1107
+ "learning_rate": 1.8111240503521325e-06,
1108
+ "loss": 0.0282,
1109
+ "step": 82000
1110
+ },
1111
+ {
1112
+ "epoch": 13.72,
1113
+ "learning_rate": 1.7002162701713526e-06,
1114
+ "loss": 0.027,
1115
+ "step": 82500
1116
+ },
1117
+ {
1118
+ "epoch": 13.81,
1119
+ "learning_rate": 1.589308489990573e-06,
1120
+ "loss": 0.0348,
1121
+ "step": 83000
1122
+ },
1123
+ {
1124
+ "epoch": 13.89,
1125
+ "learning_rate": 1.4784007098097932e-06,
1126
+ "loss": 0.0267,
1127
+ "step": 83500
1128
+ },
1129
+ {
1130
+ "epoch": 13.97,
1131
+ "learning_rate": 1.3674929296290136e-06,
1132
+ "loss": 0.0284,
1133
+ "step": 84000
1134
+ },
1135
+ {
1136
+ "epoch": 14.0,
1137
+ "eval_accuracy": 0.9918692001767565,
1138
+ "eval_loss": 0.04500320181250572,
1139
+ "eval_runtime": 183.85,
1140
+ "eval_samples_per_second": 123.089,
1141
+ "eval_steps_per_second": 3.851,
1142
+ "step": 84154
1143
+ },
1144
+ {
1145
+ "epoch": 14.06,
1146
+ "learning_rate": 1.256585149448234e-06,
1147
+ "loss": 0.0336,
1148
+ "step": 84500
1149
+ },
1150
+ {
1151
+ "epoch": 14.14,
1152
+ "learning_rate": 1.145677369267454e-06,
1153
+ "loss": 0.0285,
1154
+ "step": 85000
1155
+ },
1156
+ {
1157
+ "epoch": 14.22,
1158
+ "learning_rate": 1.0347695890866745e-06,
1159
+ "loss": 0.0265,
1160
+ "step": 85500
1161
+ },
1162
+ {
1163
+ "epoch": 14.31,
1164
+ "learning_rate": 9.238618089058949e-07,
1165
+ "loss": 0.0244,
1166
+ "step": 86000
1167
+ },
1168
+ {
1169
+ "epoch": 14.39,
1170
+ "learning_rate": 8.129540287251151e-07,
1171
+ "loss": 0.0288,
1172
+ "step": 86500
1173
+ },
1174
+ {
1175
+ "epoch": 14.47,
1176
+ "learning_rate": 7.020462485443355e-07,
1177
+ "loss": 0.0272,
1178
+ "step": 87000
1179
+ },
1180
+ {
1181
+ "epoch": 14.56,
1182
+ "learning_rate": 5.911384683635558e-07,
1183
+ "loss": 0.0293,
1184
+ "step": 87500
1185
+ },
1186
+ {
1187
+ "epoch": 14.64,
1188
+ "learning_rate": 4.802306881827761e-07,
1189
+ "loss": 0.0275,
1190
+ "step": 88000
1191
+ },
1192
+ {
1193
+ "epoch": 14.72,
1194
+ "learning_rate": 3.693229080019964e-07,
1195
+ "loss": 0.0246,
1196
+ "step": 88500
1197
+ },
1198
+ {
1199
+ "epoch": 14.81,
1200
+ "learning_rate": 2.5841512782121667e-07,
1201
+ "loss": 0.0301,
1202
+ "step": 89000
1203
+ },
1204
+ {
1205
+ "epoch": 14.89,
1206
+ "learning_rate": 1.47507347640437e-07,
1207
+ "loss": 0.0275,
1208
+ "step": 89500
1209
+ },
1210
+ {
1211
+ "epoch": 14.97,
1212
+ "learning_rate": 3.65995674596573e-08,
1213
+ "loss": 0.0223,
1214
+ "step": 90000
1215
+ },
1216
+ {
1217
+ "epoch": 15.0,
1218
+ "eval_accuracy": 0.9916482545293858,
1219
+ "eval_loss": 0.04581888020038605,
1220
+ "eval_runtime": 183.9121,
1221
+ "eval_samples_per_second": 123.048,
1222
+ "eval_steps_per_second": 3.85,
1223
+ "step": 90165
1224
+ }
1225
+ ],
1226
+ "max_steps": 90165,
1227
+ "num_train_epochs": 15,
1228
+ "total_flos": 8.488169482669056e+16,
1229
+ "trial_name": null,
1230
+ "trial_params": null
1231
+ }