Shivagowri commited on
Commit
dae6a44
·
1 Parent(s): c053326

🍻 cheers

Browse files
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_accuracy": 0.9392670157068063,
4
+ "eval_loss": 0.27537932991981506,
5
+ "eval_runtime": 14.5416,
6
+ "eval_samples_per_second": 65.674,
7
+ "eval_steps_per_second": 8.252,
8
+ "total_flos": 1.874833643725701e+18,
9
+ "train_loss": 0.2613857026712926,
10
+ "train_runtime": 1218.6852,
11
+ "train_samples_per_second": 19.849,
12
+ "train_steps_per_second": 1.243
13
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_accuracy": 0.9392670157068063,
4
+ "eval_loss": 0.27537932991981506,
5
+ "eval_runtime": 14.5416,
6
+ "eval_samples_per_second": 65.674,
7
+ "eval_steps_per_second": 8.252
8
+ }
runs/Jun30_06-08-28_fbcba3ddf03a/events.out.tfevents.1656570869.fbcba3ddf03a.74.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:978035f66abf3fc6e6e6e9c6ed511321ce64c47f4b0db5444f46c155244c3a10
3
+ size 363
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "total_flos": 1.874833643725701e+18,
4
+ "train_loss": 0.2613857026712926,
5
+ "train_runtime": 1218.6852,
6
+ "train_samples_per_second": 19.849,
7
+ "train_steps_per_second": 1.243
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1066 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.27537932991981506,
3
+ "best_model_checkpoint": "vit-snacks/checkpoint-1300",
4
+ "epoch": 5.0,
5
+ "global_step": 1515,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 0.0001986798679867987,
13
+ "loss": 2.8877,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.07,
18
+ "learning_rate": 0.00019735973597359737,
19
+ "loss": 2.6212,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.1,
24
+ "learning_rate": 0.00019603960396039606,
25
+ "loss": 2.3151,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.13,
30
+ "learning_rate": 0.00019471947194719472,
31
+ "loss": 1.9591,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.17,
36
+ "learning_rate": 0.0001933993399339934,
37
+ "loss": 1.7106,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.2,
42
+ "learning_rate": 0.00019207920792079208,
43
+ "loss": 1.367,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.23,
48
+ "learning_rate": 0.00019075907590759077,
49
+ "loss": 1.2285,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.26,
54
+ "learning_rate": 0.00018943894389438946,
55
+ "loss": 1.0703,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.3,
60
+ "learning_rate": 0.00018811881188118812,
61
+ "loss": 1.1937,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.33,
66
+ "learning_rate": 0.00018679867986798681,
67
+ "loss": 0.8724,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.33,
72
+ "eval_accuracy": 0.8670157068062827,
73
+ "eval_loss": 0.9118377566337585,
74
+ "eval_runtime": 14.6857,
75
+ "eval_samples_per_second": 65.029,
76
+ "eval_steps_per_second": 8.171,
77
+ "step": 100
78
+ },
79
+ {
80
+ "epoch": 0.36,
81
+ "learning_rate": 0.00018547854785478548,
82
+ "loss": 0.9309,
83
+ "step": 110
84
+ },
85
+ {
86
+ "epoch": 0.4,
87
+ "learning_rate": 0.00018415841584158417,
88
+ "loss": 0.8163,
89
+ "step": 120
90
+ },
91
+ {
92
+ "epoch": 0.43,
93
+ "learning_rate": 0.00018283828382838283,
94
+ "loss": 0.7416,
95
+ "step": 130
96
+ },
97
+ {
98
+ "epoch": 0.46,
99
+ "learning_rate": 0.00018151815181518153,
100
+ "loss": 0.6442,
101
+ "step": 140
102
+ },
103
+ {
104
+ "epoch": 0.5,
105
+ "learning_rate": 0.00018019801980198022,
106
+ "loss": 0.7906,
107
+ "step": 150
108
+ },
109
+ {
110
+ "epoch": 0.53,
111
+ "learning_rate": 0.0001788778877887789,
112
+ "loss": 0.6137,
113
+ "step": 160
114
+ },
115
+ {
116
+ "epoch": 0.56,
117
+ "learning_rate": 0.00017755775577557757,
118
+ "loss": 0.6342,
119
+ "step": 170
120
+ },
121
+ {
122
+ "epoch": 0.59,
123
+ "learning_rate": 0.00017623762376237624,
124
+ "loss": 0.7081,
125
+ "step": 180
126
+ },
127
+ {
128
+ "epoch": 0.63,
129
+ "learning_rate": 0.00017491749174917493,
130
+ "loss": 0.51,
131
+ "step": 190
132
+ },
133
+ {
134
+ "epoch": 0.66,
135
+ "learning_rate": 0.0001735973597359736,
136
+ "loss": 0.5628,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.66,
141
+ "eval_accuracy": 0.8471204188481676,
142
+ "eval_loss": 0.6873391270637512,
143
+ "eval_runtime": 16.1319,
144
+ "eval_samples_per_second": 59.2,
145
+ "eval_steps_per_second": 7.439,
146
+ "step": 200
147
+ },
148
+ {
149
+ "epoch": 0.69,
150
+ "learning_rate": 0.00017227722772277228,
151
+ "loss": 0.6635,
152
+ "step": 210
153
+ },
154
+ {
155
+ "epoch": 0.73,
156
+ "learning_rate": 0.00017095709570957098,
157
+ "loss": 0.51,
158
+ "step": 220
159
+ },
160
+ {
161
+ "epoch": 0.76,
162
+ "learning_rate": 0.00016963696369636967,
163
+ "loss": 0.5002,
164
+ "step": 230
165
+ },
166
+ {
167
+ "epoch": 0.79,
168
+ "learning_rate": 0.00016831683168316833,
169
+ "loss": 0.6353,
170
+ "step": 240
171
+ },
172
+ {
173
+ "epoch": 0.83,
174
+ "learning_rate": 0.000166996699669967,
175
+ "loss": 0.5326,
176
+ "step": 250
177
+ },
178
+ {
179
+ "epoch": 0.86,
180
+ "learning_rate": 0.00016567656765676569,
181
+ "loss": 0.5051,
182
+ "step": 260
183
+ },
184
+ {
185
+ "epoch": 0.89,
186
+ "learning_rate": 0.00016435643564356435,
187
+ "loss": 0.6527,
188
+ "step": 270
189
+ },
190
+ {
191
+ "epoch": 0.92,
192
+ "learning_rate": 0.00016303630363036304,
193
+ "loss": 0.4961,
194
+ "step": 280
195
+ },
196
+ {
197
+ "epoch": 0.96,
198
+ "learning_rate": 0.00016171617161716173,
199
+ "loss": 0.4981,
200
+ "step": 290
201
+ },
202
+ {
203
+ "epoch": 0.99,
204
+ "learning_rate": 0.00016039603960396042,
205
+ "loss": 0.4421,
206
+ "step": 300
207
+ },
208
+ {
209
+ "epoch": 0.99,
210
+ "eval_accuracy": 0.8691099476439791,
211
+ "eval_loss": 0.4995421767234802,
212
+ "eval_runtime": 15.1178,
213
+ "eval_samples_per_second": 63.17,
214
+ "eval_steps_per_second": 7.938,
215
+ "step": 300
216
+ },
217
+ {
218
+ "epoch": 1.02,
219
+ "learning_rate": 0.0001590759075907591,
220
+ "loss": 0.4434,
221
+ "step": 310
222
+ },
223
+ {
224
+ "epoch": 1.06,
225
+ "learning_rate": 0.00015775577557755775,
226
+ "loss": 0.2569,
227
+ "step": 320
228
+ },
229
+ {
230
+ "epoch": 1.09,
231
+ "learning_rate": 0.00015643564356435644,
232
+ "loss": 0.2208,
233
+ "step": 330
234
+ },
235
+ {
236
+ "epoch": 1.12,
237
+ "learning_rate": 0.0001551155115511551,
238
+ "loss": 0.3017,
239
+ "step": 340
240
+ },
241
+ {
242
+ "epoch": 1.16,
243
+ "learning_rate": 0.0001537953795379538,
244
+ "loss": 0.2288,
245
+ "step": 350
246
+ },
247
+ {
248
+ "epoch": 1.19,
249
+ "learning_rate": 0.0001524752475247525,
250
+ "loss": 0.1533,
251
+ "step": 360
252
+ },
253
+ {
254
+ "epoch": 1.22,
255
+ "learning_rate": 0.00015115511551155118,
256
+ "loss": 0.3527,
257
+ "step": 370
258
+ },
259
+ {
260
+ "epoch": 1.25,
261
+ "learning_rate": 0.00014983498349834985,
262
+ "loss": 0.1966,
263
+ "step": 380
264
+ },
265
+ {
266
+ "epoch": 1.29,
267
+ "learning_rate": 0.0001485148514851485,
268
+ "loss": 0.1962,
269
+ "step": 390
270
+ },
271
+ {
272
+ "epoch": 1.32,
273
+ "learning_rate": 0.0001471947194719472,
274
+ "loss": 0.2837,
275
+ "step": 400
276
+ },
277
+ {
278
+ "epoch": 1.32,
279
+ "eval_accuracy": 0.9026178010471204,
280
+ "eval_loss": 0.40083953738212585,
281
+ "eval_runtime": 14.9253,
282
+ "eval_samples_per_second": 63.985,
283
+ "eval_steps_per_second": 8.04,
284
+ "step": 400
285
+ },
286
+ {
287
+ "epoch": 1.35,
288
+ "learning_rate": 0.00014587458745874587,
289
+ "loss": 0.1951,
290
+ "step": 410
291
+ },
292
+ {
293
+ "epoch": 1.39,
294
+ "learning_rate": 0.00014455445544554456,
295
+ "loss": 0.2244,
296
+ "step": 420
297
+ },
298
+ {
299
+ "epoch": 1.42,
300
+ "learning_rate": 0.00014323432343234325,
301
+ "loss": 0.144,
302
+ "step": 430
303
+ },
304
+ {
305
+ "epoch": 1.45,
306
+ "learning_rate": 0.00014191419141914194,
307
+ "loss": 0.1707,
308
+ "step": 440
309
+ },
310
+ {
311
+ "epoch": 1.49,
312
+ "learning_rate": 0.0001405940594059406,
313
+ "loss": 0.1881,
314
+ "step": 450
315
+ },
316
+ {
317
+ "epoch": 1.52,
318
+ "learning_rate": 0.00013927392739273927,
319
+ "loss": 0.2441,
320
+ "step": 460
321
+ },
322
+ {
323
+ "epoch": 1.55,
324
+ "learning_rate": 0.00013795379537953796,
325
+ "loss": 0.2128,
326
+ "step": 470
327
+ },
328
+ {
329
+ "epoch": 1.58,
330
+ "learning_rate": 0.00013663366336633665,
331
+ "loss": 0.2773,
332
+ "step": 480
333
+ },
334
+ {
335
+ "epoch": 1.62,
336
+ "learning_rate": 0.00013531353135313532,
337
+ "loss": 0.2554,
338
+ "step": 490
339
+ },
340
+ {
341
+ "epoch": 1.65,
342
+ "learning_rate": 0.000133993399339934,
343
+ "loss": 0.1645,
344
+ "step": 500
345
+ },
346
+ {
347
+ "epoch": 1.65,
348
+ "eval_accuracy": 0.9057591623036649,
349
+ "eval_loss": 0.3701848089694977,
350
+ "eval_runtime": 14.4771,
351
+ "eval_samples_per_second": 65.966,
352
+ "eval_steps_per_second": 8.289,
353
+ "step": 500
354
+ },
355
+ {
356
+ "epoch": 1.68,
357
+ "learning_rate": 0.0001326732673267327,
358
+ "loss": 0.2479,
359
+ "step": 510
360
+ },
361
+ {
362
+ "epoch": 1.72,
363
+ "learning_rate": 0.00013135313531353136,
364
+ "loss": 0.1427,
365
+ "step": 520
366
+ },
367
+ {
368
+ "epoch": 1.75,
369
+ "learning_rate": 0.00013003300330033003,
370
+ "loss": 0.234,
371
+ "step": 530
372
+ },
373
+ {
374
+ "epoch": 1.78,
375
+ "learning_rate": 0.00012871287128712872,
376
+ "loss": 0.2,
377
+ "step": 540
378
+ },
379
+ {
380
+ "epoch": 1.82,
381
+ "learning_rate": 0.0001273927392739274,
382
+ "loss": 0.3339,
383
+ "step": 550
384
+ },
385
+ {
386
+ "epoch": 1.85,
387
+ "learning_rate": 0.00012607260726072607,
388
+ "loss": 0.1089,
389
+ "step": 560
390
+ },
391
+ {
392
+ "epoch": 1.88,
393
+ "learning_rate": 0.00012475247524752477,
394
+ "loss": 0.2647,
395
+ "step": 570
396
+ },
397
+ {
398
+ "epoch": 1.91,
399
+ "learning_rate": 0.00012343234323432346,
400
+ "loss": 0.1648,
401
+ "step": 580
402
+ },
403
+ {
404
+ "epoch": 1.95,
405
+ "learning_rate": 0.00012211221122112212,
406
+ "loss": 0.1728,
407
+ "step": 590
408
+ },
409
+ {
410
+ "epoch": 1.98,
411
+ "learning_rate": 0.0001207920792079208,
412
+ "loss": 0.1604,
413
+ "step": 600
414
+ },
415
+ {
416
+ "epoch": 1.98,
417
+ "eval_accuracy": 0.8921465968586387,
418
+ "eval_loss": 0.3981299102306366,
419
+ "eval_runtime": 14.3841,
420
+ "eval_samples_per_second": 66.393,
421
+ "eval_steps_per_second": 8.343,
422
+ "step": 600
423
+ },
424
+ {
425
+ "epoch": 2.01,
426
+ "learning_rate": 0.00011947194719471948,
427
+ "loss": 0.0877,
428
+ "step": 610
429
+ },
430
+ {
431
+ "epoch": 2.05,
432
+ "learning_rate": 0.00011815181518151817,
433
+ "loss": 0.119,
434
+ "step": 620
435
+ },
436
+ {
437
+ "epoch": 2.08,
438
+ "learning_rate": 0.00011683168316831683,
439
+ "loss": 0.0758,
440
+ "step": 630
441
+ },
442
+ {
443
+ "epoch": 2.11,
444
+ "learning_rate": 0.00011551155115511551,
445
+ "loss": 0.1496,
446
+ "step": 640
447
+ },
448
+ {
449
+ "epoch": 2.15,
450
+ "learning_rate": 0.0001141914191419142,
451
+ "loss": 0.0617,
452
+ "step": 650
453
+ },
454
+ {
455
+ "epoch": 2.18,
456
+ "learning_rate": 0.00011287128712871287,
457
+ "loss": 0.0627,
458
+ "step": 660
459
+ },
460
+ {
461
+ "epoch": 2.21,
462
+ "learning_rate": 0.00011155115511551156,
463
+ "loss": 0.0948,
464
+ "step": 670
465
+ },
466
+ {
467
+ "epoch": 2.24,
468
+ "learning_rate": 0.00011023102310231023,
469
+ "loss": 0.0987,
470
+ "step": 680
471
+ },
472
+ {
473
+ "epoch": 2.28,
474
+ "learning_rate": 0.00010891089108910893,
475
+ "loss": 0.0473,
476
+ "step": 690
477
+ },
478
+ {
479
+ "epoch": 2.31,
480
+ "learning_rate": 0.00010759075907590759,
481
+ "loss": 0.0498,
482
+ "step": 700
483
+ },
484
+ {
485
+ "epoch": 2.31,
486
+ "eval_accuracy": 0.9204188481675393,
487
+ "eval_loss": 0.31845271587371826,
488
+ "eval_runtime": 14.4624,
489
+ "eval_samples_per_second": 66.033,
490
+ "eval_steps_per_second": 8.297,
491
+ "step": 700
492
+ },
493
+ {
494
+ "epoch": 2.34,
495
+ "learning_rate": 0.00010627062706270627,
496
+ "loss": 0.0428,
497
+ "step": 710
498
+ },
499
+ {
500
+ "epoch": 2.38,
501
+ "learning_rate": 0.00010495049504950496,
502
+ "loss": 0.0848,
503
+ "step": 720
504
+ },
505
+ {
506
+ "epoch": 2.41,
507
+ "learning_rate": 0.00010363036303630365,
508
+ "loss": 0.0512,
509
+ "step": 730
510
+ },
511
+ {
512
+ "epoch": 2.44,
513
+ "learning_rate": 0.00010231023102310232,
514
+ "loss": 0.043,
515
+ "step": 740
516
+ },
517
+ {
518
+ "epoch": 2.48,
519
+ "learning_rate": 0.00010099009900990099,
520
+ "loss": 0.0372,
521
+ "step": 750
522
+ },
523
+ {
524
+ "epoch": 2.51,
525
+ "learning_rate": 9.966996699669967e-05,
526
+ "loss": 0.1442,
527
+ "step": 760
528
+ },
529
+ {
530
+ "epoch": 2.54,
531
+ "learning_rate": 9.834983498349836e-05,
532
+ "loss": 0.0426,
533
+ "step": 770
534
+ },
535
+ {
536
+ "epoch": 2.57,
537
+ "learning_rate": 9.702970297029703e-05,
538
+ "loss": 0.0511,
539
+ "step": 780
540
+ },
541
+ {
542
+ "epoch": 2.61,
543
+ "learning_rate": 9.570957095709572e-05,
544
+ "loss": 0.0634,
545
+ "step": 790
546
+ },
547
+ {
548
+ "epoch": 2.64,
549
+ "learning_rate": 9.43894389438944e-05,
550
+ "loss": 0.0406,
551
+ "step": 800
552
+ },
553
+ {
554
+ "epoch": 2.64,
555
+ "eval_accuracy": 0.9141361256544502,
556
+ "eval_loss": 0.3426617980003357,
557
+ "eval_runtime": 14.6543,
558
+ "eval_samples_per_second": 65.168,
559
+ "eval_steps_per_second": 8.189,
560
+ "step": 800
561
+ },
562
+ {
563
+ "epoch": 2.67,
564
+ "learning_rate": 9.306930693069307e-05,
565
+ "loss": 0.1104,
566
+ "step": 810
567
+ },
568
+ {
569
+ "epoch": 2.71,
570
+ "learning_rate": 9.174917491749175e-05,
571
+ "loss": 0.0778,
572
+ "step": 820
573
+ },
574
+ {
575
+ "epoch": 2.74,
576
+ "learning_rate": 9.042904290429043e-05,
577
+ "loss": 0.1189,
578
+ "step": 830
579
+ },
580
+ {
581
+ "epoch": 2.77,
582
+ "learning_rate": 8.910891089108912e-05,
583
+ "loss": 0.0873,
584
+ "step": 840
585
+ },
586
+ {
587
+ "epoch": 2.81,
588
+ "learning_rate": 8.778877887788778e-05,
589
+ "loss": 0.0414,
590
+ "step": 850
591
+ },
592
+ {
593
+ "epoch": 2.84,
594
+ "learning_rate": 8.646864686468648e-05,
595
+ "loss": 0.0723,
596
+ "step": 860
597
+ },
598
+ {
599
+ "epoch": 2.87,
600
+ "learning_rate": 8.514851485148515e-05,
601
+ "loss": 0.0495,
602
+ "step": 870
603
+ },
604
+ {
605
+ "epoch": 2.9,
606
+ "learning_rate": 8.382838283828383e-05,
607
+ "loss": 0.0698,
608
+ "step": 880
609
+ },
610
+ {
611
+ "epoch": 2.94,
612
+ "learning_rate": 8.250825082508251e-05,
613
+ "loss": 0.0636,
614
+ "step": 890
615
+ },
616
+ {
617
+ "epoch": 2.97,
618
+ "learning_rate": 8.11881188118812e-05,
619
+ "loss": 0.1049,
620
+ "step": 900
621
+ },
622
+ {
623
+ "epoch": 2.97,
624
+ "eval_accuracy": 0.9172774869109948,
625
+ "eval_loss": 0.344361275434494,
626
+ "eval_runtime": 14.443,
627
+ "eval_samples_per_second": 66.122,
628
+ "eval_steps_per_second": 8.309,
629
+ "step": 900
630
+ },
631
+ {
632
+ "epoch": 3.0,
633
+ "learning_rate": 7.986798679867988e-05,
634
+ "loss": 0.0148,
635
+ "step": 910
636
+ },
637
+ {
638
+ "epoch": 3.04,
639
+ "learning_rate": 7.854785478547854e-05,
640
+ "loss": 0.0134,
641
+ "step": 920
642
+ },
643
+ {
644
+ "epoch": 3.07,
645
+ "learning_rate": 7.722772277227723e-05,
646
+ "loss": 0.0263,
647
+ "step": 930
648
+ },
649
+ {
650
+ "epoch": 3.1,
651
+ "learning_rate": 7.590759075907591e-05,
652
+ "loss": 0.0453,
653
+ "step": 940
654
+ },
655
+ {
656
+ "epoch": 3.14,
657
+ "learning_rate": 7.458745874587459e-05,
658
+ "loss": 0.0156,
659
+ "step": 950
660
+ },
661
+ {
662
+ "epoch": 3.17,
663
+ "learning_rate": 7.326732673267327e-05,
664
+ "loss": 0.0125,
665
+ "step": 960
666
+ },
667
+ {
668
+ "epoch": 3.2,
669
+ "learning_rate": 7.194719471947196e-05,
670
+ "loss": 0.0118,
671
+ "step": 970
672
+ },
673
+ {
674
+ "epoch": 3.23,
675
+ "learning_rate": 7.062706270627064e-05,
676
+ "loss": 0.0116,
677
+ "step": 980
678
+ },
679
+ {
680
+ "epoch": 3.27,
681
+ "learning_rate": 6.93069306930693e-05,
682
+ "loss": 0.011,
683
+ "step": 990
684
+ },
685
+ {
686
+ "epoch": 3.3,
687
+ "learning_rate": 6.798679867986799e-05,
688
+ "loss": 0.0272,
689
+ "step": 1000
690
+ },
691
+ {
692
+ "epoch": 3.3,
693
+ "eval_accuracy": 0.9246073298429319,
694
+ "eval_loss": 0.31677740812301636,
695
+ "eval_runtime": 14.7709,
696
+ "eval_samples_per_second": 64.654,
697
+ "eval_steps_per_second": 8.124,
698
+ "step": 1000
699
+ },
700
+ {
701
+ "epoch": 3.33,
702
+ "learning_rate": 6.666666666666667e-05,
703
+ "loss": 0.0113,
704
+ "step": 1010
705
+ },
706
+ {
707
+ "epoch": 3.37,
708
+ "learning_rate": 6.534653465346535e-05,
709
+ "loss": 0.0287,
710
+ "step": 1020
711
+ },
712
+ {
713
+ "epoch": 3.4,
714
+ "learning_rate": 6.402640264026403e-05,
715
+ "loss": 0.0326,
716
+ "step": 1030
717
+ },
718
+ {
719
+ "epoch": 3.43,
720
+ "learning_rate": 6.270627062706272e-05,
721
+ "loss": 0.0157,
722
+ "step": 1040
723
+ },
724
+ {
725
+ "epoch": 3.47,
726
+ "learning_rate": 6.13861386138614e-05,
727
+ "loss": 0.0135,
728
+ "step": 1050
729
+ },
730
+ {
731
+ "epoch": 3.5,
732
+ "learning_rate": 6.006600660066007e-05,
733
+ "loss": 0.0154,
734
+ "step": 1060
735
+ },
736
+ {
737
+ "epoch": 3.53,
738
+ "learning_rate": 5.874587458745875e-05,
739
+ "loss": 0.0357,
740
+ "step": 1070
741
+ },
742
+ {
743
+ "epoch": 3.56,
744
+ "learning_rate": 5.742574257425742e-05,
745
+ "loss": 0.0161,
746
+ "step": 1080
747
+ },
748
+ {
749
+ "epoch": 3.6,
750
+ "learning_rate": 5.610561056105611e-05,
751
+ "loss": 0.04,
752
+ "step": 1090
753
+ },
754
+ {
755
+ "epoch": 3.63,
756
+ "learning_rate": 5.4785478547854784e-05,
757
+ "loss": 0.0186,
758
+ "step": 1100
759
+ },
760
+ {
761
+ "epoch": 3.63,
762
+ "eval_accuracy": 0.9287958115183246,
763
+ "eval_loss": 0.3141845762729645,
764
+ "eval_runtime": 14.3909,
765
+ "eval_samples_per_second": 66.361,
766
+ "eval_steps_per_second": 8.339,
767
+ "step": 1100
768
+ },
769
+ {
770
+ "epoch": 3.66,
771
+ "learning_rate": 5.346534653465347e-05,
772
+ "loss": 0.0119,
773
+ "step": 1110
774
+ },
775
+ {
776
+ "epoch": 3.7,
777
+ "learning_rate": 5.2145214521452146e-05,
778
+ "loss": 0.0089,
779
+ "step": 1120
780
+ },
781
+ {
782
+ "epoch": 3.73,
783
+ "learning_rate": 5.082508250825083e-05,
784
+ "loss": 0.0097,
785
+ "step": 1130
786
+ },
787
+ {
788
+ "epoch": 3.76,
789
+ "learning_rate": 4.950495049504951e-05,
790
+ "loss": 0.0561,
791
+ "step": 1140
792
+ },
793
+ {
794
+ "epoch": 3.8,
795
+ "learning_rate": 4.8184818481848186e-05,
796
+ "loss": 0.0457,
797
+ "step": 1150
798
+ },
799
+ {
800
+ "epoch": 3.83,
801
+ "learning_rate": 4.686468646864687e-05,
802
+ "loss": 0.0129,
803
+ "step": 1160
804
+ },
805
+ {
806
+ "epoch": 3.86,
807
+ "learning_rate": 4.554455445544555e-05,
808
+ "loss": 0.0657,
809
+ "step": 1170
810
+ },
811
+ {
812
+ "epoch": 3.89,
813
+ "learning_rate": 4.4224422442244226e-05,
814
+ "loss": 0.0081,
815
+ "step": 1180
816
+ },
817
+ {
818
+ "epoch": 3.93,
819
+ "learning_rate": 4.2904290429042904e-05,
820
+ "loss": 0.0084,
821
+ "step": 1190
822
+ },
823
+ {
824
+ "epoch": 3.96,
825
+ "learning_rate": 4.158415841584158e-05,
826
+ "loss": 0.0203,
827
+ "step": 1200
828
+ },
829
+ {
830
+ "epoch": 3.96,
831
+ "eval_accuracy": 0.9298429319371728,
832
+ "eval_loss": 0.29309454560279846,
833
+ "eval_runtime": 14.4376,
834
+ "eval_samples_per_second": 66.147,
835
+ "eval_steps_per_second": 8.312,
836
+ "step": 1200
837
+ },
838
+ {
839
+ "epoch": 3.99,
840
+ "learning_rate": 4.0264026402640266e-05,
841
+ "loss": 0.0459,
842
+ "step": 1210
843
+ },
844
+ {
845
+ "epoch": 4.03,
846
+ "learning_rate": 3.8943894389438944e-05,
847
+ "loss": 0.008,
848
+ "step": 1220
849
+ },
850
+ {
851
+ "epoch": 4.06,
852
+ "learning_rate": 3.762376237623763e-05,
853
+ "loss": 0.0078,
854
+ "step": 1230
855
+ },
856
+ {
857
+ "epoch": 4.09,
858
+ "learning_rate": 3.6303630363036307e-05,
859
+ "loss": 0.0079,
860
+ "step": 1240
861
+ },
862
+ {
863
+ "epoch": 4.13,
864
+ "learning_rate": 3.4983498349834984e-05,
865
+ "loss": 0.0087,
866
+ "step": 1250
867
+ },
868
+ {
869
+ "epoch": 4.16,
870
+ "learning_rate": 3.366336633663367e-05,
871
+ "loss": 0.008,
872
+ "step": 1260
873
+ },
874
+ {
875
+ "epoch": 4.19,
876
+ "learning_rate": 3.234323432343234e-05,
877
+ "loss": 0.0074,
878
+ "step": 1270
879
+ },
880
+ {
881
+ "epoch": 4.22,
882
+ "learning_rate": 3.1023102310231024e-05,
883
+ "loss": 0.0078,
884
+ "step": 1280
885
+ },
886
+ {
887
+ "epoch": 4.26,
888
+ "learning_rate": 2.9702970297029702e-05,
889
+ "loss": 0.0076,
890
+ "step": 1290
891
+ },
892
+ {
893
+ "epoch": 4.29,
894
+ "learning_rate": 2.8382838283828383e-05,
895
+ "loss": 0.007,
896
+ "step": 1300
897
+ },
898
+ {
899
+ "epoch": 4.29,
900
+ "eval_accuracy": 0.9392670157068063,
901
+ "eval_loss": 0.27537932991981506,
902
+ "eval_runtime": 14.7619,
903
+ "eval_samples_per_second": 64.694,
904
+ "eval_steps_per_second": 8.129,
905
+ "step": 1300
906
+ },
907
+ {
908
+ "epoch": 4.32,
909
+ "learning_rate": 2.7062706270627065e-05,
910
+ "loss": 0.0073,
911
+ "step": 1310
912
+ },
913
+ {
914
+ "epoch": 4.36,
915
+ "learning_rate": 2.5742574257425746e-05,
916
+ "loss": 0.007,
917
+ "step": 1320
918
+ },
919
+ {
920
+ "epoch": 4.39,
921
+ "learning_rate": 2.4422442244224424e-05,
922
+ "loss": 0.0071,
923
+ "step": 1330
924
+ },
925
+ {
926
+ "epoch": 4.42,
927
+ "learning_rate": 2.31023102310231e-05,
928
+ "loss": 0.0068,
929
+ "step": 1340
930
+ },
931
+ {
932
+ "epoch": 4.46,
933
+ "learning_rate": 2.1782178217821783e-05,
934
+ "loss": 0.007,
935
+ "step": 1350
936
+ },
937
+ {
938
+ "epoch": 4.49,
939
+ "learning_rate": 2.0462046204620464e-05,
940
+ "loss": 0.0068,
941
+ "step": 1360
942
+ },
943
+ {
944
+ "epoch": 4.52,
945
+ "learning_rate": 1.9141914191419145e-05,
946
+ "loss": 0.0083,
947
+ "step": 1370
948
+ },
949
+ {
950
+ "epoch": 4.55,
951
+ "learning_rate": 1.7821782178217823e-05,
952
+ "loss": 0.0167,
953
+ "step": 1380
954
+ },
955
+ {
956
+ "epoch": 4.59,
957
+ "learning_rate": 1.6501650165016504e-05,
958
+ "loss": 0.0069,
959
+ "step": 1390
960
+ },
961
+ {
962
+ "epoch": 4.62,
963
+ "learning_rate": 1.5181518151815183e-05,
964
+ "loss": 0.0072,
965
+ "step": 1400
966
+ },
967
+ {
968
+ "epoch": 4.62,
969
+ "eval_accuracy": 0.9403141361256544,
970
+ "eval_loss": 0.27776801586151123,
971
+ "eval_runtime": 14.4524,
972
+ "eval_samples_per_second": 66.079,
973
+ "eval_steps_per_second": 8.303,
974
+ "step": 1400
975
+ },
976
+ {
977
+ "epoch": 4.65,
978
+ "learning_rate": 1.3861386138613863e-05,
979
+ "loss": 0.0069,
980
+ "step": 1410
981
+ },
982
+ {
983
+ "epoch": 4.69,
984
+ "learning_rate": 1.254125412541254e-05,
985
+ "loss": 0.0067,
986
+ "step": 1420
987
+ },
988
+ {
989
+ "epoch": 4.72,
990
+ "learning_rate": 1.1221122112211222e-05,
991
+ "loss": 0.0066,
992
+ "step": 1430
993
+ },
994
+ {
995
+ "epoch": 4.75,
996
+ "learning_rate": 9.900990099009901e-06,
997
+ "loss": 0.0065,
998
+ "step": 1440
999
+ },
1000
+ {
1001
+ "epoch": 4.79,
1002
+ "learning_rate": 8.58085808580858e-06,
1003
+ "loss": 0.0342,
1004
+ "step": 1450
1005
+ },
1006
+ {
1007
+ "epoch": 4.82,
1008
+ "learning_rate": 7.260726072607261e-06,
1009
+ "loss": 0.0068,
1010
+ "step": 1460
1011
+ },
1012
+ {
1013
+ "epoch": 4.85,
1014
+ "learning_rate": 5.940594059405941e-06,
1015
+ "loss": 0.0075,
1016
+ "step": 1470
1017
+ },
1018
+ {
1019
+ "epoch": 4.88,
1020
+ "learning_rate": 4.62046204620462e-06,
1021
+ "loss": 0.0067,
1022
+ "step": 1480
1023
+ },
1024
+ {
1025
+ "epoch": 4.92,
1026
+ "learning_rate": 3.3003300330033e-06,
1027
+ "loss": 0.0066,
1028
+ "step": 1490
1029
+ },
1030
+ {
1031
+ "epoch": 4.95,
1032
+ "learning_rate": 1.9801980198019803e-06,
1033
+ "loss": 0.0073,
1034
+ "step": 1500
1035
+ },
1036
+ {
1037
+ "epoch": 4.95,
1038
+ "eval_accuracy": 0.9392670157068063,
1039
+ "eval_loss": 0.2782096564769745,
1040
+ "eval_runtime": 14.5029,
1041
+ "eval_samples_per_second": 65.849,
1042
+ "eval_steps_per_second": 8.274,
1043
+ "step": 1500
1044
+ },
1045
+ {
1046
+ "epoch": 4.98,
1047
+ "learning_rate": 6.600660066006602e-07,
1048
+ "loss": 0.0408,
1049
+ "step": 1510
1050
+ },
1051
+ {
1052
+ "epoch": 5.0,
1053
+ "step": 1515,
1054
+ "total_flos": 1.874833643725701e+18,
1055
+ "train_loss": 0.2613857026712926,
1056
+ "train_runtime": 1218.6852,
1057
+ "train_samples_per_second": 19.849,
1058
+ "train_steps_per_second": 1.243
1059
+ }
1060
+ ],
1061
+ "max_steps": 1515,
1062
+ "num_train_epochs": 5,
1063
+ "total_flos": 1.874833643725701e+18,
1064
+ "trial_name": null,
1065
+ "trial_params": null
1066
+ }