loucad commited on
Commit
742c1b5
1 Parent(s): fb7e03b

End of training

Browse files
Files changed (4) hide show
  1. all_results.json +13 -0
  2. eval_results.json +8 -0
  3. train_results.json +8 -0
  4. trainer_state.json +1255 -0
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.9507407407407408,
4
+ "eval_loss": 0.1926022469997406,
5
+ "eval_runtime": 6.9176,
6
+ "eval_samples_per_second": 390.311,
7
+ "eval_steps_per_second": 12.288,
8
+ "total_flos": 3.46193285935104e+17,
9
+ "train_loss": 0.7228629551435771,
10
+ "train_runtime": 1251.8481,
11
+ "train_samples_per_second": 194.113,
12
+ "train_steps_per_second": 1.518
13
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.9507407407407408,
4
+ "eval_loss": 0.1926022469997406,
5
+ "eval_runtime": 6.9176,
6
+ "eval_samples_per_second": 390.311,
7
+ "eval_steps_per_second": 12.288
8
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "total_flos": 3.46193285935104e+17,
4
+ "train_loss": 0.7228629551435771,
5
+ "train_runtime": 1251.8481,
6
+ "train_samples_per_second": 194.113,
7
+ "train_steps_per_second": 1.518
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1255 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9507407407407408,
3
+ "best_model_checkpoint": "mobilevit-xx-small-finetuned-eurosat/checkpoint-1520",
4
+ "epoch": 10.0,
5
+ "global_step": 1900,
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": 2.631578947368421e-06,
13
+ "loss": 2.3276,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.11,
18
+ "learning_rate": 5.263157894736842e-06,
19
+ "loss": 2.3289,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.16,
24
+ "learning_rate": 7.894736842105263e-06,
25
+ "loss": 2.3152,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.21,
30
+ "learning_rate": 1.0526315789473684e-05,
31
+ "loss": 2.3021,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.26,
36
+ "learning_rate": 1.3157894736842106e-05,
37
+ "loss": 2.2939,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.32,
42
+ "learning_rate": 1.5789473684210526e-05,
43
+ "loss": 2.2682,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.37,
48
+ "learning_rate": 1.8421052631578947e-05,
49
+ "loss": 2.2401,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.42,
54
+ "learning_rate": 2.105263157894737e-05,
55
+ "loss": 2.2106,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.47,
60
+ "learning_rate": 2.368421052631579e-05,
61
+ "loss": 2.1706,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.53,
66
+ "learning_rate": 2.6315789473684212e-05,
67
+ "loss": 2.1411,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.58,
72
+ "learning_rate": 2.8947368421052634e-05,
73
+ "loss": 2.0986,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.63,
78
+ "learning_rate": 3.157894736842105e-05,
79
+ "loss": 2.0471,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.68,
84
+ "learning_rate": 3.421052631578947e-05,
85
+ "loss": 2.0002,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.74,
90
+ "learning_rate": 3.6842105263157895e-05,
91
+ "loss": 1.9392,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.79,
96
+ "learning_rate": 3.9473684210526316e-05,
97
+ "loss": 1.8584,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.84,
102
+ "learning_rate": 4.210526315789474e-05,
103
+ "loss": 1.7827,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.89,
108
+ "learning_rate": 4.473684210526316e-05,
109
+ "loss": 1.6963,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.95,
114
+ "learning_rate": 4.736842105263158e-05,
115
+ "loss": 1.5962,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 1.0,
120
+ "learning_rate": 5e-05,
121
+ "loss": 1.5074,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 1.0,
126
+ "eval_accuracy": 0.7077777777777777,
127
+ "eval_loss": 1.3433218002319336,
128
+ "eval_runtime": 6.7484,
129
+ "eval_samples_per_second": 400.097,
130
+ "eval_steps_per_second": 12.596,
131
+ "step": 190
132
+ },
133
+ {
134
+ "epoch": 1.05,
135
+ "learning_rate": 4.970760233918128e-05,
136
+ "loss": 1.4517,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 1.11,
141
+ "learning_rate": 4.941520467836258e-05,
142
+ "loss": 1.4136,
143
+ "step": 210
144
+ },
145
+ {
146
+ "epoch": 1.16,
147
+ "learning_rate": 4.912280701754386e-05,
148
+ "loss": 1.3514,
149
+ "step": 220
150
+ },
151
+ {
152
+ "epoch": 1.21,
153
+ "learning_rate": 4.883040935672515e-05,
154
+ "loss": 1.3052,
155
+ "step": 230
156
+ },
157
+ {
158
+ "epoch": 1.26,
159
+ "learning_rate": 4.853801169590643e-05,
160
+ "loss": 1.2468,
161
+ "step": 240
162
+ },
163
+ {
164
+ "epoch": 1.32,
165
+ "learning_rate": 4.824561403508772e-05,
166
+ "loss": 1.2389,
167
+ "step": 250
168
+ },
169
+ {
170
+ "epoch": 1.37,
171
+ "learning_rate": 4.7953216374269006e-05,
172
+ "loss": 1.1687,
173
+ "step": 260
174
+ },
175
+ {
176
+ "epoch": 1.42,
177
+ "learning_rate": 4.7660818713450294e-05,
178
+ "loss": 1.1483,
179
+ "step": 270
180
+ },
181
+ {
182
+ "epoch": 1.47,
183
+ "learning_rate": 4.736842105263158e-05,
184
+ "loss": 1.1665,
185
+ "step": 280
186
+ },
187
+ {
188
+ "epoch": 1.53,
189
+ "learning_rate": 4.707602339181287e-05,
190
+ "loss": 1.0939,
191
+ "step": 290
192
+ },
193
+ {
194
+ "epoch": 1.58,
195
+ "learning_rate": 4.678362573099415e-05,
196
+ "loss": 1.0914,
197
+ "step": 300
198
+ },
199
+ {
200
+ "epoch": 1.63,
201
+ "learning_rate": 4.649122807017544e-05,
202
+ "loss": 1.0623,
203
+ "step": 310
204
+ },
205
+ {
206
+ "epoch": 1.68,
207
+ "learning_rate": 4.619883040935672e-05,
208
+ "loss": 1.0393,
209
+ "step": 320
210
+ },
211
+ {
212
+ "epoch": 1.74,
213
+ "learning_rate": 4.590643274853802e-05,
214
+ "loss": 1.0274,
215
+ "step": 330
216
+ },
217
+ {
218
+ "epoch": 1.79,
219
+ "learning_rate": 4.56140350877193e-05,
220
+ "loss": 0.9957,
221
+ "step": 340
222
+ },
223
+ {
224
+ "epoch": 1.84,
225
+ "learning_rate": 4.5321637426900585e-05,
226
+ "loss": 1.0176,
227
+ "step": 350
228
+ },
229
+ {
230
+ "epoch": 1.89,
231
+ "learning_rate": 4.502923976608187e-05,
232
+ "loss": 0.9963,
233
+ "step": 360
234
+ },
235
+ {
236
+ "epoch": 1.95,
237
+ "learning_rate": 4.473684210526316e-05,
238
+ "loss": 0.9611,
239
+ "step": 370
240
+ },
241
+ {
242
+ "epoch": 2.0,
243
+ "learning_rate": 4.4444444444444447e-05,
244
+ "loss": 0.9398,
245
+ "step": 380
246
+ },
247
+ {
248
+ "epoch": 2.0,
249
+ "eval_accuracy": 0.85,
250
+ "eval_loss": 0.717716634273529,
251
+ "eval_runtime": 6.6482,
252
+ "eval_samples_per_second": 406.123,
253
+ "eval_steps_per_second": 12.785,
254
+ "step": 380
255
+ },
256
+ {
257
+ "epoch": 2.05,
258
+ "learning_rate": 4.4152046783625734e-05,
259
+ "loss": 0.9262,
260
+ "step": 390
261
+ },
262
+ {
263
+ "epoch": 2.11,
264
+ "learning_rate": 4.3859649122807014e-05,
265
+ "loss": 0.8718,
266
+ "step": 400
267
+ },
268
+ {
269
+ "epoch": 2.16,
270
+ "learning_rate": 4.356725146198831e-05,
271
+ "loss": 0.8925,
272
+ "step": 410
273
+ },
274
+ {
275
+ "epoch": 2.21,
276
+ "learning_rate": 4.327485380116959e-05,
277
+ "loss": 0.852,
278
+ "step": 420
279
+ },
280
+ {
281
+ "epoch": 2.26,
282
+ "learning_rate": 4.298245614035088e-05,
283
+ "loss": 0.8706,
284
+ "step": 430
285
+ },
286
+ {
287
+ "epoch": 2.32,
288
+ "learning_rate": 4.269005847953216e-05,
289
+ "loss": 0.8152,
290
+ "step": 440
291
+ },
292
+ {
293
+ "epoch": 2.37,
294
+ "learning_rate": 4.239766081871345e-05,
295
+ "loss": 0.823,
296
+ "step": 450
297
+ },
298
+ {
299
+ "epoch": 2.42,
300
+ "learning_rate": 4.210526315789474e-05,
301
+ "loss": 0.8239,
302
+ "step": 460
303
+ },
304
+ {
305
+ "epoch": 2.47,
306
+ "learning_rate": 4.1812865497076025e-05,
307
+ "loss": 0.7756,
308
+ "step": 470
309
+ },
310
+ {
311
+ "epoch": 2.53,
312
+ "learning_rate": 4.152046783625731e-05,
313
+ "loss": 0.8068,
314
+ "step": 480
315
+ },
316
+ {
317
+ "epoch": 2.58,
318
+ "learning_rate": 4.12280701754386e-05,
319
+ "loss": 0.7516,
320
+ "step": 490
321
+ },
322
+ {
323
+ "epoch": 2.63,
324
+ "learning_rate": 4.093567251461988e-05,
325
+ "loss": 0.7439,
326
+ "step": 500
327
+ },
328
+ {
329
+ "epoch": 2.68,
330
+ "learning_rate": 4.0643274853801174e-05,
331
+ "loss": 0.7157,
332
+ "step": 510
333
+ },
334
+ {
335
+ "epoch": 2.74,
336
+ "learning_rate": 4.0350877192982455e-05,
337
+ "loss": 0.7353,
338
+ "step": 520
339
+ },
340
+ {
341
+ "epoch": 2.79,
342
+ "learning_rate": 4.005847953216375e-05,
343
+ "loss": 0.6941,
344
+ "step": 530
345
+ },
346
+ {
347
+ "epoch": 2.84,
348
+ "learning_rate": 3.976608187134503e-05,
349
+ "loss": 0.6756,
350
+ "step": 540
351
+ },
352
+ {
353
+ "epoch": 2.89,
354
+ "learning_rate": 3.9473684210526316e-05,
355
+ "loss": 0.6934,
356
+ "step": 550
357
+ },
358
+ {
359
+ "epoch": 2.95,
360
+ "learning_rate": 3.9181286549707604e-05,
361
+ "loss": 0.6508,
362
+ "step": 560
363
+ },
364
+ {
365
+ "epoch": 3.0,
366
+ "learning_rate": 3.888888888888889e-05,
367
+ "loss": 0.7035,
368
+ "step": 570
369
+ },
370
+ {
371
+ "epoch": 3.0,
372
+ "eval_accuracy": 0.9070370370370371,
373
+ "eval_loss": 0.4251876175403595,
374
+ "eval_runtime": 6.5087,
375
+ "eval_samples_per_second": 414.829,
376
+ "eval_steps_per_second": 13.059,
377
+ "step": 570
378
+ },
379
+ {
380
+ "epoch": 3.05,
381
+ "learning_rate": 3.859649122807018e-05,
382
+ "loss": 0.6747,
383
+ "step": 580
384
+ },
385
+ {
386
+ "epoch": 3.11,
387
+ "learning_rate": 3.8304093567251465e-05,
388
+ "loss": 0.6437,
389
+ "step": 590
390
+ },
391
+ {
392
+ "epoch": 3.16,
393
+ "learning_rate": 3.8011695906432746e-05,
394
+ "loss": 0.6677,
395
+ "step": 600
396
+ },
397
+ {
398
+ "epoch": 3.21,
399
+ "learning_rate": 3.771929824561404e-05,
400
+ "loss": 0.6333,
401
+ "step": 610
402
+ },
403
+ {
404
+ "epoch": 3.26,
405
+ "learning_rate": 3.742690058479532e-05,
406
+ "loss": 0.5999,
407
+ "step": 620
408
+ },
409
+ {
410
+ "epoch": 3.32,
411
+ "learning_rate": 3.713450292397661e-05,
412
+ "loss": 0.6074,
413
+ "step": 630
414
+ },
415
+ {
416
+ "epoch": 3.37,
417
+ "learning_rate": 3.6842105263157895e-05,
418
+ "loss": 0.6006,
419
+ "step": 640
420
+ },
421
+ {
422
+ "epoch": 3.42,
423
+ "learning_rate": 3.654970760233918e-05,
424
+ "loss": 0.5783,
425
+ "step": 650
426
+ },
427
+ {
428
+ "epoch": 3.47,
429
+ "learning_rate": 3.625730994152047e-05,
430
+ "loss": 0.6052,
431
+ "step": 660
432
+ },
433
+ {
434
+ "epoch": 3.53,
435
+ "learning_rate": 3.5964912280701756e-05,
436
+ "loss": 0.5428,
437
+ "step": 670
438
+ },
439
+ {
440
+ "epoch": 3.58,
441
+ "learning_rate": 3.5672514619883044e-05,
442
+ "loss": 0.6207,
443
+ "step": 680
444
+ },
445
+ {
446
+ "epoch": 3.63,
447
+ "learning_rate": 3.538011695906433e-05,
448
+ "loss": 0.5259,
449
+ "step": 690
450
+ },
451
+ {
452
+ "epoch": 3.68,
453
+ "learning_rate": 3.508771929824561e-05,
454
+ "loss": 0.5637,
455
+ "step": 700
456
+ },
457
+ {
458
+ "epoch": 3.74,
459
+ "learning_rate": 3.4795321637426905e-05,
460
+ "loss": 0.5595,
461
+ "step": 710
462
+ },
463
+ {
464
+ "epoch": 3.79,
465
+ "learning_rate": 3.4502923976608186e-05,
466
+ "loss": 0.5284,
467
+ "step": 720
468
+ },
469
+ {
470
+ "epoch": 3.84,
471
+ "learning_rate": 3.421052631578947e-05,
472
+ "loss": 0.5708,
473
+ "step": 730
474
+ },
475
+ {
476
+ "epoch": 3.89,
477
+ "learning_rate": 3.391812865497076e-05,
478
+ "loss": 0.568,
479
+ "step": 740
480
+ },
481
+ {
482
+ "epoch": 3.95,
483
+ "learning_rate": 3.362573099415205e-05,
484
+ "loss": 0.5766,
485
+ "step": 750
486
+ },
487
+ {
488
+ "epoch": 4.0,
489
+ "learning_rate": 3.3333333333333335e-05,
490
+ "loss": 0.5435,
491
+ "step": 760
492
+ },
493
+ {
494
+ "epoch": 4.0,
495
+ "eval_accuracy": 0.9281481481481482,
496
+ "eval_loss": 0.3079540431499481,
497
+ "eval_runtime": 6.531,
498
+ "eval_samples_per_second": 413.415,
499
+ "eval_steps_per_second": 13.015,
500
+ "step": 760
501
+ },
502
+ {
503
+ "epoch": 4.05,
504
+ "learning_rate": 3.304093567251462e-05,
505
+ "loss": 0.5273,
506
+ "step": 770
507
+ },
508
+ {
509
+ "epoch": 4.11,
510
+ "learning_rate": 3.274853801169591e-05,
511
+ "loss": 0.5117,
512
+ "step": 780
513
+ },
514
+ {
515
+ "epoch": 4.16,
516
+ "learning_rate": 3.24561403508772e-05,
517
+ "loss": 0.52,
518
+ "step": 790
519
+ },
520
+ {
521
+ "epoch": 4.21,
522
+ "learning_rate": 3.216374269005848e-05,
523
+ "loss": 0.5671,
524
+ "step": 800
525
+ },
526
+ {
527
+ "epoch": 4.26,
528
+ "learning_rate": 3.187134502923977e-05,
529
+ "loss": 0.4939,
530
+ "step": 810
531
+ },
532
+ {
533
+ "epoch": 4.32,
534
+ "learning_rate": 3.157894736842105e-05,
535
+ "loss": 0.5159,
536
+ "step": 820
537
+ },
538
+ {
539
+ "epoch": 4.37,
540
+ "learning_rate": 3.128654970760234e-05,
541
+ "loss": 0.5199,
542
+ "step": 830
543
+ },
544
+ {
545
+ "epoch": 4.42,
546
+ "learning_rate": 3.0994152046783626e-05,
547
+ "loss": 0.5391,
548
+ "step": 840
549
+ },
550
+ {
551
+ "epoch": 4.47,
552
+ "learning_rate": 3.0701754385964913e-05,
553
+ "loss": 0.5256,
554
+ "step": 850
555
+ },
556
+ {
557
+ "epoch": 4.53,
558
+ "learning_rate": 3.0409356725146197e-05,
559
+ "loss": 0.5168,
560
+ "step": 860
561
+ },
562
+ {
563
+ "epoch": 4.58,
564
+ "learning_rate": 3.0116959064327488e-05,
565
+ "loss": 0.4874,
566
+ "step": 870
567
+ },
568
+ {
569
+ "epoch": 4.63,
570
+ "learning_rate": 2.9824561403508772e-05,
571
+ "loss": 0.4711,
572
+ "step": 880
573
+ },
574
+ {
575
+ "epoch": 4.68,
576
+ "learning_rate": 2.9532163742690062e-05,
577
+ "loss": 0.4856,
578
+ "step": 890
579
+ },
580
+ {
581
+ "epoch": 4.74,
582
+ "learning_rate": 2.9239766081871346e-05,
583
+ "loss": 0.4733,
584
+ "step": 900
585
+ },
586
+ {
587
+ "epoch": 4.79,
588
+ "learning_rate": 2.8947368421052634e-05,
589
+ "loss": 0.4822,
590
+ "step": 910
591
+ },
592
+ {
593
+ "epoch": 4.84,
594
+ "learning_rate": 2.8654970760233917e-05,
595
+ "loss": 0.4633,
596
+ "step": 920
597
+ },
598
+ {
599
+ "epoch": 4.89,
600
+ "learning_rate": 2.8362573099415208e-05,
601
+ "loss": 0.4827,
602
+ "step": 930
603
+ },
604
+ {
605
+ "epoch": 4.95,
606
+ "learning_rate": 2.8070175438596492e-05,
607
+ "loss": 0.477,
608
+ "step": 940
609
+ },
610
+ {
611
+ "epoch": 5.0,
612
+ "learning_rate": 2.777777777777778e-05,
613
+ "loss": 0.5007,
614
+ "step": 950
615
+ },
616
+ {
617
+ "epoch": 5.0,
618
+ "eval_accuracy": 0.9388888888888889,
619
+ "eval_loss": 0.24653582274913788,
620
+ "eval_runtime": 7.0867,
621
+ "eval_samples_per_second": 380.997,
622
+ "eval_steps_per_second": 11.994,
623
+ "step": 950
624
+ },
625
+ {
626
+ "epoch": 5.05,
627
+ "learning_rate": 2.7485380116959063e-05,
628
+ "loss": 0.4798,
629
+ "step": 960
630
+ },
631
+ {
632
+ "epoch": 5.11,
633
+ "learning_rate": 2.7192982456140354e-05,
634
+ "loss": 0.4548,
635
+ "step": 970
636
+ },
637
+ {
638
+ "epoch": 5.16,
639
+ "learning_rate": 2.6900584795321637e-05,
640
+ "loss": 0.4981,
641
+ "step": 980
642
+ },
643
+ {
644
+ "epoch": 5.21,
645
+ "learning_rate": 2.6608187134502928e-05,
646
+ "loss": 0.4549,
647
+ "step": 990
648
+ },
649
+ {
650
+ "epoch": 5.26,
651
+ "learning_rate": 2.6315789473684212e-05,
652
+ "loss": 0.4804,
653
+ "step": 1000
654
+ },
655
+ {
656
+ "epoch": 5.32,
657
+ "learning_rate": 2.60233918128655e-05,
658
+ "loss": 0.475,
659
+ "step": 1010
660
+ },
661
+ {
662
+ "epoch": 5.37,
663
+ "learning_rate": 2.5730994152046783e-05,
664
+ "loss": 0.4739,
665
+ "step": 1020
666
+ },
667
+ {
668
+ "epoch": 5.42,
669
+ "learning_rate": 2.5438596491228074e-05,
670
+ "loss": 0.4497,
671
+ "step": 1030
672
+ },
673
+ {
674
+ "epoch": 5.47,
675
+ "learning_rate": 2.5146198830409358e-05,
676
+ "loss": 0.5059,
677
+ "step": 1040
678
+ },
679
+ {
680
+ "epoch": 5.53,
681
+ "learning_rate": 2.485380116959064e-05,
682
+ "loss": 0.4863,
683
+ "step": 1050
684
+ },
685
+ {
686
+ "epoch": 5.58,
687
+ "learning_rate": 2.456140350877193e-05,
688
+ "loss": 0.5016,
689
+ "step": 1060
690
+ },
691
+ {
692
+ "epoch": 5.63,
693
+ "learning_rate": 2.4269005847953216e-05,
694
+ "loss": 0.4152,
695
+ "step": 1070
696
+ },
697
+ {
698
+ "epoch": 5.68,
699
+ "learning_rate": 2.3976608187134503e-05,
700
+ "loss": 0.4425,
701
+ "step": 1080
702
+ },
703
+ {
704
+ "epoch": 5.74,
705
+ "learning_rate": 2.368421052631579e-05,
706
+ "loss": 0.5285,
707
+ "step": 1090
708
+ },
709
+ {
710
+ "epoch": 5.79,
711
+ "learning_rate": 2.3391812865497074e-05,
712
+ "loss": 0.4645,
713
+ "step": 1100
714
+ },
715
+ {
716
+ "epoch": 5.84,
717
+ "learning_rate": 2.309941520467836e-05,
718
+ "loss": 0.424,
719
+ "step": 1110
720
+ },
721
+ {
722
+ "epoch": 5.89,
723
+ "learning_rate": 2.280701754385965e-05,
724
+ "loss": 0.4313,
725
+ "step": 1120
726
+ },
727
+ {
728
+ "epoch": 5.95,
729
+ "learning_rate": 2.2514619883040936e-05,
730
+ "loss": 0.4525,
731
+ "step": 1130
732
+ },
733
+ {
734
+ "epoch": 6.0,
735
+ "learning_rate": 2.2222222222222223e-05,
736
+ "loss": 0.4533,
737
+ "step": 1140
738
+ },
739
+ {
740
+ "epoch": 6.0,
741
+ "eval_accuracy": 0.9444444444444444,
742
+ "eval_loss": 0.22906289994716644,
743
+ "eval_runtime": 6.6825,
744
+ "eval_samples_per_second": 404.042,
745
+ "eval_steps_per_second": 12.72,
746
+ "step": 1140
747
+ },
748
+ {
749
+ "epoch": 6.05,
750
+ "learning_rate": 2.1929824561403507e-05,
751
+ "loss": 0.4753,
752
+ "step": 1150
753
+ },
754
+ {
755
+ "epoch": 6.11,
756
+ "learning_rate": 2.1637426900584794e-05,
757
+ "loss": 0.4641,
758
+ "step": 1160
759
+ },
760
+ {
761
+ "epoch": 6.16,
762
+ "learning_rate": 2.134502923976608e-05,
763
+ "loss": 0.4654,
764
+ "step": 1170
765
+ },
766
+ {
767
+ "epoch": 6.21,
768
+ "learning_rate": 2.105263157894737e-05,
769
+ "loss": 0.4861,
770
+ "step": 1180
771
+ },
772
+ {
773
+ "epoch": 6.26,
774
+ "learning_rate": 2.0760233918128656e-05,
775
+ "loss": 0.4639,
776
+ "step": 1190
777
+ },
778
+ {
779
+ "epoch": 6.32,
780
+ "learning_rate": 2.046783625730994e-05,
781
+ "loss": 0.4197,
782
+ "step": 1200
783
+ },
784
+ {
785
+ "epoch": 6.37,
786
+ "learning_rate": 2.0175438596491227e-05,
787
+ "loss": 0.4251,
788
+ "step": 1210
789
+ },
790
+ {
791
+ "epoch": 6.42,
792
+ "learning_rate": 1.9883040935672515e-05,
793
+ "loss": 0.4716,
794
+ "step": 1220
795
+ },
796
+ {
797
+ "epoch": 6.47,
798
+ "learning_rate": 1.9590643274853802e-05,
799
+ "loss": 0.4679,
800
+ "step": 1230
801
+ },
802
+ {
803
+ "epoch": 6.53,
804
+ "learning_rate": 1.929824561403509e-05,
805
+ "loss": 0.3977,
806
+ "step": 1240
807
+ },
808
+ {
809
+ "epoch": 6.58,
810
+ "learning_rate": 1.9005847953216373e-05,
811
+ "loss": 0.4449,
812
+ "step": 1250
813
+ },
814
+ {
815
+ "epoch": 6.63,
816
+ "learning_rate": 1.871345029239766e-05,
817
+ "loss": 0.4409,
818
+ "step": 1260
819
+ },
820
+ {
821
+ "epoch": 6.68,
822
+ "learning_rate": 1.8421052631578947e-05,
823
+ "loss": 0.4587,
824
+ "step": 1270
825
+ },
826
+ {
827
+ "epoch": 6.74,
828
+ "learning_rate": 1.8128654970760235e-05,
829
+ "loss": 0.412,
830
+ "step": 1280
831
+ },
832
+ {
833
+ "epoch": 6.79,
834
+ "learning_rate": 1.7836257309941522e-05,
835
+ "loss": 0.4097,
836
+ "step": 1290
837
+ },
838
+ {
839
+ "epoch": 6.84,
840
+ "learning_rate": 1.7543859649122806e-05,
841
+ "loss": 0.4342,
842
+ "step": 1300
843
+ },
844
+ {
845
+ "epoch": 6.89,
846
+ "learning_rate": 1.7251461988304093e-05,
847
+ "loss": 0.471,
848
+ "step": 1310
849
+ },
850
+ {
851
+ "epoch": 6.95,
852
+ "learning_rate": 1.695906432748538e-05,
853
+ "loss": 0.4607,
854
+ "step": 1320
855
+ },
856
+ {
857
+ "epoch": 7.0,
858
+ "learning_rate": 1.6666666666666667e-05,
859
+ "loss": 0.3961,
860
+ "step": 1330
861
+ },
862
+ {
863
+ "epoch": 7.0,
864
+ "eval_accuracy": 0.9496296296296296,
865
+ "eval_loss": 0.19912981986999512,
866
+ "eval_runtime": 6.7692,
867
+ "eval_samples_per_second": 398.866,
868
+ "eval_steps_per_second": 12.557,
869
+ "step": 1330
870
+ },
871
+ {
872
+ "epoch": 7.05,
873
+ "learning_rate": 1.6374269005847955e-05,
874
+ "loss": 0.4264,
875
+ "step": 1340
876
+ },
877
+ {
878
+ "epoch": 7.11,
879
+ "learning_rate": 1.608187134502924e-05,
880
+ "loss": 0.387,
881
+ "step": 1350
882
+ },
883
+ {
884
+ "epoch": 7.16,
885
+ "learning_rate": 1.5789473684210526e-05,
886
+ "loss": 0.4116,
887
+ "step": 1360
888
+ },
889
+ {
890
+ "epoch": 7.21,
891
+ "learning_rate": 1.5497076023391813e-05,
892
+ "loss": 0.4089,
893
+ "step": 1370
894
+ },
895
+ {
896
+ "epoch": 7.26,
897
+ "learning_rate": 1.5204678362573099e-05,
898
+ "loss": 0.4028,
899
+ "step": 1380
900
+ },
901
+ {
902
+ "epoch": 7.32,
903
+ "learning_rate": 1.4912280701754386e-05,
904
+ "loss": 0.4375,
905
+ "step": 1390
906
+ },
907
+ {
908
+ "epoch": 7.37,
909
+ "learning_rate": 1.4619883040935673e-05,
910
+ "loss": 0.4481,
911
+ "step": 1400
912
+ },
913
+ {
914
+ "epoch": 7.42,
915
+ "learning_rate": 1.4327485380116959e-05,
916
+ "loss": 0.403,
917
+ "step": 1410
918
+ },
919
+ {
920
+ "epoch": 7.47,
921
+ "learning_rate": 1.4035087719298246e-05,
922
+ "loss": 0.4124,
923
+ "step": 1420
924
+ },
925
+ {
926
+ "epoch": 7.53,
927
+ "learning_rate": 1.3742690058479531e-05,
928
+ "loss": 0.4091,
929
+ "step": 1430
930
+ },
931
+ {
932
+ "epoch": 7.58,
933
+ "learning_rate": 1.3450292397660819e-05,
934
+ "loss": 0.4497,
935
+ "step": 1440
936
+ },
937
+ {
938
+ "epoch": 7.63,
939
+ "learning_rate": 1.3157894736842106e-05,
940
+ "loss": 0.4252,
941
+ "step": 1450
942
+ },
943
+ {
944
+ "epoch": 7.68,
945
+ "learning_rate": 1.2865497076023392e-05,
946
+ "loss": 0.4626,
947
+ "step": 1460
948
+ },
949
+ {
950
+ "epoch": 7.74,
951
+ "learning_rate": 1.2573099415204679e-05,
952
+ "loss": 0.4044,
953
+ "step": 1470
954
+ },
955
+ {
956
+ "epoch": 7.79,
957
+ "learning_rate": 1.2280701754385964e-05,
958
+ "loss": 0.4299,
959
+ "step": 1480
960
+ },
961
+ {
962
+ "epoch": 7.84,
963
+ "learning_rate": 1.1988304093567252e-05,
964
+ "loss": 0.43,
965
+ "step": 1490
966
+ },
967
+ {
968
+ "epoch": 7.89,
969
+ "learning_rate": 1.1695906432748537e-05,
970
+ "loss": 0.4418,
971
+ "step": 1500
972
+ },
973
+ {
974
+ "epoch": 7.95,
975
+ "learning_rate": 1.1403508771929824e-05,
976
+ "loss": 0.4073,
977
+ "step": 1510
978
+ },
979
+ {
980
+ "epoch": 8.0,
981
+ "learning_rate": 1.1111111111111112e-05,
982
+ "loss": 0.3949,
983
+ "step": 1520
984
+ },
985
+ {
986
+ "epoch": 8.0,
987
+ "eval_accuracy": 0.9507407407407408,
988
+ "eval_loss": 0.1926022469997406,
989
+ "eval_runtime": 6.688,
990
+ "eval_samples_per_second": 403.706,
991
+ "eval_steps_per_second": 12.709,
992
+ "step": 1520
993
+ },
994
+ {
995
+ "epoch": 8.05,
996
+ "learning_rate": 1.0818713450292397e-05,
997
+ "loss": 0.4227,
998
+ "step": 1530
999
+ },
1000
+ {
1001
+ "epoch": 8.11,
1002
+ "learning_rate": 1.0526315789473684e-05,
1003
+ "loss": 0.4243,
1004
+ "step": 1540
1005
+ },
1006
+ {
1007
+ "epoch": 8.16,
1008
+ "learning_rate": 1.023391812865497e-05,
1009
+ "loss": 0.4481,
1010
+ "step": 1550
1011
+ },
1012
+ {
1013
+ "epoch": 8.21,
1014
+ "learning_rate": 9.941520467836257e-06,
1015
+ "loss": 0.465,
1016
+ "step": 1560
1017
+ },
1018
+ {
1019
+ "epoch": 8.26,
1020
+ "learning_rate": 9.649122807017545e-06,
1021
+ "loss": 0.4156,
1022
+ "step": 1570
1023
+ },
1024
+ {
1025
+ "epoch": 8.32,
1026
+ "learning_rate": 9.35672514619883e-06,
1027
+ "loss": 0.397,
1028
+ "step": 1580
1029
+ },
1030
+ {
1031
+ "epoch": 8.37,
1032
+ "learning_rate": 9.064327485380117e-06,
1033
+ "loss": 0.4107,
1034
+ "step": 1590
1035
+ },
1036
+ {
1037
+ "epoch": 8.42,
1038
+ "learning_rate": 8.771929824561403e-06,
1039
+ "loss": 0.4445,
1040
+ "step": 1600
1041
+ },
1042
+ {
1043
+ "epoch": 8.47,
1044
+ "learning_rate": 8.47953216374269e-06,
1045
+ "loss": 0.3787,
1046
+ "step": 1610
1047
+ },
1048
+ {
1049
+ "epoch": 8.53,
1050
+ "learning_rate": 8.187134502923977e-06,
1051
+ "loss": 0.4289,
1052
+ "step": 1620
1053
+ },
1054
+ {
1055
+ "epoch": 8.58,
1056
+ "learning_rate": 7.894736842105263e-06,
1057
+ "loss": 0.398,
1058
+ "step": 1630
1059
+ },
1060
+ {
1061
+ "epoch": 8.63,
1062
+ "learning_rate": 7.602339181286549e-06,
1063
+ "loss": 0.4389,
1064
+ "step": 1640
1065
+ },
1066
+ {
1067
+ "epoch": 8.68,
1068
+ "learning_rate": 7.3099415204678366e-06,
1069
+ "loss": 0.3948,
1070
+ "step": 1650
1071
+ },
1072
+ {
1073
+ "epoch": 8.74,
1074
+ "learning_rate": 7.017543859649123e-06,
1075
+ "loss": 0.4032,
1076
+ "step": 1660
1077
+ },
1078
+ {
1079
+ "epoch": 8.79,
1080
+ "learning_rate": 6.725146198830409e-06,
1081
+ "loss": 0.4187,
1082
+ "step": 1670
1083
+ },
1084
+ {
1085
+ "epoch": 8.84,
1086
+ "learning_rate": 6.432748538011696e-06,
1087
+ "loss": 0.4195,
1088
+ "step": 1680
1089
+ },
1090
+ {
1091
+ "epoch": 8.89,
1092
+ "learning_rate": 6.140350877192982e-06,
1093
+ "loss": 0.3972,
1094
+ "step": 1690
1095
+ },
1096
+ {
1097
+ "epoch": 8.95,
1098
+ "learning_rate": 5.8479532163742686e-06,
1099
+ "loss": 0.3934,
1100
+ "step": 1700
1101
+ },
1102
+ {
1103
+ "epoch": 9.0,
1104
+ "learning_rate": 5.555555555555556e-06,
1105
+ "loss": 0.4302,
1106
+ "step": 1710
1107
+ },
1108
+ {
1109
+ "epoch": 9.0,
1110
+ "eval_accuracy": 0.95,
1111
+ "eval_loss": 0.19283165037631989,
1112
+ "eval_runtime": 6.6653,
1113
+ "eval_samples_per_second": 405.082,
1114
+ "eval_steps_per_second": 12.753,
1115
+ "step": 1710
1116
+ },
1117
+ {
1118
+ "epoch": 9.05,
1119
+ "learning_rate": 5.263157894736842e-06,
1120
+ "loss": 0.4043,
1121
+ "step": 1720
1122
+ },
1123
+ {
1124
+ "epoch": 9.11,
1125
+ "learning_rate": 4.970760233918129e-06,
1126
+ "loss": 0.4216,
1127
+ "step": 1730
1128
+ },
1129
+ {
1130
+ "epoch": 9.16,
1131
+ "learning_rate": 4.678362573099415e-06,
1132
+ "loss": 0.3846,
1133
+ "step": 1740
1134
+ },
1135
+ {
1136
+ "epoch": 9.21,
1137
+ "learning_rate": 4.3859649122807014e-06,
1138
+ "loss": 0.4307,
1139
+ "step": 1750
1140
+ },
1141
+ {
1142
+ "epoch": 9.26,
1143
+ "learning_rate": 4.093567251461989e-06,
1144
+ "loss": 0.3826,
1145
+ "step": 1760
1146
+ },
1147
+ {
1148
+ "epoch": 9.32,
1149
+ "learning_rate": 3.8011695906432747e-06,
1150
+ "loss": 0.4006,
1151
+ "step": 1770
1152
+ },
1153
+ {
1154
+ "epoch": 9.37,
1155
+ "learning_rate": 3.5087719298245615e-06,
1156
+ "loss": 0.4459,
1157
+ "step": 1780
1158
+ },
1159
+ {
1160
+ "epoch": 9.42,
1161
+ "learning_rate": 3.216374269005848e-06,
1162
+ "loss": 0.4038,
1163
+ "step": 1790
1164
+ },
1165
+ {
1166
+ "epoch": 9.47,
1167
+ "learning_rate": 2.9239766081871343e-06,
1168
+ "loss": 0.4069,
1169
+ "step": 1800
1170
+ },
1171
+ {
1172
+ "epoch": 9.53,
1173
+ "learning_rate": 2.631578947368421e-06,
1174
+ "loss": 0.3911,
1175
+ "step": 1810
1176
+ },
1177
+ {
1178
+ "epoch": 9.58,
1179
+ "learning_rate": 2.3391812865497075e-06,
1180
+ "loss": 0.4002,
1181
+ "step": 1820
1182
+ },
1183
+ {
1184
+ "epoch": 9.63,
1185
+ "learning_rate": 2.0467836257309943e-06,
1186
+ "loss": 0.3768,
1187
+ "step": 1830
1188
+ },
1189
+ {
1190
+ "epoch": 9.68,
1191
+ "learning_rate": 1.7543859649122807e-06,
1192
+ "loss": 0.4195,
1193
+ "step": 1840
1194
+ },
1195
+ {
1196
+ "epoch": 9.74,
1197
+ "learning_rate": 1.4619883040935671e-06,
1198
+ "loss": 0.4108,
1199
+ "step": 1850
1200
+ },
1201
+ {
1202
+ "epoch": 9.79,
1203
+ "learning_rate": 1.1695906432748538e-06,
1204
+ "loss": 0.3811,
1205
+ "step": 1860
1206
+ },
1207
+ {
1208
+ "epoch": 9.84,
1209
+ "learning_rate": 8.771929824561404e-07,
1210
+ "loss": 0.4,
1211
+ "step": 1870
1212
+ },
1213
+ {
1214
+ "epoch": 9.89,
1215
+ "learning_rate": 5.847953216374269e-07,
1216
+ "loss": 0.3771,
1217
+ "step": 1880
1218
+ },
1219
+ {
1220
+ "epoch": 9.95,
1221
+ "learning_rate": 2.9239766081871344e-07,
1222
+ "loss": 0.4072,
1223
+ "step": 1890
1224
+ },
1225
+ {
1226
+ "epoch": 10.0,
1227
+ "learning_rate": 0.0,
1228
+ "loss": 0.4061,
1229
+ "step": 1900
1230
+ },
1231
+ {
1232
+ "epoch": 10.0,
1233
+ "eval_accuracy": 0.9462962962962963,
1234
+ "eval_loss": 0.1930549293756485,
1235
+ "eval_runtime": 6.7634,
1236
+ "eval_samples_per_second": 399.209,
1237
+ "eval_steps_per_second": 12.568,
1238
+ "step": 1900
1239
+ },
1240
+ {
1241
+ "epoch": 10.0,
1242
+ "step": 1900,
1243
+ "total_flos": 3.46193285935104e+17,
1244
+ "train_loss": 0.7228629551435771,
1245
+ "train_runtime": 1251.8481,
1246
+ "train_samples_per_second": 194.113,
1247
+ "train_steps_per_second": 1.518
1248
+ }
1249
+ ],
1250
+ "max_steps": 1900,
1251
+ "num_train_epochs": 10,
1252
+ "total_flos": 3.46193285935104e+17,
1253
+ "trial_name": null,
1254
+ "trial_params": null
1255
+ }