therealcyberlord commited on
Commit
c360a16
1 Parent(s): b16a4e0

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +1741 -0
trainer_state.json ADDED
@@ -0,0 +1,1741 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7031954526901245,
3
+ "best_model_checkpoint": "./vit-stanford-cars/checkpoint-2500",
4
+ "epoch": 9.803921568627452,
5
+ "global_step": 2500,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.04,
12
+ "learning_rate": 1.6000000000000001e-06,
13
+ "loss": 5.3806,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.08,
18
+ "learning_rate": 3.2000000000000003e-06,
19
+ "loss": 5.4127,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.12,
24
+ "learning_rate": 4.6400000000000005e-06,
25
+ "loss": 5.3923,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.16,
30
+ "learning_rate": 6.24e-06,
31
+ "loss": 5.4098,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.2,
36
+ "learning_rate": 7.840000000000001e-06,
37
+ "loss": 5.3607,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.24,
42
+ "learning_rate": 9.440000000000001e-06,
43
+ "loss": 5.3706,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.27,
48
+ "learning_rate": 1.1040000000000001e-05,
49
+ "loss": 5.3293,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.31,
54
+ "learning_rate": 1.2640000000000001e-05,
55
+ "loss": 5.3037,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.35,
60
+ "learning_rate": 1.4240000000000001e-05,
61
+ "loss": 5.2453,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.39,
66
+ "learning_rate": 1.584e-05,
67
+ "loss": 5.2552,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.39,
72
+ "eval_accuracy": 0.0125,
73
+ "eval_loss": 5.2306718826293945,
74
+ "eval_runtime": 39.7246,
75
+ "eval_samples_per_second": 50.347,
76
+ "eval_steps_per_second": 1.586,
77
+ "step": 100
78
+ },
79
+ {
80
+ "epoch": 0.43,
81
+ "learning_rate": 1.7440000000000002e-05,
82
+ "loss": 5.1976,
83
+ "step": 110
84
+ },
85
+ {
86
+ "epoch": 0.47,
87
+ "learning_rate": 1.904e-05,
88
+ "loss": 5.1826,
89
+ "step": 120
90
+ },
91
+ {
92
+ "epoch": 0.51,
93
+ "learning_rate": 2.0640000000000002e-05,
94
+ "loss": 5.1389,
95
+ "step": 130
96
+ },
97
+ {
98
+ "epoch": 0.55,
99
+ "learning_rate": 2.2240000000000004e-05,
100
+ "loss": 5.0934,
101
+ "step": 140
102
+ },
103
+ {
104
+ "epoch": 0.59,
105
+ "learning_rate": 2.3840000000000002e-05,
106
+ "loss": 5.0457,
107
+ "step": 150
108
+ },
109
+ {
110
+ "epoch": 0.63,
111
+ "learning_rate": 2.5440000000000004e-05,
112
+ "loss": 5.0993,
113
+ "step": 160
114
+ },
115
+ {
116
+ "epoch": 0.67,
117
+ "learning_rate": 2.7040000000000005e-05,
118
+ "loss": 4.9771,
119
+ "step": 170
120
+ },
121
+ {
122
+ "epoch": 0.71,
123
+ "learning_rate": 2.864e-05,
124
+ "loss": 4.9128,
125
+ "step": 180
126
+ },
127
+ {
128
+ "epoch": 0.75,
129
+ "learning_rate": 3.0240000000000002e-05,
130
+ "loss": 4.8988,
131
+ "step": 190
132
+ },
133
+ {
134
+ "epoch": 0.78,
135
+ "learning_rate": 3.184000000000001e-05,
136
+ "loss": 4.7653,
137
+ "step": 200
138
+ },
139
+ {
140
+ "epoch": 0.78,
141
+ "eval_accuracy": 0.079,
142
+ "eval_loss": 4.759223461151123,
143
+ "eval_runtime": 40.216,
144
+ "eval_samples_per_second": 49.732,
145
+ "eval_steps_per_second": 1.567,
146
+ "step": 200
147
+ },
148
+ {
149
+ "epoch": 0.82,
150
+ "learning_rate": 3.344e-05,
151
+ "loss": 4.7304,
152
+ "step": 210
153
+ },
154
+ {
155
+ "epoch": 0.86,
156
+ "learning_rate": 3.504e-05,
157
+ "loss": 4.7613,
158
+ "step": 220
159
+ },
160
+ {
161
+ "epoch": 0.9,
162
+ "learning_rate": 3.664e-05,
163
+ "loss": 4.6521,
164
+ "step": 230
165
+ },
166
+ {
167
+ "epoch": 0.94,
168
+ "learning_rate": 3.824e-05,
169
+ "loss": 4.6012,
170
+ "step": 240
171
+ },
172
+ {
173
+ "epoch": 0.98,
174
+ "learning_rate": 3.9840000000000005e-05,
175
+ "loss": 4.4954,
176
+ "step": 250
177
+ },
178
+ {
179
+ "epoch": 1.02,
180
+ "learning_rate": 4.144e-05,
181
+ "loss": 4.3733,
182
+ "step": 260
183
+ },
184
+ {
185
+ "epoch": 1.06,
186
+ "learning_rate": 4.304000000000001e-05,
187
+ "loss": 4.1471,
188
+ "step": 270
189
+ },
190
+ {
191
+ "epoch": 1.1,
192
+ "learning_rate": 4.4640000000000006e-05,
193
+ "loss": 4.1306,
194
+ "step": 280
195
+ },
196
+ {
197
+ "epoch": 1.14,
198
+ "learning_rate": 4.624e-05,
199
+ "loss": 4.0396,
200
+ "step": 290
201
+ },
202
+ {
203
+ "epoch": 1.18,
204
+ "learning_rate": 4.784e-05,
205
+ "loss": 3.9779,
206
+ "step": 300
207
+ },
208
+ {
209
+ "epoch": 1.18,
210
+ "eval_accuracy": 0.193,
211
+ "eval_loss": 4.070125579833984,
212
+ "eval_runtime": 40.3837,
213
+ "eval_samples_per_second": 49.525,
214
+ "eval_steps_per_second": 1.56,
215
+ "step": 300
216
+ },
217
+ {
218
+ "epoch": 1.22,
219
+ "learning_rate": 4.944e-05,
220
+ "loss": 3.8729,
221
+ "step": 310
222
+ },
223
+ {
224
+ "epoch": 1.25,
225
+ "learning_rate": 5.1040000000000006e-05,
226
+ "loss": 3.7538,
227
+ "step": 320
228
+ },
229
+ {
230
+ "epoch": 1.29,
231
+ "learning_rate": 5.2640000000000004e-05,
232
+ "loss": 3.6952,
233
+ "step": 330
234
+ },
235
+ {
236
+ "epoch": 1.33,
237
+ "learning_rate": 5.424000000000001e-05,
238
+ "loss": 3.6388,
239
+ "step": 340
240
+ },
241
+ {
242
+ "epoch": 1.37,
243
+ "learning_rate": 5.584e-05,
244
+ "loss": 3.6777,
245
+ "step": 350
246
+ },
247
+ {
248
+ "epoch": 1.41,
249
+ "learning_rate": 5.7440000000000006e-05,
250
+ "loss": 3.5145,
251
+ "step": 360
252
+ },
253
+ {
254
+ "epoch": 1.45,
255
+ "learning_rate": 5.9040000000000004e-05,
256
+ "loss": 3.5315,
257
+ "step": 370
258
+ },
259
+ {
260
+ "epoch": 1.49,
261
+ "learning_rate": 6.064e-05,
262
+ "loss": 3.4288,
263
+ "step": 380
264
+ },
265
+ {
266
+ "epoch": 1.53,
267
+ "learning_rate": 6.224e-05,
268
+ "loss": 3.3541,
269
+ "step": 390
270
+ },
271
+ {
272
+ "epoch": 1.57,
273
+ "learning_rate": 6.384000000000001e-05,
274
+ "loss": 3.2754,
275
+ "step": 400
276
+ },
277
+ {
278
+ "epoch": 1.57,
279
+ "eval_accuracy": 0.3305,
280
+ "eval_loss": 3.3328330516815186,
281
+ "eval_runtime": 40.7474,
282
+ "eval_samples_per_second": 49.083,
283
+ "eval_steps_per_second": 1.546,
284
+ "step": 400
285
+ },
286
+ {
287
+ "epoch": 1.61,
288
+ "learning_rate": 6.544e-05,
289
+ "loss": 3.1604,
290
+ "step": 410
291
+ },
292
+ {
293
+ "epoch": 1.65,
294
+ "learning_rate": 6.704000000000001e-05,
295
+ "loss": 3.1043,
296
+ "step": 420
297
+ },
298
+ {
299
+ "epoch": 1.69,
300
+ "learning_rate": 6.864000000000001e-05,
301
+ "loss": 2.9978,
302
+ "step": 430
303
+ },
304
+ {
305
+ "epoch": 1.73,
306
+ "learning_rate": 7.024e-05,
307
+ "loss": 2.9723,
308
+ "step": 440
309
+ },
310
+ {
311
+ "epoch": 1.76,
312
+ "learning_rate": 7.184e-05,
313
+ "loss": 2.9747,
314
+ "step": 450
315
+ },
316
+ {
317
+ "epoch": 1.8,
318
+ "learning_rate": 7.344000000000002e-05,
319
+ "loss": 2.882,
320
+ "step": 460
321
+ },
322
+ {
323
+ "epoch": 1.84,
324
+ "learning_rate": 7.504e-05,
325
+ "loss": 2.7142,
326
+ "step": 470
327
+ },
328
+ {
329
+ "epoch": 1.88,
330
+ "learning_rate": 7.664e-05,
331
+ "loss": 2.7157,
332
+ "step": 480
333
+ },
334
+ {
335
+ "epoch": 1.92,
336
+ "learning_rate": 7.824000000000001e-05,
337
+ "loss": 2.6695,
338
+ "step": 490
339
+ },
340
+ {
341
+ "epoch": 1.96,
342
+ "learning_rate": 7.984000000000001e-05,
343
+ "loss": 2.5607,
344
+ "step": 500
345
+ },
346
+ {
347
+ "epoch": 1.96,
348
+ "eval_accuracy": 0.491,
349
+ "eval_loss": 2.6118950843811035,
350
+ "eval_runtime": 40.1814,
351
+ "eval_samples_per_second": 49.774,
352
+ "eval_steps_per_second": 1.568,
353
+ "step": 500
354
+ },
355
+ {
356
+ "epoch": 2.0,
357
+ "learning_rate": 7.964878048780489e-05,
358
+ "loss": 2.6025,
359
+ "step": 510
360
+ },
361
+ {
362
+ "epoch": 2.04,
363
+ "learning_rate": 7.925853658536585e-05,
364
+ "loss": 1.9918,
365
+ "step": 520
366
+ },
367
+ {
368
+ "epoch": 2.08,
369
+ "learning_rate": 7.886829268292684e-05,
370
+ "loss": 1.8799,
371
+ "step": 530
372
+ },
373
+ {
374
+ "epoch": 2.12,
375
+ "learning_rate": 7.84780487804878e-05,
376
+ "loss": 1.7169,
377
+ "step": 540
378
+ },
379
+ {
380
+ "epoch": 2.16,
381
+ "learning_rate": 7.808780487804879e-05,
382
+ "loss": 1.7148,
383
+ "step": 550
384
+ },
385
+ {
386
+ "epoch": 2.2,
387
+ "learning_rate": 7.769756097560977e-05,
388
+ "loss": 1.7091,
389
+ "step": 560
390
+ },
391
+ {
392
+ "epoch": 2.24,
393
+ "learning_rate": 7.730731707317074e-05,
394
+ "loss": 1.569,
395
+ "step": 570
396
+ },
397
+ {
398
+ "epoch": 2.27,
399
+ "learning_rate": 7.691707317073172e-05,
400
+ "loss": 1.5601,
401
+ "step": 580
402
+ },
403
+ {
404
+ "epoch": 2.31,
405
+ "learning_rate": 7.652682926829269e-05,
406
+ "loss": 1.5463,
407
+ "step": 590
408
+ },
409
+ {
410
+ "epoch": 2.35,
411
+ "learning_rate": 7.613658536585367e-05,
412
+ "loss": 1.5249,
413
+ "step": 600
414
+ },
415
+ {
416
+ "epoch": 2.35,
417
+ "eval_accuracy": 0.61,
418
+ "eval_loss": 2.0372462272644043,
419
+ "eval_runtime": 40.501,
420
+ "eval_samples_per_second": 49.381,
421
+ "eval_steps_per_second": 1.556,
422
+ "step": 600
423
+ },
424
+ {
425
+ "epoch": 2.39,
426
+ "learning_rate": 7.574634146341463e-05,
427
+ "loss": 1.4028,
428
+ "step": 610
429
+ },
430
+ {
431
+ "epoch": 2.43,
432
+ "learning_rate": 7.535609756097562e-05,
433
+ "loss": 1.3612,
434
+ "step": 620
435
+ },
436
+ {
437
+ "epoch": 2.47,
438
+ "learning_rate": 7.49658536585366e-05,
439
+ "loss": 1.3848,
440
+ "step": 630
441
+ },
442
+ {
443
+ "epoch": 2.51,
444
+ "learning_rate": 7.457560975609757e-05,
445
+ "loss": 1.3545,
446
+ "step": 640
447
+ },
448
+ {
449
+ "epoch": 2.55,
450
+ "learning_rate": 7.418536585365855e-05,
451
+ "loss": 1.2483,
452
+ "step": 650
453
+ },
454
+ {
455
+ "epoch": 2.59,
456
+ "learning_rate": 7.379512195121952e-05,
457
+ "loss": 1.2979,
458
+ "step": 660
459
+ },
460
+ {
461
+ "epoch": 2.63,
462
+ "learning_rate": 7.34048780487805e-05,
463
+ "loss": 1.2087,
464
+ "step": 670
465
+ },
466
+ {
467
+ "epoch": 2.67,
468
+ "learning_rate": 7.301463414634147e-05,
469
+ "loss": 1.2661,
470
+ "step": 680
471
+ },
472
+ {
473
+ "epoch": 2.71,
474
+ "learning_rate": 7.262439024390245e-05,
475
+ "loss": 1.2181,
476
+ "step": 690
477
+ },
478
+ {
479
+ "epoch": 2.75,
480
+ "learning_rate": 7.223414634146343e-05,
481
+ "loss": 1.1915,
482
+ "step": 700
483
+ },
484
+ {
485
+ "epoch": 2.75,
486
+ "eval_accuracy": 0.663,
487
+ "eval_loss": 1.6082996129989624,
488
+ "eval_runtime": 39.9953,
489
+ "eval_samples_per_second": 50.006,
490
+ "eval_steps_per_second": 1.575,
491
+ "step": 700
492
+ },
493
+ {
494
+ "epoch": 2.78,
495
+ "learning_rate": 7.18439024390244e-05,
496
+ "loss": 1.1685,
497
+ "step": 710
498
+ },
499
+ {
500
+ "epoch": 2.82,
501
+ "learning_rate": 7.145365853658538e-05,
502
+ "loss": 1.0873,
503
+ "step": 720
504
+ },
505
+ {
506
+ "epoch": 2.86,
507
+ "learning_rate": 7.106341463414635e-05,
508
+ "loss": 1.1013,
509
+ "step": 730
510
+ },
511
+ {
512
+ "epoch": 2.9,
513
+ "learning_rate": 7.067317073170733e-05,
514
+ "loss": 1.1133,
515
+ "step": 740
516
+ },
517
+ {
518
+ "epoch": 2.94,
519
+ "learning_rate": 7.02829268292683e-05,
520
+ "loss": 1.0955,
521
+ "step": 750
522
+ },
523
+ {
524
+ "epoch": 2.98,
525
+ "learning_rate": 6.989268292682928e-05,
526
+ "loss": 0.9469,
527
+ "step": 760
528
+ },
529
+ {
530
+ "epoch": 3.02,
531
+ "learning_rate": 6.950243902439025e-05,
532
+ "loss": 0.7376,
533
+ "step": 770
534
+ },
535
+ {
536
+ "epoch": 3.06,
537
+ "learning_rate": 6.911219512195123e-05,
538
+ "loss": 0.4818,
539
+ "step": 780
540
+ },
541
+ {
542
+ "epoch": 3.1,
543
+ "learning_rate": 6.87219512195122e-05,
544
+ "loss": 0.4825,
545
+ "step": 790
546
+ },
547
+ {
548
+ "epoch": 3.14,
549
+ "learning_rate": 6.833170731707318e-05,
550
+ "loss": 0.4414,
551
+ "step": 800
552
+ },
553
+ {
554
+ "epoch": 3.14,
555
+ "eval_accuracy": 0.725,
556
+ "eval_loss": 1.320522665977478,
557
+ "eval_runtime": 40.1363,
558
+ "eval_samples_per_second": 49.83,
559
+ "eval_steps_per_second": 1.57,
560
+ "step": 800
561
+ },
562
+ {
563
+ "epoch": 3.18,
564
+ "learning_rate": 6.794146341463415e-05,
565
+ "loss": 0.4289,
566
+ "step": 810
567
+ },
568
+ {
569
+ "epoch": 3.22,
570
+ "learning_rate": 6.755121951219513e-05,
571
+ "loss": 0.3919,
572
+ "step": 820
573
+ },
574
+ {
575
+ "epoch": 3.25,
576
+ "learning_rate": 6.71609756097561e-05,
577
+ "loss": 0.3715,
578
+ "step": 830
579
+ },
580
+ {
581
+ "epoch": 3.29,
582
+ "learning_rate": 6.677073170731708e-05,
583
+ "loss": 0.3895,
584
+ "step": 840
585
+ },
586
+ {
587
+ "epoch": 3.33,
588
+ "learning_rate": 6.638048780487805e-05,
589
+ "loss": 0.3804,
590
+ "step": 850
591
+ },
592
+ {
593
+ "epoch": 3.37,
594
+ "learning_rate": 6.599024390243903e-05,
595
+ "loss": 0.3973,
596
+ "step": 860
597
+ },
598
+ {
599
+ "epoch": 3.41,
600
+ "learning_rate": 6.56e-05,
601
+ "loss": 0.3987,
602
+ "step": 870
603
+ },
604
+ {
605
+ "epoch": 3.45,
606
+ "learning_rate": 6.520975609756098e-05,
607
+ "loss": 0.3903,
608
+ "step": 880
609
+ },
610
+ {
611
+ "epoch": 3.49,
612
+ "learning_rate": 6.481951219512196e-05,
613
+ "loss": 0.3469,
614
+ "step": 890
615
+ },
616
+ {
617
+ "epoch": 3.53,
618
+ "learning_rate": 6.442926829268294e-05,
619
+ "loss": 0.323,
620
+ "step": 900
621
+ },
622
+ {
623
+ "epoch": 3.53,
624
+ "eval_accuracy": 0.7615,
625
+ "eval_loss": 1.1008645296096802,
626
+ "eval_runtime": 41.4264,
627
+ "eval_samples_per_second": 48.278,
628
+ "eval_steps_per_second": 1.521,
629
+ "step": 900
630
+ },
631
+ {
632
+ "epoch": 3.57,
633
+ "learning_rate": 6.403902439024391e-05,
634
+ "loss": 0.3168,
635
+ "step": 910
636
+ },
637
+ {
638
+ "epoch": 3.61,
639
+ "learning_rate": 6.364878048780489e-05,
640
+ "loss": 0.3404,
641
+ "step": 920
642
+ },
643
+ {
644
+ "epoch": 3.65,
645
+ "learning_rate": 6.325853658536586e-05,
646
+ "loss": 0.3553,
647
+ "step": 930
648
+ },
649
+ {
650
+ "epoch": 3.69,
651
+ "learning_rate": 6.286829268292684e-05,
652
+ "loss": 0.3665,
653
+ "step": 940
654
+ },
655
+ {
656
+ "epoch": 3.73,
657
+ "learning_rate": 6.247804878048781e-05,
658
+ "loss": 0.2983,
659
+ "step": 950
660
+ },
661
+ {
662
+ "epoch": 3.76,
663
+ "learning_rate": 6.208780487804879e-05,
664
+ "loss": 0.2886,
665
+ "step": 960
666
+ },
667
+ {
668
+ "epoch": 3.8,
669
+ "learning_rate": 6.169756097560976e-05,
670
+ "loss": 0.2721,
671
+ "step": 970
672
+ },
673
+ {
674
+ "epoch": 3.84,
675
+ "learning_rate": 6.130731707317074e-05,
676
+ "loss": 0.2898,
677
+ "step": 980
678
+ },
679
+ {
680
+ "epoch": 3.88,
681
+ "learning_rate": 6.0917073170731714e-05,
682
+ "loss": 0.3222,
683
+ "step": 990
684
+ },
685
+ {
686
+ "epoch": 3.92,
687
+ "learning_rate": 6.052682926829269e-05,
688
+ "loss": 0.2993,
689
+ "step": 1000
690
+ },
691
+ {
692
+ "epoch": 3.92,
693
+ "eval_accuracy": 0.766,
694
+ "eval_loss": 1.0115569829940796,
695
+ "eval_runtime": 40.1299,
696
+ "eval_samples_per_second": 49.838,
697
+ "eval_steps_per_second": 1.57,
698
+ "step": 1000
699
+ },
700
+ {
701
+ "epoch": 3.96,
702
+ "learning_rate": 6.0136585365853664e-05,
703
+ "loss": 0.2903,
704
+ "step": 1010
705
+ },
706
+ {
707
+ "epoch": 4.0,
708
+ "learning_rate": 5.974634146341464e-05,
709
+ "loss": 0.2707,
710
+ "step": 1020
711
+ },
712
+ {
713
+ "epoch": 4.04,
714
+ "learning_rate": 5.9356097560975614e-05,
715
+ "loss": 0.0953,
716
+ "step": 1030
717
+ },
718
+ {
719
+ "epoch": 4.08,
720
+ "learning_rate": 5.896585365853659e-05,
721
+ "loss": 0.1066,
722
+ "step": 1040
723
+ },
724
+ {
725
+ "epoch": 4.12,
726
+ "learning_rate": 5.8575609756097564e-05,
727
+ "loss": 0.0944,
728
+ "step": 1050
729
+ },
730
+ {
731
+ "epoch": 4.16,
732
+ "learning_rate": 5.818536585365854e-05,
733
+ "loss": 0.0943,
734
+ "step": 1060
735
+ },
736
+ {
737
+ "epoch": 4.2,
738
+ "learning_rate": 5.779512195121951e-05,
739
+ "loss": 0.0985,
740
+ "step": 1070
741
+ },
742
+ {
743
+ "epoch": 4.24,
744
+ "learning_rate": 5.740487804878049e-05,
745
+ "loss": 0.1034,
746
+ "step": 1080
747
+ },
748
+ {
749
+ "epoch": 4.27,
750
+ "learning_rate": 5.701463414634146e-05,
751
+ "loss": 0.1015,
752
+ "step": 1090
753
+ },
754
+ {
755
+ "epoch": 4.31,
756
+ "learning_rate": 5.662439024390244e-05,
757
+ "loss": 0.1024,
758
+ "step": 1100
759
+ },
760
+ {
761
+ "epoch": 4.31,
762
+ "eval_accuracy": 0.7895,
763
+ "eval_loss": 0.9027227759361267,
764
+ "eval_runtime": 40.6004,
765
+ "eval_samples_per_second": 49.261,
766
+ "eval_steps_per_second": 1.552,
767
+ "step": 1100
768
+ },
769
+ {
770
+ "epoch": 4.35,
771
+ "learning_rate": 5.623414634146341e-05,
772
+ "loss": 0.0779,
773
+ "step": 1110
774
+ },
775
+ {
776
+ "epoch": 4.39,
777
+ "learning_rate": 5.584390243902439e-05,
778
+ "loss": 0.0754,
779
+ "step": 1120
780
+ },
781
+ {
782
+ "epoch": 4.43,
783
+ "learning_rate": 5.545365853658537e-05,
784
+ "loss": 0.0754,
785
+ "step": 1130
786
+ },
787
+ {
788
+ "epoch": 4.47,
789
+ "learning_rate": 5.5063414634146344e-05,
790
+ "loss": 0.0724,
791
+ "step": 1140
792
+ },
793
+ {
794
+ "epoch": 4.51,
795
+ "learning_rate": 5.4673170731707326e-05,
796
+ "loss": 0.0765,
797
+ "step": 1150
798
+ },
799
+ {
800
+ "epoch": 4.55,
801
+ "learning_rate": 5.42829268292683e-05,
802
+ "loss": 0.0854,
803
+ "step": 1160
804
+ },
805
+ {
806
+ "epoch": 4.59,
807
+ "learning_rate": 5.3892682926829276e-05,
808
+ "loss": 0.0619,
809
+ "step": 1170
810
+ },
811
+ {
812
+ "epoch": 4.63,
813
+ "learning_rate": 5.350243902439025e-05,
814
+ "loss": 0.0679,
815
+ "step": 1180
816
+ },
817
+ {
818
+ "epoch": 4.67,
819
+ "learning_rate": 5.3112195121951225e-05,
820
+ "loss": 0.0836,
821
+ "step": 1190
822
+ },
823
+ {
824
+ "epoch": 4.71,
825
+ "learning_rate": 5.27219512195122e-05,
826
+ "loss": 0.097,
827
+ "step": 1200
828
+ },
829
+ {
830
+ "epoch": 4.71,
831
+ "eval_accuracy": 0.7825,
832
+ "eval_loss": 0.8696740865707397,
833
+ "eval_runtime": 39.9428,
834
+ "eval_samples_per_second": 50.072,
835
+ "eval_steps_per_second": 1.577,
836
+ "step": 1200
837
+ },
838
+ {
839
+ "epoch": 4.75,
840
+ "learning_rate": 5.2331707317073175e-05,
841
+ "loss": 0.0763,
842
+ "step": 1210
843
+ },
844
+ {
845
+ "epoch": 4.78,
846
+ "learning_rate": 5.194146341463415e-05,
847
+ "loss": 0.0863,
848
+ "step": 1220
849
+ },
850
+ {
851
+ "epoch": 4.82,
852
+ "learning_rate": 5.1551219512195125e-05,
853
+ "loss": 0.0827,
854
+ "step": 1230
855
+ },
856
+ {
857
+ "epoch": 4.86,
858
+ "learning_rate": 5.11609756097561e-05,
859
+ "loss": 0.0929,
860
+ "step": 1240
861
+ },
862
+ {
863
+ "epoch": 4.9,
864
+ "learning_rate": 5.0770731707317075e-05,
865
+ "loss": 0.0619,
866
+ "step": 1250
867
+ },
868
+ {
869
+ "epoch": 4.94,
870
+ "learning_rate": 5.0380487804878056e-05,
871
+ "loss": 0.0827,
872
+ "step": 1260
873
+ },
874
+ {
875
+ "epoch": 4.98,
876
+ "learning_rate": 4.999024390243903e-05,
877
+ "loss": 0.0694,
878
+ "step": 1270
879
+ },
880
+ {
881
+ "epoch": 5.02,
882
+ "learning_rate": 4.9600000000000006e-05,
883
+ "loss": 0.0603,
884
+ "step": 1280
885
+ },
886
+ {
887
+ "epoch": 5.06,
888
+ "learning_rate": 4.920975609756098e-05,
889
+ "loss": 0.0233,
890
+ "step": 1290
891
+ },
892
+ {
893
+ "epoch": 5.1,
894
+ "learning_rate": 4.8819512195121956e-05,
895
+ "loss": 0.0273,
896
+ "step": 1300
897
+ },
898
+ {
899
+ "epoch": 5.1,
900
+ "eval_accuracy": 0.805,
901
+ "eval_loss": 0.7953792810440063,
902
+ "eval_runtime": 39.9571,
903
+ "eval_samples_per_second": 50.054,
904
+ "eval_steps_per_second": 1.577,
905
+ "step": 1300
906
+ },
907
+ {
908
+ "epoch": 5.14,
909
+ "learning_rate": 4.842926829268293e-05,
910
+ "loss": 0.0197,
911
+ "step": 1310
912
+ },
913
+ {
914
+ "epoch": 5.18,
915
+ "learning_rate": 4.8039024390243906e-05,
916
+ "loss": 0.0194,
917
+ "step": 1320
918
+ },
919
+ {
920
+ "epoch": 5.22,
921
+ "learning_rate": 4.764878048780488e-05,
922
+ "loss": 0.0326,
923
+ "step": 1330
924
+ },
925
+ {
926
+ "epoch": 5.25,
927
+ "learning_rate": 4.7258536585365856e-05,
928
+ "loss": 0.0367,
929
+ "step": 1340
930
+ },
931
+ {
932
+ "epoch": 5.29,
933
+ "learning_rate": 4.686829268292683e-05,
934
+ "loss": 0.0172,
935
+ "step": 1350
936
+ },
937
+ {
938
+ "epoch": 5.33,
939
+ "learning_rate": 4.6478048780487805e-05,
940
+ "loss": 0.0259,
941
+ "step": 1360
942
+ },
943
+ {
944
+ "epoch": 5.37,
945
+ "learning_rate": 4.608780487804878e-05,
946
+ "loss": 0.0176,
947
+ "step": 1370
948
+ },
949
+ {
950
+ "epoch": 5.41,
951
+ "learning_rate": 4.5697560975609755e-05,
952
+ "loss": 0.0233,
953
+ "step": 1380
954
+ },
955
+ {
956
+ "epoch": 5.45,
957
+ "learning_rate": 4.530731707317073e-05,
958
+ "loss": 0.0393,
959
+ "step": 1390
960
+ },
961
+ {
962
+ "epoch": 5.49,
963
+ "learning_rate": 4.491707317073172e-05,
964
+ "loss": 0.0386,
965
+ "step": 1400
966
+ },
967
+ {
968
+ "epoch": 5.49,
969
+ "eval_accuracy": 0.8095,
970
+ "eval_loss": 0.782217264175415,
971
+ "eval_runtime": 40.3652,
972
+ "eval_samples_per_second": 49.548,
973
+ "eval_steps_per_second": 1.561,
974
+ "step": 1400
975
+ },
976
+ {
977
+ "epoch": 5.53,
978
+ "learning_rate": 4.452682926829269e-05,
979
+ "loss": 0.0222,
980
+ "step": 1410
981
+ },
982
+ {
983
+ "epoch": 5.57,
984
+ "learning_rate": 4.413658536585367e-05,
985
+ "loss": 0.041,
986
+ "step": 1420
987
+ },
988
+ {
989
+ "epoch": 5.61,
990
+ "learning_rate": 4.374634146341464e-05,
991
+ "loss": 0.0188,
992
+ "step": 1430
993
+ },
994
+ {
995
+ "epoch": 5.65,
996
+ "learning_rate": 4.335609756097562e-05,
997
+ "loss": 0.032,
998
+ "step": 1440
999
+ },
1000
+ {
1001
+ "epoch": 5.69,
1002
+ "learning_rate": 4.296585365853659e-05,
1003
+ "loss": 0.0159,
1004
+ "step": 1450
1005
+ },
1006
+ {
1007
+ "epoch": 5.73,
1008
+ "learning_rate": 4.257560975609757e-05,
1009
+ "loss": 0.0433,
1010
+ "step": 1460
1011
+ },
1012
+ {
1013
+ "epoch": 5.76,
1014
+ "learning_rate": 4.218536585365854e-05,
1015
+ "loss": 0.0174,
1016
+ "step": 1470
1017
+ },
1018
+ {
1019
+ "epoch": 5.8,
1020
+ "learning_rate": 4.179512195121952e-05,
1021
+ "loss": 0.0307,
1022
+ "step": 1480
1023
+ },
1024
+ {
1025
+ "epoch": 5.84,
1026
+ "learning_rate": 4.140487804878049e-05,
1027
+ "loss": 0.0221,
1028
+ "step": 1490
1029
+ },
1030
+ {
1031
+ "epoch": 5.88,
1032
+ "learning_rate": 4.101463414634147e-05,
1033
+ "loss": 0.0304,
1034
+ "step": 1500
1035
+ },
1036
+ {
1037
+ "epoch": 5.88,
1038
+ "eval_accuracy": 0.8025,
1039
+ "eval_loss": 0.7874125242233276,
1040
+ "eval_runtime": 40.5243,
1041
+ "eval_samples_per_second": 49.353,
1042
+ "eval_steps_per_second": 1.555,
1043
+ "step": 1500
1044
+ },
1045
+ {
1046
+ "epoch": 5.92,
1047
+ "learning_rate": 4.062439024390244e-05,
1048
+ "loss": 0.0429,
1049
+ "step": 1510
1050
+ },
1051
+ {
1052
+ "epoch": 5.96,
1053
+ "learning_rate": 4.023414634146342e-05,
1054
+ "loss": 0.0244,
1055
+ "step": 1520
1056
+ },
1057
+ {
1058
+ "epoch": 6.0,
1059
+ "learning_rate": 3.984390243902439e-05,
1060
+ "loss": 0.0368,
1061
+ "step": 1530
1062
+ },
1063
+ {
1064
+ "epoch": 6.04,
1065
+ "learning_rate": 3.945365853658537e-05,
1066
+ "loss": 0.0104,
1067
+ "step": 1540
1068
+ },
1069
+ {
1070
+ "epoch": 6.08,
1071
+ "learning_rate": 3.906341463414634e-05,
1072
+ "loss": 0.0105,
1073
+ "step": 1550
1074
+ },
1075
+ {
1076
+ "epoch": 6.12,
1077
+ "learning_rate": 3.8673170731707317e-05,
1078
+ "loss": 0.0164,
1079
+ "step": 1560
1080
+ },
1081
+ {
1082
+ "epoch": 6.16,
1083
+ "learning_rate": 3.828292682926829e-05,
1084
+ "loss": 0.009,
1085
+ "step": 1570
1086
+ },
1087
+ {
1088
+ "epoch": 6.2,
1089
+ "learning_rate": 3.7892682926829266e-05,
1090
+ "loss": 0.0101,
1091
+ "step": 1580
1092
+ },
1093
+ {
1094
+ "epoch": 6.24,
1095
+ "learning_rate": 3.750243902439025e-05,
1096
+ "loss": 0.0184,
1097
+ "step": 1590
1098
+ },
1099
+ {
1100
+ "epoch": 6.27,
1101
+ "learning_rate": 3.711219512195122e-05,
1102
+ "loss": 0.0163,
1103
+ "step": 1600
1104
+ },
1105
+ {
1106
+ "epoch": 6.27,
1107
+ "eval_accuracy": 0.816,
1108
+ "eval_loss": 0.7488501071929932,
1109
+ "eval_runtime": 40.3879,
1110
+ "eval_samples_per_second": 49.52,
1111
+ "eval_steps_per_second": 1.56,
1112
+ "step": 1600
1113
+ },
1114
+ {
1115
+ "epoch": 6.31,
1116
+ "learning_rate": 3.67219512195122e-05,
1117
+ "loss": 0.0118,
1118
+ "step": 1610
1119
+ },
1120
+ {
1121
+ "epoch": 6.35,
1122
+ "learning_rate": 3.633170731707317e-05,
1123
+ "loss": 0.0082,
1124
+ "step": 1620
1125
+ },
1126
+ {
1127
+ "epoch": 6.39,
1128
+ "learning_rate": 3.594146341463415e-05,
1129
+ "loss": 0.011,
1130
+ "step": 1630
1131
+ },
1132
+ {
1133
+ "epoch": 6.43,
1134
+ "learning_rate": 3.555121951219512e-05,
1135
+ "loss": 0.0088,
1136
+ "step": 1640
1137
+ },
1138
+ {
1139
+ "epoch": 6.47,
1140
+ "learning_rate": 3.51609756097561e-05,
1141
+ "loss": 0.0403,
1142
+ "step": 1650
1143
+ },
1144
+ {
1145
+ "epoch": 6.51,
1146
+ "learning_rate": 3.477073170731708e-05,
1147
+ "loss": 0.0089,
1148
+ "step": 1660
1149
+ },
1150
+ {
1151
+ "epoch": 6.55,
1152
+ "learning_rate": 3.4380487804878054e-05,
1153
+ "loss": 0.0203,
1154
+ "step": 1670
1155
+ },
1156
+ {
1157
+ "epoch": 6.59,
1158
+ "learning_rate": 3.399024390243903e-05,
1159
+ "loss": 0.0157,
1160
+ "step": 1680
1161
+ },
1162
+ {
1163
+ "epoch": 6.63,
1164
+ "learning_rate": 3.3600000000000004e-05,
1165
+ "loss": 0.0145,
1166
+ "step": 1690
1167
+ },
1168
+ {
1169
+ "epoch": 6.67,
1170
+ "learning_rate": 3.320975609756098e-05,
1171
+ "loss": 0.0171,
1172
+ "step": 1700
1173
+ },
1174
+ {
1175
+ "epoch": 6.67,
1176
+ "eval_accuracy": 0.816,
1177
+ "eval_loss": 0.7466872930526733,
1178
+ "eval_runtime": 40.0569,
1179
+ "eval_samples_per_second": 49.929,
1180
+ "eval_steps_per_second": 1.573,
1181
+ "step": 1700
1182
+ },
1183
+ {
1184
+ "epoch": 6.71,
1185
+ "learning_rate": 3.281951219512195e-05,
1186
+ "loss": 0.0115,
1187
+ "step": 1710
1188
+ },
1189
+ {
1190
+ "epoch": 6.75,
1191
+ "learning_rate": 3.2429268292682935e-05,
1192
+ "loss": 0.0207,
1193
+ "step": 1720
1194
+ },
1195
+ {
1196
+ "epoch": 6.78,
1197
+ "learning_rate": 3.203902439024391e-05,
1198
+ "loss": 0.038,
1199
+ "step": 1730
1200
+ },
1201
+ {
1202
+ "epoch": 6.82,
1203
+ "learning_rate": 3.1648780487804885e-05,
1204
+ "loss": 0.0289,
1205
+ "step": 1740
1206
+ },
1207
+ {
1208
+ "epoch": 6.86,
1209
+ "learning_rate": 3.125853658536586e-05,
1210
+ "loss": 0.0261,
1211
+ "step": 1750
1212
+ },
1213
+ {
1214
+ "epoch": 6.9,
1215
+ "learning_rate": 3.0868292682926835e-05,
1216
+ "loss": 0.0191,
1217
+ "step": 1760
1218
+ },
1219
+ {
1220
+ "epoch": 6.94,
1221
+ "learning_rate": 3.047804878048781e-05,
1222
+ "loss": 0.008,
1223
+ "step": 1770
1224
+ },
1225
+ {
1226
+ "epoch": 6.98,
1227
+ "learning_rate": 3.0087804878048784e-05,
1228
+ "loss": 0.0076,
1229
+ "step": 1780
1230
+ },
1231
+ {
1232
+ "epoch": 7.02,
1233
+ "learning_rate": 2.969756097560976e-05,
1234
+ "loss": 0.0085,
1235
+ "step": 1790
1236
+ },
1237
+ {
1238
+ "epoch": 7.06,
1239
+ "learning_rate": 2.9307317073170734e-05,
1240
+ "loss": 0.0068,
1241
+ "step": 1800
1242
+ },
1243
+ {
1244
+ "epoch": 7.06,
1245
+ "eval_accuracy": 0.8155,
1246
+ "eval_loss": 0.7402506470680237,
1247
+ "eval_runtime": 40.0633,
1248
+ "eval_samples_per_second": 49.921,
1249
+ "eval_steps_per_second": 1.573,
1250
+ "step": 1800
1251
+ },
1252
+ {
1253
+ "epoch": 7.1,
1254
+ "learning_rate": 2.891707317073171e-05,
1255
+ "loss": 0.0132,
1256
+ "step": 1810
1257
+ },
1258
+ {
1259
+ "epoch": 7.14,
1260
+ "learning_rate": 2.8526829268292684e-05,
1261
+ "loss": 0.0136,
1262
+ "step": 1820
1263
+ },
1264
+ {
1265
+ "epoch": 7.18,
1266
+ "learning_rate": 2.813658536585366e-05,
1267
+ "loss": 0.0107,
1268
+ "step": 1830
1269
+ },
1270
+ {
1271
+ "epoch": 7.22,
1272
+ "learning_rate": 2.7746341463414634e-05,
1273
+ "loss": 0.0069,
1274
+ "step": 1840
1275
+ },
1276
+ {
1277
+ "epoch": 7.25,
1278
+ "learning_rate": 2.7356097560975615e-05,
1279
+ "loss": 0.0152,
1280
+ "step": 1850
1281
+ },
1282
+ {
1283
+ "epoch": 7.29,
1284
+ "learning_rate": 2.696585365853659e-05,
1285
+ "loss": 0.0135,
1286
+ "step": 1860
1287
+ },
1288
+ {
1289
+ "epoch": 7.33,
1290
+ "learning_rate": 2.6575609756097565e-05,
1291
+ "loss": 0.0134,
1292
+ "step": 1870
1293
+ },
1294
+ {
1295
+ "epoch": 7.37,
1296
+ "learning_rate": 2.618536585365854e-05,
1297
+ "loss": 0.0174,
1298
+ "step": 1880
1299
+ },
1300
+ {
1301
+ "epoch": 7.41,
1302
+ "learning_rate": 2.5795121951219515e-05,
1303
+ "loss": 0.0062,
1304
+ "step": 1890
1305
+ },
1306
+ {
1307
+ "epoch": 7.45,
1308
+ "learning_rate": 2.540487804878049e-05,
1309
+ "loss": 0.0064,
1310
+ "step": 1900
1311
+ },
1312
+ {
1313
+ "epoch": 7.45,
1314
+ "eval_accuracy": 0.8225,
1315
+ "eval_loss": 0.7162501215934753,
1316
+ "eval_runtime": 40.5713,
1317
+ "eval_samples_per_second": 49.296,
1318
+ "eval_steps_per_second": 1.553,
1319
+ "step": 1900
1320
+ },
1321
+ {
1322
+ "epoch": 7.49,
1323
+ "learning_rate": 2.5014634146341465e-05,
1324
+ "loss": 0.0058,
1325
+ "step": 1910
1326
+ },
1327
+ {
1328
+ "epoch": 7.53,
1329
+ "learning_rate": 2.462439024390244e-05,
1330
+ "loss": 0.0056,
1331
+ "step": 1920
1332
+ },
1333
+ {
1334
+ "epoch": 7.57,
1335
+ "learning_rate": 2.4234146341463414e-05,
1336
+ "loss": 0.006,
1337
+ "step": 1930
1338
+ },
1339
+ {
1340
+ "epoch": 7.61,
1341
+ "learning_rate": 2.3843902439024393e-05,
1342
+ "loss": 0.0175,
1343
+ "step": 1940
1344
+ },
1345
+ {
1346
+ "epoch": 7.65,
1347
+ "learning_rate": 2.3453658536585367e-05,
1348
+ "loss": 0.0062,
1349
+ "step": 1950
1350
+ },
1351
+ {
1352
+ "epoch": 7.69,
1353
+ "learning_rate": 2.3063414634146342e-05,
1354
+ "loss": 0.0062,
1355
+ "step": 1960
1356
+ },
1357
+ {
1358
+ "epoch": 7.73,
1359
+ "learning_rate": 2.2673170731707317e-05,
1360
+ "loss": 0.0166,
1361
+ "step": 1970
1362
+ },
1363
+ {
1364
+ "epoch": 7.76,
1365
+ "learning_rate": 2.2282926829268296e-05,
1366
+ "loss": 0.0121,
1367
+ "step": 1980
1368
+ },
1369
+ {
1370
+ "epoch": 7.8,
1371
+ "learning_rate": 2.189268292682927e-05,
1372
+ "loss": 0.0119,
1373
+ "step": 1990
1374
+ },
1375
+ {
1376
+ "epoch": 7.84,
1377
+ "learning_rate": 2.150243902439025e-05,
1378
+ "loss": 0.0236,
1379
+ "step": 2000
1380
+ },
1381
+ {
1382
+ "epoch": 7.84,
1383
+ "eval_accuracy": 0.824,
1384
+ "eval_loss": 0.7153184413909912,
1385
+ "eval_runtime": 40.1832,
1386
+ "eval_samples_per_second": 49.772,
1387
+ "eval_steps_per_second": 1.568,
1388
+ "step": 2000
1389
+ },
1390
+ {
1391
+ "epoch": 7.88,
1392
+ "learning_rate": 2.1112195121951224e-05,
1393
+ "loss": 0.0141,
1394
+ "step": 2010
1395
+ },
1396
+ {
1397
+ "epoch": 7.92,
1398
+ "learning_rate": 2.07219512195122e-05,
1399
+ "loss": 0.016,
1400
+ "step": 2020
1401
+ },
1402
+ {
1403
+ "epoch": 7.96,
1404
+ "learning_rate": 2.0331707317073173e-05,
1405
+ "loss": 0.0173,
1406
+ "step": 2030
1407
+ },
1408
+ {
1409
+ "epoch": 8.0,
1410
+ "learning_rate": 1.9941463414634148e-05,
1411
+ "loss": 0.0058,
1412
+ "step": 2040
1413
+ },
1414
+ {
1415
+ "epoch": 8.04,
1416
+ "learning_rate": 1.9551219512195123e-05,
1417
+ "loss": 0.007,
1418
+ "step": 2050
1419
+ },
1420
+ {
1421
+ "epoch": 8.08,
1422
+ "learning_rate": 1.9160975609756098e-05,
1423
+ "loss": 0.007,
1424
+ "step": 2060
1425
+ },
1426
+ {
1427
+ "epoch": 8.12,
1428
+ "learning_rate": 1.8770731707317073e-05,
1429
+ "loss": 0.0052,
1430
+ "step": 2070
1431
+ },
1432
+ {
1433
+ "epoch": 8.16,
1434
+ "learning_rate": 1.838048780487805e-05,
1435
+ "loss": 0.0059,
1436
+ "step": 2080
1437
+ },
1438
+ {
1439
+ "epoch": 8.2,
1440
+ "learning_rate": 1.7990243902439026e-05,
1441
+ "loss": 0.0051,
1442
+ "step": 2090
1443
+ },
1444
+ {
1445
+ "epoch": 8.24,
1446
+ "learning_rate": 1.76e-05,
1447
+ "loss": 0.0075,
1448
+ "step": 2100
1449
+ },
1450
+ {
1451
+ "epoch": 8.24,
1452
+ "eval_accuracy": 0.8255,
1453
+ "eval_loss": 0.7127741575241089,
1454
+ "eval_runtime": 40.1888,
1455
+ "eval_samples_per_second": 49.765,
1456
+ "eval_steps_per_second": 1.568,
1457
+ "step": 2100
1458
+ },
1459
+ {
1460
+ "epoch": 8.27,
1461
+ "learning_rate": 1.7209756097560976e-05,
1462
+ "loss": 0.0049,
1463
+ "step": 2110
1464
+ },
1465
+ {
1466
+ "epoch": 8.31,
1467
+ "learning_rate": 1.681951219512195e-05,
1468
+ "loss": 0.0087,
1469
+ "step": 2120
1470
+ },
1471
+ {
1472
+ "epoch": 8.35,
1473
+ "learning_rate": 1.6429268292682926e-05,
1474
+ "loss": 0.0065,
1475
+ "step": 2130
1476
+ },
1477
+ {
1478
+ "epoch": 8.39,
1479
+ "learning_rate": 1.6039024390243904e-05,
1480
+ "loss": 0.0053,
1481
+ "step": 2140
1482
+ },
1483
+ {
1484
+ "epoch": 8.43,
1485
+ "learning_rate": 1.564878048780488e-05,
1486
+ "loss": 0.0133,
1487
+ "step": 2150
1488
+ },
1489
+ {
1490
+ "epoch": 8.47,
1491
+ "learning_rate": 1.5258536585365855e-05,
1492
+ "loss": 0.0186,
1493
+ "step": 2160
1494
+ },
1495
+ {
1496
+ "epoch": 8.51,
1497
+ "learning_rate": 1.486829268292683e-05,
1498
+ "loss": 0.0178,
1499
+ "step": 2170
1500
+ },
1501
+ {
1502
+ "epoch": 8.55,
1503
+ "learning_rate": 1.4478048780487805e-05,
1504
+ "loss": 0.0051,
1505
+ "step": 2180
1506
+ },
1507
+ {
1508
+ "epoch": 8.59,
1509
+ "learning_rate": 1.4087804878048782e-05,
1510
+ "loss": 0.0051,
1511
+ "step": 2190
1512
+ },
1513
+ {
1514
+ "epoch": 8.63,
1515
+ "learning_rate": 1.3697560975609758e-05,
1516
+ "loss": 0.0154,
1517
+ "step": 2200
1518
+ },
1519
+ {
1520
+ "epoch": 8.63,
1521
+ "eval_accuracy": 0.8235,
1522
+ "eval_loss": 0.7104279398918152,
1523
+ "eval_runtime": 39.6742,
1524
+ "eval_samples_per_second": 50.411,
1525
+ "eval_steps_per_second": 1.588,
1526
+ "step": 2200
1527
+ },
1528
+ {
1529
+ "epoch": 8.67,
1530
+ "learning_rate": 1.3307317073170733e-05,
1531
+ "loss": 0.0051,
1532
+ "step": 2210
1533
+ },
1534
+ {
1535
+ "epoch": 8.71,
1536
+ "learning_rate": 1.2917073170731708e-05,
1537
+ "loss": 0.0097,
1538
+ "step": 2220
1539
+ },
1540
+ {
1541
+ "epoch": 8.75,
1542
+ "learning_rate": 1.2526829268292685e-05,
1543
+ "loss": 0.0213,
1544
+ "step": 2230
1545
+ },
1546
+ {
1547
+ "epoch": 8.78,
1548
+ "learning_rate": 1.213658536585366e-05,
1549
+ "loss": 0.0091,
1550
+ "step": 2240
1551
+ },
1552
+ {
1553
+ "epoch": 8.82,
1554
+ "learning_rate": 1.1746341463414634e-05,
1555
+ "loss": 0.0087,
1556
+ "step": 2250
1557
+ },
1558
+ {
1559
+ "epoch": 8.86,
1560
+ "learning_rate": 1.135609756097561e-05,
1561
+ "loss": 0.005,
1562
+ "step": 2260
1563
+ },
1564
+ {
1565
+ "epoch": 8.9,
1566
+ "learning_rate": 1.0965853658536587e-05,
1567
+ "loss": 0.0119,
1568
+ "step": 2270
1569
+ },
1570
+ {
1571
+ "epoch": 8.94,
1572
+ "learning_rate": 1.0575609756097562e-05,
1573
+ "loss": 0.0049,
1574
+ "step": 2280
1575
+ },
1576
+ {
1577
+ "epoch": 8.98,
1578
+ "learning_rate": 1.0185365853658537e-05,
1579
+ "loss": 0.0152,
1580
+ "step": 2290
1581
+ },
1582
+ {
1583
+ "epoch": 9.02,
1584
+ "learning_rate": 9.795121951219512e-06,
1585
+ "loss": 0.0082,
1586
+ "step": 2300
1587
+ },
1588
+ {
1589
+ "epoch": 9.02,
1590
+ "eval_accuracy": 0.824,
1591
+ "eval_loss": 0.7048487067222595,
1592
+ "eval_runtime": 39.8204,
1593
+ "eval_samples_per_second": 50.226,
1594
+ "eval_steps_per_second": 1.582,
1595
+ "step": 2300
1596
+ },
1597
+ {
1598
+ "epoch": 9.06,
1599
+ "learning_rate": 9.404878048780489e-06,
1600
+ "loss": 0.0046,
1601
+ "step": 2310
1602
+ },
1603
+ {
1604
+ "epoch": 9.1,
1605
+ "learning_rate": 9.014634146341464e-06,
1606
+ "loss": 0.0076,
1607
+ "step": 2320
1608
+ },
1609
+ {
1610
+ "epoch": 9.14,
1611
+ "learning_rate": 8.62439024390244e-06,
1612
+ "loss": 0.0107,
1613
+ "step": 2330
1614
+ },
1615
+ {
1616
+ "epoch": 9.18,
1617
+ "learning_rate": 8.234146341463415e-06,
1618
+ "loss": 0.0182,
1619
+ "step": 2340
1620
+ },
1621
+ {
1622
+ "epoch": 9.22,
1623
+ "learning_rate": 7.843902439024392e-06,
1624
+ "loss": 0.0045,
1625
+ "step": 2350
1626
+ },
1627
+ {
1628
+ "epoch": 9.25,
1629
+ "learning_rate": 7.4536585365853665e-06,
1630
+ "loss": 0.0047,
1631
+ "step": 2360
1632
+ },
1633
+ {
1634
+ "epoch": 9.29,
1635
+ "learning_rate": 7.063414634146341e-06,
1636
+ "loss": 0.0047,
1637
+ "step": 2370
1638
+ },
1639
+ {
1640
+ "epoch": 9.33,
1641
+ "learning_rate": 6.673170731707318e-06,
1642
+ "loss": 0.0045,
1643
+ "step": 2380
1644
+ },
1645
+ {
1646
+ "epoch": 9.37,
1647
+ "learning_rate": 6.282926829268293e-06,
1648
+ "loss": 0.0162,
1649
+ "step": 2390
1650
+ },
1651
+ {
1652
+ "epoch": 9.41,
1653
+ "learning_rate": 5.892682926829269e-06,
1654
+ "loss": 0.007,
1655
+ "step": 2400
1656
+ },
1657
+ {
1658
+ "epoch": 9.41,
1659
+ "eval_accuracy": 0.823,
1660
+ "eval_loss": 0.7026473879814148,
1661
+ "eval_runtime": 40.761,
1662
+ "eval_samples_per_second": 49.067,
1663
+ "eval_steps_per_second": 1.546,
1664
+ "step": 2400
1665
+ },
1666
+ {
1667
+ "epoch": 9.45,
1668
+ "learning_rate": 5.502439024390245e-06,
1669
+ "loss": 0.0074,
1670
+ "step": 2410
1671
+ },
1672
+ {
1673
+ "epoch": 9.49,
1674
+ "learning_rate": 5.11219512195122e-06,
1675
+ "loss": 0.0045,
1676
+ "step": 2420
1677
+ },
1678
+ {
1679
+ "epoch": 9.53,
1680
+ "learning_rate": 4.721951219512195e-06,
1681
+ "loss": 0.0044,
1682
+ "step": 2430
1683
+ },
1684
+ {
1685
+ "epoch": 9.57,
1686
+ "learning_rate": 4.331707317073171e-06,
1687
+ "loss": 0.0046,
1688
+ "step": 2440
1689
+ },
1690
+ {
1691
+ "epoch": 9.61,
1692
+ "learning_rate": 3.941463414634147e-06,
1693
+ "loss": 0.0066,
1694
+ "step": 2450
1695
+ },
1696
+ {
1697
+ "epoch": 9.65,
1698
+ "learning_rate": 3.551219512195122e-06,
1699
+ "loss": 0.009,
1700
+ "step": 2460
1701
+ },
1702
+ {
1703
+ "epoch": 9.69,
1704
+ "learning_rate": 3.160975609756098e-06,
1705
+ "loss": 0.0045,
1706
+ "step": 2470
1707
+ },
1708
+ {
1709
+ "epoch": 9.73,
1710
+ "learning_rate": 2.7707317073170736e-06,
1711
+ "loss": 0.0046,
1712
+ "step": 2480
1713
+ },
1714
+ {
1715
+ "epoch": 9.76,
1716
+ "learning_rate": 2.380487804878049e-06,
1717
+ "loss": 0.0147,
1718
+ "step": 2490
1719
+ },
1720
+ {
1721
+ "epoch": 9.8,
1722
+ "learning_rate": 1.9902439024390246e-06,
1723
+ "loss": 0.0045,
1724
+ "step": 2500
1725
+ },
1726
+ {
1727
+ "epoch": 9.8,
1728
+ "eval_accuracy": 0.8235,
1729
+ "eval_loss": 0.7031954526901245,
1730
+ "eval_runtime": 39.643,
1731
+ "eval_samples_per_second": 50.45,
1732
+ "eval_steps_per_second": 1.589,
1733
+ "step": 2500
1734
+ }
1735
+ ],
1736
+ "max_steps": 2550,
1737
+ "num_train_epochs": 10,
1738
+ "total_flos": 6.198960121799639e+18,
1739
+ "trial_name": null,
1740
+ "trial_params": null
1741
+ }