jalaneunos commited on
Commit
ad44ebe
1 Parent(s): 509c27e

End of training

Browse files
README.md CHANGED
@@ -22,7 +22,7 @@ model-index:
22
  metrics:
23
  - name: Accuracy
24
  type: accuracy
25
- value: 0.8655520724486242
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
32
 
33
  This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the imagefolder dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.3327
36
- - Accuracy: 0.8656
37
 
38
  ## Model description
39
 
 
22
  metrics:
23
  - name: Accuracy
24
  type: accuracy
25
+ value: 0.8683385579937304
26
  ---
27
 
28
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 
32
 
33
  This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the imagefolder dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.3321
36
+ - Accuracy: 0.8683
37
 
38
  ## Model description
39
 
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.8683385579937304,
4
+ "eval_loss": 0.3320719301700592,
5
+ "eval_runtime": 19.3598,
6
+ "eval_samples_per_second": 148.297,
7
+ "eval_steps_per_second": 4.649,
8
+ "total_flos": 2.0021605356722135e+19,
9
+ "train_loss": 0.3826778081384036,
10
+ "train_runtime": 3330.0466,
11
+ "train_samples_per_second": 77.588,
12
+ "train_steps_per_second": 0.607
13
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.8683385579937304,
4
+ "eval_loss": 0.3320719301700592,
5
+ "eval_runtime": 19.3598,
6
+ "eval_samples_per_second": 148.297,
7
+ "eval_steps_per_second": 4.649
8
+ }
runs/Jan24_07-31-05_e814a5b30138/events.out.tfevents.1706084902.e814a5b30138.1747.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ea7569ced746c4808c302cf828eaa751a86767d9bf47ada2d6c6260b3b540ea
3
+ size 411
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "total_flos": 2.0021605356722135e+19,
4
+ "train_loss": 0.3826778081384036,
5
+ "train_runtime": 3330.0466,
6
+ "train_samples_per_second": 77.588,
7
+ "train_steps_per_second": 0.607
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1332 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.8683385579937304,
3
+ "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-FER2013/checkpoint-1616",
4
+ "epoch": 10.0,
5
+ "eval_steps": 500,
6
+ "global_step": 2020,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05,
13
+ "learning_rate": 2.4752475247524753e-06,
14
+ "loss": 0.7062,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.1,
19
+ "learning_rate": 4.950495049504951e-06,
20
+ "loss": 0.6984,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.15,
25
+ "learning_rate": 7.4257425742574256e-06,
26
+ "loss": 0.6891,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.2,
31
+ "learning_rate": 9.900990099009901e-06,
32
+ "loss": 0.6814,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.25,
37
+ "learning_rate": 1.2376237623762377e-05,
38
+ "loss": 0.6688,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.3,
43
+ "learning_rate": 1.4851485148514851e-05,
44
+ "loss": 0.6468,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.35,
49
+ "learning_rate": 1.7326732673267325e-05,
50
+ "loss": 0.6276,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.4,
55
+ "learning_rate": 1.9801980198019803e-05,
56
+ "loss": 0.6003,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.45,
61
+ "learning_rate": 2.227722772277228e-05,
62
+ "loss": 0.5689,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.5,
67
+ "learning_rate": 2.4752475247524754e-05,
68
+ "loss": 0.5499,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.54,
73
+ "learning_rate": 2.722772277227723e-05,
74
+ "loss": 0.5432,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.59,
79
+ "learning_rate": 2.9702970297029702e-05,
80
+ "loss": 0.5408,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.64,
85
+ "learning_rate": 3.217821782178218e-05,
86
+ "loss": 0.535,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.69,
91
+ "learning_rate": 3.465346534653465e-05,
92
+ "loss": 0.5331,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.74,
97
+ "learning_rate": 3.712871287128713e-05,
98
+ "loss": 0.4929,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.79,
103
+ "learning_rate": 3.9603960396039605e-05,
104
+ "loss": 0.5054,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.84,
109
+ "learning_rate": 4.207920792079208e-05,
110
+ "loss": 0.4882,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.89,
115
+ "learning_rate": 4.455445544554456e-05,
116
+ "loss": 0.4844,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.94,
121
+ "learning_rate": 4.702970297029703e-05,
122
+ "loss": 0.5134,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.99,
127
+ "learning_rate": 4.950495049504951e-05,
128
+ "loss": 0.495,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 1.0,
133
+ "eval_accuracy": 0.7739463601532567,
134
+ "eval_loss": 0.46603715419769287,
135
+ "eval_runtime": 15.7268,
136
+ "eval_samples_per_second": 182.555,
137
+ "eval_steps_per_second": 5.723,
138
+ "step": 202
139
+ },
140
+ {
141
+ "epoch": 1.04,
142
+ "learning_rate": 4.977997799779978e-05,
143
+ "loss": 0.5214,
144
+ "step": 210
145
+ },
146
+ {
147
+ "epoch": 1.09,
148
+ "learning_rate": 4.950495049504951e-05,
149
+ "loss": 0.5009,
150
+ "step": 220
151
+ },
152
+ {
153
+ "epoch": 1.14,
154
+ "learning_rate": 4.9229922992299234e-05,
155
+ "loss": 0.4632,
156
+ "step": 230
157
+ },
158
+ {
159
+ "epoch": 1.19,
160
+ "learning_rate": 4.895489548954896e-05,
161
+ "loss": 0.4585,
162
+ "step": 240
163
+ },
164
+ {
165
+ "epoch": 1.24,
166
+ "learning_rate": 4.867986798679868e-05,
167
+ "loss": 0.4339,
168
+ "step": 250
169
+ },
170
+ {
171
+ "epoch": 1.29,
172
+ "learning_rate": 4.8404840484048406e-05,
173
+ "loss": 0.4703,
174
+ "step": 260
175
+ },
176
+ {
177
+ "epoch": 1.34,
178
+ "learning_rate": 4.812981298129813e-05,
179
+ "loss": 0.4748,
180
+ "step": 270
181
+ },
182
+ {
183
+ "epoch": 1.39,
184
+ "learning_rate": 4.785478547854786e-05,
185
+ "loss": 0.4688,
186
+ "step": 280
187
+ },
188
+ {
189
+ "epoch": 1.44,
190
+ "learning_rate": 4.7579757975797585e-05,
191
+ "loss": 0.4613,
192
+ "step": 290
193
+ },
194
+ {
195
+ "epoch": 1.49,
196
+ "learning_rate": 4.730473047304731e-05,
197
+ "loss": 0.4633,
198
+ "step": 300
199
+ },
200
+ {
201
+ "epoch": 1.53,
202
+ "learning_rate": 4.702970297029703e-05,
203
+ "loss": 0.4661,
204
+ "step": 310
205
+ },
206
+ {
207
+ "epoch": 1.58,
208
+ "learning_rate": 4.675467546754676e-05,
209
+ "loss": 0.4476,
210
+ "step": 320
211
+ },
212
+ {
213
+ "epoch": 1.63,
214
+ "learning_rate": 4.647964796479648e-05,
215
+ "loss": 0.458,
216
+ "step": 330
217
+ },
218
+ {
219
+ "epoch": 1.68,
220
+ "learning_rate": 4.62046204620462e-05,
221
+ "loss": 0.4606,
222
+ "step": 340
223
+ },
224
+ {
225
+ "epoch": 1.73,
226
+ "learning_rate": 4.592959295929593e-05,
227
+ "loss": 0.4433,
228
+ "step": 350
229
+ },
230
+ {
231
+ "epoch": 1.78,
232
+ "learning_rate": 4.5654565456545655e-05,
233
+ "loss": 0.4568,
234
+ "step": 360
235
+ },
236
+ {
237
+ "epoch": 1.83,
238
+ "learning_rate": 4.537953795379538e-05,
239
+ "loss": 0.4395,
240
+ "step": 370
241
+ },
242
+ {
243
+ "epoch": 1.88,
244
+ "learning_rate": 4.510451045104511e-05,
245
+ "loss": 0.448,
246
+ "step": 380
247
+ },
248
+ {
249
+ "epoch": 1.93,
250
+ "learning_rate": 4.4829482948294834e-05,
251
+ "loss": 0.4722,
252
+ "step": 390
253
+ },
254
+ {
255
+ "epoch": 1.98,
256
+ "learning_rate": 4.455445544554456e-05,
257
+ "loss": 0.4632,
258
+ "step": 400
259
+ },
260
+ {
261
+ "epoch": 2.0,
262
+ "eval_accuracy": 0.8286311389759665,
263
+ "eval_loss": 0.3819791376590729,
264
+ "eval_runtime": 15.8456,
265
+ "eval_samples_per_second": 181.185,
266
+ "eval_steps_per_second": 5.68,
267
+ "step": 404
268
+ },
269
+ {
270
+ "epoch": 2.03,
271
+ "learning_rate": 4.427942794279428e-05,
272
+ "loss": 0.4614,
273
+ "step": 410
274
+ },
275
+ {
276
+ "epoch": 2.08,
277
+ "learning_rate": 4.4004400440044006e-05,
278
+ "loss": 0.4644,
279
+ "step": 420
280
+ },
281
+ {
282
+ "epoch": 2.13,
283
+ "learning_rate": 4.372937293729373e-05,
284
+ "loss": 0.4325,
285
+ "step": 430
286
+ },
287
+ {
288
+ "epoch": 2.18,
289
+ "learning_rate": 4.345434543454346e-05,
290
+ "loss": 0.4297,
291
+ "step": 440
292
+ },
293
+ {
294
+ "epoch": 2.23,
295
+ "learning_rate": 4.3179317931793185e-05,
296
+ "loss": 0.456,
297
+ "step": 450
298
+ },
299
+ {
300
+ "epoch": 2.28,
301
+ "learning_rate": 4.2904290429042904e-05,
302
+ "loss": 0.4119,
303
+ "step": 460
304
+ },
305
+ {
306
+ "epoch": 2.33,
307
+ "learning_rate": 4.262926292629263e-05,
308
+ "loss": 0.4101,
309
+ "step": 470
310
+ },
311
+ {
312
+ "epoch": 2.38,
313
+ "learning_rate": 4.2354235423542356e-05,
314
+ "loss": 0.3827,
315
+ "step": 480
316
+ },
317
+ {
318
+ "epoch": 2.43,
319
+ "learning_rate": 4.207920792079208e-05,
320
+ "loss": 0.4177,
321
+ "step": 490
322
+ },
323
+ {
324
+ "epoch": 2.48,
325
+ "learning_rate": 4.18041804180418e-05,
326
+ "loss": 0.4286,
327
+ "step": 500
328
+ },
329
+ {
330
+ "epoch": 2.52,
331
+ "learning_rate": 4.152915291529153e-05,
332
+ "loss": 0.4471,
333
+ "step": 510
334
+ },
335
+ {
336
+ "epoch": 2.57,
337
+ "learning_rate": 4.1254125412541255e-05,
338
+ "loss": 0.4304,
339
+ "step": 520
340
+ },
341
+ {
342
+ "epoch": 2.62,
343
+ "learning_rate": 4.097909790979098e-05,
344
+ "loss": 0.4043,
345
+ "step": 530
346
+ },
347
+ {
348
+ "epoch": 2.67,
349
+ "learning_rate": 4.070407040704071e-05,
350
+ "loss": 0.4277,
351
+ "step": 540
352
+ },
353
+ {
354
+ "epoch": 2.72,
355
+ "learning_rate": 4.042904290429043e-05,
356
+ "loss": 0.403,
357
+ "step": 550
358
+ },
359
+ {
360
+ "epoch": 2.77,
361
+ "learning_rate": 4.015401540154016e-05,
362
+ "loss": 0.4138,
363
+ "step": 560
364
+ },
365
+ {
366
+ "epoch": 2.82,
367
+ "learning_rate": 3.987898789878988e-05,
368
+ "loss": 0.4005,
369
+ "step": 570
370
+ },
371
+ {
372
+ "epoch": 2.87,
373
+ "learning_rate": 3.9603960396039605e-05,
374
+ "loss": 0.407,
375
+ "step": 580
376
+ },
377
+ {
378
+ "epoch": 2.92,
379
+ "learning_rate": 3.932893289328933e-05,
380
+ "loss": 0.4124,
381
+ "step": 590
382
+ },
383
+ {
384
+ "epoch": 2.97,
385
+ "learning_rate": 3.905390539053906e-05,
386
+ "loss": 0.4013,
387
+ "step": 600
388
+ },
389
+ {
390
+ "epoch": 3.0,
391
+ "eval_accuracy": 0.8446534308603274,
392
+ "eval_loss": 0.35617050528526306,
393
+ "eval_runtime": 16.2713,
394
+ "eval_samples_per_second": 176.445,
395
+ "eval_steps_per_second": 5.531,
396
+ "step": 606
397
+ },
398
+ {
399
+ "epoch": 3.02,
400
+ "learning_rate": 3.877887788778878e-05,
401
+ "loss": 0.3881,
402
+ "step": 610
403
+ },
404
+ {
405
+ "epoch": 3.07,
406
+ "learning_rate": 3.8503850385038503e-05,
407
+ "loss": 0.3761,
408
+ "step": 620
409
+ },
410
+ {
411
+ "epoch": 3.12,
412
+ "learning_rate": 3.822882288228823e-05,
413
+ "loss": 0.4161,
414
+ "step": 630
415
+ },
416
+ {
417
+ "epoch": 3.17,
418
+ "learning_rate": 3.7953795379537956e-05,
419
+ "loss": 0.3975,
420
+ "step": 640
421
+ },
422
+ {
423
+ "epoch": 3.22,
424
+ "learning_rate": 3.767876787678768e-05,
425
+ "loss": 0.4224,
426
+ "step": 650
427
+ },
428
+ {
429
+ "epoch": 3.27,
430
+ "learning_rate": 3.74037403740374e-05,
431
+ "loss": 0.3849,
432
+ "step": 660
433
+ },
434
+ {
435
+ "epoch": 3.32,
436
+ "learning_rate": 3.712871287128713e-05,
437
+ "loss": 0.4178,
438
+ "step": 670
439
+ },
440
+ {
441
+ "epoch": 3.37,
442
+ "learning_rate": 3.6853685368536854e-05,
443
+ "loss": 0.3966,
444
+ "step": 680
445
+ },
446
+ {
447
+ "epoch": 3.42,
448
+ "learning_rate": 3.657865786578658e-05,
449
+ "loss": 0.3886,
450
+ "step": 690
451
+ },
452
+ {
453
+ "epoch": 3.47,
454
+ "learning_rate": 3.6303630363036307e-05,
455
+ "loss": 0.4021,
456
+ "step": 700
457
+ },
458
+ {
459
+ "epoch": 3.51,
460
+ "learning_rate": 3.602860286028603e-05,
461
+ "loss": 0.3913,
462
+ "step": 710
463
+ },
464
+ {
465
+ "epoch": 3.56,
466
+ "learning_rate": 3.575357535753576e-05,
467
+ "loss": 0.3886,
468
+ "step": 720
469
+ },
470
+ {
471
+ "epoch": 3.61,
472
+ "learning_rate": 3.5478547854785485e-05,
473
+ "loss": 0.4066,
474
+ "step": 730
475
+ },
476
+ {
477
+ "epoch": 3.66,
478
+ "learning_rate": 3.5203520352035205e-05,
479
+ "loss": 0.401,
480
+ "step": 740
481
+ },
482
+ {
483
+ "epoch": 3.71,
484
+ "learning_rate": 3.492849284928493e-05,
485
+ "loss": 0.4002,
486
+ "step": 750
487
+ },
488
+ {
489
+ "epoch": 3.76,
490
+ "learning_rate": 3.465346534653465e-05,
491
+ "loss": 0.3738,
492
+ "step": 760
493
+ },
494
+ {
495
+ "epoch": 3.81,
496
+ "learning_rate": 3.4378437843784377e-05,
497
+ "loss": 0.4218,
498
+ "step": 770
499
+ },
500
+ {
501
+ "epoch": 3.86,
502
+ "learning_rate": 3.41034103410341e-05,
503
+ "loss": 0.368,
504
+ "step": 780
505
+ },
506
+ {
507
+ "epoch": 3.91,
508
+ "learning_rate": 3.382838283828383e-05,
509
+ "loss": 0.3969,
510
+ "step": 790
511
+ },
512
+ {
513
+ "epoch": 3.96,
514
+ "learning_rate": 3.3553355335533555e-05,
515
+ "loss": 0.3883,
516
+ "step": 800
517
+ },
518
+ {
519
+ "epoch": 4.0,
520
+ "eval_accuracy": 0.851619644723093,
521
+ "eval_loss": 0.3426118791103363,
522
+ "eval_runtime": 15.7929,
523
+ "eval_samples_per_second": 181.79,
524
+ "eval_steps_per_second": 5.699,
525
+ "step": 808
526
+ },
527
+ {
528
+ "epoch": 4.01,
529
+ "learning_rate": 3.327832783278328e-05,
530
+ "loss": 0.3605,
531
+ "step": 810
532
+ },
533
+ {
534
+ "epoch": 4.06,
535
+ "learning_rate": 3.300330033003301e-05,
536
+ "loss": 0.3797,
537
+ "step": 820
538
+ },
539
+ {
540
+ "epoch": 4.11,
541
+ "learning_rate": 3.272827282728273e-05,
542
+ "loss": 0.3701,
543
+ "step": 830
544
+ },
545
+ {
546
+ "epoch": 4.16,
547
+ "learning_rate": 3.2453245324532453e-05,
548
+ "loss": 0.3968,
549
+ "step": 840
550
+ },
551
+ {
552
+ "epoch": 4.21,
553
+ "learning_rate": 3.217821782178218e-05,
554
+ "loss": 0.3754,
555
+ "step": 850
556
+ },
557
+ {
558
+ "epoch": 4.26,
559
+ "learning_rate": 3.1903190319031906e-05,
560
+ "loss": 0.3962,
561
+ "step": 860
562
+ },
563
+ {
564
+ "epoch": 4.31,
565
+ "learning_rate": 3.162816281628163e-05,
566
+ "loss": 0.3723,
567
+ "step": 870
568
+ },
569
+ {
570
+ "epoch": 4.36,
571
+ "learning_rate": 3.135313531353136e-05,
572
+ "loss": 0.3743,
573
+ "step": 880
574
+ },
575
+ {
576
+ "epoch": 4.41,
577
+ "learning_rate": 3.1078107810781085e-05,
578
+ "loss": 0.4,
579
+ "step": 890
580
+ },
581
+ {
582
+ "epoch": 4.46,
583
+ "learning_rate": 3.0803080308030804e-05,
584
+ "loss": 0.3827,
585
+ "step": 900
586
+ },
587
+ {
588
+ "epoch": 4.5,
589
+ "learning_rate": 3.052805280528053e-05,
590
+ "loss": 0.3488,
591
+ "step": 910
592
+ },
593
+ {
594
+ "epoch": 4.55,
595
+ "learning_rate": 3.0253025302530253e-05,
596
+ "loss": 0.3755,
597
+ "step": 920
598
+ },
599
+ {
600
+ "epoch": 4.6,
601
+ "learning_rate": 2.9977997799779976e-05,
602
+ "loss": 0.3695,
603
+ "step": 930
604
+ },
605
+ {
606
+ "epoch": 4.65,
607
+ "learning_rate": 2.9702970297029702e-05,
608
+ "loss": 0.3803,
609
+ "step": 940
610
+ },
611
+ {
612
+ "epoch": 4.7,
613
+ "learning_rate": 2.942794279427943e-05,
614
+ "loss": 0.404,
615
+ "step": 950
616
+ },
617
+ {
618
+ "epoch": 4.75,
619
+ "learning_rate": 2.9152915291529155e-05,
620
+ "loss": 0.3711,
621
+ "step": 960
622
+ },
623
+ {
624
+ "epoch": 4.8,
625
+ "learning_rate": 2.8877887788778878e-05,
626
+ "loss": 0.3495,
627
+ "step": 970
628
+ },
629
+ {
630
+ "epoch": 4.85,
631
+ "learning_rate": 2.8602860286028604e-05,
632
+ "loss": 0.3644,
633
+ "step": 980
634
+ },
635
+ {
636
+ "epoch": 4.9,
637
+ "learning_rate": 2.832783278327833e-05,
638
+ "loss": 0.3673,
639
+ "step": 990
640
+ },
641
+ {
642
+ "epoch": 4.95,
643
+ "learning_rate": 2.8052805280528056e-05,
644
+ "loss": 0.3597,
645
+ "step": 1000
646
+ },
647
+ {
648
+ "epoch": 5.0,
649
+ "learning_rate": 2.777777777777778e-05,
650
+ "loss": 0.3801,
651
+ "step": 1010
652
+ },
653
+ {
654
+ "epoch": 5.0,
655
+ "eval_accuracy": 0.8561476837338906,
656
+ "eval_loss": 0.33033424615859985,
657
+ "eval_runtime": 16.0262,
658
+ "eval_samples_per_second": 179.144,
659
+ "eval_steps_per_second": 5.616,
660
+ "step": 1010
661
+ },
662
+ {
663
+ "epoch": 5.05,
664
+ "learning_rate": 2.7502750275027505e-05,
665
+ "loss": 0.3717,
666
+ "step": 1020
667
+ },
668
+ {
669
+ "epoch": 5.1,
670
+ "learning_rate": 2.722772277227723e-05,
671
+ "loss": 0.3409,
672
+ "step": 1030
673
+ },
674
+ {
675
+ "epoch": 5.15,
676
+ "learning_rate": 2.6952695269526958e-05,
677
+ "loss": 0.3647,
678
+ "step": 1040
679
+ },
680
+ {
681
+ "epoch": 5.2,
682
+ "learning_rate": 2.667766776677668e-05,
683
+ "loss": 0.3454,
684
+ "step": 1050
685
+ },
686
+ {
687
+ "epoch": 5.25,
688
+ "learning_rate": 2.64026402640264e-05,
689
+ "loss": 0.3592,
690
+ "step": 1060
691
+ },
692
+ {
693
+ "epoch": 5.3,
694
+ "learning_rate": 2.6127612761276126e-05,
695
+ "loss": 0.3529,
696
+ "step": 1070
697
+ },
698
+ {
699
+ "epoch": 5.35,
700
+ "learning_rate": 2.5852585258525853e-05,
701
+ "loss": 0.3598,
702
+ "step": 1080
703
+ },
704
+ {
705
+ "epoch": 5.4,
706
+ "learning_rate": 2.557755775577558e-05,
707
+ "loss": 0.3922,
708
+ "step": 1090
709
+ },
710
+ {
711
+ "epoch": 5.45,
712
+ "learning_rate": 2.53025302530253e-05,
713
+ "loss": 0.3547,
714
+ "step": 1100
715
+ },
716
+ {
717
+ "epoch": 5.5,
718
+ "learning_rate": 2.5027502750275028e-05,
719
+ "loss": 0.3432,
720
+ "step": 1110
721
+ },
722
+ {
723
+ "epoch": 5.54,
724
+ "learning_rate": 2.4752475247524754e-05,
725
+ "loss": 0.3542,
726
+ "step": 1120
727
+ },
728
+ {
729
+ "epoch": 5.59,
730
+ "learning_rate": 2.447744774477448e-05,
731
+ "loss": 0.3541,
732
+ "step": 1130
733
+ },
734
+ {
735
+ "epoch": 5.64,
736
+ "learning_rate": 2.4202420242024203e-05,
737
+ "loss": 0.3475,
738
+ "step": 1140
739
+ },
740
+ {
741
+ "epoch": 5.69,
742
+ "learning_rate": 2.392739273927393e-05,
743
+ "loss": 0.3528,
744
+ "step": 1150
745
+ },
746
+ {
747
+ "epoch": 5.74,
748
+ "learning_rate": 2.3652365236523656e-05,
749
+ "loss": 0.3446,
750
+ "step": 1160
751
+ },
752
+ {
753
+ "epoch": 5.79,
754
+ "learning_rate": 2.337733773377338e-05,
755
+ "loss": 0.3264,
756
+ "step": 1170
757
+ },
758
+ {
759
+ "epoch": 5.84,
760
+ "learning_rate": 2.31023102310231e-05,
761
+ "loss": 0.3327,
762
+ "step": 1180
763
+ },
764
+ {
765
+ "epoch": 5.89,
766
+ "learning_rate": 2.2827282728272828e-05,
767
+ "loss": 0.3529,
768
+ "step": 1190
769
+ },
770
+ {
771
+ "epoch": 5.94,
772
+ "learning_rate": 2.2552255225522554e-05,
773
+ "loss": 0.3473,
774
+ "step": 1200
775
+ },
776
+ {
777
+ "epoch": 5.99,
778
+ "learning_rate": 2.227722772277228e-05,
779
+ "loss": 0.3612,
780
+ "step": 1210
781
+ },
782
+ {
783
+ "epoch": 6.0,
784
+ "eval_accuracy": 0.8557993730407524,
785
+ "eval_loss": 0.3362487852573395,
786
+ "eval_runtime": 15.9082,
787
+ "eval_samples_per_second": 180.474,
788
+ "eval_steps_per_second": 5.657,
789
+ "step": 1212
790
+ },
791
+ {
792
+ "epoch": 6.04,
793
+ "learning_rate": 2.2002200220022003e-05,
794
+ "loss": 0.3324,
795
+ "step": 1220
796
+ },
797
+ {
798
+ "epoch": 6.09,
799
+ "learning_rate": 2.172717271727173e-05,
800
+ "loss": 0.3365,
801
+ "step": 1230
802
+ },
803
+ {
804
+ "epoch": 6.14,
805
+ "learning_rate": 2.1452145214521452e-05,
806
+ "loss": 0.3212,
807
+ "step": 1240
808
+ },
809
+ {
810
+ "epoch": 6.19,
811
+ "learning_rate": 2.1177117711771178e-05,
812
+ "loss": 0.33,
813
+ "step": 1250
814
+ },
815
+ {
816
+ "epoch": 6.24,
817
+ "learning_rate": 2.09020902090209e-05,
818
+ "loss": 0.3627,
819
+ "step": 1260
820
+ },
821
+ {
822
+ "epoch": 6.29,
823
+ "learning_rate": 2.0627062706270627e-05,
824
+ "loss": 0.3322,
825
+ "step": 1270
826
+ },
827
+ {
828
+ "epoch": 6.34,
829
+ "learning_rate": 2.0352035203520354e-05,
830
+ "loss": 0.3236,
831
+ "step": 1280
832
+ },
833
+ {
834
+ "epoch": 6.39,
835
+ "learning_rate": 2.007700770077008e-05,
836
+ "loss": 0.3398,
837
+ "step": 1290
838
+ },
839
+ {
840
+ "epoch": 6.44,
841
+ "learning_rate": 1.9801980198019803e-05,
842
+ "loss": 0.3481,
843
+ "step": 1300
844
+ },
845
+ {
846
+ "epoch": 6.49,
847
+ "learning_rate": 1.952695269526953e-05,
848
+ "loss": 0.3421,
849
+ "step": 1310
850
+ },
851
+ {
852
+ "epoch": 6.53,
853
+ "learning_rate": 1.9251925192519252e-05,
854
+ "loss": 0.3136,
855
+ "step": 1320
856
+ },
857
+ {
858
+ "epoch": 6.58,
859
+ "learning_rate": 1.8976897689768978e-05,
860
+ "loss": 0.3432,
861
+ "step": 1330
862
+ },
863
+ {
864
+ "epoch": 6.63,
865
+ "learning_rate": 1.87018701870187e-05,
866
+ "loss": 0.3277,
867
+ "step": 1340
868
+ },
869
+ {
870
+ "epoch": 6.68,
871
+ "learning_rate": 1.8426842684268427e-05,
872
+ "loss": 0.3285,
873
+ "step": 1350
874
+ },
875
+ {
876
+ "epoch": 6.73,
877
+ "learning_rate": 1.8151815181518153e-05,
878
+ "loss": 0.3449,
879
+ "step": 1360
880
+ },
881
+ {
882
+ "epoch": 6.78,
883
+ "learning_rate": 1.787678767876788e-05,
884
+ "loss": 0.3433,
885
+ "step": 1370
886
+ },
887
+ {
888
+ "epoch": 6.83,
889
+ "learning_rate": 1.7601760176017602e-05,
890
+ "loss": 0.3275,
891
+ "step": 1380
892
+ },
893
+ {
894
+ "epoch": 6.88,
895
+ "learning_rate": 1.7326732673267325e-05,
896
+ "loss": 0.3351,
897
+ "step": 1390
898
+ },
899
+ {
900
+ "epoch": 6.93,
901
+ "learning_rate": 1.705170517051705e-05,
902
+ "loss": 0.2946,
903
+ "step": 1400
904
+ },
905
+ {
906
+ "epoch": 6.98,
907
+ "learning_rate": 1.6776677667766778e-05,
908
+ "loss": 0.3504,
909
+ "step": 1410
910
+ },
911
+ {
912
+ "epoch": 7.0,
913
+ "eval_accuracy": 0.8652037617554859,
914
+ "eval_loss": 0.330201655626297,
915
+ "eval_runtime": 15.6907,
916
+ "eval_samples_per_second": 182.974,
917
+ "eval_steps_per_second": 5.736,
918
+ "step": 1414
919
+ },
920
+ {
921
+ "epoch": 7.03,
922
+ "learning_rate": 1.6501650165016504e-05,
923
+ "loss": 0.3424,
924
+ "step": 1420
925
+ },
926
+ {
927
+ "epoch": 7.08,
928
+ "learning_rate": 1.6226622662266227e-05,
929
+ "loss": 0.3244,
930
+ "step": 1430
931
+ },
932
+ {
933
+ "epoch": 7.13,
934
+ "learning_rate": 1.5951595159515953e-05,
935
+ "loss": 0.3208,
936
+ "step": 1440
937
+ },
938
+ {
939
+ "epoch": 7.18,
940
+ "learning_rate": 1.567656765676568e-05,
941
+ "loss": 0.3002,
942
+ "step": 1450
943
+ },
944
+ {
945
+ "epoch": 7.23,
946
+ "learning_rate": 1.5401540154015402e-05,
947
+ "loss": 0.3007,
948
+ "step": 1460
949
+ },
950
+ {
951
+ "epoch": 7.28,
952
+ "learning_rate": 1.5126512651265127e-05,
953
+ "loss": 0.3443,
954
+ "step": 1470
955
+ },
956
+ {
957
+ "epoch": 7.33,
958
+ "learning_rate": 1.4851485148514851e-05,
959
+ "loss": 0.3192,
960
+ "step": 1480
961
+ },
962
+ {
963
+ "epoch": 7.38,
964
+ "learning_rate": 1.4576457645764577e-05,
965
+ "loss": 0.3146,
966
+ "step": 1490
967
+ },
968
+ {
969
+ "epoch": 7.43,
970
+ "learning_rate": 1.4301430143014302e-05,
971
+ "loss": 0.3147,
972
+ "step": 1500
973
+ },
974
+ {
975
+ "epoch": 7.48,
976
+ "learning_rate": 1.4026402640264028e-05,
977
+ "loss": 0.3041,
978
+ "step": 1510
979
+ },
980
+ {
981
+ "epoch": 7.52,
982
+ "learning_rate": 1.3751375137513753e-05,
983
+ "loss": 0.2958,
984
+ "step": 1520
985
+ },
986
+ {
987
+ "epoch": 7.57,
988
+ "learning_rate": 1.3476347634763479e-05,
989
+ "loss": 0.3235,
990
+ "step": 1530
991
+ },
992
+ {
993
+ "epoch": 7.62,
994
+ "learning_rate": 1.32013201320132e-05,
995
+ "loss": 0.3137,
996
+ "step": 1540
997
+ },
998
+ {
999
+ "epoch": 7.67,
1000
+ "learning_rate": 1.2926292629262926e-05,
1001
+ "loss": 0.3222,
1002
+ "step": 1550
1003
+ },
1004
+ {
1005
+ "epoch": 7.72,
1006
+ "learning_rate": 1.265126512651265e-05,
1007
+ "loss": 0.3087,
1008
+ "step": 1560
1009
+ },
1010
+ {
1011
+ "epoch": 7.77,
1012
+ "learning_rate": 1.2376237623762377e-05,
1013
+ "loss": 0.2969,
1014
+ "step": 1570
1015
+ },
1016
+ {
1017
+ "epoch": 7.82,
1018
+ "learning_rate": 1.2101210121012102e-05,
1019
+ "loss": 0.3091,
1020
+ "step": 1580
1021
+ },
1022
+ {
1023
+ "epoch": 7.87,
1024
+ "learning_rate": 1.1826182618261828e-05,
1025
+ "loss": 0.3356,
1026
+ "step": 1590
1027
+ },
1028
+ {
1029
+ "epoch": 7.92,
1030
+ "learning_rate": 1.155115511551155e-05,
1031
+ "loss": 0.2903,
1032
+ "step": 1600
1033
+ },
1034
+ {
1035
+ "epoch": 7.97,
1036
+ "learning_rate": 1.1276127612761277e-05,
1037
+ "loss": 0.3366,
1038
+ "step": 1610
1039
+ },
1040
+ {
1041
+ "epoch": 8.0,
1042
+ "eval_accuracy": 0.8683385579937304,
1043
+ "eval_loss": 0.3320719301700592,
1044
+ "eval_runtime": 15.9857,
1045
+ "eval_samples_per_second": 179.598,
1046
+ "eval_steps_per_second": 5.63,
1047
+ "step": 1616
1048
+ },
1049
+ {
1050
+ "epoch": 8.02,
1051
+ "learning_rate": 1.1001100110011001e-05,
1052
+ "loss": 0.2956,
1053
+ "step": 1620
1054
+ },
1055
+ {
1056
+ "epoch": 8.07,
1057
+ "learning_rate": 1.0726072607260726e-05,
1058
+ "loss": 0.2987,
1059
+ "step": 1630
1060
+ },
1061
+ {
1062
+ "epoch": 8.12,
1063
+ "learning_rate": 1.045104510451045e-05,
1064
+ "loss": 0.2965,
1065
+ "step": 1640
1066
+ },
1067
+ {
1068
+ "epoch": 8.17,
1069
+ "learning_rate": 1.0176017601760177e-05,
1070
+ "loss": 0.3014,
1071
+ "step": 1650
1072
+ },
1073
+ {
1074
+ "epoch": 8.22,
1075
+ "learning_rate": 9.900990099009901e-06,
1076
+ "loss": 0.3267,
1077
+ "step": 1660
1078
+ },
1079
+ {
1080
+ "epoch": 8.27,
1081
+ "learning_rate": 9.625962596259626e-06,
1082
+ "loss": 0.3243,
1083
+ "step": 1670
1084
+ },
1085
+ {
1086
+ "epoch": 8.32,
1087
+ "learning_rate": 9.35093509350935e-06,
1088
+ "loss": 0.2841,
1089
+ "step": 1680
1090
+ },
1091
+ {
1092
+ "epoch": 8.37,
1093
+ "learning_rate": 9.075907590759077e-06,
1094
+ "loss": 0.2949,
1095
+ "step": 1690
1096
+ },
1097
+ {
1098
+ "epoch": 8.42,
1099
+ "learning_rate": 8.800880088008801e-06,
1100
+ "loss": 0.2944,
1101
+ "step": 1700
1102
+ },
1103
+ {
1104
+ "epoch": 8.47,
1105
+ "learning_rate": 8.525852585258526e-06,
1106
+ "loss": 0.3137,
1107
+ "step": 1710
1108
+ },
1109
+ {
1110
+ "epoch": 8.51,
1111
+ "learning_rate": 8.250825082508252e-06,
1112
+ "loss": 0.2825,
1113
+ "step": 1720
1114
+ },
1115
+ {
1116
+ "epoch": 8.56,
1117
+ "learning_rate": 7.975797579757976e-06,
1118
+ "loss": 0.3012,
1119
+ "step": 1730
1120
+ },
1121
+ {
1122
+ "epoch": 8.61,
1123
+ "learning_rate": 7.700770077007701e-06,
1124
+ "loss": 0.3221,
1125
+ "step": 1740
1126
+ },
1127
+ {
1128
+ "epoch": 8.66,
1129
+ "learning_rate": 7.4257425742574256e-06,
1130
+ "loss": 0.2977,
1131
+ "step": 1750
1132
+ },
1133
+ {
1134
+ "epoch": 8.71,
1135
+ "learning_rate": 7.150715071507151e-06,
1136
+ "loss": 0.316,
1137
+ "step": 1760
1138
+ },
1139
+ {
1140
+ "epoch": 8.76,
1141
+ "learning_rate": 6.875687568756876e-06,
1142
+ "loss": 0.3023,
1143
+ "step": 1770
1144
+ },
1145
+ {
1146
+ "epoch": 8.81,
1147
+ "learning_rate": 6.6006600660066e-06,
1148
+ "loss": 0.2937,
1149
+ "step": 1780
1150
+ },
1151
+ {
1152
+ "epoch": 8.86,
1153
+ "learning_rate": 6.325632563256325e-06,
1154
+ "loss": 0.2825,
1155
+ "step": 1790
1156
+ },
1157
+ {
1158
+ "epoch": 8.91,
1159
+ "learning_rate": 6.050605060506051e-06,
1160
+ "loss": 0.2907,
1161
+ "step": 1800
1162
+ },
1163
+ {
1164
+ "epoch": 8.96,
1165
+ "learning_rate": 5.775577557755775e-06,
1166
+ "loss": 0.3007,
1167
+ "step": 1810
1168
+ },
1169
+ {
1170
+ "epoch": 9.0,
1171
+ "eval_accuracy": 0.866597004528039,
1172
+ "eval_loss": 0.33304381370544434,
1173
+ "eval_runtime": 16.241,
1174
+ "eval_samples_per_second": 176.775,
1175
+ "eval_steps_per_second": 5.542,
1176
+ "step": 1818
1177
+ },
1178
+ {
1179
+ "epoch": 9.01,
1180
+ "learning_rate": 5.500550055005501e-06,
1181
+ "loss": 0.297,
1182
+ "step": 1820
1183
+ },
1184
+ {
1185
+ "epoch": 9.06,
1186
+ "learning_rate": 5.225522552255225e-06,
1187
+ "loss": 0.2894,
1188
+ "step": 1830
1189
+ },
1190
+ {
1191
+ "epoch": 9.11,
1192
+ "learning_rate": 4.950495049504951e-06,
1193
+ "loss": 0.2927,
1194
+ "step": 1840
1195
+ },
1196
+ {
1197
+ "epoch": 9.16,
1198
+ "learning_rate": 4.675467546754675e-06,
1199
+ "loss": 0.2884,
1200
+ "step": 1850
1201
+ },
1202
+ {
1203
+ "epoch": 9.21,
1204
+ "learning_rate": 4.400440044004401e-06,
1205
+ "loss": 0.3217,
1206
+ "step": 1860
1207
+ },
1208
+ {
1209
+ "epoch": 9.26,
1210
+ "learning_rate": 4.125412541254126e-06,
1211
+ "loss": 0.3278,
1212
+ "step": 1870
1213
+ },
1214
+ {
1215
+ "epoch": 9.31,
1216
+ "learning_rate": 3.8503850385038505e-06,
1217
+ "loss": 0.3126,
1218
+ "step": 1880
1219
+ },
1220
+ {
1221
+ "epoch": 9.36,
1222
+ "learning_rate": 3.5753575357535755e-06,
1223
+ "loss": 0.2997,
1224
+ "step": 1890
1225
+ },
1226
+ {
1227
+ "epoch": 9.41,
1228
+ "learning_rate": 3.3003300330033e-06,
1229
+ "loss": 0.279,
1230
+ "step": 1900
1231
+ },
1232
+ {
1233
+ "epoch": 9.46,
1234
+ "learning_rate": 3.0253025302530254e-06,
1235
+ "loss": 0.2686,
1236
+ "step": 1910
1237
+ },
1238
+ {
1239
+ "epoch": 9.5,
1240
+ "learning_rate": 2.7502750275027504e-06,
1241
+ "loss": 0.2811,
1242
+ "step": 1920
1243
+ },
1244
+ {
1245
+ "epoch": 9.55,
1246
+ "learning_rate": 2.4752475247524753e-06,
1247
+ "loss": 0.2866,
1248
+ "step": 1930
1249
+ },
1250
+ {
1251
+ "epoch": 9.6,
1252
+ "learning_rate": 2.2002200220022003e-06,
1253
+ "loss": 0.2746,
1254
+ "step": 1940
1255
+ },
1256
+ {
1257
+ "epoch": 9.65,
1258
+ "learning_rate": 1.9251925192519253e-06,
1259
+ "loss": 0.2565,
1260
+ "step": 1950
1261
+ },
1262
+ {
1263
+ "epoch": 9.7,
1264
+ "learning_rate": 1.65016501650165e-06,
1265
+ "loss": 0.3004,
1266
+ "step": 1960
1267
+ },
1268
+ {
1269
+ "epoch": 9.75,
1270
+ "learning_rate": 1.3751375137513752e-06,
1271
+ "loss": 0.3053,
1272
+ "step": 1970
1273
+ },
1274
+ {
1275
+ "epoch": 9.8,
1276
+ "learning_rate": 1.1001100110011001e-06,
1277
+ "loss": 0.2805,
1278
+ "step": 1980
1279
+ },
1280
+ {
1281
+ "epoch": 9.85,
1282
+ "learning_rate": 8.25082508250825e-07,
1283
+ "loss": 0.2817,
1284
+ "step": 1990
1285
+ },
1286
+ {
1287
+ "epoch": 9.9,
1288
+ "learning_rate": 5.500550055005501e-07,
1289
+ "loss": 0.3138,
1290
+ "step": 2000
1291
+ },
1292
+ {
1293
+ "epoch": 9.95,
1294
+ "learning_rate": 2.7502750275027504e-07,
1295
+ "loss": 0.2733,
1296
+ "step": 2010
1297
+ },
1298
+ {
1299
+ "epoch": 10.0,
1300
+ "learning_rate": 0.0,
1301
+ "loss": 0.3089,
1302
+ "step": 2020
1303
+ },
1304
+ {
1305
+ "epoch": 10.0,
1306
+ "eval_accuracy": 0.8655520724486242,
1307
+ "eval_loss": 0.3326851427555084,
1308
+ "eval_runtime": 15.8375,
1309
+ "eval_samples_per_second": 181.279,
1310
+ "eval_steps_per_second": 5.683,
1311
+ "step": 2020
1312
+ },
1313
+ {
1314
+ "epoch": 10.0,
1315
+ "step": 2020,
1316
+ "total_flos": 2.0021605356722135e+19,
1317
+ "train_loss": 0.3826778081384036,
1318
+ "train_runtime": 3330.0466,
1319
+ "train_samples_per_second": 77.588,
1320
+ "train_steps_per_second": 0.607
1321
+ }
1322
+ ],
1323
+ "logging_steps": 10,
1324
+ "max_steps": 2020,
1325
+ "num_input_tokens_seen": 0,
1326
+ "num_train_epochs": 10,
1327
+ "save_steps": 500,
1328
+ "total_flos": 2.0021605356722135e+19,
1329
+ "train_batch_size": 32,
1330
+ "trial_name": null,
1331
+ "trial_params": null
1332
+ }