BilalMuftuoglu commited on
Commit
fbdc301
1 Parent(s): 632b705

End of training

Browse files
README.md CHANGED
@@ -32,7 +32,7 @@ should probably proofread and complete it, then remove this comment. -->
32
 
33
  This model is a fine-tuned version of [facebook/deit-base-distilled-patch16-224](https://huggingface.co/facebook/deit-base-distilled-patch16-224) on the imagefolder dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 2.3933
36
  - Accuracy: 0.7273
37
 
38
  ## Model description
 
32
 
33
  This model is a fine-tuned version of [facebook/deit-base-distilled-patch16-224](https://huggingface.co/facebook/deit-base-distilled-patch16-224) on the imagefolder dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.9542
36
  - Accuracy: 0.7273
37
 
38
  ## Model description
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "eval_accuracy": 0.7272727272727273,
4
+ "eval_loss": 0.9542110562324524,
5
+ "eval_runtime": 0.2165,
6
+ "eval_samples_per_second": 50.803,
7
+ "eval_steps_per_second": 4.618,
8
+ "total_flos": 4.572150213593088e+17,
9
+ "train_loss": 0.10843914225697518,
10
+ "train_runtime": 416.0402,
11
+ "train_samples_per_second": 14.181,
12
+ "train_steps_per_second": 0.24
13
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "eval_accuracy": 0.7272727272727273,
4
+ "eval_loss": 0.9542110562324524,
5
+ "eval_runtime": 0.2165,
6
+ "eval_samples_per_second": 50.803,
7
+ "eval_steps_per_second": 4.618
8
+ }
runs/May22_16-16-24_927069b479b5/events.out.tfevents.1716395023.927069b479b5.252.43 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05de8b5b6fa854bc51429bb29cf8394fc35bc04573422d6a40fb3a7a4b03b064
3
+ size 405
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "total_flos": 4.572150213593088e+17,
4
+ "train_loss": 0.10843914225697518,
5
+ "train_runtime": 416.0402,
6
+ "train_samples_per_second": 14.181,
7
+ "train_steps_per_second": 0.24
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1021 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7272727272727273,
3
+ "best_model_checkpoint": "deit-base-distilled-patch16-224-hasta-85-fold4/checkpoint-5",
4
+ "epoch": 100.0,
5
+ "eval_steps": 500,
6
+ "global_step": 100,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 1.0,
13
+ "eval_accuracy": 0.0,
14
+ "eval_loss": 2.0908453464508057,
15
+ "eval_runtime": 0.176,
16
+ "eval_samples_per_second": 62.496,
17
+ "eval_steps_per_second": 5.681,
18
+ "step": 1
19
+ },
20
+ {
21
+ "epoch": 2.0,
22
+ "eval_accuracy": 0.0,
23
+ "eval_loss": 1.8863940238952637,
24
+ "eval_runtime": 0.1665,
25
+ "eval_samples_per_second": 66.07,
26
+ "eval_steps_per_second": 6.006,
27
+ "step": 2
28
+ },
29
+ {
30
+ "epoch": 3.0,
31
+ "eval_accuracy": 0.0,
32
+ "eval_loss": 1.531255841255188,
33
+ "eval_runtime": 0.1783,
34
+ "eval_samples_per_second": 61.703,
35
+ "eval_steps_per_second": 5.609,
36
+ "step": 3
37
+ },
38
+ {
39
+ "epoch": 4.0,
40
+ "eval_accuracy": 0.36363636363636365,
41
+ "eval_loss": 1.152018427848816,
42
+ "eval_runtime": 0.1729,
43
+ "eval_samples_per_second": 63.617,
44
+ "eval_steps_per_second": 5.783,
45
+ "step": 4
46
+ },
47
+ {
48
+ "epoch": 5.0,
49
+ "eval_accuracy": 0.7272727272727273,
50
+ "eval_loss": 0.9542110562324524,
51
+ "eval_runtime": 0.1767,
52
+ "eval_samples_per_second": 62.268,
53
+ "eval_steps_per_second": 5.661,
54
+ "step": 5
55
+ },
56
+ {
57
+ "epoch": 6.0,
58
+ "eval_accuracy": 0.7272727272727273,
59
+ "eval_loss": 1.0324329137802124,
60
+ "eval_runtime": 0.17,
61
+ "eval_samples_per_second": 64.704,
62
+ "eval_steps_per_second": 5.882,
63
+ "step": 6
64
+ },
65
+ {
66
+ "epoch": 7.0,
67
+ "eval_accuracy": 0.7272727272727273,
68
+ "eval_loss": 1.1791726350784302,
69
+ "eval_runtime": 0.1753,
70
+ "eval_samples_per_second": 62.76,
71
+ "eval_steps_per_second": 5.705,
72
+ "step": 7
73
+ },
74
+ {
75
+ "epoch": 8.0,
76
+ "eval_accuracy": 0.7272727272727273,
77
+ "eval_loss": 1.2680414915084839,
78
+ "eval_runtime": 0.1732,
79
+ "eval_samples_per_second": 63.512,
80
+ "eval_steps_per_second": 5.774,
81
+ "step": 8
82
+ },
83
+ {
84
+ "epoch": 9.0,
85
+ "eval_accuracy": 0.7272727272727273,
86
+ "eval_loss": 1.2078349590301514,
87
+ "eval_runtime": 0.1753,
88
+ "eval_samples_per_second": 62.764,
89
+ "eval_steps_per_second": 5.706,
90
+ "step": 9
91
+ },
92
+ {
93
+ "epoch": 10.0,
94
+ "grad_norm": 1.6796222925186157,
95
+ "learning_rate": 5e-05,
96
+ "loss": 0.4805,
97
+ "step": 10
98
+ },
99
+ {
100
+ "epoch": 10.0,
101
+ "eval_accuracy": 0.7272727272727273,
102
+ "eval_loss": 1.159353494644165,
103
+ "eval_runtime": 0.1705,
104
+ "eval_samples_per_second": 64.535,
105
+ "eval_steps_per_second": 5.867,
106
+ "step": 10
107
+ },
108
+ {
109
+ "epoch": 11.0,
110
+ "eval_accuracy": 0.7272727272727273,
111
+ "eval_loss": 1.1603728532791138,
112
+ "eval_runtime": 0.1692,
113
+ "eval_samples_per_second": 65.006,
114
+ "eval_steps_per_second": 5.91,
115
+ "step": 11
116
+ },
117
+ {
118
+ "epoch": 12.0,
119
+ "eval_accuracy": 0.7272727272727273,
120
+ "eval_loss": 1.1109801530838013,
121
+ "eval_runtime": 0.1707,
122
+ "eval_samples_per_second": 64.449,
123
+ "eval_steps_per_second": 5.859,
124
+ "step": 12
125
+ },
126
+ {
127
+ "epoch": 13.0,
128
+ "eval_accuracy": 0.7272727272727273,
129
+ "eval_loss": 1.1073939800262451,
130
+ "eval_runtime": 0.1697,
131
+ "eval_samples_per_second": 64.813,
132
+ "eval_steps_per_second": 5.892,
133
+ "step": 13
134
+ },
135
+ {
136
+ "epoch": 14.0,
137
+ "eval_accuracy": 0.7272727272727273,
138
+ "eval_loss": 1.1010178327560425,
139
+ "eval_runtime": 0.1677,
140
+ "eval_samples_per_second": 65.604,
141
+ "eval_steps_per_second": 5.964,
142
+ "step": 14
143
+ },
144
+ {
145
+ "epoch": 15.0,
146
+ "eval_accuracy": 0.7272727272727273,
147
+ "eval_loss": 1.0594475269317627,
148
+ "eval_runtime": 0.1682,
149
+ "eval_samples_per_second": 65.406,
150
+ "eval_steps_per_second": 5.946,
151
+ "step": 15
152
+ },
153
+ {
154
+ "epoch": 16.0,
155
+ "eval_accuracy": 0.7272727272727273,
156
+ "eval_loss": 1.0675151348114014,
157
+ "eval_runtime": 0.1681,
158
+ "eval_samples_per_second": 65.438,
159
+ "eval_steps_per_second": 5.949,
160
+ "step": 16
161
+ },
162
+ {
163
+ "epoch": 17.0,
164
+ "eval_accuracy": 0.7272727272727273,
165
+ "eval_loss": 1.168864130973816,
166
+ "eval_runtime": 0.1694,
167
+ "eval_samples_per_second": 64.924,
168
+ "eval_steps_per_second": 5.902,
169
+ "step": 17
170
+ },
171
+ {
172
+ "epoch": 18.0,
173
+ "eval_accuracy": 0.7272727272727273,
174
+ "eval_loss": 1.2629767656326294,
175
+ "eval_runtime": 0.1679,
176
+ "eval_samples_per_second": 65.502,
177
+ "eval_steps_per_second": 5.955,
178
+ "step": 18
179
+ },
180
+ {
181
+ "epoch": 19.0,
182
+ "eval_accuracy": 0.7272727272727273,
183
+ "eval_loss": 1.3017388582229614,
184
+ "eval_runtime": 0.1683,
185
+ "eval_samples_per_second": 65.373,
186
+ "eval_steps_per_second": 5.943,
187
+ "step": 19
188
+ },
189
+ {
190
+ "epoch": 20.0,
191
+ "grad_norm": 2.425222158432007,
192
+ "learning_rate": 4.4444444444444447e-05,
193
+ "loss": 0.1904,
194
+ "step": 20
195
+ },
196
+ {
197
+ "epoch": 20.0,
198
+ "eval_accuracy": 0.7272727272727273,
199
+ "eval_loss": 1.3082902431488037,
200
+ "eval_runtime": 0.1663,
201
+ "eval_samples_per_second": 66.163,
202
+ "eval_steps_per_second": 6.015,
203
+ "step": 20
204
+ },
205
+ {
206
+ "epoch": 21.0,
207
+ "eval_accuracy": 0.7272727272727273,
208
+ "eval_loss": 1.3764029741287231,
209
+ "eval_runtime": 0.1644,
210
+ "eval_samples_per_second": 66.912,
211
+ "eval_steps_per_second": 6.083,
212
+ "step": 21
213
+ },
214
+ {
215
+ "epoch": 22.0,
216
+ "eval_accuracy": 0.7272727272727273,
217
+ "eval_loss": 1.4228819608688354,
218
+ "eval_runtime": 0.1685,
219
+ "eval_samples_per_second": 65.281,
220
+ "eval_steps_per_second": 5.935,
221
+ "step": 22
222
+ },
223
+ {
224
+ "epoch": 23.0,
225
+ "eval_accuracy": 0.7272727272727273,
226
+ "eval_loss": 1.4829826354980469,
227
+ "eval_runtime": 0.1675,
228
+ "eval_samples_per_second": 65.667,
229
+ "eval_steps_per_second": 5.97,
230
+ "step": 23
231
+ },
232
+ {
233
+ "epoch": 24.0,
234
+ "eval_accuracy": 0.7272727272727273,
235
+ "eval_loss": 1.4698455333709717,
236
+ "eval_runtime": 0.1727,
237
+ "eval_samples_per_second": 63.676,
238
+ "eval_steps_per_second": 5.789,
239
+ "step": 24
240
+ },
241
+ {
242
+ "epoch": 25.0,
243
+ "eval_accuracy": 0.7272727272727273,
244
+ "eval_loss": 1.4946540594100952,
245
+ "eval_runtime": 0.1741,
246
+ "eval_samples_per_second": 63.2,
247
+ "eval_steps_per_second": 5.745,
248
+ "step": 25
249
+ },
250
+ {
251
+ "epoch": 26.0,
252
+ "eval_accuracy": 0.7272727272727273,
253
+ "eval_loss": 1.5838810205459595,
254
+ "eval_runtime": 0.1781,
255
+ "eval_samples_per_second": 61.761,
256
+ "eval_steps_per_second": 5.615,
257
+ "step": 26
258
+ },
259
+ {
260
+ "epoch": 27.0,
261
+ "eval_accuracy": 0.7272727272727273,
262
+ "eval_loss": 1.7219489812850952,
263
+ "eval_runtime": 0.171,
264
+ "eval_samples_per_second": 64.317,
265
+ "eval_steps_per_second": 5.847,
266
+ "step": 27
267
+ },
268
+ {
269
+ "epoch": 28.0,
270
+ "eval_accuracy": 0.7272727272727273,
271
+ "eval_loss": 1.7473819255828857,
272
+ "eval_runtime": 0.1677,
273
+ "eval_samples_per_second": 65.576,
274
+ "eval_steps_per_second": 5.961,
275
+ "step": 28
276
+ },
277
+ {
278
+ "epoch": 29.0,
279
+ "eval_accuracy": 0.7272727272727273,
280
+ "eval_loss": 1.6644595861434937,
281
+ "eval_runtime": 0.1699,
282
+ "eval_samples_per_second": 64.736,
283
+ "eval_steps_per_second": 5.885,
284
+ "step": 29
285
+ },
286
+ {
287
+ "epoch": 30.0,
288
+ "grad_norm": 1.4546806812286377,
289
+ "learning_rate": 3.888888888888889e-05,
290
+ "loss": 0.1472,
291
+ "step": 30
292
+ },
293
+ {
294
+ "epoch": 30.0,
295
+ "eval_accuracy": 0.7272727272727273,
296
+ "eval_loss": 1.5987972021102905,
297
+ "eval_runtime": 0.1768,
298
+ "eval_samples_per_second": 62.212,
299
+ "eval_steps_per_second": 5.656,
300
+ "step": 30
301
+ },
302
+ {
303
+ "epoch": 31.0,
304
+ "eval_accuracy": 0.7272727272727273,
305
+ "eval_loss": 1.570302128791809,
306
+ "eval_runtime": 0.1813,
307
+ "eval_samples_per_second": 60.658,
308
+ "eval_steps_per_second": 5.514,
309
+ "step": 31
310
+ },
311
+ {
312
+ "epoch": 32.0,
313
+ "eval_accuracy": 0.7272727272727273,
314
+ "eval_loss": 1.5965434312820435,
315
+ "eval_runtime": 0.1753,
316
+ "eval_samples_per_second": 62.751,
317
+ "eval_steps_per_second": 5.705,
318
+ "step": 32
319
+ },
320
+ {
321
+ "epoch": 33.0,
322
+ "eval_accuracy": 0.7272727272727273,
323
+ "eval_loss": 1.661800742149353,
324
+ "eval_runtime": 0.1773,
325
+ "eval_samples_per_second": 62.046,
326
+ "eval_steps_per_second": 5.641,
327
+ "step": 33
328
+ },
329
+ {
330
+ "epoch": 34.0,
331
+ "eval_accuracy": 0.7272727272727273,
332
+ "eval_loss": 1.6431862115859985,
333
+ "eval_runtime": 0.1709,
334
+ "eval_samples_per_second": 64.381,
335
+ "eval_steps_per_second": 5.853,
336
+ "step": 34
337
+ },
338
+ {
339
+ "epoch": 35.0,
340
+ "eval_accuracy": 0.7272727272727273,
341
+ "eval_loss": 1.5392369031906128,
342
+ "eval_runtime": 0.1749,
343
+ "eval_samples_per_second": 62.897,
344
+ "eval_steps_per_second": 5.718,
345
+ "step": 35
346
+ },
347
+ {
348
+ "epoch": 36.0,
349
+ "eval_accuracy": 0.7272727272727273,
350
+ "eval_loss": 1.4367867708206177,
351
+ "eval_runtime": 0.1732,
352
+ "eval_samples_per_second": 63.525,
353
+ "eval_steps_per_second": 5.775,
354
+ "step": 36
355
+ },
356
+ {
357
+ "epoch": 37.0,
358
+ "eval_accuracy": 0.7272727272727273,
359
+ "eval_loss": 1.4018795490264893,
360
+ "eval_runtime": 0.1763,
361
+ "eval_samples_per_second": 62.392,
362
+ "eval_steps_per_second": 5.672,
363
+ "step": 37
364
+ },
365
+ {
366
+ "epoch": 38.0,
367
+ "eval_accuracy": 0.7272727272727273,
368
+ "eval_loss": 1.4714306592941284,
369
+ "eval_runtime": 0.1707,
370
+ "eval_samples_per_second": 64.459,
371
+ "eval_steps_per_second": 5.86,
372
+ "step": 38
373
+ },
374
+ {
375
+ "epoch": 39.0,
376
+ "eval_accuracy": 0.7272727272727273,
377
+ "eval_loss": 1.7523380517959595,
378
+ "eval_runtime": 0.1761,
379
+ "eval_samples_per_second": 62.479,
380
+ "eval_steps_per_second": 5.68,
381
+ "step": 39
382
+ },
383
+ {
384
+ "epoch": 40.0,
385
+ "grad_norm": 3.4498791694641113,
386
+ "learning_rate": 3.3333333333333335e-05,
387
+ "loss": 0.0741,
388
+ "step": 40
389
+ },
390
+ {
391
+ "epoch": 40.0,
392
+ "eval_accuracy": 0.7272727272727273,
393
+ "eval_loss": 1.865343451499939,
394
+ "eval_runtime": 0.1669,
395
+ "eval_samples_per_second": 65.911,
396
+ "eval_steps_per_second": 5.992,
397
+ "step": 40
398
+ },
399
+ {
400
+ "epoch": 41.0,
401
+ "eval_accuracy": 0.7272727272727273,
402
+ "eval_loss": 1.8965870141983032,
403
+ "eval_runtime": 0.1763,
404
+ "eval_samples_per_second": 62.394,
405
+ "eval_steps_per_second": 5.672,
406
+ "step": 41
407
+ },
408
+ {
409
+ "epoch": 42.0,
410
+ "eval_accuracy": 0.7272727272727273,
411
+ "eval_loss": 1.8252369165420532,
412
+ "eval_runtime": 0.1737,
413
+ "eval_samples_per_second": 63.318,
414
+ "eval_steps_per_second": 5.756,
415
+ "step": 42
416
+ },
417
+ {
418
+ "epoch": 43.0,
419
+ "eval_accuracy": 0.7272727272727273,
420
+ "eval_loss": 1.8432804346084595,
421
+ "eval_runtime": 0.175,
422
+ "eval_samples_per_second": 62.868,
423
+ "eval_steps_per_second": 5.715,
424
+ "step": 43
425
+ },
426
+ {
427
+ "epoch": 44.0,
428
+ "eval_accuracy": 0.7272727272727273,
429
+ "eval_loss": 1.7326781749725342,
430
+ "eval_runtime": 0.1711,
431
+ "eval_samples_per_second": 64.288,
432
+ "eval_steps_per_second": 5.844,
433
+ "step": 44
434
+ },
435
+ {
436
+ "epoch": 45.0,
437
+ "eval_accuracy": 0.7272727272727273,
438
+ "eval_loss": 1.745629906654358,
439
+ "eval_runtime": 0.1698,
440
+ "eval_samples_per_second": 64.769,
441
+ "eval_steps_per_second": 5.888,
442
+ "step": 45
443
+ },
444
+ {
445
+ "epoch": 46.0,
446
+ "eval_accuracy": 0.7272727272727273,
447
+ "eval_loss": 1.894829273223877,
448
+ "eval_runtime": 0.1694,
449
+ "eval_samples_per_second": 64.945,
450
+ "eval_steps_per_second": 5.904,
451
+ "step": 46
452
+ },
453
+ {
454
+ "epoch": 47.0,
455
+ "eval_accuracy": 0.7272727272727273,
456
+ "eval_loss": 1.8810946941375732,
457
+ "eval_runtime": 0.174,
458
+ "eval_samples_per_second": 63.202,
459
+ "eval_steps_per_second": 5.746,
460
+ "step": 47
461
+ },
462
+ {
463
+ "epoch": 48.0,
464
+ "eval_accuracy": 0.7272727272727273,
465
+ "eval_loss": 1.8930473327636719,
466
+ "eval_runtime": 0.1802,
467
+ "eval_samples_per_second": 61.031,
468
+ "eval_steps_per_second": 5.548,
469
+ "step": 48
470
+ },
471
+ {
472
+ "epoch": 49.0,
473
+ "eval_accuracy": 0.7272727272727273,
474
+ "eval_loss": 1.8065048456192017,
475
+ "eval_runtime": 0.1789,
476
+ "eval_samples_per_second": 61.489,
477
+ "eval_steps_per_second": 5.59,
478
+ "step": 49
479
+ },
480
+ {
481
+ "epoch": 50.0,
482
+ "grad_norm": 1.586561918258667,
483
+ "learning_rate": 2.777777777777778e-05,
484
+ "loss": 0.0506,
485
+ "step": 50
486
+ },
487
+ {
488
+ "epoch": 50.0,
489
+ "eval_accuracy": 0.7272727272727273,
490
+ "eval_loss": 1.7629249095916748,
491
+ "eval_runtime": 0.162,
492
+ "eval_samples_per_second": 67.915,
493
+ "eval_steps_per_second": 6.174,
494
+ "step": 50
495
+ },
496
+ {
497
+ "epoch": 51.0,
498
+ "eval_accuracy": 0.7272727272727273,
499
+ "eval_loss": 1.7893092632293701,
500
+ "eval_runtime": 0.1708,
501
+ "eval_samples_per_second": 64.418,
502
+ "eval_steps_per_second": 5.856,
503
+ "step": 51
504
+ },
505
+ {
506
+ "epoch": 52.0,
507
+ "eval_accuracy": 0.7272727272727273,
508
+ "eval_loss": 1.876960277557373,
509
+ "eval_runtime": 0.1703,
510
+ "eval_samples_per_second": 64.587,
511
+ "eval_steps_per_second": 5.872,
512
+ "step": 52
513
+ },
514
+ {
515
+ "epoch": 53.0,
516
+ "eval_accuracy": 0.7272727272727273,
517
+ "eval_loss": 1.9144260883331299,
518
+ "eval_runtime": 0.176,
519
+ "eval_samples_per_second": 62.497,
520
+ "eval_steps_per_second": 5.682,
521
+ "step": 53
522
+ },
523
+ {
524
+ "epoch": 54.0,
525
+ "eval_accuracy": 0.7272727272727273,
526
+ "eval_loss": 2.0357284545898438,
527
+ "eval_runtime": 0.1764,
528
+ "eval_samples_per_second": 62.356,
529
+ "eval_steps_per_second": 5.669,
530
+ "step": 54
531
+ },
532
+ {
533
+ "epoch": 55.0,
534
+ "eval_accuracy": 0.7272727272727273,
535
+ "eval_loss": 2.210261583328247,
536
+ "eval_runtime": 0.1794,
537
+ "eval_samples_per_second": 61.314,
538
+ "eval_steps_per_second": 5.574,
539
+ "step": 55
540
+ },
541
+ {
542
+ "epoch": 56.0,
543
+ "eval_accuracy": 0.7272727272727273,
544
+ "eval_loss": 2.256763219833374,
545
+ "eval_runtime": 0.1747,
546
+ "eval_samples_per_second": 62.979,
547
+ "eval_steps_per_second": 5.725,
548
+ "step": 56
549
+ },
550
+ {
551
+ "epoch": 57.0,
552
+ "eval_accuracy": 0.7272727272727273,
553
+ "eval_loss": 2.2400619983673096,
554
+ "eval_runtime": 0.1705,
555
+ "eval_samples_per_second": 64.515,
556
+ "eval_steps_per_second": 5.865,
557
+ "step": 57
558
+ },
559
+ {
560
+ "epoch": 58.0,
561
+ "eval_accuracy": 0.7272727272727273,
562
+ "eval_loss": 2.140550374984741,
563
+ "eval_runtime": 0.1738,
564
+ "eval_samples_per_second": 63.306,
565
+ "eval_steps_per_second": 5.755,
566
+ "step": 58
567
+ },
568
+ {
569
+ "epoch": 59.0,
570
+ "eval_accuracy": 0.7272727272727273,
571
+ "eval_loss": 2.027967929840088,
572
+ "eval_runtime": 0.1674,
573
+ "eval_samples_per_second": 65.703,
574
+ "eval_steps_per_second": 5.973,
575
+ "step": 59
576
+ },
577
+ {
578
+ "epoch": 60.0,
579
+ "grad_norm": 0.3207845091819763,
580
+ "learning_rate": 2.2222222222222223e-05,
581
+ "loss": 0.0355,
582
+ "step": 60
583
+ },
584
+ {
585
+ "epoch": 60.0,
586
+ "eval_accuracy": 0.7272727272727273,
587
+ "eval_loss": 1.9516960382461548,
588
+ "eval_runtime": 0.1682,
589
+ "eval_samples_per_second": 65.399,
590
+ "eval_steps_per_second": 5.945,
591
+ "step": 60
592
+ },
593
+ {
594
+ "epoch": 61.0,
595
+ "eval_accuracy": 0.7272727272727273,
596
+ "eval_loss": 1.8787117004394531,
597
+ "eval_runtime": 0.1705,
598
+ "eval_samples_per_second": 64.52,
599
+ "eval_steps_per_second": 5.865,
600
+ "step": 61
601
+ },
602
+ {
603
+ "epoch": 62.0,
604
+ "eval_accuracy": 0.7272727272727273,
605
+ "eval_loss": 1.8958450555801392,
606
+ "eval_runtime": 0.1776,
607
+ "eval_samples_per_second": 61.949,
608
+ "eval_steps_per_second": 5.632,
609
+ "step": 62
610
+ },
611
+ {
612
+ "epoch": 63.0,
613
+ "eval_accuracy": 0.7272727272727273,
614
+ "eval_loss": 1.9893440008163452,
615
+ "eval_runtime": 0.1715,
616
+ "eval_samples_per_second": 64.156,
617
+ "eval_steps_per_second": 5.832,
618
+ "step": 63
619
+ },
620
+ {
621
+ "epoch": 64.0,
622
+ "eval_accuracy": 0.7272727272727273,
623
+ "eval_loss": 2.1194653511047363,
624
+ "eval_runtime": 0.169,
625
+ "eval_samples_per_second": 65.094,
626
+ "eval_steps_per_second": 5.918,
627
+ "step": 64
628
+ },
629
+ {
630
+ "epoch": 65.0,
631
+ "eval_accuracy": 0.7272727272727273,
632
+ "eval_loss": 2.271503210067749,
633
+ "eval_runtime": 0.1759,
634
+ "eval_samples_per_second": 62.532,
635
+ "eval_steps_per_second": 5.685,
636
+ "step": 65
637
+ },
638
+ {
639
+ "epoch": 66.0,
640
+ "eval_accuracy": 0.7272727272727273,
641
+ "eval_loss": 2.364030599594116,
642
+ "eval_runtime": 0.1726,
643
+ "eval_samples_per_second": 63.725,
644
+ "eval_steps_per_second": 5.793,
645
+ "step": 66
646
+ },
647
+ {
648
+ "epoch": 67.0,
649
+ "eval_accuracy": 0.7272727272727273,
650
+ "eval_loss": 2.4390556812286377,
651
+ "eval_runtime": 0.171,
652
+ "eval_samples_per_second": 64.312,
653
+ "eval_steps_per_second": 5.847,
654
+ "step": 67
655
+ },
656
+ {
657
+ "epoch": 68.0,
658
+ "eval_accuracy": 0.7272727272727273,
659
+ "eval_loss": 2.4633097648620605,
660
+ "eval_runtime": 0.1695,
661
+ "eval_samples_per_second": 64.911,
662
+ "eval_steps_per_second": 5.901,
663
+ "step": 68
664
+ },
665
+ {
666
+ "epoch": 69.0,
667
+ "eval_accuracy": 0.7272727272727273,
668
+ "eval_loss": 2.44889235496521,
669
+ "eval_runtime": 0.1727,
670
+ "eval_samples_per_second": 63.701,
671
+ "eval_steps_per_second": 5.791,
672
+ "step": 69
673
+ },
674
+ {
675
+ "epoch": 70.0,
676
+ "grad_norm": 0.9859403371810913,
677
+ "learning_rate": 1.6666666666666667e-05,
678
+ "loss": 0.021,
679
+ "step": 70
680
+ },
681
+ {
682
+ "epoch": 70.0,
683
+ "eval_accuracy": 0.7272727272727273,
684
+ "eval_loss": 2.4176981449127197,
685
+ "eval_runtime": 0.1698,
686
+ "eval_samples_per_second": 64.773,
687
+ "eval_steps_per_second": 5.888,
688
+ "step": 70
689
+ },
690
+ {
691
+ "epoch": 71.0,
692
+ "eval_accuracy": 0.7272727272727273,
693
+ "eval_loss": 2.3674123287200928,
694
+ "eval_runtime": 0.1737,
695
+ "eval_samples_per_second": 63.324,
696
+ "eval_steps_per_second": 5.757,
697
+ "step": 71
698
+ },
699
+ {
700
+ "epoch": 72.0,
701
+ "eval_accuracy": 0.7272727272727273,
702
+ "eval_loss": 2.306504011154175,
703
+ "eval_runtime": 0.1712,
704
+ "eval_samples_per_second": 64.26,
705
+ "eval_steps_per_second": 5.842,
706
+ "step": 72
707
+ },
708
+ {
709
+ "epoch": 73.0,
710
+ "eval_accuracy": 0.7272727272727273,
711
+ "eval_loss": 2.297823190689087,
712
+ "eval_runtime": 0.1707,
713
+ "eval_samples_per_second": 64.433,
714
+ "eval_steps_per_second": 5.858,
715
+ "step": 73
716
+ },
717
+ {
718
+ "epoch": 74.0,
719
+ "eval_accuracy": 0.7272727272727273,
720
+ "eval_loss": 2.358968496322632,
721
+ "eval_runtime": 0.1759,
722
+ "eval_samples_per_second": 62.545,
723
+ "eval_steps_per_second": 5.686,
724
+ "step": 74
725
+ },
726
+ {
727
+ "epoch": 75.0,
728
+ "eval_accuracy": 0.7272727272727273,
729
+ "eval_loss": 2.4276845455169678,
730
+ "eval_runtime": 0.1709,
731
+ "eval_samples_per_second": 64.358,
732
+ "eval_steps_per_second": 5.851,
733
+ "step": 75
734
+ },
735
+ {
736
+ "epoch": 76.0,
737
+ "eval_accuracy": 0.7272727272727273,
738
+ "eval_loss": 2.5219666957855225,
739
+ "eval_runtime": 0.1666,
740
+ "eval_samples_per_second": 66.046,
741
+ "eval_steps_per_second": 6.004,
742
+ "step": 76
743
+ },
744
+ {
745
+ "epoch": 77.0,
746
+ "eval_accuracy": 0.7272727272727273,
747
+ "eval_loss": 2.6248342990875244,
748
+ "eval_runtime": 0.1704,
749
+ "eval_samples_per_second": 64.551,
750
+ "eval_steps_per_second": 5.868,
751
+ "step": 77
752
+ },
753
+ {
754
+ "epoch": 78.0,
755
+ "eval_accuracy": 0.7272727272727273,
756
+ "eval_loss": 2.6924915313720703,
757
+ "eval_runtime": 0.1781,
758
+ "eval_samples_per_second": 61.76,
759
+ "eval_steps_per_second": 5.615,
760
+ "step": 78
761
+ },
762
+ {
763
+ "epoch": 79.0,
764
+ "eval_accuracy": 0.7272727272727273,
765
+ "eval_loss": 2.7284491062164307,
766
+ "eval_runtime": 0.1728,
767
+ "eval_samples_per_second": 63.641,
768
+ "eval_steps_per_second": 5.786,
769
+ "step": 79
770
+ },
771
+ {
772
+ "epoch": 80.0,
773
+ "grad_norm": 1.2645738124847412,
774
+ "learning_rate": 1.1111111111111112e-05,
775
+ "loss": 0.0346,
776
+ "step": 80
777
+ },
778
+ {
779
+ "epoch": 80.0,
780
+ "eval_accuracy": 0.7272727272727273,
781
+ "eval_loss": 2.740863084793091,
782
+ "eval_runtime": 0.1665,
783
+ "eval_samples_per_second": 66.083,
784
+ "eval_steps_per_second": 6.008,
785
+ "step": 80
786
+ },
787
+ {
788
+ "epoch": 81.0,
789
+ "eval_accuracy": 0.7272727272727273,
790
+ "eval_loss": 2.7218360900878906,
791
+ "eval_runtime": 0.1684,
792
+ "eval_samples_per_second": 65.316,
793
+ "eval_steps_per_second": 5.938,
794
+ "step": 81
795
+ },
796
+ {
797
+ "epoch": 82.0,
798
+ "eval_accuracy": 0.7272727272727273,
799
+ "eval_loss": 2.6717374324798584,
800
+ "eval_runtime": 0.1706,
801
+ "eval_samples_per_second": 64.491,
802
+ "eval_steps_per_second": 5.863,
803
+ "step": 82
804
+ },
805
+ {
806
+ "epoch": 83.0,
807
+ "eval_accuracy": 0.7272727272727273,
808
+ "eval_loss": 2.614175796508789,
809
+ "eval_runtime": 0.1719,
810
+ "eval_samples_per_second": 64.002,
811
+ "eval_steps_per_second": 5.818,
812
+ "step": 83
813
+ },
814
+ {
815
+ "epoch": 84.0,
816
+ "eval_accuracy": 0.7272727272727273,
817
+ "eval_loss": 2.519634485244751,
818
+ "eval_runtime": 0.1662,
819
+ "eval_samples_per_second": 66.169,
820
+ "eval_steps_per_second": 6.015,
821
+ "step": 84
822
+ },
823
+ {
824
+ "epoch": 85.0,
825
+ "eval_accuracy": 0.7272727272727273,
826
+ "eval_loss": 2.4437661170959473,
827
+ "eval_runtime": 0.1688,
828
+ "eval_samples_per_second": 65.155,
829
+ "eval_steps_per_second": 5.923,
830
+ "step": 85
831
+ },
832
+ {
833
+ "epoch": 86.0,
834
+ "eval_accuracy": 0.7272727272727273,
835
+ "eval_loss": 2.381150007247925,
836
+ "eval_runtime": 0.171,
837
+ "eval_samples_per_second": 64.318,
838
+ "eval_steps_per_second": 5.847,
839
+ "step": 86
840
+ },
841
+ {
842
+ "epoch": 87.0,
843
+ "eval_accuracy": 0.7272727272727273,
844
+ "eval_loss": 2.3026392459869385,
845
+ "eval_runtime": 0.1756,
846
+ "eval_samples_per_second": 62.625,
847
+ "eval_steps_per_second": 5.693,
848
+ "step": 87
849
+ },
850
+ {
851
+ "epoch": 88.0,
852
+ "eval_accuracy": 0.7272727272727273,
853
+ "eval_loss": 2.269089460372925,
854
+ "eval_runtime": 0.1704,
855
+ "eval_samples_per_second": 64.567,
856
+ "eval_steps_per_second": 5.87,
857
+ "step": 88
858
+ },
859
+ {
860
+ "epoch": 89.0,
861
+ "eval_accuracy": 0.7272727272727273,
862
+ "eval_loss": 2.2665257453918457,
863
+ "eval_runtime": 0.1725,
864
+ "eval_samples_per_second": 63.767,
865
+ "eval_steps_per_second": 5.797,
866
+ "step": 89
867
+ },
868
+ {
869
+ "epoch": 90.0,
870
+ "grad_norm": 1.038275957107544,
871
+ "learning_rate": 5.555555555555556e-06,
872
+ "loss": 0.0219,
873
+ "step": 90
874
+ },
875
+ {
876
+ "epoch": 90.0,
877
+ "eval_accuracy": 0.7272727272727273,
878
+ "eval_loss": 2.2659270763397217,
879
+ "eval_runtime": 0.1644,
880
+ "eval_samples_per_second": 66.891,
881
+ "eval_steps_per_second": 6.081,
882
+ "step": 90
883
+ },
884
+ {
885
+ "epoch": 91.0,
886
+ "eval_accuracy": 0.7272727272727273,
887
+ "eval_loss": 2.27345609664917,
888
+ "eval_runtime": 0.172,
889
+ "eval_samples_per_second": 63.936,
890
+ "eval_steps_per_second": 5.812,
891
+ "step": 91
892
+ },
893
+ {
894
+ "epoch": 92.0,
895
+ "eval_accuracy": 0.7272727272727273,
896
+ "eval_loss": 2.300039529800415,
897
+ "eval_runtime": 0.1731,
898
+ "eval_samples_per_second": 63.552,
899
+ "eval_steps_per_second": 5.777,
900
+ "step": 92
901
+ },
902
+ {
903
+ "epoch": 93.0,
904
+ "eval_accuracy": 0.7272727272727273,
905
+ "eval_loss": 2.313608407974243,
906
+ "eval_runtime": 0.1696,
907
+ "eval_samples_per_second": 64.87,
908
+ "eval_steps_per_second": 5.897,
909
+ "step": 93
910
+ },
911
+ {
912
+ "epoch": 94.0,
913
+ "eval_accuracy": 0.7272727272727273,
914
+ "eval_loss": 2.324737071990967,
915
+ "eval_runtime": 0.1718,
916
+ "eval_samples_per_second": 64.032,
917
+ "eval_steps_per_second": 5.821,
918
+ "step": 94
919
+ },
920
+ {
921
+ "epoch": 95.0,
922
+ "eval_accuracy": 0.7272727272727273,
923
+ "eval_loss": 2.3388283252716064,
924
+ "eval_runtime": 0.1708,
925
+ "eval_samples_per_second": 64.404,
926
+ "eval_steps_per_second": 5.855,
927
+ "step": 95
928
+ },
929
+ {
930
+ "epoch": 96.0,
931
+ "eval_accuracy": 0.7272727272727273,
932
+ "eval_loss": 2.3597023487091064,
933
+ "eval_runtime": 0.1716,
934
+ "eval_samples_per_second": 64.094,
935
+ "eval_steps_per_second": 5.827,
936
+ "step": 96
937
+ },
938
+ {
939
+ "epoch": 97.0,
940
+ "eval_accuracy": 0.7272727272727273,
941
+ "eval_loss": 2.3746063709259033,
942
+ "eval_runtime": 0.1715,
943
+ "eval_samples_per_second": 64.137,
944
+ "eval_steps_per_second": 5.831,
945
+ "step": 97
946
+ },
947
+ {
948
+ "epoch": 98.0,
949
+ "eval_accuracy": 0.7272727272727273,
950
+ "eval_loss": 2.386420488357544,
951
+ "eval_runtime": 0.1732,
952
+ "eval_samples_per_second": 63.509,
953
+ "eval_steps_per_second": 5.774,
954
+ "step": 98
955
+ },
956
+ {
957
+ "epoch": 99.0,
958
+ "eval_accuracy": 0.7272727272727273,
959
+ "eval_loss": 2.391254186630249,
960
+ "eval_runtime": 0.1692,
961
+ "eval_samples_per_second": 64.995,
962
+ "eval_steps_per_second": 5.909,
963
+ "step": 99
964
+ },
965
+ {
966
+ "epoch": 100.0,
967
+ "grad_norm": 0.8429810404777527,
968
+ "learning_rate": 0.0,
969
+ "loss": 0.0286,
970
+ "step": 100
971
+ },
972
+ {
973
+ "epoch": 100.0,
974
+ "eval_accuracy": 0.7272727272727273,
975
+ "eval_loss": 2.3933160305023193,
976
+ "eval_runtime": 0.1621,
977
+ "eval_samples_per_second": 67.872,
978
+ "eval_steps_per_second": 6.17,
979
+ "step": 100
980
+ },
981
+ {
982
+ "epoch": 100.0,
983
+ "step": 100,
984
+ "total_flos": 4.572150213593088e+17,
985
+ "train_loss": 0.10843914225697518,
986
+ "train_runtime": 416.0402,
987
+ "train_samples_per_second": 14.181,
988
+ "train_steps_per_second": 0.24
989
+ },
990
+ {
991
+ "epoch": 100.0,
992
+ "eval_accuracy": 0.7272727272727273,
993
+ "eval_loss": 0.9542110562324524,
994
+ "eval_runtime": 0.2165,
995
+ "eval_samples_per_second": 50.803,
996
+ "eval_steps_per_second": 4.618,
997
+ "step": 100
998
+ }
999
+ ],
1000
+ "logging_steps": 10,
1001
+ "max_steps": 100,
1002
+ "num_input_tokens_seen": 0,
1003
+ "num_train_epochs": 100,
1004
+ "save_steps": 500,
1005
+ "stateful_callbacks": {
1006
+ "TrainerControl": {
1007
+ "args": {
1008
+ "should_epoch_stop": false,
1009
+ "should_evaluate": false,
1010
+ "should_log": false,
1011
+ "should_save": true,
1012
+ "should_training_stop": true
1013
+ },
1014
+ "attributes": {}
1015
+ }
1016
+ },
1017
+ "total_flos": 4.572150213593088e+17,
1018
+ "train_batch_size": 32,
1019
+ "trial_name": null,
1020
+ "trial_params": null
1021
+ }