vuongnhathien commited on
Commit
07cdb5c
1 Parent(s): fcccbd3

End of training

Browse files
README.md CHANGED
@@ -22,7 +22,7 @@ model-index:
22
  metrics:
23
  - name: Accuracy
24
  type: accuracy
25
- value: 0.9506958250497017
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 [facebook/convnextv2-base-22k-384](https://huggingface.co/facebook/convnextv2-base-22k-384) on the imagefolder dataset.
34
  It achieves the following results on the evaluation set:
35
- - Loss: 0.2211
36
- - Accuracy: 0.9507
37
 
38
  ## Model description
39
 
 
22
  metrics:
23
  - name: Accuracy
24
  type: accuracy
25
+ value: 0.9511904761904761
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 [facebook/convnextv2-base-22k-384](https://huggingface.co/facebook/convnextv2-base-22k-384) on the imagefolder dataset.
34
  It achieves the following results on the evaluation set:
35
+ - Loss: 0.2152
36
+ - Accuracy: 0.9512
37
 
38
  ## Model description
39
 
all_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "total_flos": 4.09349935387607e+19,
4
+ "train_loss": 0.35613980254224475,
5
+ "train_runtime": 19061.1748,
6
+ "train_samples_per_second": 9.223,
7
+ "train_steps_per_second": 0.577
8
+ }
runs/May28_13-12-51_c126bcd9d0c5/events.out.tfevents.1716921369.c126bcd9d0c5.24.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc0a1214017de5760b1805823f2509f1da96f4e5f60b9c277850f0b4a384ae35
3
+ size 411
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "total_flos": 4.09349935387607e+19,
4
+ "train_loss": 0.35613980254224475,
5
+ "train_runtime": 19061.1748,
6
+ "train_samples_per_second": 9.223,
7
+ "train_steps_per_second": 0.577
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,883 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.22113871574401855,
3
+ "best_model_checkpoint": "./convnext-base-wd-1e-8-3e-5-erasing/checkpoint-10990",
4
+ "epoch": 10.0,
5
+ "eval_steps": 500,
6
+ "global_step": 10990,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.09,
13
+ "grad_norm": 23.46108627319336,
14
+ "learning_rate": 2.999387175598269e-05,
15
+ "loss": 2.6353,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.18,
20
+ "grad_norm": 16.22706413269043,
21
+ "learning_rate": 2.9975492031314045e-05,
22
+ "loss": 1.4961,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.27,
27
+ "grad_norm": 21.055408477783203,
28
+ "learning_rate": 2.994487584405244e-05,
29
+ "loss": 1.1313,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.36,
34
+ "grad_norm": 33.7674446105957,
35
+ "learning_rate": 2.990204821066006e-05,
36
+ "loss": 0.8546,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.45,
41
+ "grad_norm": 20.148395538330078,
42
+ "learning_rate": 2.984704412556199e-05,
43
+ "loss": 0.8015,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.55,
48
+ "grad_norm": 19.732004165649414,
49
+ "learning_rate": 2.977990853255228e-05,
50
+ "loss": 0.691,
51
+ "step": 600
52
+ },
53
+ {
54
+ "epoch": 0.64,
55
+ "grad_norm": 25.433279037475586,
56
+ "learning_rate": 2.970069628807043e-05,
57
+ "loss": 0.7105,
58
+ "step": 700
59
+ },
60
+ {
61
+ "epoch": 0.73,
62
+ "grad_norm": 14.037881851196289,
63
+ "learning_rate": 2.9609472116378222e-05,
64
+ "loss": 0.64,
65
+ "step": 800
66
+ },
67
+ {
68
+ "epoch": 0.82,
69
+ "grad_norm": 10.451700210571289,
70
+ "learning_rate": 2.9506310556673573e-05,
71
+ "loss": 0.5995,
72
+ "step": 900
73
+ },
74
+ {
75
+ "epoch": 0.91,
76
+ "grad_norm": 12.546747207641602,
77
+ "learning_rate": 2.9391295902184625e-05,
78
+ "loss": 0.6062,
79
+ "step": 1000
80
+ },
81
+ {
82
+ "epoch": 1.0,
83
+ "eval_accuracy": 0.8930417495029821,
84
+ "eval_loss": 0.36765438318252563,
85
+ "eval_runtime": 105.6433,
86
+ "eval_samples_per_second": 23.807,
87
+ "eval_steps_per_second": 1.496,
88
+ "step": 1099
89
+ },
90
+ {
91
+ "epoch": 1.0,
92
+ "grad_norm": 34.490543365478516,
93
+ "learning_rate": 2.9264522131293818e-05,
94
+ "loss": 0.6367,
95
+ "step": 1100
96
+ },
97
+ {
98
+ "epoch": 1.09,
99
+ "grad_norm": 15.90696907043457,
100
+ "learning_rate": 2.9126092830748217e-05,
101
+ "loss": 0.5309,
102
+ "step": 1200
103
+ },
104
+ {
105
+ "epoch": 1.18,
106
+ "grad_norm": 28.304794311523438,
107
+ "learning_rate": 2.897612111101888e-05,
108
+ "loss": 0.515,
109
+ "step": 1300
110
+ },
111
+ {
112
+ "epoch": 1.27,
113
+ "grad_norm": 18.61859893798828,
114
+ "learning_rate": 2.8814729513878365e-05,
115
+ "loss": 0.484,
116
+ "step": 1400
117
+ },
118
+ {
119
+ "epoch": 1.36,
120
+ "grad_norm": 19.281923294067383,
121
+ "learning_rate": 2.864204991227195e-05,
122
+ "loss": 0.5082,
123
+ "step": 1500
124
+ },
125
+ {
126
+ "epoch": 1.46,
127
+ "grad_norm": 20.169174194335938,
128
+ "learning_rate": 2.8458223402564366e-05,
129
+ "loss": 0.4689,
130
+ "step": 1600
131
+ },
132
+ {
133
+ "epoch": 1.55,
134
+ "grad_norm": 13.506836891174316,
135
+ "learning_rate": 2.826340018925006e-05,
136
+ "loss": 0.5311,
137
+ "step": 1700
138
+ },
139
+ {
140
+ "epoch": 1.64,
141
+ "grad_norm": 22.1657657623291,
142
+ "learning_rate": 2.8057739462221215e-05,
143
+ "loss": 0.4909,
144
+ "step": 1800
145
+ },
146
+ {
147
+ "epoch": 1.73,
148
+ "grad_norm": 20.638866424560547,
149
+ "learning_rate": 2.7841409266693838e-05,
150
+ "loss": 0.4316,
151
+ "step": 1900
152
+ },
153
+ {
154
+ "epoch": 1.82,
155
+ "grad_norm": 15.684589385986328,
156
+ "learning_rate": 2.761458636589813e-05,
157
+ "loss": 0.4746,
158
+ "step": 2000
159
+ },
160
+ {
161
+ "epoch": 1.91,
162
+ "grad_norm": 17.279956817626953,
163
+ "learning_rate": 2.7377456096645395e-05,
164
+ "loss": 0.4975,
165
+ "step": 2100
166
+ },
167
+ {
168
+ "epoch": 2.0,
169
+ "eval_accuracy": 0.9268389662027833,
170
+ "eval_loss": 0.27913880348205566,
171
+ "eval_runtime": 105.6176,
172
+ "eval_samples_per_second": 23.812,
173
+ "eval_steps_per_second": 1.496,
174
+ "step": 2198
175
+ },
176
+ {
177
+ "epoch": 2.0,
178
+ "grad_norm": 29.95267105102539,
179
+ "learning_rate": 2.7130212217889484e-05,
180
+ "loss": 0.4716,
181
+ "step": 2200
182
+ },
183
+ {
184
+ "epoch": 2.09,
185
+ "grad_norm": 14.320356369018555,
186
+ "learning_rate": 2.6873056752406504e-05,
187
+ "loss": 0.3472,
188
+ "step": 2300
189
+ },
190
+ {
191
+ "epoch": 2.18,
192
+ "grad_norm": 24.206567764282227,
193
+ "learning_rate": 2.6606199821722166e-05,
194
+ "loss": 0.4262,
195
+ "step": 2400
196
+ },
197
+ {
198
+ "epoch": 2.27,
199
+ "grad_norm": 16.749975204467773,
200
+ "learning_rate": 2.632985947442167e-05,
201
+ "loss": 0.3718,
202
+ "step": 2500
203
+ },
204
+ {
205
+ "epoch": 2.37,
206
+ "grad_norm": 13.791542053222656,
207
+ "learning_rate": 2.6044261507982356e-05,
208
+ "loss": 0.4423,
209
+ "step": 2600
210
+ },
211
+ {
212
+ "epoch": 2.46,
213
+ "grad_norm": 39.41408157348633,
214
+ "learning_rate": 2.5749639284274782e-05,
215
+ "loss": 0.3756,
216
+ "step": 2700
217
+ },
218
+ {
219
+ "epoch": 2.55,
220
+ "grad_norm": 5.441478252410889,
221
+ "learning_rate": 2.5446233538882924e-05,
222
+ "loss": 0.3997,
223
+ "step": 2800
224
+ },
225
+ {
226
+ "epoch": 2.64,
227
+ "grad_norm": 27.294628143310547,
228
+ "learning_rate": 2.513429218439932e-05,
229
+ "loss": 0.3874,
230
+ "step": 2900
231
+ },
232
+ {
233
+ "epoch": 2.73,
234
+ "grad_norm": 18.142757415771484,
235
+ "learning_rate": 2.4814070107855878e-05,
236
+ "loss": 0.4291,
237
+ "step": 3000
238
+ },
239
+ {
240
+ "epoch": 2.82,
241
+ "grad_norm": 18.32405662536621,
242
+ "learning_rate": 2.448582896245591e-05,
243
+ "loss": 0.4085,
244
+ "step": 3100
245
+ },
246
+ {
247
+ "epoch": 2.91,
248
+ "grad_norm": 27.914020538330078,
249
+ "learning_rate": 2.4149836953777488e-05,
250
+ "loss": 0.3777,
251
+ "step": 3200
252
+ },
253
+ {
254
+ "epoch": 3.0,
255
+ "eval_accuracy": 0.9355864811133201,
256
+ "eval_loss": 0.2496386617422104,
257
+ "eval_runtime": 105.6667,
258
+ "eval_samples_per_second": 23.801,
259
+ "eval_steps_per_second": 1.495,
260
+ "step": 3297
261
+ },
262
+ {
263
+ "epoch": 3.0,
264
+ "grad_norm": 11.372156143188477,
265
+ "learning_rate": 2.3806368620622876e-05,
266
+ "loss": 0.3952,
267
+ "step": 3300
268
+ },
269
+ {
270
+ "epoch": 3.09,
271
+ "grad_norm": 6.642704963684082,
272
+ "learning_rate": 2.345570461069312e-05,
273
+ "loss": 0.3029,
274
+ "step": 3400
275
+ },
276
+ {
277
+ "epoch": 3.18,
278
+ "grad_norm": 16.000930786132812,
279
+ "learning_rate": 2.3098131451271016e-05,
280
+ "loss": 0.3644,
281
+ "step": 3500
282
+ },
283
+ {
284
+ "epoch": 3.28,
285
+ "grad_norm": 27.844514846801758,
286
+ "learning_rate": 2.2733941315099883e-05,
287
+ "loss": 0.3687,
288
+ "step": 3600
289
+ },
290
+ {
291
+ "epoch": 3.37,
292
+ "grad_norm": 27.35566520690918,
293
+ "learning_rate": 2.2363431781649483e-05,
294
+ "loss": 0.3111,
295
+ "step": 3700
296
+ },
297
+ {
298
+ "epoch": 3.46,
299
+ "grad_norm": 13.689902305603027,
300
+ "learning_rate": 2.1986905593964048e-05,
301
+ "loss": 0.3542,
302
+ "step": 3800
303
+ },
304
+ {
305
+ "epoch": 3.55,
306
+ "grad_norm": 6.648231029510498,
307
+ "learning_rate": 2.1604670411291174e-05,
308
+ "loss": 0.3471,
309
+ "step": 3900
310
+ },
311
+ {
312
+ "epoch": 3.64,
313
+ "grad_norm": 17.46345329284668,
314
+ "learning_rate": 2.121703855769373e-05,
315
+ "loss": 0.3704,
316
+ "step": 4000
317
+ },
318
+ {
319
+ "epoch": 3.73,
320
+ "grad_norm": 13.02889347076416,
321
+ "learning_rate": 2.0824326766850072e-05,
322
+ "loss": 0.3491,
323
+ "step": 4100
324
+ },
325
+ {
326
+ "epoch": 3.82,
327
+ "grad_norm": 3.550931453704834,
328
+ "learning_rate": 2.042685592325123e-05,
329
+ "loss": 0.33,
330
+ "step": 4200
331
+ },
332
+ {
333
+ "epoch": 3.91,
334
+ "grad_norm": 52.093753814697266,
335
+ "learning_rate": 2.0024950800006463e-05,
336
+ "loss": 0.3135,
337
+ "step": 4300
338
+ },
339
+ {
340
+ "epoch": 4.0,
341
+ "eval_accuracy": 0.9395626242544731,
342
+ "eval_loss": 0.22969551384449005,
343
+ "eval_runtime": 105.7529,
344
+ "eval_samples_per_second": 23.782,
345
+ "eval_steps_per_second": 1.494,
346
+ "step": 4396
347
+ },
348
+ {
349
+ "epoch": 4.0,
350
+ "grad_norm": 9.075785636901855,
351
+ "learning_rate": 1.961893979347137e-05,
352
+ "loss": 0.2703,
353
+ "step": 4400
354
+ },
355
+ {
356
+ "epoch": 4.09,
357
+ "grad_norm": 16.759511947631836,
358
+ "learning_rate": 1.9209154654915524e-05,
359
+ "loss": 0.2964,
360
+ "step": 4500
361
+ },
362
+ {
363
+ "epoch": 4.19,
364
+ "grad_norm": 34.864501953125,
365
+ "learning_rate": 1.879593021944875e-05,
366
+ "loss": 0.3281,
367
+ "step": 4600
368
+ },
369
+ {
370
+ "epoch": 4.28,
371
+ "grad_norm": 31.949420928955078,
372
+ "learning_rate": 1.837960413242765e-05,
373
+ "loss": 0.2933,
374
+ "step": 4700
375
+ },
376
+ {
377
+ "epoch": 4.37,
378
+ "grad_norm": 18.939651489257812,
379
+ "learning_rate": 1.796051657356582e-05,
380
+ "loss": 0.2837,
381
+ "step": 4800
382
+ },
383
+ {
384
+ "epoch": 4.46,
385
+ "grad_norm": 10.814738273620605,
386
+ "learning_rate": 1.7539009978973312e-05,
387
+ "loss": 0.2615,
388
+ "step": 4900
389
+ },
390
+ {
391
+ "epoch": 4.55,
392
+ "grad_norm": 19.958898544311523,
393
+ "learning_rate": 1.711542876135233e-05,
394
+ "loss": 0.2938,
395
+ "step": 5000
396
+ },
397
+ {
398
+ "epoch": 4.64,
399
+ "grad_norm": 24.847991943359375,
400
+ "learning_rate": 1.669011902857791e-05,
401
+ "loss": 0.2847,
402
+ "step": 5100
403
+ },
404
+ {
405
+ "epoch": 4.73,
406
+ "grad_norm": 31.440824508666992,
407
+ "learning_rate": 1.6263428300893422e-05,
408
+ "loss": 0.311,
409
+ "step": 5200
410
+ },
411
+ {
412
+ "epoch": 4.82,
413
+ "grad_norm": 24.64202117919922,
414
+ "learning_rate": 1.5835705226952112e-05,
415
+ "loss": 0.319,
416
+ "step": 5300
417
+ },
418
+ {
419
+ "epoch": 4.91,
420
+ "grad_norm": 20.617746353149414,
421
+ "learning_rate": 1.540729929893649e-05,
422
+ "loss": 0.3022,
423
+ "step": 5400
424
+ },
425
+ {
426
+ "epoch": 5.0,
427
+ "eval_accuracy": 0.9399602385685885,
428
+ "eval_loss": 0.24197304248809814,
429
+ "eval_runtime": 105.732,
430
+ "eval_samples_per_second": 23.787,
431
+ "eval_steps_per_second": 1.494,
432
+ "step": 5495
433
+ },
434
+ {
435
+ "epoch": 5.0,
436
+ "grad_norm": 13.16674518585205,
437
+ "learning_rate": 1.4978560566988603e-05,
438
+ "loss": 0.3003,
439
+ "step": 5500
440
+ },
441
+ {
442
+ "epoch": 5.1,
443
+ "grad_norm": 11.548135757446289,
444
+ "learning_rate": 1.454983935318433e-05,
445
+ "loss": 0.2383,
446
+ "step": 5600
447
+ },
448
+ {
449
+ "epoch": 5.19,
450
+ "grad_norm": 15.562095642089844,
451
+ "learning_rate": 1.4121485965285485e-05,
452
+ "loss": 0.235,
453
+ "step": 5700
454
+ },
455
+ {
456
+ "epoch": 5.28,
457
+ "grad_norm": 22.393648147583008,
458
+ "learning_rate": 1.3693850410503614e-05,
459
+ "loss": 0.2661,
460
+ "step": 5800
461
+ },
462
+ {
463
+ "epoch": 5.37,
464
+ "grad_norm": 21.64914894104004,
465
+ "learning_rate": 1.326728210950942e-05,
466
+ "loss": 0.2697,
467
+ "step": 5900
468
+ },
469
+ {
470
+ "epoch": 5.46,
471
+ "grad_norm": 7.229253768920898,
472
+ "learning_rate": 1.2842129610921378e-05,
473
+ "loss": 0.2556,
474
+ "step": 6000
475
+ },
476
+ {
477
+ "epoch": 5.55,
478
+ "grad_norm": 17.345558166503906,
479
+ "learning_rate": 1.2418740306506923e-05,
480
+ "loss": 0.2717,
481
+ "step": 6100
482
+ },
483
+ {
484
+ "epoch": 5.64,
485
+ "grad_norm": 32.90108871459961,
486
+ "learning_rate": 1.1997460147328984e-05,
487
+ "loss": 0.2346,
488
+ "step": 6200
489
+ },
490
+ {
491
+ "epoch": 5.73,
492
+ "grad_norm": 26.533710479736328,
493
+ "learning_rate": 1.1578633361069559e-05,
494
+ "loss": 0.2724,
495
+ "step": 6300
496
+ },
497
+ {
498
+ "epoch": 5.82,
499
+ "grad_norm": 16.73810577392578,
500
+ "learning_rate": 1.1162602170761611e-05,
501
+ "loss": 0.2518,
502
+ "step": 6400
503
+ },
504
+ {
505
+ "epoch": 5.91,
506
+ "grad_norm": 31.530790328979492,
507
+ "learning_rate": 1.0749706515158863e-05,
508
+ "loss": 0.2481,
509
+ "step": 6500
510
+ },
511
+ {
512
+ "epoch": 6.0,
513
+ "eval_accuracy": 0.945924453280318,
514
+ "eval_loss": 0.2327275574207306,
515
+ "eval_runtime": 105.9698,
516
+ "eval_samples_per_second": 23.733,
517
+ "eval_steps_per_second": 1.491,
518
+ "step": 6594
519
+ },
520
+ {
521
+ "epoch": 6.01,
522
+ "grad_norm": 17.024036407470703,
523
+ "learning_rate": 1.0340283770972167e-05,
524
+ "loss": 0.2747,
525
+ "step": 6600
526
+ },
527
+ {
528
+ "epoch": 6.1,
529
+ "grad_norm": 14.112195014953613,
530
+ "learning_rate": 9.93466847719919e-06,
531
+ "loss": 0.2236,
532
+ "step": 6700
533
+ },
534
+ {
535
+ "epoch": 6.19,
536
+ "grad_norm": 7.770081043243408,
537
+ "learning_rate": 9.533192061772919e-06,
538
+ "loss": 0.2391,
539
+ "step": 6800
540
+ },
541
+ {
542
+ "epoch": 6.28,
543
+ "grad_norm": 7.291871547698975,
544
+ "learning_rate": 9.136182570752153e-06,
545
+ "loss": 0.2501,
546
+ "step": 6900
547
+ },
548
+ {
549
+ "epoch": 6.37,
550
+ "grad_norm": 9.997650146484375,
551
+ "learning_rate": 8.743964400275304e-06,
552
+ "loss": 0.2116,
553
+ "step": 7000
554
+ },
555
+ {
556
+ "epoch": 6.46,
557
+ "grad_norm": 0.30319592356681824,
558
+ "learning_rate": 8.356858031496596e-06,
559
+ "loss": 0.2563,
560
+ "step": 7100
561
+ },
562
+ {
563
+ "epoch": 6.55,
564
+ "grad_norm": 1.384082555770874,
565
+ "learning_rate": 7.975179768721187e-06,
566
+ "loss": 0.2203,
567
+ "step": 7200
568
+ },
569
+ {
570
+ "epoch": 6.64,
571
+ "grad_norm": 17.061349868774414,
572
+ "learning_rate": 7.599241480953112e-06,
573
+ "loss": 0.2054,
574
+ "step": 7300
575
+ },
576
+ {
577
+ "epoch": 6.73,
578
+ "grad_norm": 33.148921966552734,
579
+ "learning_rate": 7.229350347067426e-06,
580
+ "loss": 0.2328,
581
+ "step": 7400
582
+ },
583
+ {
584
+ "epoch": 6.82,
585
+ "grad_norm": 8.242575645446777,
586
+ "learning_rate": 6.865808604814564e-06,
587
+ "loss": 0.2108,
588
+ "step": 7500
589
+ },
590
+ {
591
+ "epoch": 6.92,
592
+ "grad_norm": 0.4348295032978058,
593
+ "learning_rate": 6.508913303862144e-06,
594
+ "loss": 0.2439,
595
+ "step": 7600
596
+ },
597
+ {
598
+ "epoch": 7.0,
599
+ "eval_accuracy": 0.9439363817097416,
600
+ "eval_loss": 0.23284506797790527,
601
+ "eval_runtime": 105.5247,
602
+ "eval_samples_per_second": 23.833,
603
+ "eval_steps_per_second": 1.497,
604
+ "step": 7693
605
+ },
606
+ {
607
+ "epoch": 7.01,
608
+ "grad_norm": 30.08086585998535,
609
+ "learning_rate": 6.1589560630758656e-06,
610
+ "loss": 0.1879,
611
+ "step": 7700
612
+ },
613
+ {
614
+ "epoch": 7.1,
615
+ "grad_norm": 41.7615852355957,
616
+ "learning_rate": 5.8162228322380155e-06,
617
+ "loss": 0.2135,
618
+ "step": 7800
619
+ },
620
+ {
621
+ "epoch": 7.19,
622
+ "grad_norm": 8.607312202453613,
623
+ "learning_rate": 5.480993658398129e-06,
624
+ "loss": 0.1963,
625
+ "step": 7900
626
+ },
627
+ {
628
+ "epoch": 7.28,
629
+ "grad_norm": 10.731322288513184,
630
+ "learning_rate": 5.153542457046737e-06,
631
+ "loss": 0.2423,
632
+ "step": 8000
633
+ },
634
+ {
635
+ "epoch": 7.37,
636
+ "grad_norm": 10.43770980834961,
637
+ "learning_rate": 4.834136788299248e-06,
638
+ "loss": 0.2129,
639
+ "step": 8100
640
+ },
641
+ {
642
+ "epoch": 7.46,
643
+ "grad_norm": 12.690656661987305,
644
+ "learning_rate": 4.523037638272822e-06,
645
+ "loss": 0.2149,
646
+ "step": 8200
647
+ },
648
+ {
649
+ "epoch": 7.55,
650
+ "grad_norm": 18.027700424194336,
651
+ "learning_rate": 4.220499205834783e-06,
652
+ "loss": 0.185,
653
+ "step": 8300
654
+ },
655
+ {
656
+ "epoch": 7.64,
657
+ "grad_norm": 15.369023323059082,
658
+ "learning_rate": 3.926768694896931e-06,
659
+ "loss": 0.22,
660
+ "step": 8400
661
+ },
662
+ {
663
+ "epoch": 7.73,
664
+ "grad_norm": 44.863441467285156,
665
+ "learning_rate": 3.6420861124254607e-06,
666
+ "loss": 0.1803,
667
+ "step": 8500
668
+ },
669
+ {
670
+ "epoch": 7.83,
671
+ "grad_norm": 9.732752799987793,
672
+ "learning_rate": 3.3666840723314145e-06,
673
+ "loss": 0.2207,
674
+ "step": 8600
675
+ },
676
+ {
677
+ "epoch": 7.92,
678
+ "grad_norm": 1.730449914932251,
679
+ "learning_rate": 3.1007876054020724e-06,
680
+ "loss": 0.1849,
681
+ "step": 8700
682
+ },
683
+ {
684
+ "epoch": 8.0,
685
+ "eval_accuracy": 0.94831013916501,
686
+ "eval_loss": 0.2235245704650879,
687
+ "eval_runtime": 105.4928,
688
+ "eval_samples_per_second": 23.84,
689
+ "eval_steps_per_second": 1.498,
690
+ "step": 8792
691
+ },
692
+ {
693
+ "epoch": 8.01,
694
+ "grad_norm": 21.83309555053711,
695
+ "learning_rate": 2.8446139754284486e-06,
696
+ "loss": 0.2095,
697
+ "step": 8800
698
+ },
699
+ {
700
+ "epoch": 8.1,
701
+ "grad_norm": 21.83153533935547,
702
+ "learning_rate": 2.5983725016792574e-06,
703
+ "loss": 0.196,
704
+ "step": 8900
705
+ },
706
+ {
707
+ "epoch": 8.19,
708
+ "grad_norm": 1.2116672992706299,
709
+ "learning_rate": 2.36226438786627e-06,
710
+ "loss": 0.1769,
711
+ "step": 9000
712
+ },
713
+ {
714
+ "epoch": 8.28,
715
+ "grad_norm": 12.81109619140625,
716
+ "learning_rate": 2.1364825577409424e-06,
717
+ "loss": 0.1455,
718
+ "step": 9100
719
+ },
720
+ {
721
+ "epoch": 8.37,
722
+ "grad_norm": 0.36004820466041565,
723
+ "learning_rate": 1.9212114974565664e-06,
724
+ "loss": 0.2078,
725
+ "step": 9200
726
+ },
727
+ {
728
+ "epoch": 8.46,
729
+ "grad_norm": 2.3525044918060303,
730
+ "learning_rate": 1.7166271048247796e-06,
731
+ "loss": 0.1607,
732
+ "step": 9300
733
+ },
734
+ {
735
+ "epoch": 8.55,
736
+ "grad_norm": 15.3230619430542,
737
+ "learning_rate": 1.5228965455896054e-06,
738
+ "loss": 0.1831,
739
+ "step": 9400
740
+ },
741
+ {
742
+ "epoch": 8.64,
743
+ "grad_norm": 14.115558624267578,
744
+ "learning_rate": 1.3401781168364591e-06,
745
+ "loss": 0.2013,
746
+ "step": 9500
747
+ },
748
+ {
749
+ "epoch": 8.74,
750
+ "grad_norm": 2.3320865631103516,
751
+ "learning_rate": 1.1686211176477208e-06,
752
+ "loss": 0.1922,
753
+ "step": 9600
754
+ },
755
+ {
756
+ "epoch": 8.83,
757
+ "grad_norm": 1.680296778678894,
758
+ "learning_rate": 1.00836572711058e-06,
759
+ "loss": 0.2407,
760
+ "step": 9700
761
+ },
762
+ {
763
+ "epoch": 8.92,
764
+ "grad_norm": 12.502205848693848,
765
+ "learning_rate": 8.595428897768071e-07,
766
+ "loss": 0.1716,
767
+ "step": 9800
768
+ },
769
+ {
770
+ "epoch": 9.0,
771
+ "eval_accuracy": 0.9506958250497017,
772
+ "eval_loss": 0.22239501774311066,
773
+ "eval_runtime": 105.4214,
774
+ "eval_samples_per_second": 23.857,
775
+ "eval_steps_per_second": 1.499,
776
+ "step": 9891
777
+ },
778
+ {
779
+ "epoch": 9.01,
780
+ "grad_norm": 19.13417625427246,
781
+ "learning_rate": 7.222742086680756e-07,
782
+ "loss": 0.1656,
783
+ "step": 9900
784
+ },
785
+ {
786
+ "epoch": 9.1,
787
+ "grad_norm": 6.075860023498535,
788
+ "learning_rate": 5.966718459142196e-07,
789
+ "loss": 0.1705,
790
+ "step": 10000
791
+ },
792
+ {
793
+ "epoch": 9.19,
794
+ "grad_norm": 17.40691375732422,
795
+ "learning_rate": 4.82838431105655e-07,
796
+ "loss": 0.1847,
797
+ "step": 10100
798
+ },
799
+ {
800
+ "epoch": 9.28,
801
+ "grad_norm": 21.60211944580078,
802
+ "learning_rate": 3.808669774348167e-07,
803
+ "loss": 0.1602,
804
+ "step": 10200
805
+ },
806
+ {
807
+ "epoch": 9.37,
808
+ "grad_norm": 11.399230003356934,
809
+ "learning_rate": 2.908408056951578e-07,
810
+ "loss": 0.1834,
811
+ "step": 10300
812
+ },
813
+ {
814
+ "epoch": 9.46,
815
+ "grad_norm": 0.1577068567276001,
816
+ "learning_rate": 2.1283347619979243e-07,
817
+ "loss": 0.1565,
818
+ "step": 10400
819
+ },
820
+ {
821
+ "epoch": 9.55,
822
+ "grad_norm": 5.944873332977295,
823
+ "learning_rate": 1.4690872867542892e-07,
824
+ "loss": 0.1948,
825
+ "step": 10500
826
+ },
827
+ {
828
+ "epoch": 9.65,
829
+ "grad_norm": 35.5730094909668,
830
+ "learning_rate": 9.312043018067762e-08,
831
+ "loss": 0.2041,
832
+ "step": 10600
833
+ },
834
+ {
835
+ "epoch": 9.74,
836
+ "grad_norm": 31.919750213623047,
837
+ "learning_rate": 5.1512531091333914e-08,
838
+ "loss": 0.1906,
839
+ "step": 10700
840
+ },
841
+ {
842
+ "epoch": 9.83,
843
+ "grad_norm": 15.108580589294434,
844
+ "learning_rate": 2.211902918855313e-08,
845
+ "loss": 0.1973,
846
+ "step": 10800
847
+ },
848
+ {
849
+ "epoch": 9.92,
850
+ "grad_norm": 29.040531158447266,
851
+ "learning_rate": 4.963941879295164e-09,
852
+ "loss": 0.1731,
853
+ "step": 10900
854
+ },
855
+ {
856
+ "epoch": 10.0,
857
+ "eval_accuracy": 0.9506958250497017,
858
+ "eval_loss": 0.22113871574401855,
859
+ "eval_runtime": 105.6004,
860
+ "eval_samples_per_second": 23.816,
861
+ "eval_steps_per_second": 1.496,
862
+ "step": 10990
863
+ },
864
+ {
865
+ "epoch": 10.0,
866
+ "step": 10990,
867
+ "total_flos": 4.09349935387607e+19,
868
+ "train_loss": 0.35613980254224475,
869
+ "train_runtime": 19061.1748,
870
+ "train_samples_per_second": 9.223,
871
+ "train_steps_per_second": 0.577
872
+ }
873
+ ],
874
+ "logging_steps": 100,
875
+ "max_steps": 10990,
876
+ "num_input_tokens_seen": 0,
877
+ "num_train_epochs": 10,
878
+ "save_steps": 500,
879
+ "total_flos": 4.09349935387607e+19,
880
+ "train_batch_size": 16,
881
+ "trial_name": null,
882
+ "trial_params": null
883
+ }