vuongnhathien commited on
Commit
e5b40eb
1 Parent(s): 7b78f54

End of training

Browse files
README.md CHANGED
@@ -22,7 +22,7 @@ model-index:
22
  metrics:
23
  - name: Accuracy
24
  type: accuracy
25
- value: 0.9280318091451292
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: 0.3485
36
- - Accuracy: 0.9280
37
 
38
  ## Model description
39
 
 
22
  metrics:
23
  - name: Accuracy
24
  type: accuracy
25
+ value: 0.9311507936507937
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: 0.3332
36
+ - Accuracy: 0.9312
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": 0.45005689912540897,
5
+ "train_runtime": 16363.7477,
6
+ "train_samples_per_second": 10.744,
7
+ "train_steps_per_second": 0.672
8
+ }
runs/May25_17-40-02_95411d75b2af/events.out.tfevents.1716675646.95411d75b2af.25.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95c7e207ed08cc1ad6c8db10e9ff63a7108b01447e848f8f17d78ffe89be4c96
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": 0.45005689912540897,
5
+ "train_runtime": 16363.7477,
6
+ "train_samples_per_second": 10.744,
7
+ "train_steps_per_second": 0.672
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,883 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.34852299094200134,
3
+ "best_model_checkpoint": "./convnext-base-3e-4/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": 17.357587814331055,
14
+ "learning_rate": 0.00029993871755982685,
15
+ "loss": 2.1854,
16
+ "step": 100
17
+ },
18
+ {
19
+ "epoch": 0.18,
20
+ "grad_norm": 28.103832244873047,
21
+ "learning_rate": 0.0002997549203131404,
22
+ "loss": 1.4323,
23
+ "step": 200
24
+ },
25
+ {
26
+ "epoch": 0.27,
27
+ "grad_norm": 15.243412017822266,
28
+ "learning_rate": 0.0002994487584405243,
29
+ "loss": 1.3262,
30
+ "step": 300
31
+ },
32
+ {
33
+ "epoch": 0.36,
34
+ "grad_norm": 13.571104049682617,
35
+ "learning_rate": 0.00029902048210660057,
36
+ "loss": 1.164,
37
+ "step": 400
38
+ },
39
+ {
40
+ "epoch": 0.45,
41
+ "grad_norm": 11.755253791809082,
42
+ "learning_rate": 0.00029847044125561983,
43
+ "loss": 1.1175,
44
+ "step": 500
45
+ },
46
+ {
47
+ "epoch": 0.55,
48
+ "grad_norm": 8.938959121704102,
49
+ "learning_rate": 0.00029779908532552276,
50
+ "loss": 1.0117,
51
+ "step": 600
52
+ },
53
+ {
54
+ "epoch": 0.64,
55
+ "grad_norm": 8.597779273986816,
56
+ "learning_rate": 0.00029700696288070426,
57
+ "loss": 1.0719,
58
+ "step": 700
59
+ },
60
+ {
61
+ "epoch": 0.73,
62
+ "grad_norm": 12.803329467773438,
63
+ "learning_rate": 0.0002960947211637822,
64
+ "loss": 1.0533,
65
+ "step": 800
66
+ },
67
+ {
68
+ "epoch": 0.82,
69
+ "grad_norm": 12.126448631286621,
70
+ "learning_rate": 0.00029506310556673567,
71
+ "loss": 0.9138,
72
+ "step": 900
73
+ },
74
+ {
75
+ "epoch": 0.91,
76
+ "grad_norm": 8.313648223876953,
77
+ "learning_rate": 0.0002939129590218462,
78
+ "loss": 0.8947,
79
+ "step": 1000
80
+ },
81
+ {
82
+ "epoch": 1.0,
83
+ "eval_accuracy": 0.7884691848906561,
84
+ "eval_loss": 0.7363528609275818,
85
+ "eval_runtime": 103.2916,
86
+ "eval_samples_per_second": 24.349,
87
+ "eval_steps_per_second": 1.53,
88
+ "step": 1099
89
+ },
90
+ {
91
+ "epoch": 1.0,
92
+ "grad_norm": 8.587052345275879,
93
+ "learning_rate": 0.00029264522131293815,
94
+ "loss": 0.9902,
95
+ "step": 1100
96
+ },
97
+ {
98
+ "epoch": 1.09,
99
+ "grad_norm": 8.60452938079834,
100
+ "learning_rate": 0.00029126092830748215,
101
+ "loss": 0.8517,
102
+ "step": 1200
103
+ },
104
+ {
105
+ "epoch": 1.18,
106
+ "grad_norm": 14.598617553710938,
107
+ "learning_rate": 0.00028976121111018877,
108
+ "loss": 0.802,
109
+ "step": 1300
110
+ },
111
+ {
112
+ "epoch": 1.27,
113
+ "grad_norm": 7.155284881591797,
114
+ "learning_rate": 0.00028814729513878363,
115
+ "loss": 0.7962,
116
+ "step": 1400
117
+ },
118
+ {
119
+ "epoch": 1.36,
120
+ "grad_norm": 13.24478816986084,
121
+ "learning_rate": 0.00028642049912271946,
122
+ "loss": 0.7782,
123
+ "step": 1500
124
+ },
125
+ {
126
+ "epoch": 1.46,
127
+ "grad_norm": 10.02603530883789,
128
+ "learning_rate": 0.0002845822340256436,
129
+ "loss": 0.7813,
130
+ "step": 1600
131
+ },
132
+ {
133
+ "epoch": 1.55,
134
+ "grad_norm": 10.052382469177246,
135
+ "learning_rate": 0.00028263400189250057,
136
+ "loss": 0.8079,
137
+ "step": 1700
138
+ },
139
+ {
140
+ "epoch": 1.64,
141
+ "grad_norm": 8.431123733520508,
142
+ "learning_rate": 0.0002805773946222121,
143
+ "loss": 0.8041,
144
+ "step": 1800
145
+ },
146
+ {
147
+ "epoch": 1.73,
148
+ "grad_norm": 7.3062944412231445,
149
+ "learning_rate": 0.00027841409266693835,
150
+ "loss": 0.8019,
151
+ "step": 1900
152
+ },
153
+ {
154
+ "epoch": 1.82,
155
+ "grad_norm": 7.146261215209961,
156
+ "learning_rate": 0.0002761458636589813,
157
+ "loss": 0.679,
158
+ "step": 2000
159
+ },
160
+ {
161
+ "epoch": 1.91,
162
+ "grad_norm": 12.160822868347168,
163
+ "learning_rate": 0.0002737745609664539,
164
+ "loss": 0.7643,
165
+ "step": 2100
166
+ },
167
+ {
168
+ "epoch": 2.0,
169
+ "eval_accuracy": 0.8170974155069582,
170
+ "eval_loss": 0.628582775592804,
171
+ "eval_runtime": 103.2407,
172
+ "eval_samples_per_second": 24.361,
173
+ "eval_steps_per_second": 1.53,
174
+ "step": 2198
175
+ },
176
+ {
177
+ "epoch": 2.0,
178
+ "grad_norm": 14.596354484558105,
179
+ "learning_rate": 0.00027130212217889483,
180
+ "loss": 0.7511,
181
+ "step": 2200
182
+ },
183
+ {
184
+ "epoch": 2.09,
185
+ "grad_norm": 11.497203826904297,
186
+ "learning_rate": 0.000268730567524065,
187
+ "loss": 0.6527,
188
+ "step": 2300
189
+ },
190
+ {
191
+ "epoch": 2.18,
192
+ "grad_norm": 6.976894855499268,
193
+ "learning_rate": 0.00026606199821722166,
194
+ "loss": 0.6289,
195
+ "step": 2400
196
+ },
197
+ {
198
+ "epoch": 2.27,
199
+ "grad_norm": 9.718855857849121,
200
+ "learning_rate": 0.0002632985947442167,
201
+ "loss": 0.6755,
202
+ "step": 2500
203
+ },
204
+ {
205
+ "epoch": 2.37,
206
+ "grad_norm": 6.231022357940674,
207
+ "learning_rate": 0.00026044261507982355,
208
+ "loss": 0.6377,
209
+ "step": 2600
210
+ },
211
+ {
212
+ "epoch": 2.46,
213
+ "grad_norm": 15.673833847045898,
214
+ "learning_rate": 0.0002574963928427478,
215
+ "loss": 0.626,
216
+ "step": 2700
217
+ },
218
+ {
219
+ "epoch": 2.55,
220
+ "grad_norm": 5.745795726776123,
221
+ "learning_rate": 0.00025446233538882923,
222
+ "loss": 0.6276,
223
+ "step": 2800
224
+ },
225
+ {
226
+ "epoch": 2.64,
227
+ "grad_norm": 11.555608749389648,
228
+ "learning_rate": 0.00025134292184399317,
229
+ "loss": 0.695,
230
+ "step": 2900
231
+ },
232
+ {
233
+ "epoch": 2.73,
234
+ "grad_norm": 7.4662089347839355,
235
+ "learning_rate": 0.00024814070107855875,
236
+ "loss": 0.6095,
237
+ "step": 3000
238
+ },
239
+ {
240
+ "epoch": 2.82,
241
+ "grad_norm": 7.660247802734375,
242
+ "learning_rate": 0.00024485828962455907,
243
+ "loss": 0.631,
244
+ "step": 3100
245
+ },
246
+ {
247
+ "epoch": 2.91,
248
+ "grad_norm": 9.731380462646484,
249
+ "learning_rate": 0.00024149836953777485,
250
+ "loss": 0.6036,
251
+ "step": 3200
252
+ },
253
+ {
254
+ "epoch": 3.0,
255
+ "eval_accuracy": 0.8481113320079523,
256
+ "eval_loss": 0.5258452892303467,
257
+ "eval_runtime": 104.1346,
258
+ "eval_samples_per_second": 24.151,
259
+ "eval_steps_per_second": 1.517,
260
+ "step": 3297
261
+ },
262
+ {
263
+ "epoch": 3.0,
264
+ "grad_norm": 5.469219207763672,
265
+ "learning_rate": 0.00023806368620622872,
266
+ "loss": 0.5889,
267
+ "step": 3300
268
+ },
269
+ {
270
+ "epoch": 3.09,
271
+ "grad_norm": 8.50100040435791,
272
+ "learning_rate": 0.0002345570461069312,
273
+ "loss": 0.5034,
274
+ "step": 3400
275
+ },
276
+ {
277
+ "epoch": 3.18,
278
+ "grad_norm": 12.361306190490723,
279
+ "learning_rate": 0.00023098131451271015,
280
+ "loss": 0.5181,
281
+ "step": 3500
282
+ },
283
+ {
284
+ "epoch": 3.28,
285
+ "grad_norm": 7.584137439727783,
286
+ "learning_rate": 0.0002273394131509988,
287
+ "loss": 0.5336,
288
+ "step": 3600
289
+ },
290
+ {
291
+ "epoch": 3.37,
292
+ "grad_norm": 10.401506423950195,
293
+ "learning_rate": 0.0002236343178164948,
294
+ "loss": 0.5216,
295
+ "step": 3700
296
+ },
297
+ {
298
+ "epoch": 3.46,
299
+ "grad_norm": 5.352778434753418,
300
+ "learning_rate": 0.00021986905593964046,
301
+ "loss": 0.4939,
302
+ "step": 3800
303
+ },
304
+ {
305
+ "epoch": 3.55,
306
+ "grad_norm": 7.9993109703063965,
307
+ "learning_rate": 0.0002160467041129117,
308
+ "loss": 0.521,
309
+ "step": 3900
310
+ },
311
+ {
312
+ "epoch": 3.64,
313
+ "grad_norm": 9.176218032836914,
314
+ "learning_rate": 0.00021217038557693726,
315
+ "loss": 0.5288,
316
+ "step": 4000
317
+ },
318
+ {
319
+ "epoch": 3.73,
320
+ "grad_norm": 9.322657585144043,
321
+ "learning_rate": 0.0002082432676685007,
322
+ "loss": 0.5168,
323
+ "step": 4100
324
+ },
325
+ {
326
+ "epoch": 3.82,
327
+ "grad_norm": 5.861387252807617,
328
+ "learning_rate": 0.00020426855923251228,
329
+ "loss": 0.5081,
330
+ "step": 4200
331
+ },
332
+ {
333
+ "epoch": 3.91,
334
+ "grad_norm": 4.290045261383057,
335
+ "learning_rate": 0.00020024950800006462,
336
+ "loss": 0.5012,
337
+ "step": 4300
338
+ },
339
+ {
340
+ "epoch": 4.0,
341
+ "eval_accuracy": 0.869582504970179,
342
+ "eval_loss": 0.49109867215156555,
343
+ "eval_runtime": 102.6443,
344
+ "eval_samples_per_second": 24.502,
345
+ "eval_steps_per_second": 1.539,
346
+ "step": 4396
347
+ },
348
+ {
349
+ "epoch": 4.0,
350
+ "grad_norm": 4.063934803009033,
351
+ "learning_rate": 0.0001961893979347137,
352
+ "loss": 0.4754,
353
+ "step": 4400
354
+ },
355
+ {
356
+ "epoch": 4.09,
357
+ "grad_norm": 5.163869380950928,
358
+ "learning_rate": 0.00019209154654915522,
359
+ "loss": 0.4471,
360
+ "step": 4500
361
+ },
362
+ {
363
+ "epoch": 4.19,
364
+ "grad_norm": 3.733952522277832,
365
+ "learning_rate": 0.0001879593021944875,
366
+ "loss": 0.4004,
367
+ "step": 4600
368
+ },
369
+ {
370
+ "epoch": 4.28,
371
+ "grad_norm": 7.615368366241455,
372
+ "learning_rate": 0.00018379604132427648,
373
+ "loss": 0.4076,
374
+ "step": 4700
375
+ },
376
+ {
377
+ "epoch": 4.37,
378
+ "grad_norm": 9.600367546081543,
379
+ "learning_rate": 0.0001796051657356582,
380
+ "loss": 0.4035,
381
+ "step": 4800
382
+ },
383
+ {
384
+ "epoch": 4.46,
385
+ "grad_norm": 9.41919231414795,
386
+ "learning_rate": 0.0001753900997897331,
387
+ "loss": 0.4281,
388
+ "step": 4900
389
+ },
390
+ {
391
+ "epoch": 4.55,
392
+ "grad_norm": 13.647310256958008,
393
+ "learning_rate": 0.00017115428761352327,
394
+ "loss": 0.3674,
395
+ "step": 5000
396
+ },
397
+ {
398
+ "epoch": 4.64,
399
+ "grad_norm": 2.1058132648468018,
400
+ "learning_rate": 0.00016690119028577906,
401
+ "loss": 0.3917,
402
+ "step": 5100
403
+ },
404
+ {
405
+ "epoch": 4.73,
406
+ "grad_norm": 4.259520053863525,
407
+ "learning_rate": 0.0001626342830089342,
408
+ "loss": 0.3899,
409
+ "step": 5200
410
+ },
411
+ {
412
+ "epoch": 4.82,
413
+ "grad_norm": 4.13034200668335,
414
+ "learning_rate": 0.0001583570522695211,
415
+ "loss": 0.4178,
416
+ "step": 5300
417
+ },
418
+ {
419
+ "epoch": 4.91,
420
+ "grad_norm": 13.085577011108398,
421
+ "learning_rate": 0.00015407299298936486,
422
+ "loss": 0.3926,
423
+ "step": 5400
424
+ },
425
+ {
426
+ "epoch": 5.0,
427
+ "eval_accuracy": 0.8930417495029821,
428
+ "eval_loss": 0.38039031624794006,
429
+ "eval_runtime": 103.7463,
430
+ "eval_samples_per_second": 24.242,
431
+ "eval_steps_per_second": 1.523,
432
+ "step": 5495
433
+ },
434
+ {
435
+ "epoch": 5.0,
436
+ "grad_norm": 3.522939443588257,
437
+ "learning_rate": 0.000149785605669886,
438
+ "loss": 0.3671,
439
+ "step": 5500
440
+ },
441
+ {
442
+ "epoch": 5.1,
443
+ "grad_norm": 5.17425537109375,
444
+ "learning_rate": 0.00014549839353184327,
445
+ "loss": 0.3017,
446
+ "step": 5600
447
+ },
448
+ {
449
+ "epoch": 5.19,
450
+ "grad_norm": 6.327219009399414,
451
+ "learning_rate": 0.00014121485965285484,
452
+ "loss": 0.2922,
453
+ "step": 5700
454
+ },
455
+ {
456
+ "epoch": 5.28,
457
+ "grad_norm": 0.7092263102531433,
458
+ "learning_rate": 0.00013693850410503614,
459
+ "loss": 0.314,
460
+ "step": 5800
461
+ },
462
+ {
463
+ "epoch": 5.37,
464
+ "grad_norm": 0.38373059034347534,
465
+ "learning_rate": 0.0001326728210950942,
466
+ "loss": 0.3141,
467
+ "step": 5900
468
+ },
469
+ {
470
+ "epoch": 5.46,
471
+ "grad_norm": 5.553852558135986,
472
+ "learning_rate": 0.00012842129610921376,
473
+ "loss": 0.2821,
474
+ "step": 6000
475
+ },
476
+ {
477
+ "epoch": 5.55,
478
+ "grad_norm": 3.678790330886841,
479
+ "learning_rate": 0.00012418740306506922,
480
+ "loss": 0.3359,
481
+ "step": 6100
482
+ },
483
+ {
484
+ "epoch": 5.64,
485
+ "grad_norm": 4.428023338317871,
486
+ "learning_rate": 0.00011997460147328983,
487
+ "loss": 0.2825,
488
+ "step": 6200
489
+ },
490
+ {
491
+ "epoch": 5.73,
492
+ "grad_norm": 3.3043198585510254,
493
+ "learning_rate": 0.00011578633361069557,
494
+ "loss": 0.3317,
495
+ "step": 6300
496
+ },
497
+ {
498
+ "epoch": 5.82,
499
+ "grad_norm": 1.317456603050232,
500
+ "learning_rate": 0.0001116260217076161,
501
+ "loss": 0.2983,
502
+ "step": 6400
503
+ },
504
+ {
505
+ "epoch": 5.91,
506
+ "grad_norm": 8.99087142944336,
507
+ "learning_rate": 0.00010749706515158862,
508
+ "loss": 0.3348,
509
+ "step": 6500
510
+ },
511
+ {
512
+ "epoch": 6.0,
513
+ "eval_accuracy": 0.8970178926441352,
514
+ "eval_loss": 0.41324833035469055,
515
+ "eval_runtime": 103.3001,
516
+ "eval_samples_per_second": 24.347,
517
+ "eval_steps_per_second": 1.53,
518
+ "step": 6594
519
+ },
520
+ {
521
+ "epoch": 6.01,
522
+ "grad_norm": 4.097568988800049,
523
+ "learning_rate": 0.00010340283770972167,
524
+ "loss": 0.3045,
525
+ "step": 6600
526
+ },
527
+ {
528
+ "epoch": 6.1,
529
+ "grad_norm": 10.98578929901123,
530
+ "learning_rate": 9.93466847719919e-05,
531
+ "loss": 0.2327,
532
+ "step": 6700
533
+ },
534
+ {
535
+ "epoch": 6.19,
536
+ "grad_norm": 13.919866561889648,
537
+ "learning_rate": 9.533192061772917e-05,
538
+ "loss": 0.2696,
539
+ "step": 6800
540
+ },
541
+ {
542
+ "epoch": 6.28,
543
+ "grad_norm": 6.072042942047119,
544
+ "learning_rate": 9.136182570752152e-05,
545
+ "loss": 0.2258,
546
+ "step": 6900
547
+ },
548
+ {
549
+ "epoch": 6.37,
550
+ "grad_norm": 0.17451171576976776,
551
+ "learning_rate": 8.743964400275302e-05,
552
+ "loss": 0.2406,
553
+ "step": 7000
554
+ },
555
+ {
556
+ "epoch": 6.46,
557
+ "grad_norm": 0.33122172951698303,
558
+ "learning_rate": 8.356858031496595e-05,
559
+ "loss": 0.2505,
560
+ "step": 7100
561
+ },
562
+ {
563
+ "epoch": 6.55,
564
+ "grad_norm": 1.151172161102295,
565
+ "learning_rate": 7.975179768721186e-05,
566
+ "loss": 0.1903,
567
+ "step": 7200
568
+ },
569
+ {
570
+ "epoch": 6.64,
571
+ "grad_norm": 0.7031873464584351,
572
+ "learning_rate": 7.59924148095311e-05,
573
+ "loss": 0.2085,
574
+ "step": 7300
575
+ },
576
+ {
577
+ "epoch": 6.73,
578
+ "grad_norm": 6.131903171539307,
579
+ "learning_rate": 7.229350347067424e-05,
580
+ "loss": 0.2471,
581
+ "step": 7400
582
+ },
583
+ {
584
+ "epoch": 6.82,
585
+ "grad_norm": 6.110349178314209,
586
+ "learning_rate": 6.865808604814564e-05,
587
+ "loss": 0.2085,
588
+ "step": 7500
589
+ },
590
+ {
591
+ "epoch": 6.92,
592
+ "grad_norm": 4.413149833679199,
593
+ "learning_rate": 6.508913303862143e-05,
594
+ "loss": 0.2594,
595
+ "step": 7600
596
+ },
597
+ {
598
+ "epoch": 7.0,
599
+ "eval_accuracy": 0.9153081510934393,
600
+ "eval_loss": 0.3626956641674042,
601
+ "eval_runtime": 103.1814,
602
+ "eval_samples_per_second": 24.375,
603
+ "eval_steps_per_second": 1.531,
604
+ "step": 7693
605
+ },
606
+ {
607
+ "epoch": 7.01,
608
+ "grad_norm": 3.3153018951416016,
609
+ "learning_rate": 6.158956063075865e-05,
610
+ "loss": 0.1743,
611
+ "step": 7700
612
+ },
613
+ {
614
+ "epoch": 7.1,
615
+ "grad_norm": 2.5595600605010986,
616
+ "learning_rate": 5.816222832238015e-05,
617
+ "loss": 0.1699,
618
+ "step": 7800
619
+ },
620
+ {
621
+ "epoch": 7.19,
622
+ "grad_norm": 3.9605636596679688,
623
+ "learning_rate": 5.4809936583981286e-05,
624
+ "loss": 0.2036,
625
+ "step": 7900
626
+ },
627
+ {
628
+ "epoch": 7.28,
629
+ "grad_norm": 0.7597993612289429,
630
+ "learning_rate": 5.1535424570467366e-05,
631
+ "loss": 0.1829,
632
+ "step": 8000
633
+ },
634
+ {
635
+ "epoch": 7.37,
636
+ "grad_norm": 5.694293022155762,
637
+ "learning_rate": 4.834136788299248e-05,
638
+ "loss": 0.2039,
639
+ "step": 8100
640
+ },
641
+ {
642
+ "epoch": 7.46,
643
+ "grad_norm": 0.5163713097572327,
644
+ "learning_rate": 4.523037638272821e-05,
645
+ "loss": 0.1764,
646
+ "step": 8200
647
+ },
648
+ {
649
+ "epoch": 7.55,
650
+ "grad_norm": 4.396867275238037,
651
+ "learning_rate": 4.220499205834782e-05,
652
+ "loss": 0.1862,
653
+ "step": 8300
654
+ },
655
+ {
656
+ "epoch": 7.64,
657
+ "grad_norm": 0.054451316595077515,
658
+ "learning_rate": 3.926768694896931e-05,
659
+ "loss": 0.1773,
660
+ "step": 8400
661
+ },
662
+ {
663
+ "epoch": 7.73,
664
+ "grad_norm": 0.23744052648544312,
665
+ "learning_rate": 3.64208611242546e-05,
666
+ "loss": 0.1648,
667
+ "step": 8500
668
+ },
669
+ {
670
+ "epoch": 7.83,
671
+ "grad_norm": 3.540268659591675,
672
+ "learning_rate": 3.366684072331414e-05,
673
+ "loss": 0.1541,
674
+ "step": 8600
675
+ },
676
+ {
677
+ "epoch": 7.92,
678
+ "grad_norm": 0.33744722604751587,
679
+ "learning_rate": 3.100787605402072e-05,
680
+ "loss": 0.1751,
681
+ "step": 8700
682
+ },
683
+ {
684
+ "epoch": 8.0,
685
+ "eval_accuracy": 0.9308151093439364,
686
+ "eval_loss": 0.3506681025028229,
687
+ "eval_runtime": 103.8384,
688
+ "eval_samples_per_second": 24.22,
689
+ "eval_steps_per_second": 1.522,
690
+ "step": 8792
691
+ },
692
+ {
693
+ "epoch": 8.01,
694
+ "grad_norm": 0.13831892609596252,
695
+ "learning_rate": 2.844613975428448e-05,
696
+ "loss": 0.1472,
697
+ "step": 8800
698
+ },
699
+ {
700
+ "epoch": 8.1,
701
+ "grad_norm": 2.3098576068878174,
702
+ "learning_rate": 2.5983725016792572e-05,
703
+ "loss": 0.1772,
704
+ "step": 8900
705
+ },
706
+ {
707
+ "epoch": 8.19,
708
+ "grad_norm": 0.10428429394960403,
709
+ "learning_rate": 2.3622643878662696e-05,
710
+ "loss": 0.1524,
711
+ "step": 9000
712
+ },
713
+ {
714
+ "epoch": 8.28,
715
+ "grad_norm": 9.646160125732422,
716
+ "learning_rate": 2.1364825577409422e-05,
717
+ "loss": 0.1023,
718
+ "step": 9100
719
+ },
720
+ {
721
+ "epoch": 8.37,
722
+ "grad_norm": 0.3879956305027008,
723
+ "learning_rate": 1.9212114974565664e-05,
724
+ "loss": 0.1421,
725
+ "step": 9200
726
+ },
727
+ {
728
+ "epoch": 8.46,
729
+ "grad_norm": 0.022449787706136703,
730
+ "learning_rate": 1.7166271048247792e-05,
731
+ "loss": 0.1101,
732
+ "step": 9300
733
+ },
734
+ {
735
+ "epoch": 8.55,
736
+ "grad_norm": 4.682805061340332,
737
+ "learning_rate": 1.5228965455896053e-05,
738
+ "loss": 0.1355,
739
+ "step": 9400
740
+ },
741
+ {
742
+ "epoch": 8.64,
743
+ "grad_norm": 4.212618350982666,
744
+ "learning_rate": 1.3401781168364589e-05,
745
+ "loss": 0.1465,
746
+ "step": 9500
747
+ },
748
+ {
749
+ "epoch": 8.74,
750
+ "grad_norm": 0.17709462344646454,
751
+ "learning_rate": 1.1686211176477206e-05,
752
+ "loss": 0.1375,
753
+ "step": 9600
754
+ },
755
+ {
756
+ "epoch": 8.83,
757
+ "grad_norm": 7.981707572937012,
758
+ "learning_rate": 1.0083657271105799e-05,
759
+ "loss": 0.1498,
760
+ "step": 9700
761
+ },
762
+ {
763
+ "epoch": 8.92,
764
+ "grad_norm": 7.477297306060791,
765
+ "learning_rate": 8.59542889776807e-06,
766
+ "loss": 0.1613,
767
+ "step": 9800
768
+ },
769
+ {
770
+ "epoch": 9.0,
771
+ "eval_accuracy": 0.9300198807157057,
772
+ "eval_loss": 0.34880414605140686,
773
+ "eval_runtime": 103.3381,
774
+ "eval_samples_per_second": 24.338,
775
+ "eval_steps_per_second": 1.529,
776
+ "step": 9891
777
+ },
778
+ {
779
+ "epoch": 9.01,
780
+ "grad_norm": 1.4414594173431396,
781
+ "learning_rate": 7.222742086680755e-06,
782
+ "loss": 0.1335,
783
+ "step": 9900
784
+ },
785
+ {
786
+ "epoch": 9.1,
787
+ "grad_norm": 4.2091474533081055,
788
+ "learning_rate": 5.966718459142195e-06,
789
+ "loss": 0.1066,
790
+ "step": 10000
791
+ },
792
+ {
793
+ "epoch": 9.19,
794
+ "grad_norm": 18.196033477783203,
795
+ "learning_rate": 4.828384311056549e-06,
796
+ "loss": 0.125,
797
+ "step": 10100
798
+ },
799
+ {
800
+ "epoch": 9.28,
801
+ "grad_norm": 2.323212146759033,
802
+ "learning_rate": 3.8086697743481664e-06,
803
+ "loss": 0.1239,
804
+ "step": 10200
805
+ },
806
+ {
807
+ "epoch": 9.37,
808
+ "grad_norm": 0.2004556953907013,
809
+ "learning_rate": 2.9084080569515775e-06,
810
+ "loss": 0.1076,
811
+ "step": 10300
812
+ },
813
+ {
814
+ "epoch": 9.46,
815
+ "grad_norm": 12.655696868896484,
816
+ "learning_rate": 2.128334761997924e-06,
817
+ "loss": 0.1054,
818
+ "step": 10400
819
+ },
820
+ {
821
+ "epoch": 9.55,
822
+ "grad_norm": 0.03721316158771515,
823
+ "learning_rate": 1.469087286754289e-06,
824
+ "loss": 0.125,
825
+ "step": 10500
826
+ },
827
+ {
828
+ "epoch": 9.65,
829
+ "grad_norm": 3.7458605766296387,
830
+ "learning_rate": 9.31204301806776e-07,
831
+ "loss": 0.1161,
832
+ "step": 10600
833
+ },
834
+ {
835
+ "epoch": 9.74,
836
+ "grad_norm": 0.63627690076828,
837
+ "learning_rate": 5.151253109133391e-07,
838
+ "loss": 0.1342,
839
+ "step": 10700
840
+ },
841
+ {
842
+ "epoch": 9.83,
843
+ "grad_norm": 7.162803649902344,
844
+ "learning_rate": 2.211902918855313e-07,
845
+ "loss": 0.1365,
846
+ "step": 10800
847
+ },
848
+ {
849
+ "epoch": 9.92,
850
+ "grad_norm": 1.5799212455749512,
851
+ "learning_rate": 4.9639418792951634e-08,
852
+ "loss": 0.1102,
853
+ "step": 10900
854
+ },
855
+ {
856
+ "epoch": 10.0,
857
+ "eval_accuracy": 0.9280318091451292,
858
+ "eval_loss": 0.34852299094200134,
859
+ "eval_runtime": 103.4235,
860
+ "eval_samples_per_second": 24.317,
861
+ "eval_steps_per_second": 1.528,
862
+ "step": 10990
863
+ },
864
+ {
865
+ "epoch": 10.0,
866
+ "step": 10990,
867
+ "total_flos": 4.09349935387607e+19,
868
+ "train_loss": 0.45005689912540897,
869
+ "train_runtime": 16363.7477,
870
+ "train_samples_per_second": 10.744,
871
+ "train_steps_per_second": 0.672
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
+ }