Hugging Face's logo

mmomm25
/
vit-base-patch16-224-in21k-bridgedefectVIT

Safetensors
vit
Generated from Trainer
Eval Results
Model card Files Community
mmomm25 commited on
Commit
98a8ce1
·
verified ·
1 Parent(s): a01b3b6

Training in progress, epoch 6

Browse files
Files changed (25) hide show
  1. model.safetensors +1 -1
  2. tmp-checkpoint-1020/config.json +38 -0
  3. tmp-checkpoint-1020/model.safetensors +3 -0
  4. tmp-checkpoint-1020/optimizer.pt +3 -0
  5. tmp-checkpoint-1020/preprocessor_config.json +22 -0
  6. tmp-checkpoint-1020/rng_state.pth +3 -0
  7. tmp-checkpoint-1020/scheduler.pt +3 -0
  8. tmp-checkpoint-1020/trainer_state.json +713 -0
  9. tmp-checkpoint-1020/training_args.bin +3 -0
  10. tmp-checkpoint-1275/config.json +38 -0
  11. tmp-checkpoint-1275/model.safetensors +3 -0
  12. tmp-checkpoint-1275/optimizer.pt +3 -0
  13. tmp-checkpoint-1275/preprocessor_config.json +22 -0
  14. tmp-checkpoint-1275/rng_state.pth +3 -0
  15. tmp-checkpoint-1275/scheduler.pt +3 -0
  16. tmp-checkpoint-1275/trainer_state.json +883 -0
  17. tmp-checkpoint-1275/training_args.bin +3 -0
  18. tmp-checkpoint-1530/config.json +38 -0
  19. tmp-checkpoint-1530/model.safetensors +3 -0
  20. tmp-checkpoint-1530/optimizer.pt +3 -0
  21. tmp-checkpoint-1530/preprocessor_config.json +22 -0
  22. tmp-checkpoint-1530/rng_state.pth +3 -0
  23. tmp-checkpoint-1530/scheduler.pt +3 -0
  24. tmp-checkpoint-1530/trainer_state.json +1059 -0
  25. tmp-checkpoint-1530/training_args.bin +3 -0
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fc8840a55470dd2e6a6b93487d6ef8d9bc38f276aa0c1c59052c184623f26a4a
3
  size 343233204
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea04cfdf58ce0391ff8ec6cc6ea0fbc767cb6b148e76135d7b85483eee2b7176
3
  size 343233204
tmp-checkpoint-1020/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/vit-base-patch16-224-in21k",
3
+ "architectures": [
4
+ "ViTForImageClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "encoder_stride": 16,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.0,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "crack",
13
+ "1": "for review",
14
+ "2": "material deterioration",
15
+ "3": "repair",
16
+ "4": "water repair"
17
+ },
18
+ "image_size": 224,
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 3072,
21
+ "label2id": {
22
+ "crack": 0,
23
+ "for review": 1,
24
+ "material deterioration": 2,
25
+ "repair": 3,
26
+ "water repair": 4
27
+ },
28
+ "layer_norm_eps": 1e-12,
29
+ "model_type": "vit",
30
+ "num_attention_heads": 12,
31
+ "num_channels": 3,
32
+ "num_hidden_layers": 12,
33
+ "patch_size": 16,
34
+ "problem_type": "single_label_classification",
35
+ "qkv_bias": true,
36
+ "torch_dtype": "float32",
37
+ "transformers_version": "4.37.2"
38
+ }
tmp-checkpoint-1020/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c28f5475600b613b980d62c1345bbbcee26ed2afea855a027c3553ffd2deec4
3
+ size 343233204
tmp-checkpoint-1020/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f9de2568f137357c09683d52d55aa572bb6c2e359ccecf4f26dfe8b4867143a7
3
+ size 686587322
tmp-checkpoint-1020/preprocessor_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "do_rescale": true,
4
+ "do_resize": true,
5
+ "image_mean": [
6
+ 0.5,
7
+ 0.5,
8
+ 0.5
9
+ ],
10
+ "image_processor_type": "ViTImageProcessor",
11
+ "image_std": [
12
+ 0.5,
13
+ 0.5,
14
+ 0.5
15
+ ],
16
+ "resample": 2,
17
+ "rescale_factor": 0.00392156862745098,
18
+ "size": {
19
+ "height": 224,
20
+ "width": 224
21
+ }
22
+ }
tmp-checkpoint-1020/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:efa08405b576dfca9f9d6a8212f4e8fac9e5670ef1ea1213a91ac46635e29eac
3
+ size 14244
tmp-checkpoint-1020/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e4d5f20341b06629bb0a3518e11dca3077ccbd2ee5b869c3535e69ba1caead9c
3
+ size 1064
tmp-checkpoint-1020/trainer_state.json ADDED
@@ -0,0 +1,713 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.3946979343891144,
3
+ "best_model_checkpoint": "vit-base-patch16-224-in21k-bridgedefectVIT\\checkpoint-1020",
4
+ "epoch": 4.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1020,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.04,
13
+ "learning_rate": 6.535947712418301e-07,
14
+ "loss": 1.6179,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.08,
19
+ "learning_rate": 1.3071895424836602e-06,
20
+ "loss": 1.6223,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.12,
25
+ "learning_rate": 1.96078431372549e-06,
26
+ "loss": 1.6263,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.16,
31
+ "learning_rate": 2.6143790849673204e-06,
32
+ "loss": 1.601,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.2,
37
+ "learning_rate": 3.2679738562091506e-06,
38
+ "loss": 1.6074,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.24,
43
+ "learning_rate": 3.92156862745098e-06,
44
+ "loss": 1.5955,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.27,
49
+ "learning_rate": 4.5751633986928105e-06,
50
+ "loss": 1.5698,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.31,
55
+ "learning_rate": 5.228758169934641e-06,
56
+ "loss": 1.5797,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.35,
61
+ "learning_rate": 5.882352941176471e-06,
62
+ "loss": 1.5531,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.39,
67
+ "learning_rate": 6.535947712418301e-06,
68
+ "loss": 1.5434,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.43,
73
+ "learning_rate": 7.1895424836601305e-06,
74
+ "loss": 1.5433,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.47,
79
+ "learning_rate": 7.84313725490196e-06,
80
+ "loss": 1.5329,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.51,
85
+ "learning_rate": 8.496732026143791e-06,
86
+ "loss": 1.4995,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.55,
91
+ "learning_rate": 9.150326797385621e-06,
92
+ "loss": 1.4875,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.59,
97
+ "learning_rate": 9.803921568627451e-06,
98
+ "loss": 1.4863,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.63,
103
+ "learning_rate": 1.0457516339869281e-05,
104
+ "loss": 1.4209,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.67,
109
+ "learning_rate": 1.1111111111111112e-05,
110
+ "loss": 1.4354,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.71,
115
+ "learning_rate": 1.1764705882352942e-05,
116
+ "loss": 1.3572,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.75,
121
+ "learning_rate": 1.2418300653594772e-05,
122
+ "loss": 1.3714,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.78,
127
+ "learning_rate": 1.3071895424836602e-05,
128
+ "loss": 1.3406,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.82,
133
+ "learning_rate": 1.3725490196078432e-05,
134
+ "loss": 1.2574,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.86,
139
+ "learning_rate": 1.4379084967320261e-05,
140
+ "loss": 1.2965,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.9,
145
+ "learning_rate": 1.5032679738562091e-05,
146
+ "loss": 1.2021,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.94,
151
+ "learning_rate": 1.568627450980392e-05,
152
+ "loss": 1.2072,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.98,
157
+ "learning_rate": 1.6339869281045753e-05,
158
+ "loss": 1.1457,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 1.0,
163
+ "eval_accuracy": {
164
+ "accuracy": 0.6941176470588235
165
+ },
166
+ "eval_f1": {
167
+ "f1": 0.683792091280419
168
+ },
169
+ "eval_loss": 1.158827304840088,
170
+ "eval_precision": {
171
+ "precision": 0.7377494408375539
172
+ },
173
+ "eval_recall": {
174
+ "recall": 0.6941484490686618
175
+ },
176
+ "eval_runtime": 25.3262,
177
+ "eval_samples_per_second": 20.137,
178
+ "eval_steps_per_second": 10.069,
179
+ "step": 255
180
+ },
181
+ {
182
+ "epoch": 1.02,
183
+ "learning_rate": 1.6993464052287582e-05,
184
+ "loss": 1.0924,
185
+ "step": 260
186
+ },
187
+ {
188
+ "epoch": 1.06,
189
+ "learning_rate": 1.7647058823529414e-05,
190
+ "loss": 1.0007,
191
+ "step": 270
192
+ },
193
+ {
194
+ "epoch": 1.1,
195
+ "learning_rate": 1.8300653594771242e-05,
196
+ "loss": 1.0054,
197
+ "step": 280
198
+ },
199
+ {
200
+ "epoch": 1.14,
201
+ "learning_rate": 1.895424836601307e-05,
202
+ "loss": 0.9365,
203
+ "step": 290
204
+ },
205
+ {
206
+ "epoch": 1.18,
207
+ "learning_rate": 1.9607843137254903e-05,
208
+ "loss": 0.8932,
209
+ "step": 300
210
+ },
211
+ {
212
+ "epoch": 1.22,
213
+ "learning_rate": 2.0261437908496734e-05,
214
+ "loss": 0.916,
215
+ "step": 310
216
+ },
217
+ {
218
+ "epoch": 1.25,
219
+ "learning_rate": 2.0915032679738563e-05,
220
+ "loss": 0.8099,
221
+ "step": 320
222
+ },
223
+ {
224
+ "epoch": 1.29,
225
+ "learning_rate": 2.1568627450980395e-05,
226
+ "loss": 0.8211,
227
+ "step": 330
228
+ },
229
+ {
230
+ "epoch": 1.33,
231
+ "learning_rate": 2.2222222222222223e-05,
232
+ "loss": 0.6861,
233
+ "step": 340
234
+ },
235
+ {
236
+ "epoch": 1.37,
237
+ "learning_rate": 2.2875816993464052e-05,
238
+ "loss": 0.7379,
239
+ "step": 350
240
+ },
241
+ {
242
+ "epoch": 1.41,
243
+ "learning_rate": 2.3529411764705884e-05,
244
+ "loss": 0.7349,
245
+ "step": 360
246
+ },
247
+ {
248
+ "epoch": 1.45,
249
+ "learning_rate": 2.4183006535947712e-05,
250
+ "loss": 0.6907,
251
+ "step": 370
252
+ },
253
+ {
254
+ "epoch": 1.49,
255
+ "learning_rate": 2.4836601307189544e-05,
256
+ "loss": 0.7376,
257
+ "step": 380
258
+ },
259
+ {
260
+ "epoch": 1.53,
261
+ "learning_rate": 2.5490196078431373e-05,
262
+ "loss": 0.6827,
263
+ "step": 390
264
+ },
265
+ {
266
+ "epoch": 1.57,
267
+ "learning_rate": 2.6143790849673204e-05,
268
+ "loss": 0.6969,
269
+ "step": 400
270
+ },
271
+ {
272
+ "epoch": 1.61,
273
+ "learning_rate": 2.6797385620915033e-05,
274
+ "loss": 0.6247,
275
+ "step": 410
276
+ },
277
+ {
278
+ "epoch": 1.65,
279
+ "learning_rate": 2.7450980392156865e-05,
280
+ "loss": 0.5244,
281
+ "step": 420
282
+ },
283
+ {
284
+ "epoch": 1.69,
285
+ "learning_rate": 2.8104575163398693e-05,
286
+ "loss": 0.7068,
287
+ "step": 430
288
+ },
289
+ {
290
+ "epoch": 1.73,
291
+ "learning_rate": 2.8758169934640522e-05,
292
+ "loss": 0.5623,
293
+ "step": 440
294
+ },
295
+ {
296
+ "epoch": 1.76,
297
+ "learning_rate": 2.9411764705882354e-05,
298
+ "loss": 0.5931,
299
+ "step": 450
300
+ },
301
+ {
302
+ "epoch": 1.8,
303
+ "learning_rate": 3.0065359477124182e-05,
304
+ "loss": 0.6641,
305
+ "step": 460
306
+ },
307
+ {
308
+ "epoch": 1.84,
309
+ "learning_rate": 3.0718954248366014e-05,
310
+ "loss": 0.5438,
311
+ "step": 470
312
+ },
313
+ {
314
+ "epoch": 1.88,
315
+ "learning_rate": 3.137254901960784e-05,
316
+ "loss": 0.5134,
317
+ "step": 480
318
+ },
319
+ {
320
+ "epoch": 1.92,
321
+ "learning_rate": 3.202614379084967e-05,
322
+ "loss": 0.4682,
323
+ "step": 490
324
+ },
325
+ {
326
+ "epoch": 1.96,
327
+ "learning_rate": 3.2679738562091506e-05,
328
+ "loss": 0.4762,
329
+ "step": 500
330
+ },
331
+ {
332
+ "epoch": 2.0,
333
+ "learning_rate": 3.3333333333333335e-05,
334
+ "loss": 0.4558,
335
+ "step": 510
336
+ },
337
+ {
338
+ "epoch": 2.0,
339
+ "eval_accuracy": {
340
+ "accuracy": 0.8470588235294118
341
+ },
342
+ "eval_f1": {
343
+ "f1": 0.8510200857474889
344
+ },
345
+ "eval_loss": 0.498457670211792,
346
+ "eval_precision": {
347
+ "precision": 0.8766865984788879
348
+ },
349
+ "eval_recall": {
350
+ "recall": 0.8454231938274491
351
+ },
352
+ "eval_runtime": 10.0517,
353
+ "eval_samples_per_second": 50.738,
354
+ "eval_steps_per_second": 25.369,
355
+ "step": 510
356
+ },
357
+ {
358
+ "epoch": 2.04,
359
+ "learning_rate": 3.3986928104575163e-05,
360
+ "loss": 0.4716,
361
+ "step": 520
362
+ },
363
+ {
364
+ "epoch": 2.08,
365
+ "learning_rate": 3.464052287581699e-05,
366
+ "loss": 0.5375,
367
+ "step": 530
368
+ },
369
+ {
370
+ "epoch": 2.12,
371
+ "learning_rate": 3.529411764705883e-05,
372
+ "loss": 0.4725,
373
+ "step": 540
374
+ },
375
+ {
376
+ "epoch": 2.16,
377
+ "learning_rate": 3.5947712418300656e-05,
378
+ "loss": 0.4804,
379
+ "step": 550
380
+ },
381
+ {
382
+ "epoch": 2.2,
383
+ "learning_rate": 3.6601307189542484e-05,
384
+ "loss": 0.3504,
385
+ "step": 560
386
+ },
387
+ {
388
+ "epoch": 2.24,
389
+ "learning_rate": 3.725490196078432e-05,
390
+ "loss": 0.4049,
391
+ "step": 570
392
+ },
393
+ {
394
+ "epoch": 2.27,
395
+ "learning_rate": 3.790849673202614e-05,
396
+ "loss": 0.4309,
397
+ "step": 580
398
+ },
399
+ {
400
+ "epoch": 2.31,
401
+ "learning_rate": 3.8562091503267977e-05,
402
+ "loss": 0.4938,
403
+ "step": 590
404
+ },
405
+ {
406
+ "epoch": 2.35,
407
+ "learning_rate": 3.9215686274509805e-05,
408
+ "loss": 0.5427,
409
+ "step": 600
410
+ },
411
+ {
412
+ "epoch": 2.39,
413
+ "learning_rate": 3.986928104575164e-05,
414
+ "loss": 0.4719,
415
+ "step": 610
416
+ },
417
+ {
418
+ "epoch": 2.43,
419
+ "learning_rate": 4.052287581699347e-05,
420
+ "loss": 0.4627,
421
+ "step": 620
422
+ },
423
+ {
424
+ "epoch": 2.47,
425
+ "learning_rate": 4.11764705882353e-05,
426
+ "loss": 0.5277,
427
+ "step": 630
428
+ },
429
+ {
430
+ "epoch": 2.51,
431
+ "learning_rate": 4.1830065359477126e-05,
432
+ "loss": 0.3109,
433
+ "step": 640
434
+ },
435
+ {
436
+ "epoch": 2.55,
437
+ "learning_rate": 4.2483660130718954e-05,
438
+ "loss": 0.3799,
439
+ "step": 650
440
+ },
441
+ {
442
+ "epoch": 2.59,
443
+ "learning_rate": 4.313725490196079e-05,
444
+ "loss": 0.3515,
445
+ "step": 660
446
+ },
447
+ {
448
+ "epoch": 2.63,
449
+ "learning_rate": 4.379084967320262e-05,
450
+ "loss": 0.3722,
451
+ "step": 670
452
+ },
453
+ {
454
+ "epoch": 2.67,
455
+ "learning_rate": 4.4444444444444447e-05,
456
+ "loss": 0.5337,
457
+ "step": 680
458
+ },
459
+ {
460
+ "epoch": 2.71,
461
+ "learning_rate": 4.5098039215686275e-05,
462
+ "loss": 0.3263,
463
+ "step": 690
464
+ },
465
+ {
466
+ "epoch": 2.75,
467
+ "learning_rate": 4.5751633986928104e-05,
468
+ "loss": 0.3703,
469
+ "step": 700
470
+ },
471
+ {
472
+ "epoch": 2.78,
473
+ "learning_rate": 4.640522875816994e-05,
474
+ "loss": 0.4498,
475
+ "step": 710
476
+ },
477
+ {
478
+ "epoch": 2.82,
479
+ "learning_rate": 4.705882352941177e-05,
480
+ "loss": 0.4109,
481
+ "step": 720
482
+ },
483
+ {
484
+ "epoch": 2.86,
485
+ "learning_rate": 4.77124183006536e-05,
486
+ "loss": 0.2953,
487
+ "step": 730
488
+ },
489
+ {
490
+ "epoch": 2.9,
491
+ "learning_rate": 4.8366013071895424e-05,
492
+ "loss": 0.4033,
493
+ "step": 740
494
+ },
495
+ {
496
+ "epoch": 2.94,
497
+ "learning_rate": 4.901960784313725e-05,
498
+ "loss": 0.4186,
499
+ "step": 750
500
+ },
501
+ {
502
+ "epoch": 2.98,
503
+ "learning_rate": 4.967320261437909e-05,
504
+ "loss": 0.3368,
505
+ "step": 760
506
+ },
507
+ {
508
+ "epoch": 3.0,
509
+ "eval_accuracy": {
510
+ "accuracy": 0.8705882352941177
511
+ },
512
+ "eval_f1": {
513
+ "f1": 0.8713991689505157
514
+ },
515
+ "eval_loss": 0.44290485978126526,
516
+ "eval_precision": {
517
+ "precision": 0.8884985994397759
518
+ },
519
+ "eval_recall": {
520
+ "recall": 0.8711626139817629
521
+ },
522
+ "eval_runtime": 9.9506,
523
+ "eval_samples_per_second": 51.253,
524
+ "eval_steps_per_second": 25.626,
525
+ "step": 765
526
+ },
527
+ {
528
+ "epoch": 3.02,
529
+ "learning_rate": 4.9963689179375456e-05,
530
+ "loss": 0.3347,
531
+ "step": 770
532
+ },
533
+ {
534
+ "epoch": 3.06,
535
+ "learning_rate": 4.9891067538126364e-05,
536
+ "loss": 0.3339,
537
+ "step": 780
538
+ },
539
+ {
540
+ "epoch": 3.1,
541
+ "learning_rate": 4.981844589687727e-05,
542
+ "loss": 0.5573,
543
+ "step": 790
544
+ },
545
+ {
546
+ "epoch": 3.14,
547
+ "learning_rate": 4.974582425562818e-05,
548
+ "loss": 0.5463,
549
+ "step": 800
550
+ },
551
+ {
552
+ "epoch": 3.18,
553
+ "learning_rate": 4.967320261437909e-05,
554
+ "loss": 0.3453,
555
+ "step": 810
556
+ },
557
+ {
558
+ "epoch": 3.22,
559
+ "learning_rate": 4.9600580973129996e-05,
560
+ "loss": 0.3814,
561
+ "step": 820
562
+ },
563
+ {
564
+ "epoch": 3.25,
565
+ "learning_rate": 4.9527959331880904e-05,
566
+ "loss": 0.4434,
567
+ "step": 830
568
+ },
569
+ {
570
+ "epoch": 3.29,
571
+ "learning_rate": 4.945533769063181e-05,
572
+ "loss": 0.3675,
573
+ "step": 840
574
+ },
575
+ {
576
+ "epoch": 3.33,
577
+ "learning_rate": 4.938271604938271e-05,
578
+ "loss": 0.5386,
579
+ "step": 850
580
+ },
581
+ {
582
+ "epoch": 3.37,
583
+ "learning_rate": 4.931009440813363e-05,
584
+ "loss": 0.2692,
585
+ "step": 860
586
+ },
587
+ {
588
+ "epoch": 3.41,
589
+ "learning_rate": 4.9237472766884536e-05,
590
+ "loss": 0.4815,
591
+ "step": 870
592
+ },
593
+ {
594
+ "epoch": 3.45,
595
+ "learning_rate": 4.9164851125635444e-05,
596
+ "loss": 0.4178,
597
+ "step": 880
598
+ },
599
+ {
600
+ "epoch": 3.49,
601
+ "learning_rate": 4.909222948438635e-05,
602
+ "loss": 0.3618,
603
+ "step": 890
604
+ },
605
+ {
606
+ "epoch": 3.53,
607
+ "learning_rate": 4.901960784313725e-05,
608
+ "loss": 0.6841,
609
+ "step": 900
610
+ },
611
+ {
612
+ "epoch": 3.57,
613
+ "learning_rate": 4.894698620188817e-05,
614
+ "loss": 0.471,
615
+ "step": 910
616
+ },
617
+ {
618
+ "epoch": 3.61,
619
+ "learning_rate": 4.8874364560639076e-05,
620
+ "loss": 0.313,
621
+ "step": 920
622
+ },
623
+ {
624
+ "epoch": 3.65,
625
+ "learning_rate": 4.8801742919389983e-05,
626
+ "loss": 0.2437,
627
+ "step": 930
628
+ },
629
+ {
630
+ "epoch": 3.69,
631
+ "learning_rate": 4.8729121278140885e-05,
632
+ "loss": 0.4748,
633
+ "step": 940
634
+ },
635
+ {
636
+ "epoch": 3.73,
637
+ "learning_rate": 4.865649963689179e-05,
638
+ "loss": 0.5519,
639
+ "step": 950
640
+ },
641
+ {
642
+ "epoch": 3.76,
643
+ "learning_rate": 4.85838779956427e-05,
644
+ "loss": 0.3051,
645
+ "step": 960
646
+ },
647
+ {
648
+ "epoch": 3.8,
649
+ "learning_rate": 4.8511256354393615e-05,
650
+ "loss": 0.3847,
651
+ "step": 970
652
+ },
653
+ {
654
+ "epoch": 3.84,
655
+ "learning_rate": 4.843863471314452e-05,
656
+ "loss": 0.2454,
657
+ "step": 980
658
+ },
659
+ {
660
+ "epoch": 3.88,
661
+ "learning_rate": 4.8366013071895424e-05,
662
+ "loss": 0.379,
663
+ "step": 990
664
+ },
665
+ {
666
+ "epoch": 3.92,
667
+ "learning_rate": 4.829339143064633e-05,
668
+ "loss": 0.4485,
669
+ "step": 1000
670
+ },
671
+ {
672
+ "epoch": 3.96,
673
+ "learning_rate": 4.822076978939724e-05,
674
+ "loss": 0.2484,
675
+ "step": 1010
676
+ },
677
+ {
678
+ "epoch": 4.0,
679
+ "learning_rate": 4.814814814814815e-05,
680
+ "loss": 0.4402,
681
+ "step": 1020
682
+ },
683
+ {
684
+ "epoch": 4.0,
685
+ "eval_accuracy": {
686
+ "accuracy": 0.8784313725490196
687
+ },
688
+ "eval_f1": {
689
+ "f1": 0.880471094803992
690
+ },
691
+ "eval_loss": 0.3946979343891144,
692
+ "eval_precision": {
693
+ "precision": 0.8952890487430221
694
+ },
695
+ "eval_recall": {
696
+ "recall": 0.8776488582339645
697
+ },
698
+ "eval_runtime": 10.1771,
699
+ "eval_samples_per_second": 50.113,
700
+ "eval_steps_per_second": 25.056,
701
+ "step": 1020
702
+ }
703
+ ],
704
+ "logging_steps": 10,
705
+ "max_steps": 7650,
706
+ "num_input_tokens_seen": 0,
707
+ "num_train_epochs": 30,
708
+ "save_steps": 500,
709
+ "total_flos": 6.323516375010509e+17,
710
+ "train_batch_size": 2,
711
+ "trial_name": null,
712
+ "trial_params": null
713
+ }
tmp-checkpoint-1020/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:418b85f35316fefe4bc777808f9dff13bb59b607bf93a8504dc8b5709313507f
3
+ size 4728
tmp-checkpoint-1275/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/vit-base-patch16-224-in21k",
3
+ "architectures": [
4
+ "ViTForImageClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "encoder_stride": 16,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.0,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "crack",
13
+ "1": "for review",
14
+ "2": "material deterioration",
15
+ "3": "repair",
16
+ "4": "water repair"
17
+ },
18
+ "image_size": 224,
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 3072,
21
+ "label2id": {
22
+ "crack": 0,
23
+ "for review": 1,
24
+ "material deterioration": 2,
25
+ "repair": 3,
26
+ "water repair": 4
27
+ },
28
+ "layer_norm_eps": 1e-12,
29
+ "model_type": "vit",
30
+ "num_attention_heads": 12,
31
+ "num_channels": 3,
32
+ "num_hidden_layers": 12,
33
+ "patch_size": 16,
34
+ "problem_type": "single_label_classification",
35
+ "qkv_bias": true,
36
+ "torch_dtype": "float32",
37
+ "transformers_version": "4.37.2"
38
+ }
tmp-checkpoint-1275/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:230154c8efeb462794b61cca5cf8704968cd1c2dc1ffa8dceee38a6be5aad21f
3
+ size 343233204
tmp-checkpoint-1275/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edeb762066508aac76abd09da60f9434bcc0bf612df3bc23c8c491f18ef189c8
3
+ size 686587322
tmp-checkpoint-1275/preprocessor_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "do_rescale": true,
4
+ "do_resize": true,
5
+ "image_mean": [
6
+ 0.5,
7
+ 0.5,
8
+ 0.5
9
+ ],
10
+ "image_processor_type": "ViTImageProcessor",
11
+ "image_std": [
12
+ 0.5,
13
+ 0.5,
14
+ 0.5
15
+ ],
16
+ "resample": 2,
17
+ "rescale_factor": 0.00392156862745098,
18
+ "size": {
19
+ "height": 224,
20
+ "width": 224
21
+ }
22
+ }
tmp-checkpoint-1275/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19dbabbaee145ce4bc6744fe8f5fb5ba6e6dcdd62a60cc5872001492c061e1cb
3
+ size 14244
tmp-checkpoint-1275/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af0f8ccddc207376c523460fb9a4583d28ce17522740ea21fb3e9f66f2d279bf
3
+ size 1064
tmp-checkpoint-1275/trainer_state.json ADDED
@@ -0,0 +1,883 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.3946979343891144,
3
+ "best_model_checkpoint": "vit-base-patch16-224-in21k-bridgedefectVIT\\checkpoint-1020",
4
+ "epoch": 5.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1275,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.04,
13
+ "learning_rate": 6.535947712418301e-07,
14
+ "loss": 1.6179,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.08,
19
+ "learning_rate": 1.3071895424836602e-06,
20
+ "loss": 1.6223,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.12,
25
+ "learning_rate": 1.96078431372549e-06,
26
+ "loss": 1.6263,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.16,
31
+ "learning_rate": 2.6143790849673204e-06,
32
+ "loss": 1.601,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.2,
37
+ "learning_rate": 3.2679738562091506e-06,
38
+ "loss": 1.6074,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.24,
43
+ "learning_rate": 3.92156862745098e-06,
44
+ "loss": 1.5955,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.27,
49
+ "learning_rate": 4.5751633986928105e-06,
50
+ "loss": 1.5698,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.31,
55
+ "learning_rate": 5.228758169934641e-06,
56
+ "loss": 1.5797,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.35,
61
+ "learning_rate": 5.882352941176471e-06,
62
+ "loss": 1.5531,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.39,
67
+ "learning_rate": 6.535947712418301e-06,
68
+ "loss": 1.5434,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.43,
73
+ "learning_rate": 7.1895424836601305e-06,
74
+ "loss": 1.5433,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.47,
79
+ "learning_rate": 7.84313725490196e-06,
80
+ "loss": 1.5329,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.51,
85
+ "learning_rate": 8.496732026143791e-06,
86
+ "loss": 1.4995,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.55,
91
+ "learning_rate": 9.150326797385621e-06,
92
+ "loss": 1.4875,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.59,
97
+ "learning_rate": 9.803921568627451e-06,
98
+ "loss": 1.4863,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.63,
103
+ "learning_rate": 1.0457516339869281e-05,
104
+ "loss": 1.4209,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.67,
109
+ "learning_rate": 1.1111111111111112e-05,
110
+ "loss": 1.4354,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.71,
115
+ "learning_rate": 1.1764705882352942e-05,
116
+ "loss": 1.3572,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.75,
121
+ "learning_rate": 1.2418300653594772e-05,
122
+ "loss": 1.3714,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.78,
127
+ "learning_rate": 1.3071895424836602e-05,
128
+ "loss": 1.3406,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.82,
133
+ "learning_rate": 1.3725490196078432e-05,
134
+ "loss": 1.2574,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.86,
139
+ "learning_rate": 1.4379084967320261e-05,
140
+ "loss": 1.2965,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.9,
145
+ "learning_rate": 1.5032679738562091e-05,
146
+ "loss": 1.2021,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.94,
151
+ "learning_rate": 1.568627450980392e-05,
152
+ "loss": 1.2072,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.98,
157
+ "learning_rate": 1.6339869281045753e-05,
158
+ "loss": 1.1457,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 1.0,
163
+ "eval_accuracy": {
164
+ "accuracy": 0.6941176470588235
165
+ },
166
+ "eval_f1": {
167
+ "f1": 0.683792091280419
168
+ },
169
+ "eval_loss": 1.158827304840088,
170
+ "eval_precision": {
171
+ "precision": 0.7377494408375539
172
+ },
173
+ "eval_recall": {
174
+ "recall": 0.6941484490686618
175
+ },
176
+ "eval_runtime": 25.3262,
177
+ "eval_samples_per_second": 20.137,
178
+ "eval_steps_per_second": 10.069,
179
+ "step": 255
180
+ },
181
+ {
182
+ "epoch": 1.02,
183
+ "learning_rate": 1.6993464052287582e-05,
184
+ "loss": 1.0924,
185
+ "step": 260
186
+ },
187
+ {
188
+ "epoch": 1.06,
189
+ "learning_rate": 1.7647058823529414e-05,
190
+ "loss": 1.0007,
191
+ "step": 270
192
+ },
193
+ {
194
+ "epoch": 1.1,
195
+ "learning_rate": 1.8300653594771242e-05,
196
+ "loss": 1.0054,
197
+ "step": 280
198
+ },
199
+ {
200
+ "epoch": 1.14,
201
+ "learning_rate": 1.895424836601307e-05,
202
+ "loss": 0.9365,
203
+ "step": 290
204
+ },
205
+ {
206
+ "epoch": 1.18,
207
+ "learning_rate": 1.9607843137254903e-05,
208
+ "loss": 0.8932,
209
+ "step": 300
210
+ },
211
+ {
212
+ "epoch": 1.22,
213
+ "learning_rate": 2.0261437908496734e-05,
214
+ "loss": 0.916,
215
+ "step": 310
216
+ },
217
+ {
218
+ "epoch": 1.25,
219
+ "learning_rate": 2.0915032679738563e-05,
220
+ "loss": 0.8099,
221
+ "step": 320
222
+ },
223
+ {
224
+ "epoch": 1.29,
225
+ "learning_rate": 2.1568627450980395e-05,
226
+ "loss": 0.8211,
227
+ "step": 330
228
+ },
229
+ {
230
+ "epoch": 1.33,
231
+ "learning_rate": 2.2222222222222223e-05,
232
+ "loss": 0.6861,
233
+ "step": 340
234
+ },
235
+ {
236
+ "epoch": 1.37,
237
+ "learning_rate": 2.2875816993464052e-05,
238
+ "loss": 0.7379,
239
+ "step": 350
240
+ },
241
+ {
242
+ "epoch": 1.41,
243
+ "learning_rate": 2.3529411764705884e-05,
244
+ "loss": 0.7349,
245
+ "step": 360
246
+ },
247
+ {
248
+ "epoch": 1.45,
249
+ "learning_rate": 2.4183006535947712e-05,
250
+ "loss": 0.6907,
251
+ "step": 370
252
+ },
253
+ {
254
+ "epoch": 1.49,
255
+ "learning_rate": 2.4836601307189544e-05,
256
+ "loss": 0.7376,
257
+ "step": 380
258
+ },
259
+ {
260
+ "epoch": 1.53,
261
+ "learning_rate": 2.5490196078431373e-05,
262
+ "loss": 0.6827,
263
+ "step": 390
264
+ },
265
+ {
266
+ "epoch": 1.57,
267
+ "learning_rate": 2.6143790849673204e-05,
268
+ "loss": 0.6969,
269
+ "step": 400
270
+ },
271
+ {
272
+ "epoch": 1.61,
273
+ "learning_rate": 2.6797385620915033e-05,
274
+ "loss": 0.6247,
275
+ "step": 410
276
+ },
277
+ {
278
+ "epoch": 1.65,
279
+ "learning_rate": 2.7450980392156865e-05,
280
+ "loss": 0.5244,
281
+ "step": 420
282
+ },
283
+ {
284
+ "epoch": 1.69,
285
+ "learning_rate": 2.8104575163398693e-05,
286
+ "loss": 0.7068,
287
+ "step": 430
288
+ },
289
+ {
290
+ "epoch": 1.73,
291
+ "learning_rate": 2.8758169934640522e-05,
292
+ "loss": 0.5623,
293
+ "step": 440
294
+ },
295
+ {
296
+ "epoch": 1.76,
297
+ "learning_rate": 2.9411764705882354e-05,
298
+ "loss": 0.5931,
299
+ "step": 450
300
+ },
301
+ {
302
+ "epoch": 1.8,
303
+ "learning_rate": 3.0065359477124182e-05,
304
+ "loss": 0.6641,
305
+ "step": 460
306
+ },
307
+ {
308
+ "epoch": 1.84,
309
+ "learning_rate": 3.0718954248366014e-05,
310
+ "loss": 0.5438,
311
+ "step": 470
312
+ },
313
+ {
314
+ "epoch": 1.88,
315
+ "learning_rate": 3.137254901960784e-05,
316
+ "loss": 0.5134,
317
+ "step": 480
318
+ },
319
+ {
320
+ "epoch": 1.92,
321
+ "learning_rate": 3.202614379084967e-05,
322
+ "loss": 0.4682,
323
+ "step": 490
324
+ },
325
+ {
326
+ "epoch": 1.96,
327
+ "learning_rate": 3.2679738562091506e-05,
328
+ "loss": 0.4762,
329
+ "step": 500
330
+ },
331
+ {
332
+ "epoch": 2.0,
333
+ "learning_rate": 3.3333333333333335e-05,
334
+ "loss": 0.4558,
335
+ "step": 510
336
+ },
337
+ {
338
+ "epoch": 2.0,
339
+ "eval_accuracy": {
340
+ "accuracy": 0.8470588235294118
341
+ },
342
+ "eval_f1": {
343
+ "f1": 0.8510200857474889
344
+ },
345
+ "eval_loss": 0.498457670211792,
346
+ "eval_precision": {
347
+ "precision": 0.8766865984788879
348
+ },
349
+ "eval_recall": {
350
+ "recall": 0.8454231938274491
351
+ },
352
+ "eval_runtime": 10.0517,
353
+ "eval_samples_per_second": 50.738,
354
+ "eval_steps_per_second": 25.369,
355
+ "step": 510
356
+ },
357
+ {
358
+ "epoch": 2.04,
359
+ "learning_rate": 3.3986928104575163e-05,
360
+ "loss": 0.4716,
361
+ "step": 520
362
+ },
363
+ {
364
+ "epoch": 2.08,
365
+ "learning_rate": 3.464052287581699e-05,
366
+ "loss": 0.5375,
367
+ "step": 530
368
+ },
369
+ {
370
+ "epoch": 2.12,
371
+ "learning_rate": 3.529411764705883e-05,
372
+ "loss": 0.4725,
373
+ "step": 540
374
+ },
375
+ {
376
+ "epoch": 2.16,
377
+ "learning_rate": 3.5947712418300656e-05,
378
+ "loss": 0.4804,
379
+ "step": 550
380
+ },
381
+ {
382
+ "epoch": 2.2,
383
+ "learning_rate": 3.6601307189542484e-05,
384
+ "loss": 0.3504,
385
+ "step": 560
386
+ },
387
+ {
388
+ "epoch": 2.24,
389
+ "learning_rate": 3.725490196078432e-05,
390
+ "loss": 0.4049,
391
+ "step": 570
392
+ },
393
+ {
394
+ "epoch": 2.27,
395
+ "learning_rate": 3.790849673202614e-05,
396
+ "loss": 0.4309,
397
+ "step": 580
398
+ },
399
+ {
400
+ "epoch": 2.31,
401
+ "learning_rate": 3.8562091503267977e-05,
402
+ "loss": 0.4938,
403
+ "step": 590
404
+ },
405
+ {
406
+ "epoch": 2.35,
407
+ "learning_rate": 3.9215686274509805e-05,
408
+ "loss": 0.5427,
409
+ "step": 600
410
+ },
411
+ {
412
+ "epoch": 2.39,
413
+ "learning_rate": 3.986928104575164e-05,
414
+ "loss": 0.4719,
415
+ "step": 610
416
+ },
417
+ {
418
+ "epoch": 2.43,
419
+ "learning_rate": 4.052287581699347e-05,
420
+ "loss": 0.4627,
421
+ "step": 620
422
+ },
423
+ {
424
+ "epoch": 2.47,
425
+ "learning_rate": 4.11764705882353e-05,
426
+ "loss": 0.5277,
427
+ "step": 630
428
+ },
429
+ {
430
+ "epoch": 2.51,
431
+ "learning_rate": 4.1830065359477126e-05,
432
+ "loss": 0.3109,
433
+ "step": 640
434
+ },
435
+ {
436
+ "epoch": 2.55,
437
+ "learning_rate": 4.2483660130718954e-05,
438
+ "loss": 0.3799,
439
+ "step": 650
440
+ },
441
+ {
442
+ "epoch": 2.59,
443
+ "learning_rate": 4.313725490196079e-05,
444
+ "loss": 0.3515,
445
+ "step": 660
446
+ },
447
+ {
448
+ "epoch": 2.63,
449
+ "learning_rate": 4.379084967320262e-05,
450
+ "loss": 0.3722,
451
+ "step": 670
452
+ },
453
+ {
454
+ "epoch": 2.67,
455
+ "learning_rate": 4.4444444444444447e-05,
456
+ "loss": 0.5337,
457
+ "step": 680
458
+ },
459
+ {
460
+ "epoch": 2.71,
461
+ "learning_rate": 4.5098039215686275e-05,
462
+ "loss": 0.3263,
463
+ "step": 690
464
+ },
465
+ {
466
+ "epoch": 2.75,
467
+ "learning_rate": 4.5751633986928104e-05,
468
+ "loss": 0.3703,
469
+ "step": 700
470
+ },
471
+ {
472
+ "epoch": 2.78,
473
+ "learning_rate": 4.640522875816994e-05,
474
+ "loss": 0.4498,
475
+ "step": 710
476
+ },
477
+ {
478
+ "epoch": 2.82,
479
+ "learning_rate": 4.705882352941177e-05,
480
+ "loss": 0.4109,
481
+ "step": 720
482
+ },
483
+ {
484
+ "epoch": 2.86,
485
+ "learning_rate": 4.77124183006536e-05,
486
+ "loss": 0.2953,
487
+ "step": 730
488
+ },
489
+ {
490
+ "epoch": 2.9,
491
+ "learning_rate": 4.8366013071895424e-05,
492
+ "loss": 0.4033,
493
+ "step": 740
494
+ },
495
+ {
496
+ "epoch": 2.94,
497
+ "learning_rate": 4.901960784313725e-05,
498
+ "loss": 0.4186,
499
+ "step": 750
500
+ },
501
+ {
502
+ "epoch": 2.98,
503
+ "learning_rate": 4.967320261437909e-05,
504
+ "loss": 0.3368,
505
+ "step": 760
506
+ },
507
+ {
508
+ "epoch": 3.0,
509
+ "eval_accuracy": {
510
+ "accuracy": 0.8705882352941177
511
+ },
512
+ "eval_f1": {
513
+ "f1": 0.8713991689505157
514
+ },
515
+ "eval_loss": 0.44290485978126526,
516
+ "eval_precision": {
517
+ "precision": 0.8884985994397759
518
+ },
519
+ "eval_recall": {
520
+ "recall": 0.8711626139817629
521
+ },
522
+ "eval_runtime": 9.9506,
523
+ "eval_samples_per_second": 51.253,
524
+ "eval_steps_per_second": 25.626,
525
+ "step": 765
526
+ },
527
+ {
528
+ "epoch": 3.02,
529
+ "learning_rate": 4.9963689179375456e-05,
530
+ "loss": 0.3347,
531
+ "step": 770
532
+ },
533
+ {
534
+ "epoch": 3.06,
535
+ "learning_rate": 4.9891067538126364e-05,
536
+ "loss": 0.3339,
537
+ "step": 780
538
+ },
539
+ {
540
+ "epoch": 3.1,
541
+ "learning_rate": 4.981844589687727e-05,
542
+ "loss": 0.5573,
543
+ "step": 790
544
+ },
545
+ {
546
+ "epoch": 3.14,
547
+ "learning_rate": 4.974582425562818e-05,
548
+ "loss": 0.5463,
549
+ "step": 800
550
+ },
551
+ {
552
+ "epoch": 3.18,
553
+ "learning_rate": 4.967320261437909e-05,
554
+ "loss": 0.3453,
555
+ "step": 810
556
+ },
557
+ {
558
+ "epoch": 3.22,
559
+ "learning_rate": 4.9600580973129996e-05,
560
+ "loss": 0.3814,
561
+ "step": 820
562
+ },
563
+ {
564
+ "epoch": 3.25,
565
+ "learning_rate": 4.9527959331880904e-05,
566
+ "loss": 0.4434,
567
+ "step": 830
568
+ },
569
+ {
570
+ "epoch": 3.29,
571
+ "learning_rate": 4.945533769063181e-05,
572
+ "loss": 0.3675,
573
+ "step": 840
574
+ },
575
+ {
576
+ "epoch": 3.33,
577
+ "learning_rate": 4.938271604938271e-05,
578
+ "loss": 0.5386,
579
+ "step": 850
580
+ },
581
+ {
582
+ "epoch": 3.37,
583
+ "learning_rate": 4.931009440813363e-05,
584
+ "loss": 0.2692,
585
+ "step": 860
586
+ },
587
+ {
588
+ "epoch": 3.41,
589
+ "learning_rate": 4.9237472766884536e-05,
590
+ "loss": 0.4815,
591
+ "step": 870
592
+ },
593
+ {
594
+ "epoch": 3.45,
595
+ "learning_rate": 4.9164851125635444e-05,
596
+ "loss": 0.4178,
597
+ "step": 880
598
+ },
599
+ {
600
+ "epoch": 3.49,
601
+ "learning_rate": 4.909222948438635e-05,
602
+ "loss": 0.3618,
603
+ "step": 890
604
+ },
605
+ {
606
+ "epoch": 3.53,
607
+ "learning_rate": 4.901960784313725e-05,
608
+ "loss": 0.6841,
609
+ "step": 900
610
+ },
611
+ {
612
+ "epoch": 3.57,
613
+ "learning_rate": 4.894698620188817e-05,
614
+ "loss": 0.471,
615
+ "step": 910
616
+ },
617
+ {
618
+ "epoch": 3.61,
619
+ "learning_rate": 4.8874364560639076e-05,
620
+ "loss": 0.313,
621
+ "step": 920
622
+ },
623
+ {
624
+ "epoch": 3.65,
625
+ "learning_rate": 4.8801742919389983e-05,
626
+ "loss": 0.2437,
627
+ "step": 930
628
+ },
629
+ {
630
+ "epoch": 3.69,
631
+ "learning_rate": 4.8729121278140885e-05,
632
+ "loss": 0.4748,
633
+ "step": 940
634
+ },
635
+ {
636
+ "epoch": 3.73,
637
+ "learning_rate": 4.865649963689179e-05,
638
+ "loss": 0.5519,
639
+ "step": 950
640
+ },
641
+ {
642
+ "epoch": 3.76,
643
+ "learning_rate": 4.85838779956427e-05,
644
+ "loss": 0.3051,
645
+ "step": 960
646
+ },
647
+ {
648
+ "epoch": 3.8,
649
+ "learning_rate": 4.8511256354393615e-05,
650
+ "loss": 0.3847,
651
+ "step": 970
652
+ },
653
+ {
654
+ "epoch": 3.84,
655
+ "learning_rate": 4.843863471314452e-05,
656
+ "loss": 0.2454,
657
+ "step": 980
658
+ },
659
+ {
660
+ "epoch": 3.88,
661
+ "learning_rate": 4.8366013071895424e-05,
662
+ "loss": 0.379,
663
+ "step": 990
664
+ },
665
+ {
666
+ "epoch": 3.92,
667
+ "learning_rate": 4.829339143064633e-05,
668
+ "loss": 0.4485,
669
+ "step": 1000
670
+ },
671
+ {
672
+ "epoch": 3.96,
673
+ "learning_rate": 4.822076978939724e-05,
674
+ "loss": 0.2484,
675
+ "step": 1010
676
+ },
677
+ {
678
+ "epoch": 4.0,
679
+ "learning_rate": 4.814814814814815e-05,
680
+ "loss": 0.4402,
681
+ "step": 1020
682
+ },
683
+ {
684
+ "epoch": 4.0,
685
+ "eval_accuracy": {
686
+ "accuracy": 0.8784313725490196
687
+ },
688
+ "eval_f1": {
689
+ "f1": 0.880471094803992
690
+ },
691
+ "eval_loss": 0.3946979343891144,
692
+ "eval_precision": {
693
+ "precision": 0.8952890487430221
694
+ },
695
+ "eval_recall": {
696
+ "recall": 0.8776488582339645
697
+ },
698
+ "eval_runtime": 10.1771,
699
+ "eval_samples_per_second": 50.113,
700
+ "eval_steps_per_second": 25.056,
701
+ "step": 1020
702
+ },
703
+ {
704
+ "epoch": 4.04,
705
+ "learning_rate": 4.807552650689906e-05,
706
+ "loss": 0.3494,
707
+ "step": 1030
708
+ },
709
+ {
710
+ "epoch": 4.08,
711
+ "learning_rate": 4.8002904865649964e-05,
712
+ "loss": 0.3327,
713
+ "step": 1040
714
+ },
715
+ {
716
+ "epoch": 4.12,
717
+ "learning_rate": 4.793028322440087e-05,
718
+ "loss": 0.2898,
719
+ "step": 1050
720
+ },
721
+ {
722
+ "epoch": 4.16,
723
+ "learning_rate": 4.785766158315178e-05,
724
+ "loss": 0.3965,
725
+ "step": 1060
726
+ },
727
+ {
728
+ "epoch": 4.2,
729
+ "learning_rate": 4.778503994190269e-05,
730
+ "loss": 0.3607,
731
+ "step": 1070
732
+ },
733
+ {
734
+ "epoch": 4.24,
735
+ "learning_rate": 4.77124183006536e-05,
736
+ "loss": 0.3714,
737
+ "step": 1080
738
+ },
739
+ {
740
+ "epoch": 4.27,
741
+ "learning_rate": 4.7639796659404504e-05,
742
+ "loss": 0.1796,
743
+ "step": 1090
744
+ },
745
+ {
746
+ "epoch": 4.31,
747
+ "learning_rate": 4.756717501815541e-05,
748
+ "loss": 0.1874,
749
+ "step": 1100
750
+ },
751
+ {
752
+ "epoch": 4.35,
753
+ "learning_rate": 4.749455337690632e-05,
754
+ "loss": 0.3462,
755
+ "step": 1110
756
+ },
757
+ {
758
+ "epoch": 4.39,
759
+ "learning_rate": 4.742193173565723e-05,
760
+ "loss": 0.2082,
761
+ "step": 1120
762
+ },
763
+ {
764
+ "epoch": 4.43,
765
+ "learning_rate": 4.7349310094408136e-05,
766
+ "loss": 0.4531,
767
+ "step": 1130
768
+ },
769
+ {
770
+ "epoch": 4.47,
771
+ "learning_rate": 4.7276688453159044e-05,
772
+ "loss": 0.3951,
773
+ "step": 1140
774
+ },
775
+ {
776
+ "epoch": 4.51,
777
+ "learning_rate": 4.720406681190995e-05,
778
+ "loss": 0.2337,
779
+ "step": 1150
780
+ },
781
+ {
782
+ "epoch": 4.55,
783
+ "learning_rate": 4.713144517066086e-05,
784
+ "loss": 0.2118,
785
+ "step": 1160
786
+ },
787
+ {
788
+ "epoch": 4.59,
789
+ "learning_rate": 4.705882352941177e-05,
790
+ "loss": 0.373,
791
+ "step": 1170
792
+ },
793
+ {
794
+ "epoch": 4.63,
795
+ "learning_rate": 4.6986201888162675e-05,
796
+ "loss": 0.1484,
797
+ "step": 1180
798
+ },
799
+ {
800
+ "epoch": 4.67,
801
+ "learning_rate": 4.691358024691358e-05,
802
+ "loss": 0.2568,
803
+ "step": 1190
804
+ },
805
+ {
806
+ "epoch": 4.71,
807
+ "learning_rate": 4.684095860566449e-05,
808
+ "loss": 0.3067,
809
+ "step": 1200
810
+ },
811
+ {
812
+ "epoch": 4.75,
813
+ "learning_rate": 4.67683369644154e-05,
814
+ "loss": 0.2007,
815
+ "step": 1210
816
+ },
817
+ {
818
+ "epoch": 4.78,
819
+ "learning_rate": 4.669571532316631e-05,
820
+ "loss": 0.2493,
821
+ "step": 1220
822
+ },
823
+ {
824
+ "epoch": 4.82,
825
+ "learning_rate": 4.6623093681917215e-05,
826
+ "loss": 0.2717,
827
+ "step": 1230
828
+ },
829
+ {
830
+ "epoch": 4.86,
831
+ "learning_rate": 4.6550472040668116e-05,
832
+ "loss": 0.4134,
833
+ "step": 1240
834
+ },
835
+ {
836
+ "epoch": 4.9,
837
+ "learning_rate": 4.647785039941903e-05,
838
+ "loss": 0.2417,
839
+ "step": 1250
840
+ },
841
+ {
842
+ "epoch": 4.94,
843
+ "learning_rate": 4.640522875816994e-05,
844
+ "loss": 0.2573,
845
+ "step": 1260
846
+ },
847
+ {
848
+ "epoch": 4.98,
849
+ "learning_rate": 4.633260711692085e-05,
850
+ "loss": 0.3937,
851
+ "step": 1270
852
+ },
853
+ {
854
+ "epoch": 5.0,
855
+ "eval_accuracy": {
856
+ "accuracy": 0.8725490196078431
857
+ },
858
+ "eval_f1": {
859
+ "f1": 0.8703780158419334
860
+ },
861
+ "eval_loss": 0.4377364218235016,
862
+ "eval_precision": {
863
+ "precision": 0.8900760323128502
864
+ },
865
+ "eval_recall": {
866
+ "recall": 0.8691138648585458
867
+ },
868
+ "eval_runtime": 10.1762,
869
+ "eval_samples_per_second": 50.117,
870
+ "eval_steps_per_second": 25.058,
871
+ "step": 1275
872
+ }
873
+ ],
874
+ "logging_steps": 10,
875
+ "max_steps": 7650,
876
+ "num_input_tokens_seen": 0,
877
+ "num_train_epochs": 30,
878
+ "save_steps": 500,
879
+ "total_flos": 7.904395468763136e+17,
880
+ "train_batch_size": 2,
881
+ "trial_name": null,
882
+ "trial_params": null
883
+ }
tmp-checkpoint-1275/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:418b85f35316fefe4bc777808f9dff13bb59b607bf93a8504dc8b5709313507f
3
+ size 4728
tmp-checkpoint-1530/config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "google/vit-base-patch16-224-in21k",
3
+ "architectures": [
4
+ "ViTForImageClassification"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.0,
7
+ "encoder_stride": 16,
8
+ "hidden_act": "gelu",
9
+ "hidden_dropout_prob": 0.0,
10
+ "hidden_size": 768,
11
+ "id2label": {
12
+ "0": "crack",
13
+ "1": "for review",
14
+ "2": "material deterioration",
15
+ "3": "repair",
16
+ "4": "water repair"
17
+ },
18
+ "image_size": 224,
19
+ "initializer_range": 0.02,
20
+ "intermediate_size": 3072,
21
+ "label2id": {
22
+ "crack": 0,
23
+ "for review": 1,
24
+ "material deterioration": 2,
25
+ "repair": 3,
26
+ "water repair": 4
27
+ },
28
+ "layer_norm_eps": 1e-12,
29
+ "model_type": "vit",
30
+ "num_attention_heads": 12,
31
+ "num_channels": 3,
32
+ "num_hidden_layers": 12,
33
+ "patch_size": 16,
34
+ "problem_type": "single_label_classification",
35
+ "qkv_bias": true,
36
+ "torch_dtype": "float32",
37
+ "transformers_version": "4.37.2"
38
+ }
tmp-checkpoint-1530/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ea04cfdf58ce0391ff8ec6cc6ea0fbc767cb6b148e76135d7b85483eee2b7176
3
+ size 343233204
tmp-checkpoint-1530/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2b065348e6d8d1eb3fd043e132b4f93d5efed58536fd4389139742f0c77b115
3
+ size 686587322
tmp-checkpoint-1530/preprocessor_config.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_normalize": true,
3
+ "do_rescale": true,
4
+ "do_resize": true,
5
+ "image_mean": [
6
+ 0.5,
7
+ 0.5,
8
+ 0.5
9
+ ],
10
+ "image_processor_type": "ViTImageProcessor",
11
+ "image_std": [
12
+ 0.5,
13
+ 0.5,
14
+ 0.5
15
+ ],
16
+ "resample": 2,
17
+ "rescale_factor": 0.00392156862745098,
18
+ "size": {
19
+ "height": 224,
20
+ "width": 224
21
+ }
22
+ }
tmp-checkpoint-1530/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e905055e8762148ecf744353c482d10d20c583368eb35e5a7ac3cd8f7944400
3
+ size 14244
tmp-checkpoint-1530/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:194c2f23ffdb54956b0790744f1f7ce59418a91fa51f52c53f80e670d249fe95
3
+ size 1064
tmp-checkpoint-1530/trainer_state.json ADDED
@@ -0,0 +1,1059 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.3946979343891144,
3
+ "best_model_checkpoint": "vit-base-patch16-224-in21k-bridgedefectVIT\\checkpoint-1020",
4
+ "epoch": 6.0,
5
+ "eval_steps": 500,
6
+ "global_step": 1530,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.04,
13
+ "learning_rate": 6.535947712418301e-07,
14
+ "loss": 1.6179,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.08,
19
+ "learning_rate": 1.3071895424836602e-06,
20
+ "loss": 1.6223,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.12,
25
+ "learning_rate": 1.96078431372549e-06,
26
+ "loss": 1.6263,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.16,
31
+ "learning_rate": 2.6143790849673204e-06,
32
+ "loss": 1.601,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.2,
37
+ "learning_rate": 3.2679738562091506e-06,
38
+ "loss": 1.6074,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.24,
43
+ "learning_rate": 3.92156862745098e-06,
44
+ "loss": 1.5955,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.27,
49
+ "learning_rate": 4.5751633986928105e-06,
50
+ "loss": 1.5698,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.31,
55
+ "learning_rate": 5.228758169934641e-06,
56
+ "loss": 1.5797,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.35,
61
+ "learning_rate": 5.882352941176471e-06,
62
+ "loss": 1.5531,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.39,
67
+ "learning_rate": 6.535947712418301e-06,
68
+ "loss": 1.5434,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.43,
73
+ "learning_rate": 7.1895424836601305e-06,
74
+ "loss": 1.5433,
75
+ "step": 110
76
+ },
77
+ {
78
+ "epoch": 0.47,
79
+ "learning_rate": 7.84313725490196e-06,
80
+ "loss": 1.5329,
81
+ "step": 120
82
+ },
83
+ {
84
+ "epoch": 0.51,
85
+ "learning_rate": 8.496732026143791e-06,
86
+ "loss": 1.4995,
87
+ "step": 130
88
+ },
89
+ {
90
+ "epoch": 0.55,
91
+ "learning_rate": 9.150326797385621e-06,
92
+ "loss": 1.4875,
93
+ "step": 140
94
+ },
95
+ {
96
+ "epoch": 0.59,
97
+ "learning_rate": 9.803921568627451e-06,
98
+ "loss": 1.4863,
99
+ "step": 150
100
+ },
101
+ {
102
+ "epoch": 0.63,
103
+ "learning_rate": 1.0457516339869281e-05,
104
+ "loss": 1.4209,
105
+ "step": 160
106
+ },
107
+ {
108
+ "epoch": 0.67,
109
+ "learning_rate": 1.1111111111111112e-05,
110
+ "loss": 1.4354,
111
+ "step": 170
112
+ },
113
+ {
114
+ "epoch": 0.71,
115
+ "learning_rate": 1.1764705882352942e-05,
116
+ "loss": 1.3572,
117
+ "step": 180
118
+ },
119
+ {
120
+ "epoch": 0.75,
121
+ "learning_rate": 1.2418300653594772e-05,
122
+ "loss": 1.3714,
123
+ "step": 190
124
+ },
125
+ {
126
+ "epoch": 0.78,
127
+ "learning_rate": 1.3071895424836602e-05,
128
+ "loss": 1.3406,
129
+ "step": 200
130
+ },
131
+ {
132
+ "epoch": 0.82,
133
+ "learning_rate": 1.3725490196078432e-05,
134
+ "loss": 1.2574,
135
+ "step": 210
136
+ },
137
+ {
138
+ "epoch": 0.86,
139
+ "learning_rate": 1.4379084967320261e-05,
140
+ "loss": 1.2965,
141
+ "step": 220
142
+ },
143
+ {
144
+ "epoch": 0.9,
145
+ "learning_rate": 1.5032679738562091e-05,
146
+ "loss": 1.2021,
147
+ "step": 230
148
+ },
149
+ {
150
+ "epoch": 0.94,
151
+ "learning_rate": 1.568627450980392e-05,
152
+ "loss": 1.2072,
153
+ "step": 240
154
+ },
155
+ {
156
+ "epoch": 0.98,
157
+ "learning_rate": 1.6339869281045753e-05,
158
+ "loss": 1.1457,
159
+ "step": 250
160
+ },
161
+ {
162
+ "epoch": 1.0,
163
+ "eval_accuracy": {
164
+ "accuracy": 0.6941176470588235
165
+ },
166
+ "eval_f1": {
167
+ "f1": 0.683792091280419
168
+ },
169
+ "eval_loss": 1.158827304840088,
170
+ "eval_precision": {
171
+ "precision": 0.7377494408375539
172
+ },
173
+ "eval_recall": {
174
+ "recall": 0.6941484490686618
175
+ },
176
+ "eval_runtime": 25.3262,
177
+ "eval_samples_per_second": 20.137,
178
+ "eval_steps_per_second": 10.069,
179
+ "step": 255
180
+ },
181
+ {
182
+ "epoch": 1.02,
183
+ "learning_rate": 1.6993464052287582e-05,
184
+ "loss": 1.0924,
185
+ "step": 260
186
+ },
187
+ {
188
+ "epoch": 1.06,
189
+ "learning_rate": 1.7647058823529414e-05,
190
+ "loss": 1.0007,
191
+ "step": 270
192
+ },
193
+ {
194
+ "epoch": 1.1,
195
+ "learning_rate": 1.8300653594771242e-05,
196
+ "loss": 1.0054,
197
+ "step": 280
198
+ },
199
+ {
200
+ "epoch": 1.14,
201
+ "learning_rate": 1.895424836601307e-05,
202
+ "loss": 0.9365,
203
+ "step": 290
204
+ },
205
+ {
206
+ "epoch": 1.18,
207
+ "learning_rate": 1.9607843137254903e-05,
208
+ "loss": 0.8932,
209
+ "step": 300
210
+ },
211
+ {
212
+ "epoch": 1.22,
213
+ "learning_rate": 2.0261437908496734e-05,
214
+ "loss": 0.916,
215
+ "step": 310
216
+ },
217
+ {
218
+ "epoch": 1.25,
219
+ "learning_rate": 2.0915032679738563e-05,
220
+ "loss": 0.8099,
221
+ "step": 320
222
+ },
223
+ {
224
+ "epoch": 1.29,
225
+ "learning_rate": 2.1568627450980395e-05,
226
+ "loss": 0.8211,
227
+ "step": 330
228
+ },
229
+ {
230
+ "epoch": 1.33,
231
+ "learning_rate": 2.2222222222222223e-05,
232
+ "loss": 0.6861,
233
+ "step": 340
234
+ },
235
+ {
236
+ "epoch": 1.37,
237
+ "learning_rate": 2.2875816993464052e-05,
238
+ "loss": 0.7379,
239
+ "step": 350
240
+ },
241
+ {
242
+ "epoch": 1.41,
243
+ "learning_rate": 2.3529411764705884e-05,
244
+ "loss": 0.7349,
245
+ "step": 360
246
+ },
247
+ {
248
+ "epoch": 1.45,
249
+ "learning_rate": 2.4183006535947712e-05,
250
+ "loss": 0.6907,
251
+ "step": 370
252
+ },
253
+ {
254
+ "epoch": 1.49,
255
+ "learning_rate": 2.4836601307189544e-05,
256
+ "loss": 0.7376,
257
+ "step": 380
258
+ },
259
+ {
260
+ "epoch": 1.53,
261
+ "learning_rate": 2.5490196078431373e-05,
262
+ "loss": 0.6827,
263
+ "step": 390
264
+ },
265
+ {
266
+ "epoch": 1.57,
267
+ "learning_rate": 2.6143790849673204e-05,
268
+ "loss": 0.6969,
269
+ "step": 400
270
+ },
271
+ {
272
+ "epoch": 1.61,
273
+ "learning_rate": 2.6797385620915033e-05,
274
+ "loss": 0.6247,
275
+ "step": 410
276
+ },
277
+ {
278
+ "epoch": 1.65,
279
+ "learning_rate": 2.7450980392156865e-05,
280
+ "loss": 0.5244,
281
+ "step": 420
282
+ },
283
+ {
284
+ "epoch": 1.69,
285
+ "learning_rate": 2.8104575163398693e-05,
286
+ "loss": 0.7068,
287
+ "step": 430
288
+ },
289
+ {
290
+ "epoch": 1.73,
291
+ "learning_rate": 2.8758169934640522e-05,
292
+ "loss": 0.5623,
293
+ "step": 440
294
+ },
295
+ {
296
+ "epoch": 1.76,
297
+ "learning_rate": 2.9411764705882354e-05,
298
+ "loss": 0.5931,
299
+ "step": 450
300
+ },
301
+ {
302
+ "epoch": 1.8,
303
+ "learning_rate": 3.0065359477124182e-05,
304
+ "loss": 0.6641,
305
+ "step": 460
306
+ },
307
+ {
308
+ "epoch": 1.84,
309
+ "learning_rate": 3.0718954248366014e-05,
310
+ "loss": 0.5438,
311
+ "step": 470
312
+ },
313
+ {
314
+ "epoch": 1.88,
315
+ "learning_rate": 3.137254901960784e-05,
316
+ "loss": 0.5134,
317
+ "step": 480
318
+ },
319
+ {
320
+ "epoch": 1.92,
321
+ "learning_rate": 3.202614379084967e-05,
322
+ "loss": 0.4682,
323
+ "step": 490
324
+ },
325
+ {
326
+ "epoch": 1.96,
327
+ "learning_rate": 3.2679738562091506e-05,
328
+ "loss": 0.4762,
329
+ "step": 500
330
+ },
331
+ {
332
+ "epoch": 2.0,
333
+ "learning_rate": 3.3333333333333335e-05,
334
+ "loss": 0.4558,
335
+ "step": 510
336
+ },
337
+ {
338
+ "epoch": 2.0,
339
+ "eval_accuracy": {
340
+ "accuracy": 0.8470588235294118
341
+ },
342
+ "eval_f1": {
343
+ "f1": 0.8510200857474889
344
+ },
345
+ "eval_loss": 0.498457670211792,
346
+ "eval_precision": {
347
+ "precision": 0.8766865984788879
348
+ },
349
+ "eval_recall": {
350
+ "recall": 0.8454231938274491
351
+ },
352
+ "eval_runtime": 10.0517,
353
+ "eval_samples_per_second": 50.738,
354
+ "eval_steps_per_second": 25.369,
355
+ "step": 510
356
+ },
357
+ {
358
+ "epoch": 2.04,
359
+ "learning_rate": 3.3986928104575163e-05,
360
+ "loss": 0.4716,
361
+ "step": 520
362
+ },
363
+ {
364
+ "epoch": 2.08,
365
+ "learning_rate": 3.464052287581699e-05,
366
+ "loss": 0.5375,
367
+ "step": 530
368
+ },
369
+ {
370
+ "epoch": 2.12,
371
+ "learning_rate": 3.529411764705883e-05,
372
+ "loss": 0.4725,
373
+ "step": 540
374
+ },
375
+ {
376
+ "epoch": 2.16,
377
+ "learning_rate": 3.5947712418300656e-05,
378
+ "loss": 0.4804,
379
+ "step": 550
380
+ },
381
+ {
382
+ "epoch": 2.2,
383
+ "learning_rate": 3.6601307189542484e-05,
384
+ "loss": 0.3504,
385
+ "step": 560
386
+ },
387
+ {
388
+ "epoch": 2.24,
389
+ "learning_rate": 3.725490196078432e-05,
390
+ "loss": 0.4049,
391
+ "step": 570
392
+ },
393
+ {
394
+ "epoch": 2.27,
395
+ "learning_rate": 3.790849673202614e-05,
396
+ "loss": 0.4309,
397
+ "step": 580
398
+ },
399
+ {
400
+ "epoch": 2.31,
401
+ "learning_rate": 3.8562091503267977e-05,
402
+ "loss": 0.4938,
403
+ "step": 590
404
+ },
405
+ {
406
+ "epoch": 2.35,
407
+ "learning_rate": 3.9215686274509805e-05,
408
+ "loss": 0.5427,
409
+ "step": 600
410
+ },
411
+ {
412
+ "epoch": 2.39,
413
+ "learning_rate": 3.986928104575164e-05,
414
+ "loss": 0.4719,
415
+ "step": 610
416
+ },
417
+ {
418
+ "epoch": 2.43,
419
+ "learning_rate": 4.052287581699347e-05,
420
+ "loss": 0.4627,
421
+ "step": 620
422
+ },
423
+ {
424
+ "epoch": 2.47,
425
+ "learning_rate": 4.11764705882353e-05,
426
+ "loss": 0.5277,
427
+ "step": 630
428
+ },
429
+ {
430
+ "epoch": 2.51,
431
+ "learning_rate": 4.1830065359477126e-05,
432
+ "loss": 0.3109,
433
+ "step": 640
434
+ },
435
+ {
436
+ "epoch": 2.55,
437
+ "learning_rate": 4.2483660130718954e-05,
438
+ "loss": 0.3799,
439
+ "step": 650
440
+ },
441
+ {
442
+ "epoch": 2.59,
443
+ "learning_rate": 4.313725490196079e-05,
444
+ "loss": 0.3515,
445
+ "step": 660
446
+ },
447
+ {
448
+ "epoch": 2.63,
449
+ "learning_rate": 4.379084967320262e-05,
450
+ "loss": 0.3722,
451
+ "step": 670
452
+ },
453
+ {
454
+ "epoch": 2.67,
455
+ "learning_rate": 4.4444444444444447e-05,
456
+ "loss": 0.5337,
457
+ "step": 680
458
+ },
459
+ {
460
+ "epoch": 2.71,
461
+ "learning_rate": 4.5098039215686275e-05,
462
+ "loss": 0.3263,
463
+ "step": 690
464
+ },
465
+ {
466
+ "epoch": 2.75,
467
+ "learning_rate": 4.5751633986928104e-05,
468
+ "loss": 0.3703,
469
+ "step": 700
470
+ },
471
+ {
472
+ "epoch": 2.78,
473
+ "learning_rate": 4.640522875816994e-05,
474
+ "loss": 0.4498,
475
+ "step": 710
476
+ },
477
+ {
478
+ "epoch": 2.82,
479
+ "learning_rate": 4.705882352941177e-05,
480
+ "loss": 0.4109,
481
+ "step": 720
482
+ },
483
+ {
484
+ "epoch": 2.86,
485
+ "learning_rate": 4.77124183006536e-05,
486
+ "loss": 0.2953,
487
+ "step": 730
488
+ },
489
+ {
490
+ "epoch": 2.9,
491
+ "learning_rate": 4.8366013071895424e-05,
492
+ "loss": 0.4033,
493
+ "step": 740
494
+ },
495
+ {
496
+ "epoch": 2.94,
497
+ "learning_rate": 4.901960784313725e-05,
498
+ "loss": 0.4186,
499
+ "step": 750
500
+ },
501
+ {
502
+ "epoch": 2.98,
503
+ "learning_rate": 4.967320261437909e-05,
504
+ "loss": 0.3368,
505
+ "step": 760
506
+ },
507
+ {
508
+ "epoch": 3.0,
509
+ "eval_accuracy": {
510
+ "accuracy": 0.8705882352941177
511
+ },
512
+ "eval_f1": {
513
+ "f1": 0.8713991689505157
514
+ },
515
+ "eval_loss": 0.44290485978126526,
516
+ "eval_precision": {
517
+ "precision": 0.8884985994397759
518
+ },
519
+ "eval_recall": {
520
+ "recall": 0.8711626139817629
521
+ },
522
+ "eval_runtime": 9.9506,
523
+ "eval_samples_per_second": 51.253,
524
+ "eval_steps_per_second": 25.626,
525
+ "step": 765
526
+ },
527
+ {
528
+ "epoch": 3.02,
529
+ "learning_rate": 4.9963689179375456e-05,
530
+ "loss": 0.3347,
531
+ "step": 770
532
+ },
533
+ {
534
+ "epoch": 3.06,
535
+ "learning_rate": 4.9891067538126364e-05,
536
+ "loss": 0.3339,
537
+ "step": 780
538
+ },
539
+ {
540
+ "epoch": 3.1,
541
+ "learning_rate": 4.981844589687727e-05,
542
+ "loss": 0.5573,
543
+ "step": 790
544
+ },
545
+ {
546
+ "epoch": 3.14,
547
+ "learning_rate": 4.974582425562818e-05,
548
+ "loss": 0.5463,
549
+ "step": 800
550
+ },
551
+ {
552
+ "epoch": 3.18,
553
+ "learning_rate": 4.967320261437909e-05,
554
+ "loss": 0.3453,
555
+ "step": 810
556
+ },
557
+ {
558
+ "epoch": 3.22,
559
+ "learning_rate": 4.9600580973129996e-05,
560
+ "loss": 0.3814,
561
+ "step": 820
562
+ },
563
+ {
564
+ "epoch": 3.25,
565
+ "learning_rate": 4.9527959331880904e-05,
566
+ "loss": 0.4434,
567
+ "step": 830
568
+ },
569
+ {
570
+ "epoch": 3.29,
571
+ "learning_rate": 4.945533769063181e-05,
572
+ "loss": 0.3675,
573
+ "step": 840
574
+ },
575
+ {
576
+ "epoch": 3.33,
577
+ "learning_rate": 4.938271604938271e-05,
578
+ "loss": 0.5386,
579
+ "step": 850
580
+ },
581
+ {
582
+ "epoch": 3.37,
583
+ "learning_rate": 4.931009440813363e-05,
584
+ "loss": 0.2692,
585
+ "step": 860
586
+ },
587
+ {
588
+ "epoch": 3.41,
589
+ "learning_rate": 4.9237472766884536e-05,
590
+ "loss": 0.4815,
591
+ "step": 870
592
+ },
593
+ {
594
+ "epoch": 3.45,
595
+ "learning_rate": 4.9164851125635444e-05,
596
+ "loss": 0.4178,
597
+ "step": 880
598
+ },
599
+ {
600
+ "epoch": 3.49,
601
+ "learning_rate": 4.909222948438635e-05,
602
+ "loss": 0.3618,
603
+ "step": 890
604
+ },
605
+ {
606
+ "epoch": 3.53,
607
+ "learning_rate": 4.901960784313725e-05,
608
+ "loss": 0.6841,
609
+ "step": 900
610
+ },
611
+ {
612
+ "epoch": 3.57,
613
+ "learning_rate": 4.894698620188817e-05,
614
+ "loss": 0.471,
615
+ "step": 910
616
+ },
617
+ {
618
+ "epoch": 3.61,
619
+ "learning_rate": 4.8874364560639076e-05,
620
+ "loss": 0.313,
621
+ "step": 920
622
+ },
623
+ {
624
+ "epoch": 3.65,
625
+ "learning_rate": 4.8801742919389983e-05,
626
+ "loss": 0.2437,
627
+ "step": 930
628
+ },
629
+ {
630
+ "epoch": 3.69,
631
+ "learning_rate": 4.8729121278140885e-05,
632
+ "loss": 0.4748,
633
+ "step": 940
634
+ },
635
+ {
636
+ "epoch": 3.73,
637
+ "learning_rate": 4.865649963689179e-05,
638
+ "loss": 0.5519,
639
+ "step": 950
640
+ },
641
+ {
642
+ "epoch": 3.76,
643
+ "learning_rate": 4.85838779956427e-05,
644
+ "loss": 0.3051,
645
+ "step": 960
646
+ },
647
+ {
648
+ "epoch": 3.8,
649
+ "learning_rate": 4.8511256354393615e-05,
650
+ "loss": 0.3847,
651
+ "step": 970
652
+ },
653
+ {
654
+ "epoch": 3.84,
655
+ "learning_rate": 4.843863471314452e-05,
656
+ "loss": 0.2454,
657
+ "step": 980
658
+ },
659
+ {
660
+ "epoch": 3.88,
661
+ "learning_rate": 4.8366013071895424e-05,
662
+ "loss": 0.379,
663
+ "step": 990
664
+ },
665
+ {
666
+ "epoch": 3.92,
667
+ "learning_rate": 4.829339143064633e-05,
668
+ "loss": 0.4485,
669
+ "step": 1000
670
+ },
671
+ {
672
+ "epoch": 3.96,
673
+ "learning_rate": 4.822076978939724e-05,
674
+ "loss": 0.2484,
675
+ "step": 1010
676
+ },
677
+ {
678
+ "epoch": 4.0,
679
+ "learning_rate": 4.814814814814815e-05,
680
+ "loss": 0.4402,
681
+ "step": 1020
682
+ },
683
+ {
684
+ "epoch": 4.0,
685
+ "eval_accuracy": {
686
+ "accuracy": 0.8784313725490196
687
+ },
688
+ "eval_f1": {
689
+ "f1": 0.880471094803992
690
+ },
691
+ "eval_loss": 0.3946979343891144,
692
+ "eval_precision": {
693
+ "precision": 0.8952890487430221
694
+ },
695
+ "eval_recall": {
696
+ "recall": 0.8776488582339645
697
+ },
698
+ "eval_runtime": 10.1771,
699
+ "eval_samples_per_second": 50.113,
700
+ "eval_steps_per_second": 25.056,
701
+ "step": 1020
702
+ },
703
+ {
704
+ "epoch": 4.04,
705
+ "learning_rate": 4.807552650689906e-05,
706
+ "loss": 0.3494,
707
+ "step": 1030
708
+ },
709
+ {
710
+ "epoch": 4.08,
711
+ "learning_rate": 4.8002904865649964e-05,
712
+ "loss": 0.3327,
713
+ "step": 1040
714
+ },
715
+ {
716
+ "epoch": 4.12,
717
+ "learning_rate": 4.793028322440087e-05,
718
+ "loss": 0.2898,
719
+ "step": 1050
720
+ },
721
+ {
722
+ "epoch": 4.16,
723
+ "learning_rate": 4.785766158315178e-05,
724
+ "loss": 0.3965,
725
+ "step": 1060
726
+ },
727
+ {
728
+ "epoch": 4.2,
729
+ "learning_rate": 4.778503994190269e-05,
730
+ "loss": 0.3607,
731
+ "step": 1070
732
+ },
733
+ {
734
+ "epoch": 4.24,
735
+ "learning_rate": 4.77124183006536e-05,
736
+ "loss": 0.3714,
737
+ "step": 1080
738
+ },
739
+ {
740
+ "epoch": 4.27,
741
+ "learning_rate": 4.7639796659404504e-05,
742
+ "loss": 0.1796,
743
+ "step": 1090
744
+ },
745
+ {
746
+ "epoch": 4.31,
747
+ "learning_rate": 4.756717501815541e-05,
748
+ "loss": 0.1874,
749
+ "step": 1100
750
+ },
751
+ {
752
+ "epoch": 4.35,
753
+ "learning_rate": 4.749455337690632e-05,
754
+ "loss": 0.3462,
755
+ "step": 1110
756
+ },
757
+ {
758
+ "epoch": 4.39,
759
+ "learning_rate": 4.742193173565723e-05,
760
+ "loss": 0.2082,
761
+ "step": 1120
762
+ },
763
+ {
764
+ "epoch": 4.43,
765
+ "learning_rate": 4.7349310094408136e-05,
766
+ "loss": 0.4531,
767
+ "step": 1130
768
+ },
769
+ {
770
+ "epoch": 4.47,
771
+ "learning_rate": 4.7276688453159044e-05,
772
+ "loss": 0.3951,
773
+ "step": 1140
774
+ },
775
+ {
776
+ "epoch": 4.51,
777
+ "learning_rate": 4.720406681190995e-05,
778
+ "loss": 0.2337,
779
+ "step": 1150
780
+ },
781
+ {
782
+ "epoch": 4.55,
783
+ "learning_rate": 4.713144517066086e-05,
784
+ "loss": 0.2118,
785
+ "step": 1160
786
+ },
787
+ {
788
+ "epoch": 4.59,
789
+ "learning_rate": 4.705882352941177e-05,
790
+ "loss": 0.373,
791
+ "step": 1170
792
+ },
793
+ {
794
+ "epoch": 4.63,
795
+ "learning_rate": 4.6986201888162675e-05,
796
+ "loss": 0.1484,
797
+ "step": 1180
798
+ },
799
+ {
800
+ "epoch": 4.67,
801
+ "learning_rate": 4.691358024691358e-05,
802
+ "loss": 0.2568,
803
+ "step": 1190
804
+ },
805
+ {
806
+ "epoch": 4.71,
807
+ "learning_rate": 4.684095860566449e-05,
808
+ "loss": 0.3067,
809
+ "step": 1200
810
+ },
811
+ {
812
+ "epoch": 4.75,
813
+ "learning_rate": 4.67683369644154e-05,
814
+ "loss": 0.2007,
815
+ "step": 1210
816
+ },
817
+ {
818
+ "epoch": 4.78,
819
+ "learning_rate": 4.669571532316631e-05,
820
+ "loss": 0.2493,
821
+ "step": 1220
822
+ },
823
+ {
824
+ "epoch": 4.82,
825
+ "learning_rate": 4.6623093681917215e-05,
826
+ "loss": 0.2717,
827
+ "step": 1230
828
+ },
829
+ {
830
+ "epoch": 4.86,
831
+ "learning_rate": 4.6550472040668116e-05,
832
+ "loss": 0.4134,
833
+ "step": 1240
834
+ },
835
+ {
836
+ "epoch": 4.9,
837
+ "learning_rate": 4.647785039941903e-05,
838
+ "loss": 0.2417,
839
+ "step": 1250
840
+ },
841
+ {
842
+ "epoch": 4.94,
843
+ "learning_rate": 4.640522875816994e-05,
844
+ "loss": 0.2573,
845
+ "step": 1260
846
+ },
847
+ {
848
+ "epoch": 4.98,
849
+ "learning_rate": 4.633260711692085e-05,
850
+ "loss": 0.3937,
851
+ "step": 1270
852
+ },
853
+ {
854
+ "epoch": 5.0,
855
+ "eval_accuracy": {
856
+ "accuracy": 0.8725490196078431
857
+ },
858
+ "eval_f1": {
859
+ "f1": 0.8703780158419334
860
+ },
861
+ "eval_loss": 0.4377364218235016,
862
+ "eval_precision": {
863
+ "precision": 0.8900760323128502
864
+ },
865
+ "eval_recall": {
866
+ "recall": 0.8691138648585458
867
+ },
868
+ "eval_runtime": 10.1762,
869
+ "eval_samples_per_second": 50.117,
870
+ "eval_steps_per_second": 25.058,
871
+ "step": 1275
872
+ },
873
+ {
874
+ "epoch": 5.02,
875
+ "learning_rate": 4.6259985475671755e-05,
876
+ "loss": 0.165,
877
+ "step": 1280
878
+ },
879
+ {
880
+ "epoch": 5.06,
881
+ "learning_rate": 4.6187363834422656e-05,
882
+ "loss": 0.1452,
883
+ "step": 1290
884
+ },
885
+ {
886
+ "epoch": 5.1,
887
+ "learning_rate": 4.6114742193173564e-05,
888
+ "loss": 0.1548,
889
+ "step": 1300
890
+ },
891
+ {
892
+ "epoch": 5.14,
893
+ "learning_rate": 4.604212055192448e-05,
894
+ "loss": 0.4443,
895
+ "step": 1310
896
+ },
897
+ {
898
+ "epoch": 5.18,
899
+ "learning_rate": 4.5969498910675387e-05,
900
+ "loss": 0.4322,
901
+ "step": 1320
902
+ },
903
+ {
904
+ "epoch": 5.22,
905
+ "learning_rate": 4.5896877269426294e-05,
906
+ "loss": 0.2355,
907
+ "step": 1330
908
+ },
909
+ {
910
+ "epoch": 5.25,
911
+ "learning_rate": 4.5824255628177196e-05,
912
+ "loss": 0.1949,
913
+ "step": 1340
914
+ },
915
+ {
916
+ "epoch": 5.29,
917
+ "learning_rate": 4.5751633986928104e-05,
918
+ "loss": 0.1959,
919
+ "step": 1350
920
+ },
921
+ {
922
+ "epoch": 5.33,
923
+ "learning_rate": 4.567901234567901e-05,
924
+ "loss": 0.3841,
925
+ "step": 1360
926
+ },
927
+ {
928
+ "epoch": 5.37,
929
+ "learning_rate": 4.5606390704429926e-05,
930
+ "loss": 0.1401,
931
+ "step": 1370
932
+ },
933
+ {
934
+ "epoch": 5.41,
935
+ "learning_rate": 4.5533769063180834e-05,
936
+ "loss": 0.1148,
937
+ "step": 1380
938
+ },
939
+ {
940
+ "epoch": 5.45,
941
+ "learning_rate": 4.5461147421931735e-05,
942
+ "loss": 0.234,
943
+ "step": 1390
944
+ },
945
+ {
946
+ "epoch": 5.49,
947
+ "learning_rate": 4.538852578068264e-05,
948
+ "loss": 0.3899,
949
+ "step": 1400
950
+ },
951
+ {
952
+ "epoch": 5.53,
953
+ "learning_rate": 4.531590413943355e-05,
954
+ "loss": 0.4873,
955
+ "step": 1410
956
+ },
957
+ {
958
+ "epoch": 5.57,
959
+ "learning_rate": 4.5243282498184466e-05,
960
+ "loss": 0.2483,
961
+ "step": 1420
962
+ },
963
+ {
964
+ "epoch": 5.61,
965
+ "learning_rate": 4.5170660856935374e-05,
966
+ "loss": 0.2654,
967
+ "step": 1430
968
+ },
969
+ {
970
+ "epoch": 5.65,
971
+ "learning_rate": 4.5098039215686275e-05,
972
+ "loss": 0.2035,
973
+ "step": 1440
974
+ },
975
+ {
976
+ "epoch": 5.69,
977
+ "learning_rate": 4.502541757443718e-05,
978
+ "loss": 0.3744,
979
+ "step": 1450
980
+ },
981
+ {
982
+ "epoch": 5.73,
983
+ "learning_rate": 4.495279593318809e-05,
984
+ "loss": 0.2139,
985
+ "step": 1460
986
+ },
987
+ {
988
+ "epoch": 5.76,
989
+ "learning_rate": 4.4880174291939e-05,
990
+ "loss": 0.2212,
991
+ "step": 1470
992
+ },
993
+ {
994
+ "epoch": 5.8,
995
+ "learning_rate": 4.480755265068991e-05,
996
+ "loss": 0.376,
997
+ "step": 1480
998
+ },
999
+ {
1000
+ "epoch": 5.84,
1001
+ "learning_rate": 4.4734931009440815e-05,
1002
+ "loss": 0.2399,
1003
+ "step": 1490
1004
+ },
1005
+ {
1006
+ "epoch": 5.88,
1007
+ "learning_rate": 4.466230936819172e-05,
1008
+ "loss": 0.3534,
1009
+ "step": 1500
1010
+ },
1011
+ {
1012
+ "epoch": 5.92,
1013
+ "learning_rate": 4.458968772694263e-05,
1014
+ "loss": 0.3667,
1015
+ "step": 1510
1016
+ },
1017
+ {
1018
+ "epoch": 5.96,
1019
+ "learning_rate": 4.451706608569354e-05,
1020
+ "loss": 0.496,
1021
+ "step": 1520
1022
+ },
1023
+ {
1024
+ "epoch": 6.0,
1025
+ "learning_rate": 4.4444444444444447e-05,
1026
+ "loss": 0.4042,
1027
+ "step": 1530
1028
+ },
1029
+ {
1030
+ "epoch": 6.0,
1031
+ "eval_accuracy": {
1032
+ "accuracy": 0.8862745098039215
1033
+ },
1034
+ "eval_f1": {
1035
+ "f1": 0.8859196308355288
1036
+ },
1037
+ "eval_loss": 0.39891302585601807,
1038
+ "eval_precision": {
1039
+ "precision": 0.8972093144987638
1040
+ },
1041
+ "eval_recall": {
1042
+ "recall": 0.8834955381497934
1043
+ },
1044
+ "eval_runtime": 10.1174,
1045
+ "eval_samples_per_second": 50.408,
1046
+ "eval_steps_per_second": 25.204,
1047
+ "step": 1530
1048
+ }
1049
+ ],
1050
+ "logging_steps": 10,
1051
+ "max_steps": 7650,
1052
+ "num_input_tokens_seen": 0,
1053
+ "num_train_epochs": 30,
1054
+ "save_steps": 500,
1055
+ "total_flos": 9.485274562515763e+17,
1056
+ "train_batch_size": 2,
1057
+ "trial_name": null,
1058
+ "trial_params": null
1059
+ }
tmp-checkpoint-1530/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:418b85f35316fefe4bc777808f9dff13bb59b607bf93a8504dc8b5709313507f
3
+ size 4728