lrzjason commited on
Commit
bbc7266
1 Parent(s): 5f9fbe7

Upload 8 files

Browse files
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.9536585365853658,
4
+ "eval_loss": 0.1909320205450058,
5
+ "eval_runtime": 19.2055,
6
+ "eval_samples_per_second": 21.348,
7
+ "eval_steps_per_second": 2.708,
8
+ "train_loss": 0.16907700259904354,
9
+ "train_runtime": 2197.7206,
10
+ "train_samples_per_second": 15.016,
11
+ "train_steps_per_second": 0.942
12
+ }
config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "fullbody",
13
+ "1": "head_only",
14
+ "2": "knee_level",
15
+ "3": "upperbody"
16
+ },
17
+ "image_size": 224,
18
+ "initializer_range": 0.02,
19
+ "intermediate_size": 3072,
20
+ "label2id": {
21
+ "fullbody": "0",
22
+ "head_only": "1",
23
+ "knee_level": "2",
24
+ "upperbody": "3"
25
+ },
26
+ "layer_norm_eps": 1e-12,
27
+ "model_type": "vit",
28
+ "num_attention_heads": 12,
29
+ "num_channels": 3,
30
+ "num_hidden_layers": 12,
31
+ "patch_size": 16,
32
+ "problem_type": "single_label_classification",
33
+ "qkv_bias": true,
34
+ "torch_dtype": "float32",
35
+ "transformers_version": "4.36.2"
36
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.9536585365853658,
4
+ "eval_loss": 0.1909320205450058,
5
+ "eval_runtime": 19.2055,
6
+ "eval_samples_per_second": 21.348,
7
+ "eval_steps_per_second": 2.708
8
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:622424d7e1e79bda11abb27fa9834fd865901d026b5582eec59a19b5ce58087b
3
+ size 343230128
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
+ }
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "train_loss": 0.16907700259904354,
4
+ "train_runtime": 2197.7206,
5
+ "train_samples_per_second": 15.016,
6
+ "train_steps_per_second": 0.942
7
+ }
trainer_state.json ADDED
@@ -0,0 +1,1452 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.1909320205450058,
3
+ "best_model_checkpoint": "./vit-base-fullbody-classifier-6\\checkpoint-1100",
4
+ "epoch": 10.0,
5
+ "eval_steps": 100,
6
+ "global_step": 2070,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05,
13
+ "learning_rate": 1.0000000000000002e-06,
14
+ "loss": 1.3746,
15
+ "step": 10
16
+ },
17
+ {
18
+ "epoch": 0.1,
19
+ "learning_rate": 2.0000000000000003e-06,
20
+ "loss": 1.3829,
21
+ "step": 20
22
+ },
23
+ {
24
+ "epoch": 0.14,
25
+ "learning_rate": 3e-06,
26
+ "loss": 1.3757,
27
+ "step": 30
28
+ },
29
+ {
30
+ "epoch": 0.19,
31
+ "learning_rate": 4.000000000000001e-06,
32
+ "loss": 1.3656,
33
+ "step": 40
34
+ },
35
+ {
36
+ "epoch": 0.24,
37
+ "learning_rate": 5e-06,
38
+ "loss": 1.3448,
39
+ "step": 50
40
+ },
41
+ {
42
+ "epoch": 0.29,
43
+ "learning_rate": 6e-06,
44
+ "loss": 1.3373,
45
+ "step": 60
46
+ },
47
+ {
48
+ "epoch": 0.34,
49
+ "learning_rate": 7.000000000000001e-06,
50
+ "loss": 1.3232,
51
+ "step": 70
52
+ },
53
+ {
54
+ "epoch": 0.39,
55
+ "learning_rate": 8.000000000000001e-06,
56
+ "loss": 1.2872,
57
+ "step": 80
58
+ },
59
+ {
60
+ "epoch": 0.43,
61
+ "learning_rate": 9e-06,
62
+ "loss": 1.2638,
63
+ "step": 90
64
+ },
65
+ {
66
+ "epoch": 0.48,
67
+ "learning_rate": 1e-05,
68
+ "loss": 1.25,
69
+ "step": 100
70
+ },
71
+ {
72
+ "epoch": 0.48,
73
+ "eval_accuracy": 0.5926829268292683,
74
+ "eval_loss": 1.2367020845413208,
75
+ "eval_runtime": 20.8647,
76
+ "eval_samples_per_second": 19.65,
77
+ "eval_steps_per_second": 2.492,
78
+ "step": 100
79
+ },
80
+ {
81
+ "epoch": 0.53,
82
+ "learning_rate": 1.1000000000000001e-05,
83
+ "loss": 1.1748,
84
+ "step": 110
85
+ },
86
+ {
87
+ "epoch": 0.58,
88
+ "learning_rate": 1.2e-05,
89
+ "loss": 1.1491,
90
+ "step": 120
91
+ },
92
+ {
93
+ "epoch": 0.63,
94
+ "learning_rate": 1.3000000000000001e-05,
95
+ "loss": 1.0737,
96
+ "step": 130
97
+ },
98
+ {
99
+ "epoch": 0.68,
100
+ "learning_rate": 1.4000000000000001e-05,
101
+ "loss": 1.0358,
102
+ "step": 140
103
+ },
104
+ {
105
+ "epoch": 0.72,
106
+ "learning_rate": 1.5e-05,
107
+ "loss": 0.9602,
108
+ "step": 150
109
+ },
110
+ {
111
+ "epoch": 0.77,
112
+ "learning_rate": 1.6000000000000003e-05,
113
+ "loss": 0.9154,
114
+ "step": 160
115
+ },
116
+ {
117
+ "epoch": 0.82,
118
+ "learning_rate": 1.7000000000000003e-05,
119
+ "loss": 0.8374,
120
+ "step": 170
121
+ },
122
+ {
123
+ "epoch": 0.87,
124
+ "learning_rate": 1.8e-05,
125
+ "loss": 0.7678,
126
+ "step": 180
127
+ },
128
+ {
129
+ "epoch": 0.92,
130
+ "learning_rate": 1.9e-05,
131
+ "loss": 0.7105,
132
+ "step": 190
133
+ },
134
+ {
135
+ "epoch": 0.97,
136
+ "learning_rate": 2e-05,
137
+ "loss": 0.6594,
138
+ "step": 200
139
+ },
140
+ {
141
+ "epoch": 0.97,
142
+ "eval_accuracy": 0.8731707317073171,
143
+ "eval_loss": 0.64021235704422,
144
+ "eval_runtime": 21.6651,
145
+ "eval_samples_per_second": 18.924,
146
+ "eval_steps_per_second": 2.4,
147
+ "step": 200
148
+ },
149
+ {
150
+ "epoch": 1.01,
151
+ "learning_rate": 2.1e-05,
152
+ "loss": 0.6217,
153
+ "step": 210
154
+ },
155
+ {
156
+ "epoch": 1.06,
157
+ "learning_rate": 2.2000000000000003e-05,
158
+ "loss": 0.6355,
159
+ "step": 220
160
+ },
161
+ {
162
+ "epoch": 1.11,
163
+ "learning_rate": 2.3000000000000003e-05,
164
+ "loss": 0.5103,
165
+ "step": 230
166
+ },
167
+ {
168
+ "epoch": 1.16,
169
+ "learning_rate": 2.4e-05,
170
+ "loss": 0.4564,
171
+ "step": 240
172
+ },
173
+ {
174
+ "epoch": 1.21,
175
+ "learning_rate": 2.5e-05,
176
+ "loss": 0.3947,
177
+ "step": 250
178
+ },
179
+ {
180
+ "epoch": 1.26,
181
+ "learning_rate": 2.6000000000000002e-05,
182
+ "loss": 0.3605,
183
+ "step": 260
184
+ },
185
+ {
186
+ "epoch": 1.3,
187
+ "learning_rate": 2.7000000000000002e-05,
188
+ "loss": 0.3896,
189
+ "step": 270
190
+ },
191
+ {
192
+ "epoch": 1.35,
193
+ "learning_rate": 2.8000000000000003e-05,
194
+ "loss": 0.3193,
195
+ "step": 280
196
+ },
197
+ {
198
+ "epoch": 1.4,
199
+ "learning_rate": 2.9e-05,
200
+ "loss": 0.334,
201
+ "step": 290
202
+ },
203
+ {
204
+ "epoch": 1.45,
205
+ "learning_rate": 3e-05,
206
+ "loss": 0.2512,
207
+ "step": 300
208
+ },
209
+ {
210
+ "epoch": 1.45,
211
+ "eval_accuracy": 0.948780487804878,
212
+ "eval_loss": 0.28558439016342163,
213
+ "eval_runtime": 22.0969,
214
+ "eval_samples_per_second": 18.555,
215
+ "eval_steps_per_second": 2.353,
216
+ "step": 300
217
+ },
218
+ {
219
+ "epoch": 1.5,
220
+ "learning_rate": 3.1e-05,
221
+ "loss": 0.257,
222
+ "step": 310
223
+ },
224
+ {
225
+ "epoch": 1.55,
226
+ "learning_rate": 3.2000000000000005e-05,
227
+ "loss": 0.2081,
228
+ "step": 320
229
+ },
230
+ {
231
+ "epoch": 1.59,
232
+ "learning_rate": 3.3e-05,
233
+ "loss": 0.2643,
234
+ "step": 330
235
+ },
236
+ {
237
+ "epoch": 1.64,
238
+ "learning_rate": 3.4000000000000007e-05,
239
+ "loss": 0.271,
240
+ "step": 340
241
+ },
242
+ {
243
+ "epoch": 1.69,
244
+ "learning_rate": 3.5e-05,
245
+ "loss": 0.2353,
246
+ "step": 350
247
+ },
248
+ {
249
+ "epoch": 1.74,
250
+ "learning_rate": 3.6e-05,
251
+ "loss": 0.316,
252
+ "step": 360
253
+ },
254
+ {
255
+ "epoch": 1.79,
256
+ "learning_rate": 3.7e-05,
257
+ "loss": 0.1995,
258
+ "step": 370
259
+ },
260
+ {
261
+ "epoch": 1.84,
262
+ "learning_rate": 3.8e-05,
263
+ "loss": 0.3356,
264
+ "step": 380
265
+ },
266
+ {
267
+ "epoch": 1.88,
268
+ "learning_rate": 3.9000000000000006e-05,
269
+ "loss": 0.2513,
270
+ "step": 390
271
+ },
272
+ {
273
+ "epoch": 1.93,
274
+ "learning_rate": 4e-05,
275
+ "loss": 0.2175,
276
+ "step": 400
277
+ },
278
+ {
279
+ "epoch": 1.93,
280
+ "eval_accuracy": 0.8975609756097561,
281
+ "eval_loss": 0.3272535502910614,
282
+ "eval_runtime": 21.8078,
283
+ "eval_samples_per_second": 18.801,
284
+ "eval_steps_per_second": 2.384,
285
+ "step": 400
286
+ },
287
+ {
288
+ "epoch": 1.98,
289
+ "learning_rate": 4.1e-05,
290
+ "loss": 0.281,
291
+ "step": 410
292
+ },
293
+ {
294
+ "epoch": 2.03,
295
+ "learning_rate": 4.2e-05,
296
+ "loss": 0.2321,
297
+ "step": 420
298
+ },
299
+ {
300
+ "epoch": 2.08,
301
+ "learning_rate": 4.29e-05,
302
+ "loss": 0.1141,
303
+ "step": 430
304
+ },
305
+ {
306
+ "epoch": 2.13,
307
+ "learning_rate": 4.39e-05,
308
+ "loss": 0.1023,
309
+ "step": 440
310
+ },
311
+ {
312
+ "epoch": 2.17,
313
+ "learning_rate": 4.49e-05,
314
+ "loss": 0.1081,
315
+ "step": 450
316
+ },
317
+ {
318
+ "epoch": 2.22,
319
+ "learning_rate": 4.5900000000000004e-05,
320
+ "loss": 0.0969,
321
+ "step": 460
322
+ },
323
+ {
324
+ "epoch": 2.27,
325
+ "learning_rate": 4.69e-05,
326
+ "loss": 0.0822,
327
+ "step": 470
328
+ },
329
+ {
330
+ "epoch": 2.32,
331
+ "learning_rate": 4.79e-05,
332
+ "loss": 0.1229,
333
+ "step": 480
334
+ },
335
+ {
336
+ "epoch": 2.37,
337
+ "learning_rate": 4.89e-05,
338
+ "loss": 0.1556,
339
+ "step": 490
340
+ },
341
+ {
342
+ "epoch": 2.42,
343
+ "learning_rate": 4.99e-05,
344
+ "loss": 0.1313,
345
+ "step": 500
346
+ },
347
+ {
348
+ "epoch": 2.42,
349
+ "eval_accuracy": 0.9317073170731708,
350
+ "eval_loss": 0.229569673538208,
351
+ "eval_runtime": 20.4104,
352
+ "eval_samples_per_second": 20.088,
353
+ "eval_steps_per_second": 2.548,
354
+ "step": 500
355
+ },
356
+ {
357
+ "epoch": 2.46,
358
+ "learning_rate": 4.9713375796178345e-05,
359
+ "loss": 0.1338,
360
+ "step": 510
361
+ },
362
+ {
363
+ "epoch": 2.51,
364
+ "learning_rate": 4.939490445859873e-05,
365
+ "loss": 0.1007,
366
+ "step": 520
367
+ },
368
+ {
369
+ "epoch": 2.56,
370
+ "learning_rate": 4.907643312101911e-05,
371
+ "loss": 0.1014,
372
+ "step": 530
373
+ },
374
+ {
375
+ "epoch": 2.61,
376
+ "learning_rate": 4.875796178343949e-05,
377
+ "loss": 0.1331,
378
+ "step": 540
379
+ },
380
+ {
381
+ "epoch": 2.66,
382
+ "learning_rate": 4.8439490445859877e-05,
383
+ "loss": 0.1637,
384
+ "step": 550
385
+ },
386
+ {
387
+ "epoch": 2.71,
388
+ "learning_rate": 4.8121019108280254e-05,
389
+ "loss": 0.1037,
390
+ "step": 560
391
+ },
392
+ {
393
+ "epoch": 2.75,
394
+ "learning_rate": 4.780254777070064e-05,
395
+ "loss": 0.1222,
396
+ "step": 570
397
+ },
398
+ {
399
+ "epoch": 2.8,
400
+ "learning_rate": 4.7484076433121024e-05,
401
+ "loss": 0.0777,
402
+ "step": 580
403
+ },
404
+ {
405
+ "epoch": 2.85,
406
+ "learning_rate": 4.71656050955414e-05,
407
+ "loss": 0.1369,
408
+ "step": 590
409
+ },
410
+ {
411
+ "epoch": 2.9,
412
+ "learning_rate": 4.6847133757961786e-05,
413
+ "loss": 0.0698,
414
+ "step": 600
415
+ },
416
+ {
417
+ "epoch": 2.9,
418
+ "eval_accuracy": 0.9414634146341463,
419
+ "eval_loss": 0.2208263874053955,
420
+ "eval_runtime": 20.9683,
421
+ "eval_samples_per_second": 19.553,
422
+ "eval_steps_per_second": 2.48,
423
+ "step": 600
424
+ },
425
+ {
426
+ "epoch": 2.95,
427
+ "learning_rate": 4.652866242038217e-05,
428
+ "loss": 0.1643,
429
+ "step": 610
430
+ },
431
+ {
432
+ "epoch": 3.0,
433
+ "learning_rate": 4.621019108280255e-05,
434
+ "loss": 0.1105,
435
+ "step": 620
436
+ },
437
+ {
438
+ "epoch": 3.04,
439
+ "learning_rate": 4.589171974522293e-05,
440
+ "loss": 0.0556,
441
+ "step": 630
442
+ },
443
+ {
444
+ "epoch": 3.09,
445
+ "learning_rate": 4.557324840764332e-05,
446
+ "loss": 0.0482,
447
+ "step": 640
448
+ },
449
+ {
450
+ "epoch": 3.14,
451
+ "learning_rate": 4.5254777070063695e-05,
452
+ "loss": 0.0553,
453
+ "step": 650
454
+ },
455
+ {
456
+ "epoch": 3.19,
457
+ "learning_rate": 4.493630573248408e-05,
458
+ "loss": 0.0677,
459
+ "step": 660
460
+ },
461
+ {
462
+ "epoch": 3.24,
463
+ "learning_rate": 4.4617834394904464e-05,
464
+ "loss": 0.0365,
465
+ "step": 670
466
+ },
467
+ {
468
+ "epoch": 3.29,
469
+ "learning_rate": 4.429936305732484e-05,
470
+ "loss": 0.0529,
471
+ "step": 680
472
+ },
473
+ {
474
+ "epoch": 3.33,
475
+ "learning_rate": 4.398089171974523e-05,
476
+ "loss": 0.041,
477
+ "step": 690
478
+ },
479
+ {
480
+ "epoch": 3.38,
481
+ "learning_rate": 4.366242038216561e-05,
482
+ "loss": 0.0553,
483
+ "step": 700
484
+ },
485
+ {
486
+ "epoch": 3.38,
487
+ "eval_accuracy": 0.9512195121951219,
488
+ "eval_loss": 0.1926226019859314,
489
+ "eval_runtime": 21.809,
490
+ "eval_samples_per_second": 18.8,
491
+ "eval_steps_per_second": 2.384,
492
+ "step": 700
493
+ },
494
+ {
495
+ "epoch": 3.43,
496
+ "learning_rate": 4.334394904458599e-05,
497
+ "loss": 0.0574,
498
+ "step": 710
499
+ },
500
+ {
501
+ "epoch": 3.48,
502
+ "learning_rate": 4.3025477707006374e-05,
503
+ "loss": 0.0681,
504
+ "step": 720
505
+ },
506
+ {
507
+ "epoch": 3.53,
508
+ "learning_rate": 4.270700636942676e-05,
509
+ "loss": 0.0541,
510
+ "step": 730
511
+ },
512
+ {
513
+ "epoch": 3.57,
514
+ "learning_rate": 4.2388535031847136e-05,
515
+ "loss": 0.0574,
516
+ "step": 740
517
+ },
518
+ {
519
+ "epoch": 3.62,
520
+ "learning_rate": 4.207006369426752e-05,
521
+ "loss": 0.0447,
522
+ "step": 750
523
+ },
524
+ {
525
+ "epoch": 3.67,
526
+ "learning_rate": 4.1751592356687905e-05,
527
+ "loss": 0.0461,
528
+ "step": 760
529
+ },
530
+ {
531
+ "epoch": 3.72,
532
+ "learning_rate": 4.143312101910828e-05,
533
+ "loss": 0.0228,
534
+ "step": 770
535
+ },
536
+ {
537
+ "epoch": 3.77,
538
+ "learning_rate": 4.111464968152867e-05,
539
+ "loss": 0.034,
540
+ "step": 780
541
+ },
542
+ {
543
+ "epoch": 3.82,
544
+ "learning_rate": 4.0796178343949045e-05,
545
+ "loss": 0.0557,
546
+ "step": 790
547
+ },
548
+ {
549
+ "epoch": 3.86,
550
+ "learning_rate": 4.047770700636943e-05,
551
+ "loss": 0.0239,
552
+ "step": 800
553
+ },
554
+ {
555
+ "epoch": 3.86,
556
+ "eval_accuracy": 0.9390243902439024,
557
+ "eval_loss": 0.19786834716796875,
558
+ "eval_runtime": 21.3483,
559
+ "eval_samples_per_second": 19.205,
560
+ "eval_steps_per_second": 2.436,
561
+ "step": 800
562
+ },
563
+ {
564
+ "epoch": 3.91,
565
+ "learning_rate": 4.0159235668789814e-05,
566
+ "loss": 0.0534,
567
+ "step": 810
568
+ },
569
+ {
570
+ "epoch": 3.96,
571
+ "learning_rate": 3.984076433121019e-05,
572
+ "loss": 0.0521,
573
+ "step": 820
574
+ },
575
+ {
576
+ "epoch": 4.01,
577
+ "learning_rate": 3.952229299363058e-05,
578
+ "loss": 0.0423,
579
+ "step": 830
580
+ },
581
+ {
582
+ "epoch": 4.06,
583
+ "learning_rate": 3.920382165605096e-05,
584
+ "loss": 0.0197,
585
+ "step": 840
586
+ },
587
+ {
588
+ "epoch": 4.11,
589
+ "learning_rate": 3.888535031847134e-05,
590
+ "loss": 0.0584,
591
+ "step": 850
592
+ },
593
+ {
594
+ "epoch": 4.15,
595
+ "learning_rate": 3.8566878980891724e-05,
596
+ "loss": 0.0695,
597
+ "step": 860
598
+ },
599
+ {
600
+ "epoch": 4.2,
601
+ "learning_rate": 3.824840764331211e-05,
602
+ "loss": 0.039,
603
+ "step": 870
604
+ },
605
+ {
606
+ "epoch": 4.25,
607
+ "learning_rate": 3.7929936305732486e-05,
608
+ "loss": 0.0183,
609
+ "step": 880
610
+ },
611
+ {
612
+ "epoch": 4.3,
613
+ "learning_rate": 3.761146496815287e-05,
614
+ "loss": 0.0175,
615
+ "step": 890
616
+ },
617
+ {
618
+ "epoch": 4.35,
619
+ "learning_rate": 3.729299363057325e-05,
620
+ "loss": 0.0386,
621
+ "step": 900
622
+ },
623
+ {
624
+ "epoch": 4.35,
625
+ "eval_accuracy": 0.9365853658536586,
626
+ "eval_loss": 0.22525091469287872,
627
+ "eval_runtime": 21.4031,
628
+ "eval_samples_per_second": 19.156,
629
+ "eval_steps_per_second": 2.43,
630
+ "step": 900
631
+ },
632
+ {
633
+ "epoch": 4.4,
634
+ "learning_rate": 3.697452229299363e-05,
635
+ "loss": 0.0164,
636
+ "step": 910
637
+ },
638
+ {
639
+ "epoch": 4.44,
640
+ "learning_rate": 3.665605095541401e-05,
641
+ "loss": 0.0161,
642
+ "step": 920
643
+ },
644
+ {
645
+ "epoch": 4.49,
646
+ "learning_rate": 3.6337579617834395e-05,
647
+ "loss": 0.0209,
648
+ "step": 930
649
+ },
650
+ {
651
+ "epoch": 4.54,
652
+ "learning_rate": 3.601910828025477e-05,
653
+ "loss": 0.0668,
654
+ "step": 940
655
+ },
656
+ {
657
+ "epoch": 4.59,
658
+ "learning_rate": 3.570063694267516e-05,
659
+ "loss": 0.0697,
660
+ "step": 950
661
+ },
662
+ {
663
+ "epoch": 4.64,
664
+ "learning_rate": 3.538216560509554e-05,
665
+ "loss": 0.0161,
666
+ "step": 960
667
+ },
668
+ {
669
+ "epoch": 4.69,
670
+ "learning_rate": 3.506369426751592e-05,
671
+ "loss": 0.0783,
672
+ "step": 970
673
+ },
674
+ {
675
+ "epoch": 4.73,
676
+ "learning_rate": 3.4745222929936305e-05,
677
+ "loss": 0.0152,
678
+ "step": 980
679
+ },
680
+ {
681
+ "epoch": 4.78,
682
+ "learning_rate": 3.442675159235669e-05,
683
+ "loss": 0.0146,
684
+ "step": 990
685
+ },
686
+ {
687
+ "epoch": 4.83,
688
+ "learning_rate": 3.410828025477707e-05,
689
+ "loss": 0.0143,
690
+ "step": 1000
691
+ },
692
+ {
693
+ "epoch": 4.83,
694
+ "eval_accuracy": 0.9463414634146341,
695
+ "eval_loss": 0.19345715641975403,
696
+ "eval_runtime": 21.1331,
697
+ "eval_samples_per_second": 19.401,
698
+ "eval_steps_per_second": 2.461,
699
+ "step": 1000
700
+ },
701
+ {
702
+ "epoch": 4.88,
703
+ "learning_rate": 3.378980891719745e-05,
704
+ "loss": 0.0143,
705
+ "step": 1010
706
+ },
707
+ {
708
+ "epoch": 4.93,
709
+ "learning_rate": 3.3471337579617836e-05,
710
+ "loss": 0.0136,
711
+ "step": 1020
712
+ },
713
+ {
714
+ "epoch": 4.98,
715
+ "learning_rate": 3.3152866242038214e-05,
716
+ "loss": 0.0386,
717
+ "step": 1030
718
+ },
719
+ {
720
+ "epoch": 5.02,
721
+ "learning_rate": 3.28343949044586e-05,
722
+ "loss": 0.0336,
723
+ "step": 1040
724
+ },
725
+ {
726
+ "epoch": 5.07,
727
+ "learning_rate": 3.251592356687898e-05,
728
+ "loss": 0.013,
729
+ "step": 1050
730
+ },
731
+ {
732
+ "epoch": 5.12,
733
+ "learning_rate": 3.219745222929936e-05,
734
+ "loss": 0.0126,
735
+ "step": 1060
736
+ },
737
+ {
738
+ "epoch": 5.17,
739
+ "learning_rate": 3.1878980891719745e-05,
740
+ "loss": 0.0128,
741
+ "step": 1070
742
+ },
743
+ {
744
+ "epoch": 5.22,
745
+ "learning_rate": 3.156050955414013e-05,
746
+ "loss": 0.0122,
747
+ "step": 1080
748
+ },
749
+ {
750
+ "epoch": 5.27,
751
+ "learning_rate": 3.124203821656051e-05,
752
+ "loss": 0.0136,
753
+ "step": 1090
754
+ },
755
+ {
756
+ "epoch": 5.31,
757
+ "learning_rate": 3.092356687898089e-05,
758
+ "loss": 0.0606,
759
+ "step": 1100
760
+ },
761
+ {
762
+ "epoch": 5.31,
763
+ "eval_accuracy": 0.9536585365853658,
764
+ "eval_loss": 0.1909320205450058,
765
+ "eval_runtime": 20.2405,
766
+ "eval_samples_per_second": 20.256,
767
+ "eval_steps_per_second": 2.569,
768
+ "step": 1100
769
+ },
770
+ {
771
+ "epoch": 5.36,
772
+ "learning_rate": 3.060509554140128e-05,
773
+ "loss": 0.0368,
774
+ "step": 1110
775
+ },
776
+ {
777
+ "epoch": 5.41,
778
+ "learning_rate": 3.0286624203821655e-05,
779
+ "loss": 0.0116,
780
+ "step": 1120
781
+ },
782
+ {
783
+ "epoch": 5.46,
784
+ "learning_rate": 2.996815286624204e-05,
785
+ "loss": 0.0335,
786
+ "step": 1130
787
+ },
788
+ {
789
+ "epoch": 5.51,
790
+ "learning_rate": 2.964968152866242e-05,
791
+ "loss": 0.0211,
792
+ "step": 1140
793
+ },
794
+ {
795
+ "epoch": 5.56,
796
+ "learning_rate": 2.93312101910828e-05,
797
+ "loss": 0.0114,
798
+ "step": 1150
799
+ },
800
+ {
801
+ "epoch": 5.6,
802
+ "learning_rate": 2.9012738853503186e-05,
803
+ "loss": 0.0119,
804
+ "step": 1160
805
+ },
806
+ {
807
+ "epoch": 5.65,
808
+ "learning_rate": 2.8694267515923567e-05,
809
+ "loss": 0.0167,
810
+ "step": 1170
811
+ },
812
+ {
813
+ "epoch": 5.7,
814
+ "learning_rate": 2.837579617834395e-05,
815
+ "loss": 0.0111,
816
+ "step": 1180
817
+ },
818
+ {
819
+ "epoch": 5.75,
820
+ "learning_rate": 2.8057324840764333e-05,
821
+ "loss": 0.011,
822
+ "step": 1190
823
+ },
824
+ {
825
+ "epoch": 5.8,
826
+ "learning_rate": 2.7738853503184714e-05,
827
+ "loss": 0.0105,
828
+ "step": 1200
829
+ },
830
+ {
831
+ "epoch": 5.8,
832
+ "eval_accuracy": 0.9390243902439024,
833
+ "eval_loss": 0.24669073522090912,
834
+ "eval_runtime": 20.6758,
835
+ "eval_samples_per_second": 19.83,
836
+ "eval_steps_per_second": 2.515,
837
+ "step": 1200
838
+ },
839
+ {
840
+ "epoch": 5.85,
841
+ "learning_rate": 2.7420382165605095e-05,
842
+ "loss": 0.0104,
843
+ "step": 1210
844
+ },
845
+ {
846
+ "epoch": 5.89,
847
+ "learning_rate": 2.710191082802548e-05,
848
+ "loss": 0.0106,
849
+ "step": 1220
850
+ },
851
+ {
852
+ "epoch": 5.94,
853
+ "learning_rate": 2.678343949044586e-05,
854
+ "loss": 0.0101,
855
+ "step": 1230
856
+ },
857
+ {
858
+ "epoch": 5.99,
859
+ "learning_rate": 2.6464968152866242e-05,
860
+ "loss": 0.0395,
861
+ "step": 1240
862
+ },
863
+ {
864
+ "epoch": 6.04,
865
+ "learning_rate": 2.6146496815286624e-05,
866
+ "loss": 0.0098,
867
+ "step": 1250
868
+ },
869
+ {
870
+ "epoch": 6.09,
871
+ "learning_rate": 2.5828025477707008e-05,
872
+ "loss": 0.0099,
873
+ "step": 1260
874
+ },
875
+ {
876
+ "epoch": 6.14,
877
+ "learning_rate": 2.550955414012739e-05,
878
+ "loss": 0.0097,
879
+ "step": 1270
880
+ },
881
+ {
882
+ "epoch": 6.18,
883
+ "learning_rate": 2.519108280254777e-05,
884
+ "loss": 0.0095,
885
+ "step": 1280
886
+ },
887
+ {
888
+ "epoch": 6.23,
889
+ "learning_rate": 2.4872611464968155e-05,
890
+ "loss": 0.0094,
891
+ "step": 1290
892
+ },
893
+ {
894
+ "epoch": 6.28,
895
+ "learning_rate": 2.4554140127388536e-05,
896
+ "loss": 0.0092,
897
+ "step": 1300
898
+ },
899
+ {
900
+ "epoch": 6.28,
901
+ "eval_accuracy": 0.9512195121951219,
902
+ "eval_loss": 0.19856394827365875,
903
+ "eval_runtime": 19.7104,
904
+ "eval_samples_per_second": 20.801,
905
+ "eval_steps_per_second": 2.638,
906
+ "step": 1300
907
+ },
908
+ {
909
+ "epoch": 6.33,
910
+ "learning_rate": 2.4235668789808917e-05,
911
+ "loss": 0.0091,
912
+ "step": 1310
913
+ },
914
+ {
915
+ "epoch": 6.38,
916
+ "learning_rate": 2.3917197452229302e-05,
917
+ "loss": 0.009,
918
+ "step": 1320
919
+ },
920
+ {
921
+ "epoch": 6.43,
922
+ "learning_rate": 2.3598726114649683e-05,
923
+ "loss": 0.0089,
924
+ "step": 1330
925
+ },
926
+ {
927
+ "epoch": 6.47,
928
+ "learning_rate": 2.3280254777070064e-05,
929
+ "loss": 0.0092,
930
+ "step": 1340
931
+ },
932
+ {
933
+ "epoch": 6.52,
934
+ "learning_rate": 2.296178343949045e-05,
935
+ "loss": 0.0087,
936
+ "step": 1350
937
+ },
938
+ {
939
+ "epoch": 6.57,
940
+ "learning_rate": 2.264331210191083e-05,
941
+ "loss": 0.0086,
942
+ "step": 1360
943
+ },
944
+ {
945
+ "epoch": 6.62,
946
+ "learning_rate": 2.232484076433121e-05,
947
+ "loss": 0.029,
948
+ "step": 1370
949
+ },
950
+ {
951
+ "epoch": 6.67,
952
+ "learning_rate": 2.2006369426751592e-05,
953
+ "loss": 0.0087,
954
+ "step": 1380
955
+ },
956
+ {
957
+ "epoch": 6.71,
958
+ "learning_rate": 2.1687898089171977e-05,
959
+ "loss": 0.0087,
960
+ "step": 1390
961
+ },
962
+ {
963
+ "epoch": 6.76,
964
+ "learning_rate": 2.1369426751592358e-05,
965
+ "loss": 0.0085,
966
+ "step": 1400
967
+ },
968
+ {
969
+ "epoch": 6.76,
970
+ "eval_accuracy": 0.9512195121951219,
971
+ "eval_loss": 0.20551267266273499,
972
+ "eval_runtime": 21.7651,
973
+ "eval_samples_per_second": 18.837,
974
+ "eval_steps_per_second": 2.389,
975
+ "step": 1400
976
+ },
977
+ {
978
+ "epoch": 6.81,
979
+ "learning_rate": 2.105095541401274e-05,
980
+ "loss": 0.0085,
981
+ "step": 1410
982
+ },
983
+ {
984
+ "epoch": 6.86,
985
+ "learning_rate": 2.0732484076433124e-05,
986
+ "loss": 0.0084,
987
+ "step": 1420
988
+ },
989
+ {
990
+ "epoch": 6.91,
991
+ "learning_rate": 2.0414012738853505e-05,
992
+ "loss": 0.0081,
993
+ "step": 1430
994
+ },
995
+ {
996
+ "epoch": 6.96,
997
+ "learning_rate": 2.0095541401273886e-05,
998
+ "loss": 0.008,
999
+ "step": 1440
1000
+ },
1001
+ {
1002
+ "epoch": 7.0,
1003
+ "learning_rate": 1.977707006369427e-05,
1004
+ "loss": 0.0112,
1005
+ "step": 1450
1006
+ },
1007
+ {
1008
+ "epoch": 7.05,
1009
+ "learning_rate": 1.9458598726114652e-05,
1010
+ "loss": 0.0079,
1011
+ "step": 1460
1012
+ },
1013
+ {
1014
+ "epoch": 7.1,
1015
+ "learning_rate": 1.9140127388535033e-05,
1016
+ "loss": 0.0078,
1017
+ "step": 1470
1018
+ },
1019
+ {
1020
+ "epoch": 7.15,
1021
+ "learning_rate": 1.8821656050955418e-05,
1022
+ "loss": 0.0078,
1023
+ "step": 1480
1024
+ },
1025
+ {
1026
+ "epoch": 7.2,
1027
+ "learning_rate": 1.8503184713375796e-05,
1028
+ "loss": 0.0076,
1029
+ "step": 1490
1030
+ },
1031
+ {
1032
+ "epoch": 7.25,
1033
+ "learning_rate": 1.8184713375796177e-05,
1034
+ "loss": 0.0076,
1035
+ "step": 1500
1036
+ },
1037
+ {
1038
+ "epoch": 7.25,
1039
+ "eval_accuracy": 0.9463414634146341,
1040
+ "eval_loss": 0.2132251262664795,
1041
+ "eval_runtime": 21.8557,
1042
+ "eval_samples_per_second": 18.759,
1043
+ "eval_steps_per_second": 2.379,
1044
+ "step": 1500
1045
+ },
1046
+ {
1047
+ "epoch": 7.29,
1048
+ "learning_rate": 1.786624203821656e-05,
1049
+ "loss": 0.0075,
1050
+ "step": 1510
1051
+ },
1052
+ {
1053
+ "epoch": 7.34,
1054
+ "learning_rate": 1.7547770700636942e-05,
1055
+ "loss": 0.0075,
1056
+ "step": 1520
1057
+ },
1058
+ {
1059
+ "epoch": 7.39,
1060
+ "learning_rate": 1.7229299363057324e-05,
1061
+ "loss": 0.0075,
1062
+ "step": 1530
1063
+ },
1064
+ {
1065
+ "epoch": 7.44,
1066
+ "learning_rate": 1.6910828025477708e-05,
1067
+ "loss": 0.0074,
1068
+ "step": 1540
1069
+ },
1070
+ {
1071
+ "epoch": 7.49,
1072
+ "learning_rate": 1.659235668789809e-05,
1073
+ "loss": 0.0073,
1074
+ "step": 1550
1075
+ },
1076
+ {
1077
+ "epoch": 7.54,
1078
+ "learning_rate": 1.627388535031847e-05,
1079
+ "loss": 0.0073,
1080
+ "step": 1560
1081
+ },
1082
+ {
1083
+ "epoch": 7.58,
1084
+ "learning_rate": 1.5955414012738855e-05,
1085
+ "loss": 0.0073,
1086
+ "step": 1570
1087
+ },
1088
+ {
1089
+ "epoch": 7.63,
1090
+ "learning_rate": 1.5636942675159236e-05,
1091
+ "loss": 0.0072,
1092
+ "step": 1580
1093
+ },
1094
+ {
1095
+ "epoch": 7.68,
1096
+ "learning_rate": 1.5318471337579618e-05,
1097
+ "loss": 0.0072,
1098
+ "step": 1590
1099
+ },
1100
+ {
1101
+ "epoch": 7.73,
1102
+ "learning_rate": 1.5e-05,
1103
+ "loss": 0.0071,
1104
+ "step": 1600
1105
+ },
1106
+ {
1107
+ "epoch": 7.73,
1108
+ "eval_accuracy": 0.9463414634146341,
1109
+ "eval_loss": 0.21747459471225739,
1110
+ "eval_runtime": 20.4277,
1111
+ "eval_samples_per_second": 20.071,
1112
+ "eval_steps_per_second": 2.546,
1113
+ "step": 1600
1114
+ },
1115
+ {
1116
+ "epoch": 7.78,
1117
+ "learning_rate": 1.4681528662420382e-05,
1118
+ "loss": 0.007,
1119
+ "step": 1610
1120
+ },
1121
+ {
1122
+ "epoch": 7.83,
1123
+ "learning_rate": 1.4363057324840764e-05,
1124
+ "loss": 0.007,
1125
+ "step": 1620
1126
+ },
1127
+ {
1128
+ "epoch": 7.87,
1129
+ "learning_rate": 1.4044585987261147e-05,
1130
+ "loss": 0.0069,
1131
+ "step": 1630
1132
+ },
1133
+ {
1134
+ "epoch": 7.92,
1135
+ "learning_rate": 1.3726114649681528e-05,
1136
+ "loss": 0.0069,
1137
+ "step": 1640
1138
+ },
1139
+ {
1140
+ "epoch": 7.97,
1141
+ "learning_rate": 1.3407643312101911e-05,
1142
+ "loss": 0.0171,
1143
+ "step": 1650
1144
+ },
1145
+ {
1146
+ "epoch": 8.02,
1147
+ "learning_rate": 1.3089171974522294e-05,
1148
+ "loss": 0.0068,
1149
+ "step": 1660
1150
+ },
1151
+ {
1152
+ "epoch": 8.07,
1153
+ "learning_rate": 1.2770700636942675e-05,
1154
+ "loss": 0.0068,
1155
+ "step": 1670
1156
+ },
1157
+ {
1158
+ "epoch": 8.12,
1159
+ "learning_rate": 1.2452229299363058e-05,
1160
+ "loss": 0.0068,
1161
+ "step": 1680
1162
+ },
1163
+ {
1164
+ "epoch": 8.16,
1165
+ "learning_rate": 1.213375796178344e-05,
1166
+ "loss": 0.0067,
1167
+ "step": 1690
1168
+ },
1169
+ {
1170
+ "epoch": 8.21,
1171
+ "learning_rate": 1.1815286624203822e-05,
1172
+ "loss": 0.0067,
1173
+ "step": 1700
1174
+ },
1175
+ {
1176
+ "epoch": 8.21,
1177
+ "eval_accuracy": 0.948780487804878,
1178
+ "eval_loss": 0.2226085364818573,
1179
+ "eval_runtime": 21.3206,
1180
+ "eval_samples_per_second": 19.23,
1181
+ "eval_steps_per_second": 2.439,
1182
+ "step": 1700
1183
+ },
1184
+ {
1185
+ "epoch": 8.26,
1186
+ "learning_rate": 1.1496815286624205e-05,
1187
+ "loss": 0.0067,
1188
+ "step": 1710
1189
+ },
1190
+ {
1191
+ "epoch": 8.31,
1192
+ "learning_rate": 1.1178343949044586e-05,
1193
+ "loss": 0.0066,
1194
+ "step": 1720
1195
+ },
1196
+ {
1197
+ "epoch": 8.36,
1198
+ "learning_rate": 1.085987261146497e-05,
1199
+ "loss": 0.0066,
1200
+ "step": 1730
1201
+ },
1202
+ {
1203
+ "epoch": 8.41,
1204
+ "learning_rate": 1.054140127388535e-05,
1205
+ "loss": 0.0065,
1206
+ "step": 1740
1207
+ },
1208
+ {
1209
+ "epoch": 8.45,
1210
+ "learning_rate": 1.0222929936305733e-05,
1211
+ "loss": 0.0065,
1212
+ "step": 1750
1213
+ },
1214
+ {
1215
+ "epoch": 8.5,
1216
+ "learning_rate": 9.904458598726116e-06,
1217
+ "loss": 0.0065,
1218
+ "step": 1760
1219
+ },
1220
+ {
1221
+ "epoch": 8.55,
1222
+ "learning_rate": 9.585987261146497e-06,
1223
+ "loss": 0.0065,
1224
+ "step": 1770
1225
+ },
1226
+ {
1227
+ "epoch": 8.6,
1228
+ "learning_rate": 9.267515923566879e-06,
1229
+ "loss": 0.0064,
1230
+ "step": 1780
1231
+ },
1232
+ {
1233
+ "epoch": 8.65,
1234
+ "learning_rate": 8.949044585987261e-06,
1235
+ "loss": 0.0064,
1236
+ "step": 1790
1237
+ },
1238
+ {
1239
+ "epoch": 8.7,
1240
+ "learning_rate": 8.630573248407643e-06,
1241
+ "loss": 0.0064,
1242
+ "step": 1800
1243
+ },
1244
+ {
1245
+ "epoch": 8.7,
1246
+ "eval_accuracy": 0.9463414634146341,
1247
+ "eval_loss": 0.22321389615535736,
1248
+ "eval_runtime": 19.9115,
1249
+ "eval_samples_per_second": 20.591,
1250
+ "eval_steps_per_second": 2.612,
1251
+ "step": 1800
1252
+ },
1253
+ {
1254
+ "epoch": 8.74,
1255
+ "learning_rate": 8.312101910828025e-06,
1256
+ "loss": 0.0064,
1257
+ "step": 1810
1258
+ },
1259
+ {
1260
+ "epoch": 8.79,
1261
+ "learning_rate": 7.993630573248408e-06,
1262
+ "loss": 0.0106,
1263
+ "step": 1820
1264
+ },
1265
+ {
1266
+ "epoch": 8.84,
1267
+ "learning_rate": 7.67515923566879e-06,
1268
+ "loss": 0.0064,
1269
+ "step": 1830
1270
+ },
1271
+ {
1272
+ "epoch": 8.89,
1273
+ "learning_rate": 7.356687898089172e-06,
1274
+ "loss": 0.0063,
1275
+ "step": 1840
1276
+ },
1277
+ {
1278
+ "epoch": 8.94,
1279
+ "learning_rate": 7.0382165605095544e-06,
1280
+ "loss": 0.0063,
1281
+ "step": 1850
1282
+ },
1283
+ {
1284
+ "epoch": 8.99,
1285
+ "learning_rate": 6.7197452229299365e-06,
1286
+ "loss": 0.013,
1287
+ "step": 1860
1288
+ },
1289
+ {
1290
+ "epoch": 9.03,
1291
+ "learning_rate": 6.4012738853503185e-06,
1292
+ "loss": 0.0062,
1293
+ "step": 1870
1294
+ },
1295
+ {
1296
+ "epoch": 9.08,
1297
+ "learning_rate": 6.082802547770701e-06,
1298
+ "loss": 0.0062,
1299
+ "step": 1880
1300
+ },
1301
+ {
1302
+ "epoch": 9.13,
1303
+ "learning_rate": 5.764331210191083e-06,
1304
+ "loss": 0.0062,
1305
+ "step": 1890
1306
+ },
1307
+ {
1308
+ "epoch": 9.18,
1309
+ "learning_rate": 5.445859872611465e-06,
1310
+ "loss": 0.0062,
1311
+ "step": 1900
1312
+ },
1313
+ {
1314
+ "epoch": 9.18,
1315
+ "eval_accuracy": 0.9463414634146341,
1316
+ "eval_loss": 0.2225855439901352,
1317
+ "eval_runtime": 18.1543,
1318
+ "eval_samples_per_second": 22.584,
1319
+ "eval_steps_per_second": 2.864,
1320
+ "step": 1900
1321
+ },
1322
+ {
1323
+ "epoch": 9.23,
1324
+ "learning_rate": 5.1273885350318474e-06,
1325
+ "loss": 0.0061,
1326
+ "step": 1910
1327
+ },
1328
+ {
1329
+ "epoch": 9.28,
1330
+ "learning_rate": 4.8089171974522295e-06,
1331
+ "loss": 0.0062,
1332
+ "step": 1920
1333
+ },
1334
+ {
1335
+ "epoch": 9.32,
1336
+ "learning_rate": 4.4904458598726115e-06,
1337
+ "loss": 0.0061,
1338
+ "step": 1930
1339
+ },
1340
+ {
1341
+ "epoch": 9.37,
1342
+ "learning_rate": 4.1719745222929935e-06,
1343
+ "loss": 0.0062,
1344
+ "step": 1940
1345
+ },
1346
+ {
1347
+ "epoch": 9.42,
1348
+ "learning_rate": 3.8535031847133755e-06,
1349
+ "loss": 0.0061,
1350
+ "step": 1950
1351
+ },
1352
+ {
1353
+ "epoch": 9.47,
1354
+ "learning_rate": 3.535031847133758e-06,
1355
+ "loss": 0.0061,
1356
+ "step": 1960
1357
+ },
1358
+ {
1359
+ "epoch": 9.52,
1360
+ "learning_rate": 3.2165605095541404e-06,
1361
+ "loss": 0.0061,
1362
+ "step": 1970
1363
+ },
1364
+ {
1365
+ "epoch": 9.57,
1366
+ "learning_rate": 2.8980891719745225e-06,
1367
+ "loss": 0.0061,
1368
+ "step": 1980
1369
+ },
1370
+ {
1371
+ "epoch": 9.61,
1372
+ "learning_rate": 2.579617834394905e-06,
1373
+ "loss": 0.0061,
1374
+ "step": 1990
1375
+ },
1376
+ {
1377
+ "epoch": 9.66,
1378
+ "learning_rate": 2.2611464968152865e-06,
1379
+ "loss": 0.0061,
1380
+ "step": 2000
1381
+ },
1382
+ {
1383
+ "epoch": 9.66,
1384
+ "eval_accuracy": 0.9463414634146341,
1385
+ "eval_loss": 0.2239292562007904,
1386
+ "eval_runtime": 17.7456,
1387
+ "eval_samples_per_second": 23.104,
1388
+ "eval_steps_per_second": 2.93,
1389
+ "step": 2000
1390
+ },
1391
+ {
1392
+ "epoch": 9.71,
1393
+ "learning_rate": 1.942675159235669e-06,
1394
+ "loss": 0.0061,
1395
+ "step": 2010
1396
+ },
1397
+ {
1398
+ "epoch": 9.76,
1399
+ "learning_rate": 1.624203821656051e-06,
1400
+ "loss": 0.0061,
1401
+ "step": 2020
1402
+ },
1403
+ {
1404
+ "epoch": 9.81,
1405
+ "learning_rate": 1.3057324840764332e-06,
1406
+ "loss": 0.0097,
1407
+ "step": 2030
1408
+ },
1409
+ {
1410
+ "epoch": 9.86,
1411
+ "learning_rate": 9.872611464968153e-07,
1412
+ "loss": 0.0061,
1413
+ "step": 2040
1414
+ },
1415
+ {
1416
+ "epoch": 9.9,
1417
+ "learning_rate": 6.687898089171975e-07,
1418
+ "loss": 0.0061,
1419
+ "step": 2050
1420
+ },
1421
+ {
1422
+ "epoch": 9.95,
1423
+ "learning_rate": 3.503184713375796e-07,
1424
+ "loss": 0.0061,
1425
+ "step": 2060
1426
+ },
1427
+ {
1428
+ "epoch": 10.0,
1429
+ "learning_rate": 3.184713375796178e-08,
1430
+ "loss": 0.0112,
1431
+ "step": 2070
1432
+ },
1433
+ {
1434
+ "epoch": 10.0,
1435
+ "step": 2070,
1436
+ "total_flos": 2.557281496670208e+18,
1437
+ "train_loss": 0.16907700259904354,
1438
+ "train_runtime": 2197.7206,
1439
+ "train_samples_per_second": 15.016,
1440
+ "train_steps_per_second": 0.942
1441
+ }
1442
+ ],
1443
+ "logging_steps": 10,
1444
+ "max_steps": 2070,
1445
+ "num_input_tokens_seen": 0,
1446
+ "num_train_epochs": 10,
1447
+ "save_steps": 100,
1448
+ "total_flos": 2.557281496670208e+18,
1449
+ "train_batch_size": 16,
1450
+ "trial_name": null,
1451
+ "trial_params": null
1452
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:39e9d9bf5502b258df1714174834e3e322f83b454ec3f695145ab519bf000096
3
+ size 4728