ismgar01 commited on
Commit
93d58c8
1 Parent(s): 23e93f5

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_accuracy": 0.9937357630979499,
4
+ "eval_loss": 0.0182377677410841,
5
+ "eval_runtime": 26.022,
6
+ "eval_samples_per_second": 134.963,
7
+ "eval_steps_per_second": 4.227,
8
+ "train_loss": 0.06854689962371369,
9
+ "train_runtime": 1208.8541,
10
+ "train_samples_per_second": 82.301,
11
+ "train_steps_per_second": 2.573
12
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "eval_accuracy": 0.9937357630979499,
4
+ "eval_loss": 0.0182377677410841,
5
+ "eval_runtime": 26.022,
6
+ "eval_samples_per_second": 134.963,
7
+ "eval_steps_per_second": 4.227
8
+ }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8b87f3d03cfea0b086b5c7dab60a9630efef7ee4133cdac59f8b1dea2770798a
3
  size 343284983
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be4fcdf6ed10d6b11b211b62ef1c18598f4e0d6c88612abcc59ffa729d8ed57a
3
  size 343284983
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 5.0,
3
+ "train_loss": 0.06854689962371369,
4
+ "train_runtime": 1208.8541,
5
+ "train_samples_per_second": 82.301,
6
+ "train_steps_per_second": 2.573
7
+ }
trainer_state.json ADDED
@@ -0,0 +1,1936 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.0182377677410841,
3
+ "best_model_checkpoint": "./cats_vs_dogs_outputs/checkpoint-3110",
4
+ "epoch": 5.0,
5
+ "global_step": 3110,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.02,
12
+ "learning_rate": 0.00019935691318327976,
13
+ "loss": 0.3497,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.03,
18
+ "learning_rate": 0.00019871382636655948,
19
+ "loss": 0.1232,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.05,
24
+ "learning_rate": 0.00019807073954983923,
25
+ "loss": 0.1623,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.06,
30
+ "learning_rate": 0.00019742765273311899,
31
+ "loss": 0.1004,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.08,
36
+ "learning_rate": 0.00019678456591639874,
37
+ "loss": 0.0815,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.1,
42
+ "learning_rate": 0.00019614147909967846,
43
+ "loss": 0.161,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.11,
48
+ "learning_rate": 0.0001954983922829582,
49
+ "loss": 0.168,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.13,
54
+ "learning_rate": 0.00019485530546623796,
55
+ "loss": 0.1196,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.14,
60
+ "learning_rate": 0.00019421221864951768,
61
+ "loss": 0.1474,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.16,
66
+ "learning_rate": 0.00019356913183279743,
67
+ "loss": 0.1664,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.18,
72
+ "learning_rate": 0.00019292604501607719,
73
+ "loss": 0.1165,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.19,
78
+ "learning_rate": 0.00019228295819935694,
79
+ "loss": 0.0778,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.21,
84
+ "learning_rate": 0.00019163987138263666,
85
+ "loss": 0.0898,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.23,
90
+ "learning_rate": 0.0001909967845659164,
91
+ "loss": 0.1072,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.24,
96
+ "learning_rate": 0.00019035369774919616,
97
+ "loss": 0.1722,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.26,
102
+ "learning_rate": 0.00018971061093247588,
103
+ "loss": 0.1277,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.27,
108
+ "learning_rate": 0.00018906752411575564,
109
+ "loss": 0.1044,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.29,
114
+ "learning_rate": 0.00018842443729903539,
115
+ "loss": 0.1282,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.31,
120
+ "learning_rate": 0.00018778135048231514,
121
+ "loss": 0.1034,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.32,
126
+ "learning_rate": 0.00018713826366559486,
127
+ "loss": 0.1608,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.34,
132
+ "learning_rate": 0.0001864951768488746,
133
+ "loss": 0.108,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.35,
138
+ "learning_rate": 0.00018585209003215436,
139
+ "loss": 0.0853,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.37,
144
+ "learning_rate": 0.00018520900321543409,
145
+ "loss": 0.178,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.39,
150
+ "learning_rate": 0.00018456591639871384,
151
+ "loss": 0.1595,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.4,
156
+ "learning_rate": 0.0001839228295819936,
157
+ "loss": 0.1134,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.42,
162
+ "learning_rate": 0.0001832797427652733,
163
+ "loss": 0.0914,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.43,
168
+ "learning_rate": 0.00018263665594855306,
169
+ "loss": 0.107,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.45,
174
+ "learning_rate": 0.0001819935691318328,
175
+ "loss": 0.078,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.47,
180
+ "learning_rate": 0.00018135048231511256,
181
+ "loss": 0.0715,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.48,
186
+ "learning_rate": 0.00018070739549839229,
187
+ "loss": 0.055,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.5,
192
+ "learning_rate": 0.00018006430868167204,
193
+ "loss": 0.1244,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.51,
198
+ "learning_rate": 0.0001794212218649518,
199
+ "loss": 0.1074,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.53,
204
+ "learning_rate": 0.0001787781350482315,
205
+ "loss": 0.1277,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.55,
210
+ "learning_rate": 0.00017813504823151126,
211
+ "loss": 0.1298,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.56,
216
+ "learning_rate": 0.000177491961414791,
217
+ "loss": 0.1037,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.58,
222
+ "learning_rate": 0.00017684887459807076,
223
+ "loss": 0.1061,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.59,
228
+ "learning_rate": 0.00017620578778135049,
229
+ "loss": 0.1455,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.61,
234
+ "learning_rate": 0.00017556270096463024,
235
+ "loss": 0.1469,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.63,
240
+ "learning_rate": 0.00017491961414791,
241
+ "loss": 0.127,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.64,
246
+ "learning_rate": 0.0001742765273311897,
247
+ "loss": 0.1033,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.66,
252
+ "learning_rate": 0.00017363344051446946,
253
+ "loss": 0.1018,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.68,
258
+ "learning_rate": 0.0001729903536977492,
259
+ "loss": 0.0866,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.69,
264
+ "learning_rate": 0.00017234726688102896,
265
+ "loss": 0.079,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.71,
270
+ "learning_rate": 0.0001717041800643087,
271
+ "loss": 0.0837,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.72,
276
+ "learning_rate": 0.00017106109324758844,
277
+ "loss": 0.1862,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.74,
282
+ "learning_rate": 0.0001704180064308682,
283
+ "loss": 0.1114,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.76,
288
+ "learning_rate": 0.0001697749196141479,
289
+ "loss": 0.071,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.77,
294
+ "learning_rate": 0.00016913183279742766,
295
+ "loss": 0.1379,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.79,
300
+ "learning_rate": 0.0001684887459807074,
301
+ "loss": 0.0764,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.8,
306
+ "learning_rate": 0.00016784565916398716,
307
+ "loss": 0.115,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.82,
312
+ "learning_rate": 0.0001672025723472669,
313
+ "loss": 0.0704,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.84,
318
+ "learning_rate": 0.00016655948553054664,
319
+ "loss": 0.0867,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.85,
324
+ "learning_rate": 0.0001659163987138264,
325
+ "loss": 0.0827,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 0.87,
330
+ "learning_rate": 0.0001652733118971061,
331
+ "loss": 0.0583,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 0.88,
336
+ "learning_rate": 0.00016463022508038586,
337
+ "loss": 0.1185,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 0.9,
342
+ "learning_rate": 0.0001639871382636656,
343
+ "loss": 0.1053,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 0.92,
348
+ "learning_rate": 0.00016334405144694536,
349
+ "loss": 0.0695,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 0.93,
354
+ "learning_rate": 0.0001627009646302251,
355
+ "loss": 0.0818,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 0.95,
360
+ "learning_rate": 0.00016205787781350484,
361
+ "loss": 0.1222,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 0.96,
366
+ "learning_rate": 0.0001614147909967846,
367
+ "loss": 0.2004,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 0.98,
372
+ "learning_rate": 0.0001607717041800643,
373
+ "loss": 0.1069,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 1.0,
378
+ "learning_rate": 0.00016012861736334406,
379
+ "loss": 0.1177,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 1.0,
384
+ "eval_accuracy": 0.9832004555808656,
385
+ "eval_loss": 0.04725871980190277,
386
+ "eval_runtime": 25.5362,
387
+ "eval_samples_per_second": 137.53,
388
+ "eval_steps_per_second": 4.308,
389
+ "step": 622
390
+ },
391
+ {
392
+ "epoch": 1.01,
393
+ "learning_rate": 0.00015948553054662381,
394
+ "loss": 0.1198,
395
+ "step": 630
396
+ },
397
+ {
398
+ "epoch": 1.03,
399
+ "learning_rate": 0.00015884244372990354,
400
+ "loss": 0.0857,
401
+ "step": 640
402
+ },
403
+ {
404
+ "epoch": 1.05,
405
+ "learning_rate": 0.0001581993569131833,
406
+ "loss": 0.1011,
407
+ "step": 650
408
+ },
409
+ {
410
+ "epoch": 1.06,
411
+ "learning_rate": 0.00015755627009646304,
412
+ "loss": 0.0743,
413
+ "step": 660
414
+ },
415
+ {
416
+ "epoch": 1.08,
417
+ "learning_rate": 0.0001569131832797428,
418
+ "loss": 0.0917,
419
+ "step": 670
420
+ },
421
+ {
422
+ "epoch": 1.09,
423
+ "learning_rate": 0.0001562700964630225,
424
+ "loss": 0.0922,
425
+ "step": 680
426
+ },
427
+ {
428
+ "epoch": 1.11,
429
+ "learning_rate": 0.00015562700964630226,
430
+ "loss": 0.1241,
431
+ "step": 690
432
+ },
433
+ {
434
+ "epoch": 1.13,
435
+ "learning_rate": 0.00015498392282958201,
436
+ "loss": 0.087,
437
+ "step": 700
438
+ },
439
+ {
440
+ "epoch": 1.14,
441
+ "learning_rate": 0.00015434083601286174,
442
+ "loss": 0.0675,
443
+ "step": 710
444
+ },
445
+ {
446
+ "epoch": 1.16,
447
+ "learning_rate": 0.0001536977491961415,
448
+ "loss": 0.0733,
449
+ "step": 720
450
+ },
451
+ {
452
+ "epoch": 1.17,
453
+ "learning_rate": 0.00015305466237942124,
454
+ "loss": 0.0816,
455
+ "step": 730
456
+ },
457
+ {
458
+ "epoch": 1.19,
459
+ "learning_rate": 0.000152411575562701,
460
+ "loss": 0.1012,
461
+ "step": 740
462
+ },
463
+ {
464
+ "epoch": 1.21,
465
+ "learning_rate": 0.0001517684887459807,
466
+ "loss": 0.0797,
467
+ "step": 750
468
+ },
469
+ {
470
+ "epoch": 1.22,
471
+ "learning_rate": 0.00015112540192926046,
472
+ "loss": 0.0804,
473
+ "step": 760
474
+ },
475
+ {
476
+ "epoch": 1.24,
477
+ "learning_rate": 0.00015048231511254021,
478
+ "loss": 0.0917,
479
+ "step": 770
480
+ },
481
+ {
482
+ "epoch": 1.25,
483
+ "learning_rate": 0.00014983922829581994,
484
+ "loss": 0.1153,
485
+ "step": 780
486
+ },
487
+ {
488
+ "epoch": 1.27,
489
+ "learning_rate": 0.0001491961414790997,
490
+ "loss": 0.0476,
491
+ "step": 790
492
+ },
493
+ {
494
+ "epoch": 1.29,
495
+ "learning_rate": 0.00014855305466237944,
496
+ "loss": 0.0813,
497
+ "step": 800
498
+ },
499
+ {
500
+ "epoch": 1.3,
501
+ "learning_rate": 0.0001479099678456592,
502
+ "loss": 0.0923,
503
+ "step": 810
504
+ },
505
+ {
506
+ "epoch": 1.32,
507
+ "learning_rate": 0.00014726688102893891,
508
+ "loss": 0.0822,
509
+ "step": 820
510
+ },
511
+ {
512
+ "epoch": 1.33,
513
+ "learning_rate": 0.00014662379421221866,
514
+ "loss": 0.0579,
515
+ "step": 830
516
+ },
517
+ {
518
+ "epoch": 1.35,
519
+ "learning_rate": 0.00014598070739549841,
520
+ "loss": 0.1083,
521
+ "step": 840
522
+ },
523
+ {
524
+ "epoch": 1.37,
525
+ "learning_rate": 0.00014533762057877814,
526
+ "loss": 0.1207,
527
+ "step": 850
528
+ },
529
+ {
530
+ "epoch": 1.38,
531
+ "learning_rate": 0.0001446945337620579,
532
+ "loss": 0.0672,
533
+ "step": 860
534
+ },
535
+ {
536
+ "epoch": 1.4,
537
+ "learning_rate": 0.00014405144694533764,
538
+ "loss": 0.0978,
539
+ "step": 870
540
+ },
541
+ {
542
+ "epoch": 1.41,
543
+ "learning_rate": 0.0001434083601286174,
544
+ "loss": 0.1004,
545
+ "step": 880
546
+ },
547
+ {
548
+ "epoch": 1.43,
549
+ "learning_rate": 0.00014276527331189711,
550
+ "loss": 0.0673,
551
+ "step": 890
552
+ },
553
+ {
554
+ "epoch": 1.45,
555
+ "learning_rate": 0.00014212218649517686,
556
+ "loss": 0.0641,
557
+ "step": 900
558
+ },
559
+ {
560
+ "epoch": 1.46,
561
+ "learning_rate": 0.00014147909967845662,
562
+ "loss": 0.045,
563
+ "step": 910
564
+ },
565
+ {
566
+ "epoch": 1.48,
567
+ "learning_rate": 0.00014083601286173634,
568
+ "loss": 0.0643,
569
+ "step": 920
570
+ },
571
+ {
572
+ "epoch": 1.5,
573
+ "learning_rate": 0.0001401929260450161,
574
+ "loss": 0.0643,
575
+ "step": 930
576
+ },
577
+ {
578
+ "epoch": 1.51,
579
+ "learning_rate": 0.00013954983922829584,
580
+ "loss": 0.1317,
581
+ "step": 940
582
+ },
583
+ {
584
+ "epoch": 1.53,
585
+ "learning_rate": 0.0001389067524115756,
586
+ "loss": 0.0433,
587
+ "step": 950
588
+ },
589
+ {
590
+ "epoch": 1.54,
591
+ "learning_rate": 0.00013826366559485531,
592
+ "loss": 0.0862,
593
+ "step": 960
594
+ },
595
+ {
596
+ "epoch": 1.56,
597
+ "learning_rate": 0.00013762057877813507,
598
+ "loss": 0.1031,
599
+ "step": 970
600
+ },
601
+ {
602
+ "epoch": 1.58,
603
+ "learning_rate": 0.00013697749196141482,
604
+ "loss": 0.0849,
605
+ "step": 980
606
+ },
607
+ {
608
+ "epoch": 1.59,
609
+ "learning_rate": 0.00013633440514469454,
610
+ "loss": 0.0924,
611
+ "step": 990
612
+ },
613
+ {
614
+ "epoch": 1.61,
615
+ "learning_rate": 0.0001356913183279743,
616
+ "loss": 0.0773,
617
+ "step": 1000
618
+ },
619
+ {
620
+ "epoch": 1.62,
621
+ "learning_rate": 0.00013504823151125404,
622
+ "loss": 0.0823,
623
+ "step": 1010
624
+ },
625
+ {
626
+ "epoch": 1.64,
627
+ "learning_rate": 0.00013440514469453376,
628
+ "loss": 0.0473,
629
+ "step": 1020
630
+ },
631
+ {
632
+ "epoch": 1.66,
633
+ "learning_rate": 0.00013376205787781351,
634
+ "loss": 0.1218,
635
+ "step": 1030
636
+ },
637
+ {
638
+ "epoch": 1.67,
639
+ "learning_rate": 0.00013311897106109327,
640
+ "loss": 0.0687,
641
+ "step": 1040
642
+ },
643
+ {
644
+ "epoch": 1.69,
645
+ "learning_rate": 0.00013247588424437302,
646
+ "loss": 0.0692,
647
+ "step": 1050
648
+ },
649
+ {
650
+ "epoch": 1.7,
651
+ "learning_rate": 0.00013183279742765274,
652
+ "loss": 0.0502,
653
+ "step": 1060
654
+ },
655
+ {
656
+ "epoch": 1.72,
657
+ "learning_rate": 0.0001311897106109325,
658
+ "loss": 0.0593,
659
+ "step": 1070
660
+ },
661
+ {
662
+ "epoch": 1.74,
663
+ "learning_rate": 0.00013054662379421224,
664
+ "loss": 0.0846,
665
+ "step": 1080
666
+ },
667
+ {
668
+ "epoch": 1.75,
669
+ "learning_rate": 0.00012990353697749196,
670
+ "loss": 0.063,
671
+ "step": 1090
672
+ },
673
+ {
674
+ "epoch": 1.77,
675
+ "learning_rate": 0.00012926045016077172,
676
+ "loss": 0.0941,
677
+ "step": 1100
678
+ },
679
+ {
680
+ "epoch": 1.78,
681
+ "learning_rate": 0.00012861736334405147,
682
+ "loss": 0.0884,
683
+ "step": 1110
684
+ },
685
+ {
686
+ "epoch": 1.8,
687
+ "learning_rate": 0.00012797427652733122,
688
+ "loss": 0.0647,
689
+ "step": 1120
690
+ },
691
+ {
692
+ "epoch": 1.82,
693
+ "learning_rate": 0.00012733118971061094,
694
+ "loss": 0.1016,
695
+ "step": 1130
696
+ },
697
+ {
698
+ "epoch": 1.83,
699
+ "learning_rate": 0.0001266881028938907,
700
+ "loss": 0.0821,
701
+ "step": 1140
702
+ },
703
+ {
704
+ "epoch": 1.85,
705
+ "learning_rate": 0.00012604501607717044,
706
+ "loss": 0.0653,
707
+ "step": 1150
708
+ },
709
+ {
710
+ "epoch": 1.86,
711
+ "learning_rate": 0.00012540192926045017,
712
+ "loss": 0.103,
713
+ "step": 1160
714
+ },
715
+ {
716
+ "epoch": 1.88,
717
+ "learning_rate": 0.00012475884244372992,
718
+ "loss": 0.1374,
719
+ "step": 1170
720
+ },
721
+ {
722
+ "epoch": 1.9,
723
+ "learning_rate": 0.00012411575562700967,
724
+ "loss": 0.0682,
725
+ "step": 1180
726
+ },
727
+ {
728
+ "epoch": 1.91,
729
+ "learning_rate": 0.00012347266881028942,
730
+ "loss": 0.0811,
731
+ "step": 1190
732
+ },
733
+ {
734
+ "epoch": 1.93,
735
+ "learning_rate": 0.00012282958199356914,
736
+ "loss": 0.0448,
737
+ "step": 1200
738
+ },
739
+ {
740
+ "epoch": 1.95,
741
+ "learning_rate": 0.0001221864951768489,
742
+ "loss": 0.1038,
743
+ "step": 1210
744
+ },
745
+ {
746
+ "epoch": 1.96,
747
+ "learning_rate": 0.00012154340836012863,
748
+ "loss": 0.0777,
749
+ "step": 1220
750
+ },
751
+ {
752
+ "epoch": 1.98,
753
+ "learning_rate": 0.00012090032154340835,
754
+ "loss": 0.0588,
755
+ "step": 1230
756
+ },
757
+ {
758
+ "epoch": 1.99,
759
+ "learning_rate": 0.0001202572347266881,
760
+ "loss": 0.057,
761
+ "step": 1240
762
+ },
763
+ {
764
+ "epoch": 2.0,
765
+ "eval_accuracy": 0.9883257403189066,
766
+ "eval_loss": 0.03622407466173172,
767
+ "eval_runtime": 25.825,
768
+ "eval_samples_per_second": 135.992,
769
+ "eval_steps_per_second": 4.259,
770
+ "step": 1244
771
+ },
772
+ {
773
+ "epoch": 2.01,
774
+ "learning_rate": 0.00011961414790996785,
775
+ "loss": 0.0314,
776
+ "step": 1250
777
+ },
778
+ {
779
+ "epoch": 2.03,
780
+ "learning_rate": 0.0001189710610932476,
781
+ "loss": 0.0833,
782
+ "step": 1260
783
+ },
784
+ {
785
+ "epoch": 2.04,
786
+ "learning_rate": 0.00011832797427652733,
787
+ "loss": 0.0474,
788
+ "step": 1270
789
+ },
790
+ {
791
+ "epoch": 2.06,
792
+ "learning_rate": 0.00011768488745980708,
793
+ "loss": 0.0801,
794
+ "step": 1280
795
+ },
796
+ {
797
+ "epoch": 2.07,
798
+ "learning_rate": 0.00011704180064308683,
799
+ "loss": 0.0721,
800
+ "step": 1290
801
+ },
802
+ {
803
+ "epoch": 2.09,
804
+ "learning_rate": 0.00011639871382636655,
805
+ "loss": 0.0558,
806
+ "step": 1300
807
+ },
808
+ {
809
+ "epoch": 2.11,
810
+ "learning_rate": 0.0001157556270096463,
811
+ "loss": 0.0731,
812
+ "step": 1310
813
+ },
814
+ {
815
+ "epoch": 2.12,
816
+ "learning_rate": 0.00011511254019292605,
817
+ "loss": 0.0449,
818
+ "step": 1320
819
+ },
820
+ {
821
+ "epoch": 2.14,
822
+ "learning_rate": 0.0001144694533762058,
823
+ "loss": 0.0765,
824
+ "step": 1330
825
+ },
826
+ {
827
+ "epoch": 2.15,
828
+ "learning_rate": 0.00011382636655948553,
829
+ "loss": 0.0505,
830
+ "step": 1340
831
+ },
832
+ {
833
+ "epoch": 2.17,
834
+ "learning_rate": 0.00011318327974276528,
835
+ "loss": 0.0473,
836
+ "step": 1350
837
+ },
838
+ {
839
+ "epoch": 2.19,
840
+ "learning_rate": 0.00011254019292604503,
841
+ "loss": 0.0544,
842
+ "step": 1360
843
+ },
844
+ {
845
+ "epoch": 2.2,
846
+ "learning_rate": 0.00011189710610932475,
847
+ "loss": 0.0928,
848
+ "step": 1370
849
+ },
850
+ {
851
+ "epoch": 2.22,
852
+ "learning_rate": 0.0001112540192926045,
853
+ "loss": 0.0518,
854
+ "step": 1380
855
+ },
856
+ {
857
+ "epoch": 2.23,
858
+ "learning_rate": 0.00011061093247588425,
859
+ "loss": 0.1067,
860
+ "step": 1390
861
+ },
862
+ {
863
+ "epoch": 2.25,
864
+ "learning_rate": 0.00010996784565916398,
865
+ "loss": 0.0802,
866
+ "step": 1400
867
+ },
868
+ {
869
+ "epoch": 2.27,
870
+ "learning_rate": 0.00010932475884244373,
871
+ "loss": 0.0658,
872
+ "step": 1410
873
+ },
874
+ {
875
+ "epoch": 2.28,
876
+ "learning_rate": 0.00010868167202572348,
877
+ "loss": 0.0342,
878
+ "step": 1420
879
+ },
880
+ {
881
+ "epoch": 2.3,
882
+ "learning_rate": 0.00010803858520900323,
883
+ "loss": 0.0765,
884
+ "step": 1430
885
+ },
886
+ {
887
+ "epoch": 2.32,
888
+ "learning_rate": 0.00010739549839228295,
889
+ "loss": 0.0315,
890
+ "step": 1440
891
+ },
892
+ {
893
+ "epoch": 2.33,
894
+ "learning_rate": 0.0001067524115755627,
895
+ "loss": 0.0399,
896
+ "step": 1450
897
+ },
898
+ {
899
+ "epoch": 2.35,
900
+ "learning_rate": 0.00010610932475884245,
901
+ "loss": 0.092,
902
+ "step": 1460
903
+ },
904
+ {
905
+ "epoch": 2.36,
906
+ "learning_rate": 0.00010546623794212218,
907
+ "loss": 0.0543,
908
+ "step": 1470
909
+ },
910
+ {
911
+ "epoch": 2.38,
912
+ "learning_rate": 0.00010482315112540193,
913
+ "loss": 0.0485,
914
+ "step": 1480
915
+ },
916
+ {
917
+ "epoch": 2.4,
918
+ "learning_rate": 0.00010418006430868168,
919
+ "loss": 0.0629,
920
+ "step": 1490
921
+ },
922
+ {
923
+ "epoch": 2.41,
924
+ "learning_rate": 0.00010353697749196143,
925
+ "loss": 0.0724,
926
+ "step": 1500
927
+ },
928
+ {
929
+ "epoch": 2.43,
930
+ "learning_rate": 0.00010289389067524115,
931
+ "loss": 0.0591,
932
+ "step": 1510
933
+ },
934
+ {
935
+ "epoch": 2.44,
936
+ "learning_rate": 0.0001022508038585209,
937
+ "loss": 0.0704,
938
+ "step": 1520
939
+ },
940
+ {
941
+ "epoch": 2.46,
942
+ "learning_rate": 0.00010160771704180066,
943
+ "loss": 0.1163,
944
+ "step": 1530
945
+ },
946
+ {
947
+ "epoch": 2.48,
948
+ "learning_rate": 0.00010096463022508038,
949
+ "loss": 0.0403,
950
+ "step": 1540
951
+ },
952
+ {
953
+ "epoch": 2.49,
954
+ "learning_rate": 0.00010032154340836013,
955
+ "loss": 0.0641,
956
+ "step": 1550
957
+ },
958
+ {
959
+ "epoch": 2.51,
960
+ "learning_rate": 9.967845659163988e-05,
961
+ "loss": 0.0684,
962
+ "step": 1560
963
+ },
964
+ {
965
+ "epoch": 2.52,
966
+ "learning_rate": 9.903536977491962e-05,
967
+ "loss": 0.0534,
968
+ "step": 1570
969
+ },
970
+ {
971
+ "epoch": 2.54,
972
+ "learning_rate": 9.839228295819937e-05,
973
+ "loss": 0.0922,
974
+ "step": 1580
975
+ },
976
+ {
977
+ "epoch": 2.56,
978
+ "learning_rate": 9.77491961414791e-05,
979
+ "loss": 0.0368,
980
+ "step": 1590
981
+ },
982
+ {
983
+ "epoch": 2.57,
984
+ "learning_rate": 9.710610932475884e-05,
985
+ "loss": 0.0348,
986
+ "step": 1600
987
+ },
988
+ {
989
+ "epoch": 2.59,
990
+ "learning_rate": 9.646302250803859e-05,
991
+ "loss": 0.0355,
992
+ "step": 1610
993
+ },
994
+ {
995
+ "epoch": 2.6,
996
+ "learning_rate": 9.581993569131833e-05,
997
+ "loss": 0.0416,
998
+ "step": 1620
999
+ },
1000
+ {
1001
+ "epoch": 2.62,
1002
+ "learning_rate": 9.517684887459808e-05,
1003
+ "loss": 0.0874,
1004
+ "step": 1630
1005
+ },
1006
+ {
1007
+ "epoch": 2.64,
1008
+ "learning_rate": 9.453376205787782e-05,
1009
+ "loss": 0.0913,
1010
+ "step": 1640
1011
+ },
1012
+ {
1013
+ "epoch": 2.65,
1014
+ "learning_rate": 9.389067524115757e-05,
1015
+ "loss": 0.0588,
1016
+ "step": 1650
1017
+ },
1018
+ {
1019
+ "epoch": 2.67,
1020
+ "learning_rate": 9.32475884244373e-05,
1021
+ "loss": 0.0639,
1022
+ "step": 1660
1023
+ },
1024
+ {
1025
+ "epoch": 2.68,
1026
+ "learning_rate": 9.260450160771704e-05,
1027
+ "loss": 0.0471,
1028
+ "step": 1670
1029
+ },
1030
+ {
1031
+ "epoch": 2.7,
1032
+ "learning_rate": 9.19614147909968e-05,
1033
+ "loss": 0.0636,
1034
+ "step": 1680
1035
+ },
1036
+ {
1037
+ "epoch": 2.72,
1038
+ "learning_rate": 9.131832797427653e-05,
1039
+ "loss": 0.12,
1040
+ "step": 1690
1041
+ },
1042
+ {
1043
+ "epoch": 2.73,
1044
+ "learning_rate": 9.067524115755628e-05,
1045
+ "loss": 0.04,
1046
+ "step": 1700
1047
+ },
1048
+ {
1049
+ "epoch": 2.75,
1050
+ "learning_rate": 9.003215434083602e-05,
1051
+ "loss": 0.0316,
1052
+ "step": 1710
1053
+ },
1054
+ {
1055
+ "epoch": 2.77,
1056
+ "learning_rate": 8.938906752411576e-05,
1057
+ "loss": 0.0506,
1058
+ "step": 1720
1059
+ },
1060
+ {
1061
+ "epoch": 2.78,
1062
+ "learning_rate": 8.87459807073955e-05,
1063
+ "loss": 0.0425,
1064
+ "step": 1730
1065
+ },
1066
+ {
1067
+ "epoch": 2.8,
1068
+ "learning_rate": 8.810289389067524e-05,
1069
+ "loss": 0.0131,
1070
+ "step": 1740
1071
+ },
1072
+ {
1073
+ "epoch": 2.81,
1074
+ "learning_rate": 8.7459807073955e-05,
1075
+ "loss": 0.0486,
1076
+ "step": 1750
1077
+ },
1078
+ {
1079
+ "epoch": 2.83,
1080
+ "learning_rate": 8.681672025723473e-05,
1081
+ "loss": 0.042,
1082
+ "step": 1760
1083
+ },
1084
+ {
1085
+ "epoch": 2.85,
1086
+ "learning_rate": 8.617363344051448e-05,
1087
+ "loss": 0.0524,
1088
+ "step": 1770
1089
+ },
1090
+ {
1091
+ "epoch": 2.86,
1092
+ "learning_rate": 8.553054662379422e-05,
1093
+ "loss": 0.0415,
1094
+ "step": 1780
1095
+ },
1096
+ {
1097
+ "epoch": 2.88,
1098
+ "learning_rate": 8.488745980707396e-05,
1099
+ "loss": 0.0543,
1100
+ "step": 1790
1101
+ },
1102
+ {
1103
+ "epoch": 2.89,
1104
+ "learning_rate": 8.42443729903537e-05,
1105
+ "loss": 0.0679,
1106
+ "step": 1800
1107
+ },
1108
+ {
1109
+ "epoch": 2.91,
1110
+ "learning_rate": 8.360128617363344e-05,
1111
+ "loss": 0.1043,
1112
+ "step": 1810
1113
+ },
1114
+ {
1115
+ "epoch": 2.93,
1116
+ "learning_rate": 8.29581993569132e-05,
1117
+ "loss": 0.0508,
1118
+ "step": 1820
1119
+ },
1120
+ {
1121
+ "epoch": 2.94,
1122
+ "learning_rate": 8.231511254019293e-05,
1123
+ "loss": 0.0701,
1124
+ "step": 1830
1125
+ },
1126
+ {
1127
+ "epoch": 2.96,
1128
+ "learning_rate": 8.167202572347268e-05,
1129
+ "loss": 0.095,
1130
+ "step": 1840
1131
+ },
1132
+ {
1133
+ "epoch": 2.97,
1134
+ "learning_rate": 8.102893890675242e-05,
1135
+ "loss": 0.0853,
1136
+ "step": 1850
1137
+ },
1138
+ {
1139
+ "epoch": 2.99,
1140
+ "learning_rate": 8.038585209003216e-05,
1141
+ "loss": 0.0449,
1142
+ "step": 1860
1143
+ },
1144
+ {
1145
+ "epoch": 3.0,
1146
+ "eval_accuracy": 0.9886104783599089,
1147
+ "eval_loss": 0.02609114721417427,
1148
+ "eval_runtime": 25.2517,
1149
+ "eval_samples_per_second": 139.08,
1150
+ "eval_steps_per_second": 4.356,
1151
+ "step": 1866
1152
+ },
1153
+ {
1154
+ "epoch": 3.01,
1155
+ "learning_rate": 7.974276527331191e-05,
1156
+ "loss": 0.0794,
1157
+ "step": 1870
1158
+ },
1159
+ {
1160
+ "epoch": 3.02,
1161
+ "learning_rate": 7.909967845659164e-05,
1162
+ "loss": 0.03,
1163
+ "step": 1880
1164
+ },
1165
+ {
1166
+ "epoch": 3.04,
1167
+ "learning_rate": 7.84565916398714e-05,
1168
+ "loss": 0.0552,
1169
+ "step": 1890
1170
+ },
1171
+ {
1172
+ "epoch": 3.05,
1173
+ "learning_rate": 7.781350482315113e-05,
1174
+ "loss": 0.0577,
1175
+ "step": 1900
1176
+ },
1177
+ {
1178
+ "epoch": 3.07,
1179
+ "learning_rate": 7.717041800643087e-05,
1180
+ "loss": 0.054,
1181
+ "step": 1910
1182
+ },
1183
+ {
1184
+ "epoch": 3.09,
1185
+ "learning_rate": 7.652733118971062e-05,
1186
+ "loss": 0.0141,
1187
+ "step": 1920
1188
+ },
1189
+ {
1190
+ "epoch": 3.1,
1191
+ "learning_rate": 7.588424437299036e-05,
1192
+ "loss": 0.0579,
1193
+ "step": 1930
1194
+ },
1195
+ {
1196
+ "epoch": 3.12,
1197
+ "learning_rate": 7.524115755627011e-05,
1198
+ "loss": 0.0463,
1199
+ "step": 1940
1200
+ },
1201
+ {
1202
+ "epoch": 3.14,
1203
+ "learning_rate": 7.459807073954984e-05,
1204
+ "loss": 0.0389,
1205
+ "step": 1950
1206
+ },
1207
+ {
1208
+ "epoch": 3.15,
1209
+ "learning_rate": 7.39549839228296e-05,
1210
+ "loss": 0.0487,
1211
+ "step": 1960
1212
+ },
1213
+ {
1214
+ "epoch": 3.17,
1215
+ "learning_rate": 7.331189710610933e-05,
1216
+ "loss": 0.0246,
1217
+ "step": 1970
1218
+ },
1219
+ {
1220
+ "epoch": 3.18,
1221
+ "learning_rate": 7.266881028938907e-05,
1222
+ "loss": 0.0718,
1223
+ "step": 1980
1224
+ },
1225
+ {
1226
+ "epoch": 3.2,
1227
+ "learning_rate": 7.202572347266882e-05,
1228
+ "loss": 0.1136,
1229
+ "step": 1990
1230
+ },
1231
+ {
1232
+ "epoch": 3.22,
1233
+ "learning_rate": 7.138263665594856e-05,
1234
+ "loss": 0.0631,
1235
+ "step": 2000
1236
+ },
1237
+ {
1238
+ "epoch": 3.23,
1239
+ "learning_rate": 7.073954983922831e-05,
1240
+ "loss": 0.0363,
1241
+ "step": 2010
1242
+ },
1243
+ {
1244
+ "epoch": 3.25,
1245
+ "learning_rate": 7.009646302250804e-05,
1246
+ "loss": 0.075,
1247
+ "step": 2020
1248
+ },
1249
+ {
1250
+ "epoch": 3.26,
1251
+ "learning_rate": 6.94533762057878e-05,
1252
+ "loss": 0.0735,
1253
+ "step": 2030
1254
+ },
1255
+ {
1256
+ "epoch": 3.28,
1257
+ "learning_rate": 6.881028938906753e-05,
1258
+ "loss": 0.0461,
1259
+ "step": 2040
1260
+ },
1261
+ {
1262
+ "epoch": 3.3,
1263
+ "learning_rate": 6.816720257234727e-05,
1264
+ "loss": 0.0299,
1265
+ "step": 2050
1266
+ },
1267
+ {
1268
+ "epoch": 3.31,
1269
+ "learning_rate": 6.752411575562702e-05,
1270
+ "loss": 0.0526,
1271
+ "step": 2060
1272
+ },
1273
+ {
1274
+ "epoch": 3.33,
1275
+ "learning_rate": 6.688102893890676e-05,
1276
+ "loss": 0.0355,
1277
+ "step": 2070
1278
+ },
1279
+ {
1280
+ "epoch": 3.34,
1281
+ "learning_rate": 6.623794212218651e-05,
1282
+ "loss": 0.0244,
1283
+ "step": 2080
1284
+ },
1285
+ {
1286
+ "epoch": 3.36,
1287
+ "learning_rate": 6.559485530546625e-05,
1288
+ "loss": 0.0637,
1289
+ "step": 2090
1290
+ },
1291
+ {
1292
+ "epoch": 3.38,
1293
+ "learning_rate": 6.495176848874598e-05,
1294
+ "loss": 0.0595,
1295
+ "step": 2100
1296
+ },
1297
+ {
1298
+ "epoch": 3.39,
1299
+ "learning_rate": 6.430868167202573e-05,
1300
+ "loss": 0.0506,
1301
+ "step": 2110
1302
+ },
1303
+ {
1304
+ "epoch": 3.41,
1305
+ "learning_rate": 6.366559485530547e-05,
1306
+ "loss": 0.0212,
1307
+ "step": 2120
1308
+ },
1309
+ {
1310
+ "epoch": 3.42,
1311
+ "learning_rate": 6.302250803858522e-05,
1312
+ "loss": 0.0255,
1313
+ "step": 2130
1314
+ },
1315
+ {
1316
+ "epoch": 3.44,
1317
+ "learning_rate": 6.237942122186496e-05,
1318
+ "loss": 0.0224,
1319
+ "step": 2140
1320
+ },
1321
+ {
1322
+ "epoch": 3.46,
1323
+ "learning_rate": 6.173633440514471e-05,
1324
+ "loss": 0.0633,
1325
+ "step": 2150
1326
+ },
1327
+ {
1328
+ "epoch": 3.47,
1329
+ "learning_rate": 6.109324758842445e-05,
1330
+ "loss": 0.0554,
1331
+ "step": 2160
1332
+ },
1333
+ {
1334
+ "epoch": 3.49,
1335
+ "learning_rate": 6.0450160771704176e-05,
1336
+ "loss": 0.0653,
1337
+ "step": 2170
1338
+ },
1339
+ {
1340
+ "epoch": 3.5,
1341
+ "learning_rate": 5.980707395498393e-05,
1342
+ "loss": 0.02,
1343
+ "step": 2180
1344
+ },
1345
+ {
1346
+ "epoch": 3.52,
1347
+ "learning_rate": 5.9163987138263664e-05,
1348
+ "loss": 0.0668,
1349
+ "step": 2190
1350
+ },
1351
+ {
1352
+ "epoch": 3.54,
1353
+ "learning_rate": 5.8520900321543414e-05,
1354
+ "loss": 0.027,
1355
+ "step": 2200
1356
+ },
1357
+ {
1358
+ "epoch": 3.55,
1359
+ "learning_rate": 5.787781350482315e-05,
1360
+ "loss": 0.0567,
1361
+ "step": 2210
1362
+ },
1363
+ {
1364
+ "epoch": 3.57,
1365
+ "learning_rate": 5.72347266881029e-05,
1366
+ "loss": 0.0434,
1367
+ "step": 2220
1368
+ },
1369
+ {
1370
+ "epoch": 3.59,
1371
+ "learning_rate": 5.659163987138264e-05,
1372
+ "loss": 0.0457,
1373
+ "step": 2230
1374
+ },
1375
+ {
1376
+ "epoch": 3.6,
1377
+ "learning_rate": 5.5948553054662377e-05,
1378
+ "loss": 0.0263,
1379
+ "step": 2240
1380
+ },
1381
+ {
1382
+ "epoch": 3.62,
1383
+ "learning_rate": 5.530546623794213e-05,
1384
+ "loss": 0.0426,
1385
+ "step": 2250
1386
+ },
1387
+ {
1388
+ "epoch": 3.63,
1389
+ "learning_rate": 5.4662379421221864e-05,
1390
+ "loss": 0.0397,
1391
+ "step": 2260
1392
+ },
1393
+ {
1394
+ "epoch": 3.65,
1395
+ "learning_rate": 5.4019292604501615e-05,
1396
+ "loss": 0.0393,
1397
+ "step": 2270
1398
+ },
1399
+ {
1400
+ "epoch": 3.67,
1401
+ "learning_rate": 5.337620578778135e-05,
1402
+ "loss": 0.0453,
1403
+ "step": 2280
1404
+ },
1405
+ {
1406
+ "epoch": 3.68,
1407
+ "learning_rate": 5.273311897106109e-05,
1408
+ "loss": 0.064,
1409
+ "step": 2290
1410
+ },
1411
+ {
1412
+ "epoch": 3.7,
1413
+ "learning_rate": 5.209003215434084e-05,
1414
+ "loss": 0.0587,
1415
+ "step": 2300
1416
+ },
1417
+ {
1418
+ "epoch": 3.71,
1419
+ "learning_rate": 5.144694533762058e-05,
1420
+ "loss": 0.0478,
1421
+ "step": 2310
1422
+ },
1423
+ {
1424
+ "epoch": 3.73,
1425
+ "learning_rate": 5.080385852090033e-05,
1426
+ "loss": 0.038,
1427
+ "step": 2320
1428
+ },
1429
+ {
1430
+ "epoch": 3.75,
1431
+ "learning_rate": 5.0160771704180065e-05,
1432
+ "loss": 0.0292,
1433
+ "step": 2330
1434
+ },
1435
+ {
1436
+ "epoch": 3.76,
1437
+ "learning_rate": 4.951768488745981e-05,
1438
+ "loss": 0.0412,
1439
+ "step": 2340
1440
+ },
1441
+ {
1442
+ "epoch": 3.78,
1443
+ "learning_rate": 4.887459807073955e-05,
1444
+ "loss": 0.0404,
1445
+ "step": 2350
1446
+ },
1447
+ {
1448
+ "epoch": 3.79,
1449
+ "learning_rate": 4.8231511254019296e-05,
1450
+ "loss": 0.0485,
1451
+ "step": 2360
1452
+ },
1453
+ {
1454
+ "epoch": 3.81,
1455
+ "learning_rate": 4.758842443729904e-05,
1456
+ "loss": 0.0373,
1457
+ "step": 2370
1458
+ },
1459
+ {
1460
+ "epoch": 3.83,
1461
+ "learning_rate": 4.6945337620578784e-05,
1462
+ "loss": 0.0291,
1463
+ "step": 2380
1464
+ },
1465
+ {
1466
+ "epoch": 3.84,
1467
+ "learning_rate": 4.630225080385852e-05,
1468
+ "loss": 0.0588,
1469
+ "step": 2390
1470
+ },
1471
+ {
1472
+ "epoch": 3.86,
1473
+ "learning_rate": 4.5659163987138265e-05,
1474
+ "loss": 0.0631,
1475
+ "step": 2400
1476
+ },
1477
+ {
1478
+ "epoch": 3.87,
1479
+ "learning_rate": 4.501607717041801e-05,
1480
+ "loss": 0.0338,
1481
+ "step": 2410
1482
+ },
1483
+ {
1484
+ "epoch": 3.89,
1485
+ "learning_rate": 4.437299035369775e-05,
1486
+ "loss": 0.1125,
1487
+ "step": 2420
1488
+ },
1489
+ {
1490
+ "epoch": 3.91,
1491
+ "learning_rate": 4.37299035369775e-05,
1492
+ "loss": 0.026,
1493
+ "step": 2430
1494
+ },
1495
+ {
1496
+ "epoch": 3.92,
1497
+ "learning_rate": 4.308681672025724e-05,
1498
+ "loss": 0.0389,
1499
+ "step": 2440
1500
+ },
1501
+ {
1502
+ "epoch": 3.94,
1503
+ "learning_rate": 4.244372990353698e-05,
1504
+ "loss": 0.0567,
1505
+ "step": 2450
1506
+ },
1507
+ {
1508
+ "epoch": 3.95,
1509
+ "learning_rate": 4.180064308681672e-05,
1510
+ "loss": 0.023,
1511
+ "step": 2460
1512
+ },
1513
+ {
1514
+ "epoch": 3.97,
1515
+ "learning_rate": 4.1157556270096466e-05,
1516
+ "loss": 0.0353,
1517
+ "step": 2470
1518
+ },
1519
+ {
1520
+ "epoch": 3.99,
1521
+ "learning_rate": 4.051446945337621e-05,
1522
+ "loss": 0.066,
1523
+ "step": 2480
1524
+ },
1525
+ {
1526
+ "epoch": 4.0,
1527
+ "eval_accuracy": 0.9923120728929385,
1528
+ "eval_loss": 0.02482709474861622,
1529
+ "eval_runtime": 25.9153,
1530
+ "eval_samples_per_second": 135.518,
1531
+ "eval_steps_per_second": 4.245,
1532
+ "step": 2488
1533
+ },
1534
+ {
1535
+ "epoch": 4.0,
1536
+ "learning_rate": 3.9871382636655953e-05,
1537
+ "loss": 0.055,
1538
+ "step": 2490
1539
+ },
1540
+ {
1541
+ "epoch": 4.02,
1542
+ "learning_rate": 3.92282958199357e-05,
1543
+ "loss": 0.0344,
1544
+ "step": 2500
1545
+ },
1546
+ {
1547
+ "epoch": 4.04,
1548
+ "learning_rate": 3.8585209003215434e-05,
1549
+ "loss": 0.0635,
1550
+ "step": 2510
1551
+ },
1552
+ {
1553
+ "epoch": 4.05,
1554
+ "learning_rate": 3.794212218649518e-05,
1555
+ "loss": 0.0198,
1556
+ "step": 2520
1557
+ },
1558
+ {
1559
+ "epoch": 4.07,
1560
+ "learning_rate": 3.729903536977492e-05,
1561
+ "loss": 0.0268,
1562
+ "step": 2530
1563
+ },
1564
+ {
1565
+ "epoch": 4.08,
1566
+ "learning_rate": 3.6655948553054666e-05,
1567
+ "loss": 0.0193,
1568
+ "step": 2540
1569
+ },
1570
+ {
1571
+ "epoch": 4.1,
1572
+ "learning_rate": 3.601286173633441e-05,
1573
+ "loss": 0.0285,
1574
+ "step": 2550
1575
+ },
1576
+ {
1577
+ "epoch": 4.12,
1578
+ "learning_rate": 3.5369774919614154e-05,
1579
+ "loss": 0.0128,
1580
+ "step": 2560
1581
+ },
1582
+ {
1583
+ "epoch": 4.13,
1584
+ "learning_rate": 3.47266881028939e-05,
1585
+ "loss": 0.0252,
1586
+ "step": 2570
1587
+ },
1588
+ {
1589
+ "epoch": 4.15,
1590
+ "learning_rate": 3.4083601286173635e-05,
1591
+ "loss": 0.0078,
1592
+ "step": 2580
1593
+ },
1594
+ {
1595
+ "epoch": 4.16,
1596
+ "learning_rate": 3.344051446945338e-05,
1597
+ "loss": 0.0494,
1598
+ "step": 2590
1599
+ },
1600
+ {
1601
+ "epoch": 4.18,
1602
+ "learning_rate": 3.279742765273312e-05,
1603
+ "loss": 0.0459,
1604
+ "step": 2600
1605
+ },
1606
+ {
1607
+ "epoch": 4.2,
1608
+ "learning_rate": 3.2154340836012867e-05,
1609
+ "loss": 0.0284,
1610
+ "step": 2610
1611
+ },
1612
+ {
1613
+ "epoch": 4.21,
1614
+ "learning_rate": 3.151125401929261e-05,
1615
+ "loss": 0.0423,
1616
+ "step": 2620
1617
+ },
1618
+ {
1619
+ "epoch": 4.23,
1620
+ "learning_rate": 3.0868167202572354e-05,
1621
+ "loss": 0.0094,
1622
+ "step": 2630
1623
+ },
1624
+ {
1625
+ "epoch": 4.24,
1626
+ "learning_rate": 3.0225080385852088e-05,
1627
+ "loss": 0.0371,
1628
+ "step": 2640
1629
+ },
1630
+ {
1631
+ "epoch": 4.26,
1632
+ "learning_rate": 2.9581993569131832e-05,
1633
+ "loss": 0.0241,
1634
+ "step": 2650
1635
+ },
1636
+ {
1637
+ "epoch": 4.28,
1638
+ "learning_rate": 2.8938906752411576e-05,
1639
+ "loss": 0.0497,
1640
+ "step": 2660
1641
+ },
1642
+ {
1643
+ "epoch": 4.29,
1644
+ "learning_rate": 2.829581993569132e-05,
1645
+ "loss": 0.026,
1646
+ "step": 2670
1647
+ },
1648
+ {
1649
+ "epoch": 4.31,
1650
+ "learning_rate": 2.7652733118971064e-05,
1651
+ "loss": 0.0638,
1652
+ "step": 2680
1653
+ },
1654
+ {
1655
+ "epoch": 4.32,
1656
+ "learning_rate": 2.7009646302250807e-05,
1657
+ "loss": 0.0201,
1658
+ "step": 2690
1659
+ },
1660
+ {
1661
+ "epoch": 4.34,
1662
+ "learning_rate": 2.6366559485530545e-05,
1663
+ "loss": 0.0348,
1664
+ "step": 2700
1665
+ },
1666
+ {
1667
+ "epoch": 4.36,
1668
+ "learning_rate": 2.5787781350482317e-05,
1669
+ "loss": 0.043,
1670
+ "step": 2710
1671
+ },
1672
+ {
1673
+ "epoch": 4.37,
1674
+ "learning_rate": 2.5144694533762057e-05,
1675
+ "loss": 0.0371,
1676
+ "step": 2720
1677
+ },
1678
+ {
1679
+ "epoch": 4.39,
1680
+ "learning_rate": 2.45016077170418e-05,
1681
+ "loss": 0.0164,
1682
+ "step": 2730
1683
+ },
1684
+ {
1685
+ "epoch": 4.41,
1686
+ "learning_rate": 2.385852090032154e-05,
1687
+ "loss": 0.0359,
1688
+ "step": 2740
1689
+ },
1690
+ {
1691
+ "epoch": 4.42,
1692
+ "learning_rate": 2.3215434083601286e-05,
1693
+ "loss": 0.0556,
1694
+ "step": 2750
1695
+ },
1696
+ {
1697
+ "epoch": 4.44,
1698
+ "learning_rate": 2.257234726688103e-05,
1699
+ "loss": 0.0247,
1700
+ "step": 2760
1701
+ },
1702
+ {
1703
+ "epoch": 4.45,
1704
+ "learning_rate": 2.1929260450160773e-05,
1705
+ "loss": 0.0477,
1706
+ "step": 2770
1707
+ },
1708
+ {
1709
+ "epoch": 4.47,
1710
+ "learning_rate": 2.1286173633440514e-05,
1711
+ "loss": 0.0499,
1712
+ "step": 2780
1713
+ },
1714
+ {
1715
+ "epoch": 4.49,
1716
+ "learning_rate": 2.0643086816720258e-05,
1717
+ "loss": 0.0736,
1718
+ "step": 2790
1719
+ },
1720
+ {
1721
+ "epoch": 4.5,
1722
+ "learning_rate": 2e-05,
1723
+ "loss": 0.0235,
1724
+ "step": 2800
1725
+ },
1726
+ {
1727
+ "epoch": 4.52,
1728
+ "learning_rate": 1.9356913183279742e-05,
1729
+ "loss": 0.0368,
1730
+ "step": 2810
1731
+ },
1732
+ {
1733
+ "epoch": 4.53,
1734
+ "learning_rate": 1.8713826366559486e-05,
1735
+ "loss": 0.013,
1736
+ "step": 2820
1737
+ },
1738
+ {
1739
+ "epoch": 4.55,
1740
+ "learning_rate": 1.807073954983923e-05,
1741
+ "loss": 0.0573,
1742
+ "step": 2830
1743
+ },
1744
+ {
1745
+ "epoch": 4.57,
1746
+ "learning_rate": 1.742765273311897e-05,
1747
+ "loss": 0.0268,
1748
+ "step": 2840
1749
+ },
1750
+ {
1751
+ "epoch": 4.58,
1752
+ "learning_rate": 1.6784565916398714e-05,
1753
+ "loss": 0.0229,
1754
+ "step": 2850
1755
+ },
1756
+ {
1757
+ "epoch": 4.6,
1758
+ "learning_rate": 1.6141479099678458e-05,
1759
+ "loss": 0.0294,
1760
+ "step": 2860
1761
+ },
1762
+ {
1763
+ "epoch": 4.61,
1764
+ "learning_rate": 1.54983922829582e-05,
1765
+ "loss": 0.0619,
1766
+ "step": 2870
1767
+ },
1768
+ {
1769
+ "epoch": 4.63,
1770
+ "learning_rate": 1.4855305466237943e-05,
1771
+ "loss": 0.0232,
1772
+ "step": 2880
1773
+ },
1774
+ {
1775
+ "epoch": 4.65,
1776
+ "learning_rate": 1.4212218649517686e-05,
1777
+ "loss": 0.0388,
1778
+ "step": 2890
1779
+ },
1780
+ {
1781
+ "epoch": 4.66,
1782
+ "learning_rate": 1.3569131832797427e-05,
1783
+ "loss": 0.0133,
1784
+ "step": 2900
1785
+ },
1786
+ {
1787
+ "epoch": 4.68,
1788
+ "learning_rate": 1.2926045016077171e-05,
1789
+ "loss": 0.0588,
1790
+ "step": 2910
1791
+ },
1792
+ {
1793
+ "epoch": 4.69,
1794
+ "learning_rate": 1.2282958199356915e-05,
1795
+ "loss": 0.0394,
1796
+ "step": 2920
1797
+ },
1798
+ {
1799
+ "epoch": 4.71,
1800
+ "learning_rate": 1.1639871382636657e-05,
1801
+ "loss": 0.0253,
1802
+ "step": 2930
1803
+ },
1804
+ {
1805
+ "epoch": 4.73,
1806
+ "learning_rate": 1.0996784565916399e-05,
1807
+ "loss": 0.0292,
1808
+ "step": 2940
1809
+ },
1810
+ {
1811
+ "epoch": 4.74,
1812
+ "learning_rate": 1.0353697749196143e-05,
1813
+ "loss": 0.027,
1814
+ "step": 2950
1815
+ },
1816
+ {
1817
+ "epoch": 4.76,
1818
+ "learning_rate": 9.710610932475885e-06,
1819
+ "loss": 0.0317,
1820
+ "step": 2960
1821
+ },
1822
+ {
1823
+ "epoch": 4.77,
1824
+ "learning_rate": 9.067524115755627e-06,
1825
+ "loss": 0.0326,
1826
+ "step": 2970
1827
+ },
1828
+ {
1829
+ "epoch": 4.79,
1830
+ "learning_rate": 8.42443729903537e-06,
1831
+ "loss": 0.0508,
1832
+ "step": 2980
1833
+ },
1834
+ {
1835
+ "epoch": 4.81,
1836
+ "learning_rate": 7.781350482315112e-06,
1837
+ "loss": 0.0182,
1838
+ "step": 2990
1839
+ },
1840
+ {
1841
+ "epoch": 4.82,
1842
+ "learning_rate": 7.1382636655948566e-06,
1843
+ "loss": 0.0255,
1844
+ "step": 3000
1845
+ },
1846
+ {
1847
+ "epoch": 4.84,
1848
+ "learning_rate": 6.495176848874599e-06,
1849
+ "loss": 0.0199,
1850
+ "step": 3010
1851
+ },
1852
+ {
1853
+ "epoch": 4.86,
1854
+ "learning_rate": 5.852090032154341e-06,
1855
+ "loss": 0.0206,
1856
+ "step": 3020
1857
+ },
1858
+ {
1859
+ "epoch": 4.87,
1860
+ "learning_rate": 5.209003215434084e-06,
1861
+ "loss": 0.0352,
1862
+ "step": 3030
1863
+ },
1864
+ {
1865
+ "epoch": 4.89,
1866
+ "learning_rate": 4.565916398713826e-06,
1867
+ "loss": 0.0362,
1868
+ "step": 3040
1869
+ },
1870
+ {
1871
+ "epoch": 4.9,
1872
+ "learning_rate": 3.922829581993569e-06,
1873
+ "loss": 0.0373,
1874
+ "step": 3050
1875
+ },
1876
+ {
1877
+ "epoch": 4.92,
1878
+ "learning_rate": 3.2797427652733123e-06,
1879
+ "loss": 0.0279,
1880
+ "step": 3060
1881
+ },
1882
+ {
1883
+ "epoch": 4.94,
1884
+ "learning_rate": 2.636655948553055e-06,
1885
+ "loss": 0.0286,
1886
+ "step": 3070
1887
+ },
1888
+ {
1889
+ "epoch": 4.95,
1890
+ "learning_rate": 1.9935691318327975e-06,
1891
+ "loss": 0.0359,
1892
+ "step": 3080
1893
+ },
1894
+ {
1895
+ "epoch": 4.97,
1896
+ "learning_rate": 1.3504823151125401e-06,
1897
+ "loss": 0.0382,
1898
+ "step": 3090
1899
+ },
1900
+ {
1901
+ "epoch": 4.98,
1902
+ "learning_rate": 7.073954983922831e-07,
1903
+ "loss": 0.0229,
1904
+ "step": 3100
1905
+ },
1906
+ {
1907
+ "epoch": 5.0,
1908
+ "learning_rate": 6.430868167202573e-08,
1909
+ "loss": 0.0328,
1910
+ "step": 3110
1911
+ },
1912
+ {
1913
+ "epoch": 5.0,
1914
+ "eval_accuracy": 0.9937357630979499,
1915
+ "eval_loss": 0.0182377677410841,
1916
+ "eval_runtime": 25.6312,
1917
+ "eval_samples_per_second": 137.021,
1918
+ "eval_steps_per_second": 4.292,
1919
+ "step": 3110
1920
+ },
1921
+ {
1922
+ "epoch": 5.0,
1923
+ "step": 3110,
1924
+ "total_flos": 0.0,
1925
+ "train_loss": 0.06854689962371369,
1926
+ "train_runtime": 1208.8541,
1927
+ "train_samples_per_second": 82.301,
1928
+ "train_steps_per_second": 2.573
1929
+ }
1930
+ ],
1931
+ "max_steps": 3110,
1932
+ "num_train_epochs": 5,
1933
+ "total_flos": 0.0,
1934
+ "trial_name": null,
1935
+ "trial_params": null
1936
+ }