c14kevincardenas commited on
Commit
9b93c03
·
verified ·
1 Parent(s): f3a2982

End of training

Browse files
Files changed (5) hide show
  1. README.md +7 -3
  2. all_results.json +13 -0
  3. eval_results.json +8 -0
  4. train_results.json +8 -0
  5. trainer_state.json +1139 -0
README.md CHANGED
@@ -1,6 +1,10 @@
1
  ---
2
  library_name: transformers
 
 
3
  tags:
 
 
4
  - generated_from_trainer
5
  metrics:
6
  - accuracy
@@ -14,10 +18,10 @@ should probably proofread and complete it, then remove this comment. -->
14
 
15
  # swin-tiny-patch4-window7-224_alpha0.7_temp3.0_t3
16
 
17
- This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 0.7854
20
- - Accuracy: 0.7885
21
 
22
  ## Model description
23
 
 
1
  ---
2
  library_name: transformers
3
+ license: apache-2.0
4
+ base_model: c14kevincardenas/ClimBEiT-t3
5
  tags:
6
+ - knowledge_distillation
7
+ - vision
8
  - generated_from_trainer
9
  metrics:
10
  - accuracy
 
18
 
19
  # swin-tiny-patch4-window7-224_alpha0.7_temp3.0_t3
20
 
21
+ This model is a fine-tuned version of [c14kevincardenas/ClimBEiT-t3](https://huggingface.co/c14kevincardenas/ClimBEiT-t3) on the c14kevincardenas/beta_caller_284_person_crop_seq_withlimb dataset.
22
  It achieves the following results on the evaluation set:
23
+ - Loss: 0.7700
24
+ - Accuracy: 0.7961
25
 
26
  ## Model description
27
 
all_results.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 20.0,
3
+ "eval_accuracy": 0.7960954446854663,
4
+ "eval_loss": 0.7699832320213318,
5
+ "eval_runtime": 41.9259,
6
+ "eval_samples_per_second": 21.991,
7
+ "eval_steps_per_second": 0.692,
8
+ "total_flos": 0.0,
9
+ "train_loss": 0.29519598901998706,
10
+ "train_runtime": 14634.1552,
11
+ "train_samples_per_second": 7.139,
12
+ "train_steps_per_second": 0.224
13
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 20.0,
3
+ "eval_accuracy": 0.7960954446854663,
4
+ "eval_loss": 0.7699832320213318,
5
+ "eval_runtime": 41.9259,
6
+ "eval_samples_per_second": 21.991,
7
+ "eval_steps_per_second": 0.692
8
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 20.0,
3
+ "total_flos": 0.0,
4
+ "train_loss": 0.29519598901998706,
5
+ "train_runtime": 14634.1552,
6
+ "train_samples_per_second": 7.139,
7
+ "train_steps_per_second": 0.224
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.7699832320213318,
3
+ "best_model_checkpoint": "./kd_results/microsoft/swin-tiny-patch4-window7-224_alpha0.7_temp3.0_t3/checkpoint-2460",
4
+ "epoch": 20.0,
5
+ "eval_steps": 500,
6
+ "global_step": 3280,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.1524390243902439,
13
+ "grad_norm": 4.97396993637085,
14
+ "learning_rate": 3.8109756097560976e-06,
15
+ "loss": 0.5849,
16
+ "step": 25
17
+ },
18
+ {
19
+ "epoch": 0.3048780487804878,
20
+ "grad_norm": 3.1068673133850098,
21
+ "learning_rate": 7.621951219512195e-06,
22
+ "loss": 0.5864,
23
+ "step": 50
24
+ },
25
+ {
26
+ "epoch": 0.4573170731707317,
27
+ "grad_norm": 4.72998046875,
28
+ "learning_rate": 1.1432926829268294e-05,
29
+ "loss": 0.5742,
30
+ "step": 75
31
+ },
32
+ {
33
+ "epoch": 0.6097560975609756,
34
+ "grad_norm": 2.589412212371826,
35
+ "learning_rate": 1.524390243902439e-05,
36
+ "loss": 0.5647,
37
+ "step": 100
38
+ },
39
+ {
40
+ "epoch": 0.7621951219512195,
41
+ "grad_norm": 3.89925479888916,
42
+ "learning_rate": 1.9054878048780488e-05,
43
+ "loss": 0.5744,
44
+ "step": 125
45
+ },
46
+ {
47
+ "epoch": 0.9146341463414634,
48
+ "grad_norm": 2.6995656490325928,
49
+ "learning_rate": 2.286585365853659e-05,
50
+ "loss": 0.5659,
51
+ "step": 150
52
+ },
53
+ {
54
+ "epoch": 1.0,
55
+ "eval_accuracy": 0.2635574837310195,
56
+ "eval_loss": 1.4279733896255493,
57
+ "eval_runtime": 41.2596,
58
+ "eval_samples_per_second": 22.346,
59
+ "eval_steps_per_second": 0.703,
60
+ "step": 164
61
+ },
62
+ {
63
+ "epoch": 1.0670731707317074,
64
+ "grad_norm": 2.971930980682373,
65
+ "learning_rate": 2.6676829268292686e-05,
66
+ "loss": 0.5771,
67
+ "step": 175
68
+ },
69
+ {
70
+ "epoch": 1.2195121951219512,
71
+ "grad_norm": 3.9521827697753906,
72
+ "learning_rate": 3.048780487804878e-05,
73
+ "loss": 0.5651,
74
+ "step": 200
75
+ },
76
+ {
77
+ "epoch": 1.3719512195121952,
78
+ "grad_norm": 3.7694523334503174,
79
+ "learning_rate": 3.429878048780488e-05,
80
+ "loss": 0.5488,
81
+ "step": 225
82
+ },
83
+ {
84
+ "epoch": 1.524390243902439,
85
+ "grad_norm": 3.5128374099731445,
86
+ "learning_rate": 3.8109756097560976e-05,
87
+ "loss": 0.5654,
88
+ "step": 250
89
+ },
90
+ {
91
+ "epoch": 1.6768292682926829,
92
+ "grad_norm": 5.130870342254639,
93
+ "learning_rate": 4.1920731707317077e-05,
94
+ "loss": 0.5424,
95
+ "step": 275
96
+ },
97
+ {
98
+ "epoch": 1.8292682926829267,
99
+ "grad_norm": 4.611592769622803,
100
+ "learning_rate": 4.573170731707318e-05,
101
+ "loss": 0.5299,
102
+ "step": 300
103
+ },
104
+ {
105
+ "epoch": 1.9817073170731707,
106
+ "grad_norm": 2.701108932495117,
107
+ "learning_rate": 4.954268292682927e-05,
108
+ "loss": 0.5111,
109
+ "step": 325
110
+ },
111
+ {
112
+ "epoch": 2.0,
113
+ "eval_accuracy": 0.3720173535791757,
114
+ "eval_loss": 1.3039730787277222,
115
+ "eval_runtime": 41.7449,
116
+ "eval_samples_per_second": 22.087,
117
+ "eval_steps_per_second": 0.695,
118
+ "step": 328
119
+ },
120
+ {
121
+ "epoch": 2.1341463414634148,
122
+ "grad_norm": 3.373603582382202,
123
+ "learning_rate": 4.962737127371274e-05,
124
+ "loss": 0.4844,
125
+ "step": 350
126
+ },
127
+ {
128
+ "epoch": 2.2865853658536586,
129
+ "grad_norm": 2.5607736110687256,
130
+ "learning_rate": 4.92039295392954e-05,
131
+ "loss": 0.4813,
132
+ "step": 375
133
+ },
134
+ {
135
+ "epoch": 2.4390243902439024,
136
+ "grad_norm": 4.668150901794434,
137
+ "learning_rate": 4.878048780487805e-05,
138
+ "loss": 0.4456,
139
+ "step": 400
140
+ },
141
+ {
142
+ "epoch": 2.591463414634146,
143
+ "grad_norm": 6.863236427307129,
144
+ "learning_rate": 4.8357046070460704e-05,
145
+ "loss": 0.4369,
146
+ "step": 425
147
+ },
148
+ {
149
+ "epoch": 2.7439024390243905,
150
+ "grad_norm": 4.940046310424805,
151
+ "learning_rate": 4.793360433604336e-05,
152
+ "loss": 0.4509,
153
+ "step": 450
154
+ },
155
+ {
156
+ "epoch": 2.8963414634146343,
157
+ "grad_norm": 3.610745906829834,
158
+ "learning_rate": 4.7510162601626016e-05,
159
+ "loss": 0.4093,
160
+ "step": 475
161
+ },
162
+ {
163
+ "epoch": 3.0,
164
+ "eval_accuracy": 0.6355748373101953,
165
+ "eval_loss": 0.9967854022979736,
166
+ "eval_runtime": 41.3413,
167
+ "eval_samples_per_second": 22.302,
168
+ "eval_steps_per_second": 0.701,
169
+ "step": 492
170
+ },
171
+ {
172
+ "epoch": 3.048780487804878,
173
+ "grad_norm": 4.550442218780518,
174
+ "learning_rate": 4.7086720867208675e-05,
175
+ "loss": 0.4146,
176
+ "step": 500
177
+ },
178
+ {
179
+ "epoch": 3.201219512195122,
180
+ "grad_norm": 4.069523811340332,
181
+ "learning_rate": 4.6663279132791334e-05,
182
+ "loss": 0.3872,
183
+ "step": 525
184
+ },
185
+ {
186
+ "epoch": 3.3536585365853657,
187
+ "grad_norm": 3.338857889175415,
188
+ "learning_rate": 4.6239837398373986e-05,
189
+ "loss": 0.3766,
190
+ "step": 550
191
+ },
192
+ {
193
+ "epoch": 3.5060975609756095,
194
+ "grad_norm": 2.963149070739746,
195
+ "learning_rate": 4.581639566395664e-05,
196
+ "loss": 0.3767,
197
+ "step": 575
198
+ },
199
+ {
200
+ "epoch": 3.658536585365854,
201
+ "grad_norm": 3.808338165283203,
202
+ "learning_rate": 4.53929539295393e-05,
203
+ "loss": 0.3587,
204
+ "step": 600
205
+ },
206
+ {
207
+ "epoch": 3.8109756097560976,
208
+ "grad_norm": 3.5818405151367188,
209
+ "learning_rate": 4.496951219512196e-05,
210
+ "loss": 0.3624,
211
+ "step": 625
212
+ },
213
+ {
214
+ "epoch": 3.9634146341463414,
215
+ "grad_norm": 3.399240493774414,
216
+ "learning_rate": 4.454607046070461e-05,
217
+ "loss": 0.3598,
218
+ "step": 650
219
+ },
220
+ {
221
+ "epoch": 4.0,
222
+ "eval_accuracy": 0.6442516268980477,
223
+ "eval_loss": 0.9534724950790405,
224
+ "eval_runtime": 40.9142,
225
+ "eval_samples_per_second": 22.535,
226
+ "eval_steps_per_second": 0.709,
227
+ "step": 656
228
+ },
229
+ {
230
+ "epoch": 4.115853658536586,
231
+ "grad_norm": 3.605801820755005,
232
+ "learning_rate": 4.412262872628726e-05,
233
+ "loss": 0.3402,
234
+ "step": 675
235
+ },
236
+ {
237
+ "epoch": 4.2682926829268295,
238
+ "grad_norm": 2.1427693367004395,
239
+ "learning_rate": 4.369918699186992e-05,
240
+ "loss": 0.3177,
241
+ "step": 700
242
+ },
243
+ {
244
+ "epoch": 4.420731707317073,
245
+ "grad_norm": 2.742725133895874,
246
+ "learning_rate": 4.3275745257452574e-05,
247
+ "loss": 0.3274,
248
+ "step": 725
249
+ },
250
+ {
251
+ "epoch": 4.573170731707317,
252
+ "grad_norm": 2.6349594593048096,
253
+ "learning_rate": 4.285230352303523e-05,
254
+ "loss": 0.3243,
255
+ "step": 750
256
+ },
257
+ {
258
+ "epoch": 4.725609756097561,
259
+ "grad_norm": 2.6220862865448,
260
+ "learning_rate": 4.242886178861789e-05,
261
+ "loss": 0.3133,
262
+ "step": 775
263
+ },
264
+ {
265
+ "epoch": 4.878048780487805,
266
+ "grad_norm": 2.377204656600952,
267
+ "learning_rate": 4.2005420054200545e-05,
268
+ "loss": 0.3284,
269
+ "step": 800
270
+ },
271
+ {
272
+ "epoch": 5.0,
273
+ "eval_accuracy": 0.7255965292841648,
274
+ "eval_loss": 0.84259432554245,
275
+ "eval_runtime": 41.4402,
276
+ "eval_samples_per_second": 22.249,
277
+ "eval_steps_per_second": 0.7,
278
+ "step": 820
279
+ },
280
+ {
281
+ "epoch": 5.030487804878049,
282
+ "grad_norm": 2.5129756927490234,
283
+ "learning_rate": 4.15819783197832e-05,
284
+ "loss": 0.3208,
285
+ "step": 825
286
+ },
287
+ {
288
+ "epoch": 5.182926829268292,
289
+ "grad_norm": 3.733022928237915,
290
+ "learning_rate": 4.1158536585365856e-05,
291
+ "loss": 0.2929,
292
+ "step": 850
293
+ },
294
+ {
295
+ "epoch": 5.335365853658536,
296
+ "grad_norm": 1.9925252199172974,
297
+ "learning_rate": 4.0735094850948516e-05,
298
+ "loss": 0.2878,
299
+ "step": 875
300
+ },
301
+ {
302
+ "epoch": 5.487804878048781,
303
+ "grad_norm": 1.9667726755142212,
304
+ "learning_rate": 4.031165311653117e-05,
305
+ "loss": 0.2964,
306
+ "step": 900
307
+ },
308
+ {
309
+ "epoch": 5.640243902439025,
310
+ "grad_norm": 3.29734206199646,
311
+ "learning_rate": 3.988821138211382e-05,
312
+ "loss": 0.2913,
313
+ "step": 925
314
+ },
315
+ {
316
+ "epoch": 5.7926829268292686,
317
+ "grad_norm": 1.7537786960601807,
318
+ "learning_rate": 3.946476964769648e-05,
319
+ "loss": 0.29,
320
+ "step": 950
321
+ },
322
+ {
323
+ "epoch": 5.945121951219512,
324
+ "grad_norm": 2.272477149963379,
325
+ "learning_rate": 3.904132791327913e-05,
326
+ "loss": 0.2933,
327
+ "step": 975
328
+ },
329
+ {
330
+ "epoch": 6.0,
331
+ "eval_accuracy": 0.7657266811279827,
332
+ "eval_loss": 0.8268624544143677,
333
+ "eval_runtime": 41.5751,
334
+ "eval_samples_per_second": 22.177,
335
+ "eval_steps_per_second": 0.698,
336
+ "step": 984
337
+ },
338
+ {
339
+ "epoch": 6.097560975609756,
340
+ "grad_norm": 2.4584200382232666,
341
+ "learning_rate": 3.861788617886179e-05,
342
+ "loss": 0.2829,
343
+ "step": 1000
344
+ },
345
+ {
346
+ "epoch": 6.25,
347
+ "grad_norm": 1.7212045192718506,
348
+ "learning_rate": 3.8194444444444444e-05,
349
+ "loss": 0.2738,
350
+ "step": 1025
351
+ },
352
+ {
353
+ "epoch": 6.402439024390244,
354
+ "grad_norm": 1.7124000787734985,
355
+ "learning_rate": 3.77710027100271e-05,
356
+ "loss": 0.2768,
357
+ "step": 1050
358
+ },
359
+ {
360
+ "epoch": 6.554878048780488,
361
+ "grad_norm": 1.7901511192321777,
362
+ "learning_rate": 3.7347560975609755e-05,
363
+ "loss": 0.2715,
364
+ "step": 1075
365
+ },
366
+ {
367
+ "epoch": 6.7073170731707314,
368
+ "grad_norm": 1.9946818351745605,
369
+ "learning_rate": 3.6924119241192415e-05,
370
+ "loss": 0.2751,
371
+ "step": 1100
372
+ },
373
+ {
374
+ "epoch": 6.859756097560975,
375
+ "grad_norm": 2.2430810928344727,
376
+ "learning_rate": 3.6500677506775074e-05,
377
+ "loss": 0.2734,
378
+ "step": 1125
379
+ },
380
+ {
381
+ "epoch": 7.0,
382
+ "eval_accuracy": 0.6952277657266811,
383
+ "eval_loss": 0.8815469145774841,
384
+ "eval_runtime": 41.5411,
385
+ "eval_samples_per_second": 22.195,
386
+ "eval_steps_per_second": 0.698,
387
+ "step": 1148
388
+ },
389
+ {
390
+ "epoch": 7.012195121951219,
391
+ "grad_norm": 2.6901626586914062,
392
+ "learning_rate": 3.6077235772357726e-05,
393
+ "loss": 0.269,
394
+ "step": 1150
395
+ },
396
+ {
397
+ "epoch": 7.164634146341464,
398
+ "grad_norm": 3.2748658657073975,
399
+ "learning_rate": 3.565379403794038e-05,
400
+ "loss": 0.2626,
401
+ "step": 1175
402
+ },
403
+ {
404
+ "epoch": 7.317073170731708,
405
+ "grad_norm": 2.778024435043335,
406
+ "learning_rate": 3.523035230352303e-05,
407
+ "loss": 0.256,
408
+ "step": 1200
409
+ },
410
+ {
411
+ "epoch": 7.469512195121951,
412
+ "grad_norm": 1.585498571395874,
413
+ "learning_rate": 3.48069105691057e-05,
414
+ "loss": 0.2555,
415
+ "step": 1225
416
+ },
417
+ {
418
+ "epoch": 7.621951219512195,
419
+ "grad_norm": 2.2903597354888916,
420
+ "learning_rate": 3.438346883468835e-05,
421
+ "loss": 0.2553,
422
+ "step": 1250
423
+ },
424
+ {
425
+ "epoch": 7.774390243902439,
426
+ "grad_norm": 2.6786723136901855,
427
+ "learning_rate": 3.3960027100271e-05,
428
+ "loss": 0.2654,
429
+ "step": 1275
430
+ },
431
+ {
432
+ "epoch": 7.926829268292683,
433
+ "grad_norm": 1.9107999801635742,
434
+ "learning_rate": 3.353658536585366e-05,
435
+ "loss": 0.2679,
436
+ "step": 1300
437
+ },
438
+ {
439
+ "epoch": 8.0,
440
+ "eval_accuracy": 0.7678958785249458,
441
+ "eval_loss": 0.8078573346138,
442
+ "eval_runtime": 41.0894,
443
+ "eval_samples_per_second": 22.439,
444
+ "eval_steps_per_second": 0.706,
445
+ "step": 1312
446
+ },
447
+ {
448
+ "epoch": 8.079268292682928,
449
+ "grad_norm": 1.494566798210144,
450
+ "learning_rate": 3.3113143631436314e-05,
451
+ "loss": 0.2518,
452
+ "step": 1325
453
+ },
454
+ {
455
+ "epoch": 8.231707317073171,
456
+ "grad_norm": 1.2574859857559204,
457
+ "learning_rate": 3.268970189701897e-05,
458
+ "loss": 0.2431,
459
+ "step": 1350
460
+ },
461
+ {
462
+ "epoch": 8.384146341463415,
463
+ "grad_norm": 1.92410409450531,
464
+ "learning_rate": 3.226626016260163e-05,
465
+ "loss": 0.257,
466
+ "step": 1375
467
+ },
468
+ {
469
+ "epoch": 8.536585365853659,
470
+ "grad_norm": 1.3554410934448242,
471
+ "learning_rate": 3.1842818428184285e-05,
472
+ "loss": 0.2503,
473
+ "step": 1400
474
+ },
475
+ {
476
+ "epoch": 8.689024390243903,
477
+ "grad_norm": 2.443579912185669,
478
+ "learning_rate": 3.141937669376694e-05,
479
+ "loss": 0.2534,
480
+ "step": 1425
481
+ },
482
+ {
483
+ "epoch": 8.841463414634147,
484
+ "grad_norm": 1.6061345338821411,
485
+ "learning_rate": 3.0995934959349596e-05,
486
+ "loss": 0.2483,
487
+ "step": 1450
488
+ },
489
+ {
490
+ "epoch": 8.99390243902439,
491
+ "grad_norm": 1.1513770818710327,
492
+ "learning_rate": 3.0572493224932255e-05,
493
+ "loss": 0.2574,
494
+ "step": 1475
495
+ },
496
+ {
497
+ "epoch": 9.0,
498
+ "eval_accuracy": 0.7863340563991323,
499
+ "eval_loss": 0.7823293805122375,
500
+ "eval_runtime": 41.8128,
501
+ "eval_samples_per_second": 22.051,
502
+ "eval_steps_per_second": 0.694,
503
+ "step": 1476
504
+ },
505
+ {
506
+ "epoch": 9.146341463414634,
507
+ "grad_norm": 1.9186493158340454,
508
+ "learning_rate": 3.0149051490514908e-05,
509
+ "loss": 0.247,
510
+ "step": 1500
511
+ },
512
+ {
513
+ "epoch": 9.298780487804878,
514
+ "grad_norm": 1.1330264806747437,
515
+ "learning_rate": 2.972560975609756e-05,
516
+ "loss": 0.2536,
517
+ "step": 1525
518
+ },
519
+ {
520
+ "epoch": 9.451219512195122,
521
+ "grad_norm": 1.8567355871200562,
522
+ "learning_rate": 2.930216802168022e-05,
523
+ "loss": 0.2458,
524
+ "step": 1550
525
+ },
526
+ {
527
+ "epoch": 9.603658536585366,
528
+ "grad_norm": 1.351461410522461,
529
+ "learning_rate": 2.8878726287262875e-05,
530
+ "loss": 0.2442,
531
+ "step": 1575
532
+ },
533
+ {
534
+ "epoch": 9.75609756097561,
535
+ "grad_norm": 2.0739259719848633,
536
+ "learning_rate": 2.8455284552845528e-05,
537
+ "loss": 0.25,
538
+ "step": 1600
539
+ },
540
+ {
541
+ "epoch": 9.908536585365853,
542
+ "grad_norm": 0.737319827079773,
543
+ "learning_rate": 2.8031842818428184e-05,
544
+ "loss": 0.2403,
545
+ "step": 1625
546
+ },
547
+ {
548
+ "epoch": 10.0,
549
+ "eval_accuracy": 0.7906724511930586,
550
+ "eval_loss": 0.7832654118537903,
551
+ "eval_runtime": 41.3078,
552
+ "eval_samples_per_second": 22.32,
553
+ "eval_steps_per_second": 0.702,
554
+ "step": 1640
555
+ },
556
+ {
557
+ "epoch": 10.060975609756097,
558
+ "grad_norm": 1.5469945669174194,
559
+ "learning_rate": 2.7608401084010843e-05,
560
+ "loss": 0.2422,
561
+ "step": 1650
562
+ },
563
+ {
564
+ "epoch": 10.213414634146341,
565
+ "grad_norm": 1.8835623264312744,
566
+ "learning_rate": 2.71849593495935e-05,
567
+ "loss": 0.2409,
568
+ "step": 1675
569
+ },
570
+ {
571
+ "epoch": 10.365853658536585,
572
+ "grad_norm": 0.858368456363678,
573
+ "learning_rate": 2.676151761517615e-05,
574
+ "loss": 0.2451,
575
+ "step": 1700
576
+ },
577
+ {
578
+ "epoch": 10.518292682926829,
579
+ "grad_norm": 1.7335875034332275,
580
+ "learning_rate": 2.633807588075881e-05,
581
+ "loss": 0.2423,
582
+ "step": 1725
583
+ },
584
+ {
585
+ "epoch": 10.670731707317072,
586
+ "grad_norm": 1.0270367860794067,
587
+ "learning_rate": 2.5914634146341466e-05,
588
+ "loss": 0.2398,
589
+ "step": 1750
590
+ },
591
+ {
592
+ "epoch": 10.823170731707316,
593
+ "grad_norm": 1.3634330034255981,
594
+ "learning_rate": 2.549119241192412e-05,
595
+ "loss": 0.2428,
596
+ "step": 1775
597
+ },
598
+ {
599
+ "epoch": 10.975609756097562,
600
+ "grad_norm": 0.8540765047073364,
601
+ "learning_rate": 2.5067750677506774e-05,
602
+ "loss": 0.2376,
603
+ "step": 1800
604
+ },
605
+ {
606
+ "epoch": 11.0,
607
+ "eval_accuracy": 0.7852494577006508,
608
+ "eval_loss": 0.7851491570472717,
609
+ "eval_runtime": 41.4503,
610
+ "eval_samples_per_second": 22.243,
611
+ "eval_steps_per_second": 0.7,
612
+ "step": 1804
613
+ },
614
+ {
615
+ "epoch": 11.128048780487806,
616
+ "grad_norm": 2.378941059112549,
617
+ "learning_rate": 2.464430894308943e-05,
618
+ "loss": 0.2435,
619
+ "step": 1825
620
+ },
621
+ {
622
+ "epoch": 11.28048780487805,
623
+ "grad_norm": 1.6375097036361694,
624
+ "learning_rate": 2.422086720867209e-05,
625
+ "loss": 0.2358,
626
+ "step": 1850
627
+ },
628
+ {
629
+ "epoch": 11.432926829268293,
630
+ "grad_norm": 1.5316449403762817,
631
+ "learning_rate": 2.3797425474254745e-05,
632
+ "loss": 0.2367,
633
+ "step": 1875
634
+ },
635
+ {
636
+ "epoch": 11.585365853658537,
637
+ "grad_norm": 1.458227276802063,
638
+ "learning_rate": 2.3373983739837398e-05,
639
+ "loss": 0.237,
640
+ "step": 1900
641
+ },
642
+ {
643
+ "epoch": 11.737804878048781,
644
+ "grad_norm": 1.1965981721878052,
645
+ "learning_rate": 2.2950542005420057e-05,
646
+ "loss": 0.2451,
647
+ "step": 1925
648
+ },
649
+ {
650
+ "epoch": 11.890243902439025,
651
+ "grad_norm": 1.0848238468170166,
652
+ "learning_rate": 2.252710027100271e-05,
653
+ "loss": 0.2399,
654
+ "step": 1950
655
+ },
656
+ {
657
+ "epoch": 12.0,
658
+ "eval_accuracy": 0.7939262472885033,
659
+ "eval_loss": 0.7965609431266785,
660
+ "eval_runtime": 41.2064,
661
+ "eval_samples_per_second": 22.375,
662
+ "eval_steps_per_second": 0.704,
663
+ "step": 1968
664
+ },
665
+ {
666
+ "epoch": 12.042682926829269,
667
+ "grad_norm": 0.900860071182251,
668
+ "learning_rate": 2.210365853658537e-05,
669
+ "loss": 0.2385,
670
+ "step": 1975
671
+ },
672
+ {
673
+ "epoch": 12.195121951219512,
674
+ "grad_norm": 0.6957654356956482,
675
+ "learning_rate": 2.1680216802168024e-05,
676
+ "loss": 0.235,
677
+ "step": 2000
678
+ },
679
+ {
680
+ "epoch": 12.347560975609756,
681
+ "grad_norm": 0.9133132100105286,
682
+ "learning_rate": 2.1256775067750677e-05,
683
+ "loss": 0.2397,
684
+ "step": 2025
685
+ },
686
+ {
687
+ "epoch": 12.5,
688
+ "grad_norm": 1.3867614269256592,
689
+ "learning_rate": 2.0833333333333336e-05,
690
+ "loss": 0.2356,
691
+ "step": 2050
692
+ },
693
+ {
694
+ "epoch": 12.652439024390244,
695
+ "grad_norm": 1.484837532043457,
696
+ "learning_rate": 2.040989159891599e-05,
697
+ "loss": 0.2359,
698
+ "step": 2075
699
+ },
700
+ {
701
+ "epoch": 12.804878048780488,
702
+ "grad_norm": 1.6344926357269287,
703
+ "learning_rate": 1.9986449864498648e-05,
704
+ "loss": 0.2322,
705
+ "step": 2100
706
+ },
707
+ {
708
+ "epoch": 12.957317073170731,
709
+ "grad_norm": 0.8600764870643616,
710
+ "learning_rate": 1.95630081300813e-05,
711
+ "loss": 0.231,
712
+ "step": 2125
713
+ },
714
+ {
715
+ "epoch": 13.0,
716
+ "eval_accuracy": 0.7765726681127982,
717
+ "eval_loss": 0.7956083416938782,
718
+ "eval_runtime": 41.0434,
719
+ "eval_samples_per_second": 22.464,
720
+ "eval_steps_per_second": 0.707,
721
+ "step": 2132
722
+ },
723
+ {
724
+ "epoch": 13.109756097560975,
725
+ "grad_norm": 1.2406657934188843,
726
+ "learning_rate": 1.9139566395663956e-05,
727
+ "loss": 0.2327,
728
+ "step": 2150
729
+ },
730
+ {
731
+ "epoch": 13.262195121951219,
732
+ "grad_norm": 0.9290794134140015,
733
+ "learning_rate": 1.8716124661246615e-05,
734
+ "loss": 0.238,
735
+ "step": 2175
736
+ },
737
+ {
738
+ "epoch": 13.414634146341463,
739
+ "grad_norm": 0.9094896912574768,
740
+ "learning_rate": 1.8292682926829268e-05,
741
+ "loss": 0.2332,
742
+ "step": 2200
743
+ },
744
+ {
745
+ "epoch": 13.567073170731707,
746
+ "grad_norm": 0.9954293370246887,
747
+ "learning_rate": 1.7869241192411927e-05,
748
+ "loss": 0.2273,
749
+ "step": 2225
750
+ },
751
+ {
752
+ "epoch": 13.71951219512195,
753
+ "grad_norm": 0.9158800840377808,
754
+ "learning_rate": 1.744579945799458e-05,
755
+ "loss": 0.2294,
756
+ "step": 2250
757
+ },
758
+ {
759
+ "epoch": 13.871951219512194,
760
+ "grad_norm": 0.9772486686706543,
761
+ "learning_rate": 1.702235772357724e-05,
762
+ "loss": 0.2351,
763
+ "step": 2275
764
+ },
765
+ {
766
+ "epoch": 14.0,
767
+ "eval_accuracy": 0.7917570498915402,
768
+ "eval_loss": 0.7792609930038452,
769
+ "eval_runtime": 41.3252,
770
+ "eval_samples_per_second": 22.311,
771
+ "eval_steps_per_second": 0.702,
772
+ "step": 2296
773
+ },
774
+ {
775
+ "epoch": 14.024390243902438,
776
+ "grad_norm": 0.6987525820732117,
777
+ "learning_rate": 1.659891598915989e-05,
778
+ "loss": 0.2319,
779
+ "step": 2300
780
+ },
781
+ {
782
+ "epoch": 14.176829268292684,
783
+ "grad_norm": 1.195956826210022,
784
+ "learning_rate": 1.6175474254742547e-05,
785
+ "loss": 0.2333,
786
+ "step": 2325
787
+ },
788
+ {
789
+ "epoch": 14.329268292682928,
790
+ "grad_norm": 1.15598464012146,
791
+ "learning_rate": 1.5752032520325206e-05,
792
+ "loss": 0.2314,
793
+ "step": 2350
794
+ },
795
+ {
796
+ "epoch": 14.481707317073171,
797
+ "grad_norm": 0.5450630784034729,
798
+ "learning_rate": 1.532859078590786e-05,
799
+ "loss": 0.2317,
800
+ "step": 2375
801
+ },
802
+ {
803
+ "epoch": 14.634146341463415,
804
+ "grad_norm": 1.0192309617996216,
805
+ "learning_rate": 1.4905149051490516e-05,
806
+ "loss": 0.2286,
807
+ "step": 2400
808
+ },
809
+ {
810
+ "epoch": 14.786585365853659,
811
+ "grad_norm": 0.8558110594749451,
812
+ "learning_rate": 1.448170731707317e-05,
813
+ "loss": 0.2285,
814
+ "step": 2425
815
+ },
816
+ {
817
+ "epoch": 14.939024390243903,
818
+ "grad_norm": 0.7634503245353699,
819
+ "learning_rate": 1.4058265582655828e-05,
820
+ "loss": 0.2344,
821
+ "step": 2450
822
+ },
823
+ {
824
+ "epoch": 15.0,
825
+ "eval_accuracy": 0.7960954446854663,
826
+ "eval_loss": 0.7699832320213318,
827
+ "eval_runtime": 40.7623,
828
+ "eval_samples_per_second": 22.619,
829
+ "eval_steps_per_second": 0.711,
830
+ "step": 2460
831
+ },
832
+ {
833
+ "epoch": 15.091463414634147,
834
+ "grad_norm": 1.0911827087402344,
835
+ "learning_rate": 1.3634823848238485e-05,
836
+ "loss": 0.2216,
837
+ "step": 2475
838
+ },
839
+ {
840
+ "epoch": 15.24390243902439,
841
+ "grad_norm": 0.665216863155365,
842
+ "learning_rate": 1.321138211382114e-05,
843
+ "loss": 0.2384,
844
+ "step": 2500
845
+ },
846
+ {
847
+ "epoch": 15.396341463414634,
848
+ "grad_norm": 0.8301494717597961,
849
+ "learning_rate": 1.2787940379403795e-05,
850
+ "loss": 0.2296,
851
+ "step": 2525
852
+ },
853
+ {
854
+ "epoch": 15.548780487804878,
855
+ "grad_norm": 0.9643315076828003,
856
+ "learning_rate": 1.2364498644986451e-05,
857
+ "loss": 0.2277,
858
+ "step": 2550
859
+ },
860
+ {
861
+ "epoch": 15.701219512195122,
862
+ "grad_norm": 0.8478797674179077,
863
+ "learning_rate": 1.1941056910569107e-05,
864
+ "loss": 0.2275,
865
+ "step": 2575
866
+ },
867
+ {
868
+ "epoch": 15.853658536585366,
869
+ "grad_norm": 0.6218520998954773,
870
+ "learning_rate": 1.1517615176151763e-05,
871
+ "loss": 0.232,
872
+ "step": 2600
873
+ },
874
+ {
875
+ "epoch": 16.0,
876
+ "eval_accuracy": 0.7906724511930586,
877
+ "eval_loss": 0.7844656705856323,
878
+ "eval_runtime": 40.6353,
879
+ "eval_samples_per_second": 22.69,
880
+ "eval_steps_per_second": 0.714,
881
+ "step": 2624
882
+ },
883
+ {
884
+ "epoch": 16.00609756097561,
885
+ "grad_norm": 1.0007075071334839,
886
+ "learning_rate": 1.1094173441734418e-05,
887
+ "loss": 0.2316,
888
+ "step": 2625
889
+ },
890
+ {
891
+ "epoch": 16.158536585365855,
892
+ "grad_norm": 0.5642134547233582,
893
+ "learning_rate": 1.0670731707317074e-05,
894
+ "loss": 0.2275,
895
+ "step": 2650
896
+ },
897
+ {
898
+ "epoch": 16.3109756097561,
899
+ "grad_norm": 0.5428956151008606,
900
+ "learning_rate": 1.0247289972899728e-05,
901
+ "loss": 0.2265,
902
+ "step": 2675
903
+ },
904
+ {
905
+ "epoch": 16.463414634146343,
906
+ "grad_norm": 0.49864181876182556,
907
+ "learning_rate": 9.823848238482384e-06,
908
+ "loss": 0.2246,
909
+ "step": 2700
910
+ },
911
+ {
912
+ "epoch": 16.615853658536587,
913
+ "grad_norm": 0.8356211185455322,
914
+ "learning_rate": 9.400406504065042e-06,
915
+ "loss": 0.2287,
916
+ "step": 2725
917
+ },
918
+ {
919
+ "epoch": 16.76829268292683,
920
+ "grad_norm": 0.8411365151405334,
921
+ "learning_rate": 8.976964769647697e-06,
922
+ "loss": 0.2235,
923
+ "step": 2750
924
+ },
925
+ {
926
+ "epoch": 16.920731707317074,
927
+ "grad_norm": 0.6885209083557129,
928
+ "learning_rate": 8.553523035230353e-06,
929
+ "loss": 0.2318,
930
+ "step": 2775
931
+ },
932
+ {
933
+ "epoch": 17.0,
934
+ "eval_accuracy": 0.7917570498915402,
935
+ "eval_loss": 0.7930197715759277,
936
+ "eval_runtime": 41.1598,
937
+ "eval_samples_per_second": 22.401,
938
+ "eval_steps_per_second": 0.705,
939
+ "step": 2788
940
+ },
941
+ {
942
+ "epoch": 17.073170731707318,
943
+ "grad_norm": 0.5789854526519775,
944
+ "learning_rate": 8.130081300813009e-06,
945
+ "loss": 0.2305,
946
+ "step": 2800
947
+ },
948
+ {
949
+ "epoch": 17.225609756097562,
950
+ "grad_norm": 0.8740878105163574,
951
+ "learning_rate": 7.706639566395663e-06,
952
+ "loss": 0.2261,
953
+ "step": 2825
954
+ },
955
+ {
956
+ "epoch": 17.378048780487806,
957
+ "grad_norm": 0.7720724940299988,
958
+ "learning_rate": 7.283197831978319e-06,
959
+ "loss": 0.2333,
960
+ "step": 2850
961
+ },
962
+ {
963
+ "epoch": 17.53048780487805,
964
+ "grad_norm": 0.8361527919769287,
965
+ "learning_rate": 6.859756097560977e-06,
966
+ "loss": 0.2336,
967
+ "step": 2875
968
+ },
969
+ {
970
+ "epoch": 17.682926829268293,
971
+ "grad_norm": 0.8507777452468872,
972
+ "learning_rate": 6.4363143631436324e-06,
973
+ "loss": 0.2257,
974
+ "step": 2900
975
+ },
976
+ {
977
+ "epoch": 17.835365853658537,
978
+ "grad_norm": 1.3327784538269043,
979
+ "learning_rate": 6.0128726287262874e-06,
980
+ "loss": 0.2223,
981
+ "step": 2925
982
+ },
983
+ {
984
+ "epoch": 17.98780487804878,
985
+ "grad_norm": 1.0190765857696533,
986
+ "learning_rate": 5.589430894308943e-06,
987
+ "loss": 0.2241,
988
+ "step": 2950
989
+ },
990
+ {
991
+ "epoch": 18.0,
992
+ "eval_accuracy": 0.7885032537960954,
993
+ "eval_loss": 0.7814083695411682,
994
+ "eval_runtime": 40.5372,
995
+ "eval_samples_per_second": 22.745,
996
+ "eval_steps_per_second": 0.715,
997
+ "step": 2952
998
+ },
999
+ {
1000
+ "epoch": 18.140243902439025,
1001
+ "grad_norm": 0.5855722427368164,
1002
+ "learning_rate": 5.165989159891599e-06,
1003
+ "loss": 0.2259,
1004
+ "step": 2975
1005
+ },
1006
+ {
1007
+ "epoch": 18.29268292682927,
1008
+ "grad_norm": 0.8351572751998901,
1009
+ "learning_rate": 4.742547425474255e-06,
1010
+ "loss": 0.2256,
1011
+ "step": 3000
1012
+ },
1013
+ {
1014
+ "epoch": 18.445121951219512,
1015
+ "grad_norm": 0.5234180688858032,
1016
+ "learning_rate": 4.319105691056911e-06,
1017
+ "loss": 0.2229,
1018
+ "step": 3025
1019
+ },
1020
+ {
1021
+ "epoch": 18.597560975609756,
1022
+ "grad_norm": 0.6411917209625244,
1023
+ "learning_rate": 3.8956639566395666e-06,
1024
+ "loss": 0.2278,
1025
+ "step": 3050
1026
+ },
1027
+ {
1028
+ "epoch": 18.75,
1029
+ "grad_norm": 0.6386851668357849,
1030
+ "learning_rate": 3.4722222222222224e-06,
1031
+ "loss": 0.2238,
1032
+ "step": 3075
1033
+ },
1034
+ {
1035
+ "epoch": 18.902439024390244,
1036
+ "grad_norm": 0.6463262438774109,
1037
+ "learning_rate": 3.0487804878048782e-06,
1038
+ "loss": 0.2291,
1039
+ "step": 3100
1040
+ },
1041
+ {
1042
+ "epoch": 19.0,
1043
+ "eval_accuracy": 0.7819956616052061,
1044
+ "eval_loss": 0.7901421785354614,
1045
+ "eval_runtime": 41.6462,
1046
+ "eval_samples_per_second": 22.139,
1047
+ "eval_steps_per_second": 0.696,
1048
+ "step": 3116
1049
+ },
1050
+ {
1051
+ "epoch": 19.054878048780488,
1052
+ "grad_norm": 0.45505502820014954,
1053
+ "learning_rate": 2.625338753387534e-06,
1054
+ "loss": 0.2267,
1055
+ "step": 3125
1056
+ },
1057
+ {
1058
+ "epoch": 19.20731707317073,
1059
+ "grad_norm": 0.4252510964870453,
1060
+ "learning_rate": 2.2018970189701894e-06,
1061
+ "loss": 0.2326,
1062
+ "step": 3150
1063
+ },
1064
+ {
1065
+ "epoch": 19.359756097560975,
1066
+ "grad_norm": 0.7417110204696655,
1067
+ "learning_rate": 1.7784552845528455e-06,
1068
+ "loss": 0.231,
1069
+ "step": 3175
1070
+ },
1071
+ {
1072
+ "epoch": 19.51219512195122,
1073
+ "grad_norm": 0.6761830449104309,
1074
+ "learning_rate": 1.3550135501355015e-06,
1075
+ "loss": 0.2263,
1076
+ "step": 3200
1077
+ },
1078
+ {
1079
+ "epoch": 19.664634146341463,
1080
+ "grad_norm": 0.6491538286209106,
1081
+ "learning_rate": 9.315718157181572e-07,
1082
+ "loss": 0.2268,
1083
+ "step": 3225
1084
+ },
1085
+ {
1086
+ "epoch": 19.817073170731707,
1087
+ "grad_norm": 0.6726053357124329,
1088
+ "learning_rate": 5.081300813008131e-07,
1089
+ "loss": 0.2231,
1090
+ "step": 3250
1091
+ },
1092
+ {
1093
+ "epoch": 19.96951219512195,
1094
+ "grad_norm": 0.7665335536003113,
1095
+ "learning_rate": 8.468834688346885e-08,
1096
+ "loss": 0.2227,
1097
+ "step": 3275
1098
+ },
1099
+ {
1100
+ "epoch": 20.0,
1101
+ "eval_accuracy": 0.7885032537960954,
1102
+ "eval_loss": 0.7853749990463257,
1103
+ "eval_runtime": 41.5331,
1104
+ "eval_samples_per_second": 22.199,
1105
+ "eval_steps_per_second": 0.698,
1106
+ "step": 3280
1107
+ },
1108
+ {
1109
+ "epoch": 20.0,
1110
+ "step": 3280,
1111
+ "total_flos": 0.0,
1112
+ "train_loss": 0.29519598901998706,
1113
+ "train_runtime": 14634.1552,
1114
+ "train_samples_per_second": 7.139,
1115
+ "train_steps_per_second": 0.224
1116
+ }
1117
+ ],
1118
+ "logging_steps": 25,
1119
+ "max_steps": 3280,
1120
+ "num_input_tokens_seen": 0,
1121
+ "num_train_epochs": 20,
1122
+ "save_steps": 500,
1123
+ "stateful_callbacks": {
1124
+ "TrainerControl": {
1125
+ "args": {
1126
+ "should_epoch_stop": false,
1127
+ "should_evaluate": false,
1128
+ "should_log": false,
1129
+ "should_save": true,
1130
+ "should_training_stop": true
1131
+ },
1132
+ "attributes": {}
1133
+ }
1134
+ },
1135
+ "total_flos": 0.0,
1136
+ "train_batch_size": 32,
1137
+ "trial_name": null,
1138
+ "trial_params": null
1139
+ }