tmnam20 commited on
Commit
5800aca
1 Parent(s): c647027

Training in progress, step 46000, checkpoint

Browse files
checkpoint-46000/added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "<bos>": 40478,
3
+ "<eos>": 40479,
4
+ "<pad>": 40480
5
+ }
checkpoint-46000/config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "openai-gpt",
3
+ "afn": "gelu",
4
+ "architectures": [
5
+ "OpenAIGPTForSequenceClassification"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "embd_pdrop": 0.1,
9
+ "id2label": {
10
+ "0": "entailment",
11
+ "1": "neutral",
12
+ "2": "contradiction"
13
+ },
14
+ "initializer_range": 0.02,
15
+ "label2id": {
16
+ "contradiction": 2,
17
+ "entailment": 0,
18
+ "neutral": 1
19
+ },
20
+ "layer_norm_epsilon": 1e-05,
21
+ "model_type": "openai-gpt",
22
+ "n_ctx": 512,
23
+ "n_embd": 768,
24
+ "n_head": 12,
25
+ "n_layer": 12,
26
+ "n_positions": 512,
27
+ "n_special": 0,
28
+ "pad_token_id": 40480,
29
+ "predict_special_tokens": true,
30
+ "problem_type": "single_label_classification",
31
+ "resid_pdrop": 0.1,
32
+ "summary_activation": null,
33
+ "summary_first_dropout": 0.1,
34
+ "summary_proj_to_labels": true,
35
+ "summary_type": "last",
36
+ "summary_use_proj": true,
37
+ "task_specific_params": {
38
+ "text-generation": {
39
+ "do_sample": true,
40
+ "max_length": 50
41
+ }
42
+ },
43
+ "torch_dtype": "float32",
44
+ "transformers_version": "4.33.0.dev0",
45
+ "vocab_size": 40481
46
+ }
checkpoint-46000/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-46000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:44e01ea4f016213d6583feda10af752cf6040ce872fd56daf24b84b2a5b98ce8
3
+ size 932401285
checkpoint-46000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60efa4ed5e74be5f557d514a6cc74dfe7092588bea3e792abebdcd4bc35f2c37
3
+ size 466205145
checkpoint-46000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:830893e99d99973755fb2a56a922249ce0f9afac803b3fd5a62f7b108dd3b34a
3
+ size 14575
checkpoint-46000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2eb57ec33a2cc765c451f70f66ff224d39cafd5dd2eb8495ecf81985caa9b5e3
3
+ size 627
checkpoint-46000/special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<bos>",
3
+ "eos_token": "<eos>",
4
+ "pad_token": "<pad>",
5
+ "unk_token": "<unk>"
6
+ }
checkpoint-46000/tokenizer_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "clean_up_tokenization_spaces": true,
3
+ "model_max_length": 512,
4
+ "padding_side": "left",
5
+ "tokenizer_class": "OpenAIGPTTokenizer",
6
+ "unk_token": "<unk>"
7
+ }
checkpoint-46000/trainer_state.json ADDED
@@ -0,0 +1,715 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9766710996627808,
3
+ "best_model_checkpoint": "/kaggle/output/checkpoint-46000",
4
+ "epoch": 1.8741851368970013,
5
+ "eval_steps": 1000,
6
+ "global_step": 46000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0,
13
+ "learning_rate": 2.7777777777777777e-11,
14
+ "loss": 1.029,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.04,
19
+ "learning_rate": 2.7638888888888893e-08,
20
+ "loss": 1.181,
21
+ "step": 1000
22
+ },
23
+ {
24
+ "epoch": 0.04,
25
+ "eval_accuracy": 0.3273453093812375,
26
+ "eval_loss": 1.1529844999313354,
27
+ "eval_runtime": 54.2837,
28
+ "eval_samples_per_second": 92.293,
29
+ "eval_steps_per_second": 11.55,
30
+ "step": 1000
31
+ },
32
+ {
33
+ "epoch": 0.08,
34
+ "learning_rate": 5.541666666666667e-08,
35
+ "loss": 1.1527,
36
+ "step": 2000
37
+ },
38
+ {
39
+ "epoch": 0.08,
40
+ "eval_accuracy": 0.33013972055888224,
41
+ "eval_loss": 1.1351025104522705,
42
+ "eval_runtime": 54.0518,
43
+ "eval_samples_per_second": 92.689,
44
+ "eval_steps_per_second": 11.6,
45
+ "step": 2000
46
+ },
47
+ {
48
+ "epoch": 0.12,
49
+ "learning_rate": 8.316666666666666e-08,
50
+ "loss": 1.142,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.12,
55
+ "eval_accuracy": 0.3317365269461078,
56
+ "eval_loss": 1.127414345741272,
57
+ "eval_runtime": 54.0871,
58
+ "eval_samples_per_second": 92.628,
59
+ "eval_steps_per_second": 11.592,
60
+ "step": 3000
61
+ },
62
+ {
63
+ "epoch": 0.16,
64
+ "learning_rate": 1.1091666666666668e-07,
65
+ "loss": 1.1371,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.16,
70
+ "eval_accuracy": 0.331936127744511,
71
+ "eval_loss": 1.1194497346878052,
72
+ "eval_runtime": 54.3907,
73
+ "eval_samples_per_second": 92.111,
74
+ "eval_steps_per_second": 11.528,
75
+ "step": 4000
76
+ },
77
+ {
78
+ "epoch": 0.2,
79
+ "learning_rate": 1.3869444444444447e-07,
80
+ "loss": 1.1246,
81
+ "step": 5000
82
+ },
83
+ {
84
+ "epoch": 0.2,
85
+ "eval_accuracy": 0.3345309381237525,
86
+ "eval_loss": 1.1131775379180908,
87
+ "eval_runtime": 54.3492,
88
+ "eval_samples_per_second": 92.182,
89
+ "eval_steps_per_second": 11.537,
90
+ "step": 5000
91
+ },
92
+ {
93
+ "epoch": 0.24,
94
+ "learning_rate": 1.664722222222222e-07,
95
+ "loss": 1.1201,
96
+ "step": 6000
97
+ },
98
+ {
99
+ "epoch": 0.24,
100
+ "eval_accuracy": 0.34151696606786425,
101
+ "eval_loss": 1.1093512773513794,
102
+ "eval_runtime": 54.3536,
103
+ "eval_samples_per_second": 92.174,
104
+ "eval_steps_per_second": 11.536,
105
+ "step": 6000
106
+ },
107
+ {
108
+ "epoch": 0.29,
109
+ "learning_rate": 1.9425e-07,
110
+ "loss": 1.115,
111
+ "step": 7000
112
+ },
113
+ {
114
+ "epoch": 0.29,
115
+ "eval_accuracy": 0.34510978043912177,
116
+ "eval_loss": 1.1066845655441284,
117
+ "eval_runtime": 53.9977,
118
+ "eval_samples_per_second": 92.782,
119
+ "eval_steps_per_second": 11.612,
120
+ "step": 7000
121
+ },
122
+ {
123
+ "epoch": 0.33,
124
+ "learning_rate": 2.2202777777777778e-07,
125
+ "loss": 1.1125,
126
+ "step": 8000
127
+ },
128
+ {
129
+ "epoch": 0.33,
130
+ "eval_accuracy": 0.3449101796407186,
131
+ "eval_loss": 1.1031057834625244,
132
+ "eval_runtime": 54.2562,
133
+ "eval_samples_per_second": 92.34,
134
+ "eval_steps_per_second": 11.556,
135
+ "step": 8000
136
+ },
137
+ {
138
+ "epoch": 0.37,
139
+ "learning_rate": 2.498055555555556e-07,
140
+ "loss": 1.1072,
141
+ "step": 9000
142
+ },
143
+ {
144
+ "epoch": 0.37,
145
+ "eval_accuracy": 0.34870259481037924,
146
+ "eval_loss": 1.1044224500656128,
147
+ "eval_runtime": 54.0795,
148
+ "eval_samples_per_second": 92.641,
149
+ "eval_steps_per_second": 11.594,
150
+ "step": 9000
151
+ },
152
+ {
153
+ "epoch": 0.41,
154
+ "learning_rate": 2.7755555555555554e-07,
155
+ "loss": 1.1064,
156
+ "step": 10000
157
+ },
158
+ {
159
+ "epoch": 0.41,
160
+ "eval_accuracy": 0.3499001996007984,
161
+ "eval_loss": 1.1001031398773193,
162
+ "eval_runtime": 54.0869,
163
+ "eval_samples_per_second": 92.629,
164
+ "eval_steps_per_second": 11.592,
165
+ "step": 10000
166
+ },
167
+ {
168
+ "epoch": 0.45,
169
+ "learning_rate": 3.053333333333333e-07,
170
+ "loss": 1.1053,
171
+ "step": 11000
172
+ },
173
+ {
174
+ "epoch": 0.45,
175
+ "eval_accuracy": 0.36087824351297404,
176
+ "eval_loss": 1.0987293720245361,
177
+ "eval_runtime": 53.9688,
178
+ "eval_samples_per_second": 92.831,
179
+ "eval_steps_per_second": 11.618,
180
+ "step": 11000
181
+ },
182
+ {
183
+ "epoch": 0.49,
184
+ "learning_rate": 3.3311111111111114e-07,
185
+ "loss": 1.104,
186
+ "step": 12000
187
+ },
188
+ {
189
+ "epoch": 0.49,
190
+ "eval_accuracy": 0.3594810379241517,
191
+ "eval_loss": 1.0968631505966187,
192
+ "eval_runtime": 54.0936,
193
+ "eval_samples_per_second": 92.617,
194
+ "eval_steps_per_second": 11.591,
195
+ "step": 12000
196
+ },
197
+ {
198
+ "epoch": 0.53,
199
+ "learning_rate": 3.608888888888889e-07,
200
+ "loss": 1.1022,
201
+ "step": 13000
202
+ },
203
+ {
204
+ "epoch": 0.53,
205
+ "eval_accuracy": 0.36087824351297404,
206
+ "eval_loss": 1.0962979793548584,
207
+ "eval_runtime": 54.0403,
208
+ "eval_samples_per_second": 92.709,
209
+ "eval_steps_per_second": 11.602,
210
+ "step": 13000
211
+ },
212
+ {
213
+ "epoch": 0.57,
214
+ "learning_rate": 3.886666666666667e-07,
215
+ "loss": 1.1029,
216
+ "step": 14000
217
+ },
218
+ {
219
+ "epoch": 0.57,
220
+ "eval_accuracy": 0.3528942115768463,
221
+ "eval_loss": 1.0987964868545532,
222
+ "eval_runtime": 54.06,
223
+ "eval_samples_per_second": 92.675,
224
+ "eval_steps_per_second": 11.598,
225
+ "step": 14000
226
+ },
227
+ {
228
+ "epoch": 0.61,
229
+ "learning_rate": 4.163611111111111e-07,
230
+ "loss": 1.1011,
231
+ "step": 15000
232
+ },
233
+ {
234
+ "epoch": 0.61,
235
+ "eval_accuracy": 0.36926147704590817,
236
+ "eval_loss": 1.0941977500915527,
237
+ "eval_runtime": 54.0807,
238
+ "eval_samples_per_second": 92.639,
239
+ "eval_steps_per_second": 11.594,
240
+ "step": 15000
241
+ },
242
+ {
243
+ "epoch": 0.65,
244
+ "learning_rate": 4.441388888888889e-07,
245
+ "loss": 1.1004,
246
+ "step": 16000
247
+ },
248
+ {
249
+ "epoch": 0.65,
250
+ "eval_accuracy": 0.36866267465069863,
251
+ "eval_loss": 1.093828558921814,
252
+ "eval_runtime": 54.2208,
253
+ "eval_samples_per_second": 92.4,
254
+ "eval_steps_per_second": 11.564,
255
+ "step": 16000
256
+ },
257
+ {
258
+ "epoch": 0.69,
259
+ "learning_rate": 4.719166666666667e-07,
260
+ "loss": 1.0994,
261
+ "step": 17000
262
+ },
263
+ {
264
+ "epoch": 0.69,
265
+ "eval_accuracy": 0.36726546906187624,
266
+ "eval_loss": 1.093470811843872,
267
+ "eval_runtime": 54.2584,
268
+ "eval_samples_per_second": 92.336,
269
+ "eval_steps_per_second": 11.556,
270
+ "step": 17000
271
+ },
272
+ {
273
+ "epoch": 0.73,
274
+ "learning_rate": 4.996944444444445e-07,
275
+ "loss": 1.0993,
276
+ "step": 18000
277
+ },
278
+ {
279
+ "epoch": 0.73,
280
+ "eval_accuracy": 0.36327345309381237,
281
+ "eval_loss": 1.0935499668121338,
282
+ "eval_runtime": 54.0554,
283
+ "eval_samples_per_second": 92.683,
284
+ "eval_steps_per_second": 11.599,
285
+ "step": 18000
286
+ },
287
+ {
288
+ "epoch": 0.77,
289
+ "learning_rate": 5.274722222222222e-07,
290
+ "loss": 1.0981,
291
+ "step": 19000
292
+ },
293
+ {
294
+ "epoch": 0.77,
295
+ "eval_accuracy": 0.38682634730538923,
296
+ "eval_loss": 1.089295506477356,
297
+ "eval_runtime": 54.0761,
298
+ "eval_samples_per_second": 92.647,
299
+ "eval_steps_per_second": 11.595,
300
+ "step": 19000
301
+ },
302
+ {
303
+ "epoch": 0.81,
304
+ "learning_rate": 5.5525e-07,
305
+ "loss": 1.0979,
306
+ "step": 20000
307
+ },
308
+ {
309
+ "epoch": 0.81,
310
+ "eval_accuracy": 0.4167664670658683,
311
+ "eval_loss": 1.084443211555481,
312
+ "eval_runtime": 54.0394,
313
+ "eval_samples_per_second": 92.71,
314
+ "eval_steps_per_second": 11.603,
315
+ "step": 20000
316
+ },
317
+ {
318
+ "epoch": 0.86,
319
+ "learning_rate": 5.830000000000001e-07,
320
+ "loss": 1.0946,
321
+ "step": 21000
322
+ },
323
+ {
324
+ "epoch": 0.86,
325
+ "eval_accuracy": 0.4285429141716567,
326
+ "eval_loss": 1.073406457901001,
327
+ "eval_runtime": 54.1561,
328
+ "eval_samples_per_second": 92.51,
329
+ "eval_steps_per_second": 11.578,
330
+ "step": 21000
331
+ },
332
+ {
333
+ "epoch": 0.9,
334
+ "learning_rate": 6.107777777777778e-07,
335
+ "loss": 1.0882,
336
+ "step": 22000
337
+ },
338
+ {
339
+ "epoch": 0.9,
340
+ "eval_accuracy": 0.4413173652694611,
341
+ "eval_loss": 1.0576634407043457,
342
+ "eval_runtime": 53.9715,
343
+ "eval_samples_per_second": 92.827,
344
+ "eval_steps_per_second": 11.617,
345
+ "step": 22000
346
+ },
347
+ {
348
+ "epoch": 0.94,
349
+ "learning_rate": 6.385277777777778e-07,
350
+ "loss": 1.0811,
351
+ "step": 23000
352
+ },
353
+ {
354
+ "epoch": 0.94,
355
+ "eval_accuracy": 0.4377245508982036,
356
+ "eval_loss": 1.063384771347046,
357
+ "eval_runtime": 54.4655,
358
+ "eval_samples_per_second": 91.985,
359
+ "eval_steps_per_second": 11.512,
360
+ "step": 23000
361
+ },
362
+ {
363
+ "epoch": 0.98,
364
+ "learning_rate": 6.662777777777778e-07,
365
+ "loss": 1.0773,
366
+ "step": 24000
367
+ },
368
+ {
369
+ "epoch": 0.98,
370
+ "eval_accuracy": 0.4347305389221557,
371
+ "eval_loss": 1.065600037574768,
372
+ "eval_runtime": 54.0545,
373
+ "eval_samples_per_second": 92.684,
374
+ "eval_steps_per_second": 11.599,
375
+ "step": 24000
376
+ },
377
+ {
378
+ "epoch": 1.02,
379
+ "learning_rate": 6.940555555555557e-07,
380
+ "loss": 1.0749,
381
+ "step": 25000
382
+ },
383
+ {
384
+ "epoch": 1.02,
385
+ "eval_accuracy": 0.4499001996007984,
386
+ "eval_loss": 1.0491468906402588,
387
+ "eval_runtime": 53.9397,
388
+ "eval_samples_per_second": 92.882,
389
+ "eval_steps_per_second": 11.624,
390
+ "step": 25000
391
+ },
392
+ {
393
+ "epoch": 1.06,
394
+ "learning_rate": 7.218055555555556e-07,
395
+ "loss": 1.0687,
396
+ "step": 26000
397
+ },
398
+ {
399
+ "epoch": 1.06,
400
+ "eval_accuracy": 0.44471057884231535,
401
+ "eval_loss": 1.0531450510025024,
402
+ "eval_runtime": 53.996,
403
+ "eval_samples_per_second": 92.785,
404
+ "eval_steps_per_second": 11.612,
405
+ "step": 26000
406
+ },
407
+ {
408
+ "epoch": 1.1,
409
+ "learning_rate": 7.495833333333334e-07,
410
+ "loss": 1.07,
411
+ "step": 27000
412
+ },
413
+ {
414
+ "epoch": 1.1,
415
+ "eval_accuracy": 0.4475049900199601,
416
+ "eval_loss": 1.0514291524887085,
417
+ "eval_runtime": 54.0862,
418
+ "eval_samples_per_second": 92.63,
419
+ "eval_steps_per_second": 11.593,
420
+ "step": 27000
421
+ },
422
+ {
423
+ "epoch": 1.14,
424
+ "learning_rate": 7.773611111111112e-07,
425
+ "loss": 1.0619,
426
+ "step": 28000
427
+ },
428
+ {
429
+ "epoch": 1.14,
430
+ "eval_accuracy": 0.4483033932135729,
431
+ "eval_loss": 1.0422637462615967,
432
+ "eval_runtime": 54.3061,
433
+ "eval_samples_per_second": 92.255,
434
+ "eval_steps_per_second": 11.546,
435
+ "step": 28000
436
+ },
437
+ {
438
+ "epoch": 1.18,
439
+ "learning_rate": 8.051388888888889e-07,
440
+ "loss": 1.0586,
441
+ "step": 29000
442
+ },
443
+ {
444
+ "epoch": 1.18,
445
+ "eval_accuracy": 0.4544910179640719,
446
+ "eval_loss": 1.0364656448364258,
447
+ "eval_runtime": 54.1221,
448
+ "eval_samples_per_second": 92.568,
449
+ "eval_steps_per_second": 11.585,
450
+ "step": 29000
451
+ },
452
+ {
453
+ "epoch": 1.22,
454
+ "learning_rate": 8.32888888888889e-07,
455
+ "loss": 1.0571,
456
+ "step": 30000
457
+ },
458
+ {
459
+ "epoch": 1.22,
460
+ "eval_accuracy": 0.4594810379241517,
461
+ "eval_loss": 1.0300606489181519,
462
+ "eval_runtime": 54.1093,
463
+ "eval_samples_per_second": 92.59,
464
+ "eval_steps_per_second": 11.588,
465
+ "step": 30000
466
+ },
467
+ {
468
+ "epoch": 1.26,
469
+ "learning_rate": 8.606666666666668e-07,
470
+ "loss": 1.0564,
471
+ "step": 31000
472
+ },
473
+ {
474
+ "epoch": 1.26,
475
+ "eval_accuracy": 0.4666666666666667,
476
+ "eval_loss": 1.0317378044128418,
477
+ "eval_runtime": 54.0415,
478
+ "eval_samples_per_second": 92.707,
479
+ "eval_steps_per_second": 11.602,
480
+ "step": 31000
481
+ },
482
+ {
483
+ "epoch": 1.3,
484
+ "learning_rate": 8.884444444444445e-07,
485
+ "loss": 1.0541,
486
+ "step": 32000
487
+ },
488
+ {
489
+ "epoch": 1.3,
490
+ "eval_accuracy": 0.47065868263473054,
491
+ "eval_loss": 1.0293121337890625,
492
+ "eval_runtime": 54.1157,
493
+ "eval_samples_per_second": 92.579,
494
+ "eval_steps_per_second": 11.586,
495
+ "step": 32000
496
+ },
497
+ {
498
+ "epoch": 1.34,
499
+ "learning_rate": 9.161944444444445e-07,
500
+ "loss": 1.0485,
501
+ "step": 33000
502
+ },
503
+ {
504
+ "epoch": 1.34,
505
+ "eval_accuracy": 0.480439121756487,
506
+ "eval_loss": 1.0269454717636108,
507
+ "eval_runtime": 54.0627,
508
+ "eval_samples_per_second": 92.67,
509
+ "eval_steps_per_second": 11.598,
510
+ "step": 33000
511
+ },
512
+ {
513
+ "epoch": 1.39,
514
+ "learning_rate": 9.439722222222223e-07,
515
+ "loss": 1.0469,
516
+ "step": 34000
517
+ },
518
+ {
519
+ "epoch": 1.39,
520
+ "eval_accuracy": 0.4780439121756487,
521
+ "eval_loss": 1.026614785194397,
522
+ "eval_runtime": 54.0605,
523
+ "eval_samples_per_second": 92.674,
524
+ "eval_steps_per_second": 11.598,
525
+ "step": 34000
526
+ },
527
+ {
528
+ "epoch": 1.43,
529
+ "learning_rate": 9.7175e-07,
530
+ "loss": 1.0479,
531
+ "step": 35000
532
+ },
533
+ {
534
+ "epoch": 1.43,
535
+ "eval_accuracy": 0.48423153692614773,
536
+ "eval_loss": 1.0213700532913208,
537
+ "eval_runtime": 54.1341,
538
+ "eval_samples_per_second": 92.548,
539
+ "eval_steps_per_second": 11.582,
540
+ "step": 35000
541
+ },
542
+ {
543
+ "epoch": 1.47,
544
+ "learning_rate": 9.995277777777778e-07,
545
+ "loss": 1.0433,
546
+ "step": 36000
547
+ },
548
+ {
549
+ "epoch": 1.47,
550
+ "eval_accuracy": 0.49101796407185627,
551
+ "eval_loss": 1.0118540525436401,
552
+ "eval_runtime": 54.1201,
553
+ "eval_samples_per_second": 92.572,
554
+ "eval_steps_per_second": 11.585,
555
+ "step": 36000
556
+ },
557
+ {
558
+ "epoch": 1.51,
559
+ "learning_rate": 1.0273055555555556e-06,
560
+ "loss": 1.0368,
561
+ "step": 37000
562
+ },
563
+ {
564
+ "epoch": 1.51,
565
+ "eval_accuracy": 0.49101796407185627,
566
+ "eval_loss": 1.0219311714172363,
567
+ "eval_runtime": 54.1135,
568
+ "eval_samples_per_second": 92.583,
569
+ "eval_steps_per_second": 11.587,
570
+ "step": 37000
571
+ },
572
+ {
573
+ "epoch": 1.55,
574
+ "learning_rate": 1.0550555555555557e-06,
575
+ "loss": 1.0401,
576
+ "step": 38000
577
+ },
578
+ {
579
+ "epoch": 1.55,
580
+ "eval_accuracy": 0.49720558882235527,
581
+ "eval_loss": 1.013930320739746,
582
+ "eval_runtime": 54.0309,
583
+ "eval_samples_per_second": 92.725,
584
+ "eval_steps_per_second": 11.604,
585
+ "step": 38000
586
+ },
587
+ {
588
+ "epoch": 1.59,
589
+ "learning_rate": 1.0828333333333334e-06,
590
+ "loss": 1.0341,
591
+ "step": 39000
592
+ },
593
+ {
594
+ "epoch": 1.59,
595
+ "eval_accuracy": 0.4992015968063872,
596
+ "eval_loss": 1.0132442712783813,
597
+ "eval_runtime": 54.0478,
598
+ "eval_samples_per_second": 92.696,
599
+ "eval_steps_per_second": 11.601,
600
+ "step": 39000
601
+ },
602
+ {
603
+ "epoch": 1.63,
604
+ "learning_rate": 1.1106111111111112e-06,
605
+ "loss": 1.0254,
606
+ "step": 40000
607
+ },
608
+ {
609
+ "epoch": 1.63,
610
+ "eval_accuracy": 0.5095808383233533,
611
+ "eval_loss": 1.0034514665603638,
612
+ "eval_runtime": 54.1138,
613
+ "eval_samples_per_second": 92.583,
614
+ "eval_steps_per_second": 11.587,
615
+ "step": 40000
616
+ },
617
+ {
618
+ "epoch": 1.67,
619
+ "learning_rate": 1.1383611111111113e-06,
620
+ "loss": 1.0264,
621
+ "step": 41000
622
+ },
623
+ {
624
+ "epoch": 1.67,
625
+ "eval_accuracy": 0.5159680638722555,
626
+ "eval_loss": 1.0014362335205078,
627
+ "eval_runtime": 54.2855,
628
+ "eval_samples_per_second": 92.29,
629
+ "eval_steps_per_second": 11.55,
630
+ "step": 41000
631
+ },
632
+ {
633
+ "epoch": 1.71,
634
+ "learning_rate": 1.166138888888889e-06,
635
+ "loss": 1.0222,
636
+ "step": 42000
637
+ },
638
+ {
639
+ "epoch": 1.71,
640
+ "eval_accuracy": 0.5183632734530939,
641
+ "eval_loss": 0.9864664077758789,
642
+ "eval_runtime": 54.1055,
643
+ "eval_samples_per_second": 92.597,
644
+ "eval_steps_per_second": 11.588,
645
+ "step": 42000
646
+ },
647
+ {
648
+ "epoch": 1.75,
649
+ "learning_rate": 1.193888888888889e-06,
650
+ "loss": 1.017,
651
+ "step": 43000
652
+ },
653
+ {
654
+ "epoch": 1.75,
655
+ "eval_accuracy": 0.5153692614770459,
656
+ "eval_loss": 0.995494544506073,
657
+ "eval_runtime": 54.2838,
658
+ "eval_samples_per_second": 92.293,
659
+ "eval_steps_per_second": 11.55,
660
+ "step": 43000
661
+ },
662
+ {
663
+ "epoch": 1.79,
664
+ "learning_rate": 1.2216666666666667e-06,
665
+ "loss": 1.0225,
666
+ "step": 44000
667
+ },
668
+ {
669
+ "epoch": 1.79,
670
+ "eval_accuracy": 0.5165668662674651,
671
+ "eval_loss": 0.9881771802902222,
672
+ "eval_runtime": 54.6114,
673
+ "eval_samples_per_second": 91.739,
674
+ "eval_steps_per_second": 11.481,
675
+ "step": 44000
676
+ },
677
+ {
678
+ "epoch": 1.83,
679
+ "learning_rate": 1.2494444444444447e-06,
680
+ "loss": 1.0149,
681
+ "step": 45000
682
+ },
683
+ {
684
+ "epoch": 1.83,
685
+ "eval_accuracy": 0.5265469061876248,
686
+ "eval_loss": 0.9803952574729919,
687
+ "eval_runtime": 54.6026,
688
+ "eval_samples_per_second": 91.754,
689
+ "eval_steps_per_second": 11.483,
690
+ "step": 45000
691
+ },
692
+ {
693
+ "epoch": 1.87,
694
+ "learning_rate": 1.2772222222222224e-06,
695
+ "loss": 1.017,
696
+ "step": 46000
697
+ },
698
+ {
699
+ "epoch": 1.87,
700
+ "eval_accuracy": 0.5331337325349301,
701
+ "eval_loss": 0.9766710996627808,
702
+ "eval_runtime": 54.449,
703
+ "eval_samples_per_second": 92.013,
704
+ "eval_steps_per_second": 11.515,
705
+ "step": 46000
706
+ }
707
+ ],
708
+ "logging_steps": 1000,
709
+ "max_steps": 10000000,
710
+ "num_train_epochs": 408,
711
+ "save_steps": 1000,
712
+ "total_flos": 9.61560747025367e+16,
713
+ "trial_name": null,
714
+ "trial_params": null
715
+ }
checkpoint-46000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5520446b18da09f4a43173b75839457981a344c0e709fd4bbcb249e3f60c859e
3
+ size 4091
checkpoint-46000/vocab.json ADDED
The diff for this file is too large to render. See raw diff