tmnam20 commited on
Commit
3218049
1 Parent(s): 9da50bf

Training in progress, step 79000, checkpoint

Browse files
checkpoint-79000/added_tokens.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "<bos>": 40478,
3
+ "<eos>": 40479,
4
+ "<pad>": 40480
5
+ }
checkpoint-79000/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": 6,
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": "cls_index",
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-79000/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-79000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:211f6c012485054292d59bacc2e6a28ba00d90724add15c24223f36112f657e9
3
+ size 932401605
checkpoint-79000/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8729582a37fac674f0315d918e8b9fb5d8d0756fc5a9a4665fd9f3b6f63cbc31
3
+ size 466205145
checkpoint-79000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d3ed3209c4688737222e13b4a9fa908e7a28f4c366b614209aee1b515dad314
3
+ size 14575
checkpoint-79000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11120de6fb20a07ab4757070c51873e81681c27a22c66e2413b4a31876d92d9b
3
+ size 627
checkpoint-79000/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-79000/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-79000/trainer_state.json ADDED
@@ -0,0 +1,1210 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.9642478823661804,
3
+ "best_model_checkpoint": "/kaggle/output/checkpoint-78000",
4
+ "epoch": 3.21870925684485,
5
+ "eval_steps": 1000,
6
+ "global_step": 79000,
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": 1.2076,
15
+ "step": 1
16
+ },
17
+ {
18
+ "epoch": 0.04,
19
+ "learning_rate": 2.7750000000000004e-08,
20
+ "loss": 1.1301,
21
+ "step": 1000
22
+ },
23
+ {
24
+ "epoch": 0.04,
25
+ "eval_accuracy": 0.3321357285429142,
26
+ "eval_loss": 1.1047232151031494,
27
+ "eval_runtime": 50.8344,
28
+ "eval_samples_per_second": 98.555,
29
+ "eval_steps_per_second": 12.334,
30
+ "step": 1000
31
+ },
32
+ {
33
+ "epoch": 0.08,
34
+ "learning_rate": 5.5527777777777784e-08,
35
+ "loss": 1.1286,
36
+ "step": 2000
37
+ },
38
+ {
39
+ "epoch": 0.08,
40
+ "eval_accuracy": 0.3377245508982036,
41
+ "eval_loss": 1.1038333177566528,
42
+ "eval_runtime": 50.8114,
43
+ "eval_samples_per_second": 98.6,
44
+ "eval_steps_per_second": 12.34,
45
+ "step": 2000
46
+ },
47
+ {
48
+ "epoch": 0.12,
49
+ "learning_rate": 8.327777777777778e-08,
50
+ "loss": 1.1214,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.12,
55
+ "eval_accuracy": 0.3409181636726547,
56
+ "eval_loss": 1.103529930114746,
57
+ "eval_runtime": 50.8258,
58
+ "eval_samples_per_second": 98.572,
59
+ "eval_steps_per_second": 12.336,
60
+ "step": 3000
61
+ },
62
+ {
63
+ "epoch": 0.16,
64
+ "learning_rate": 1.1105555555555557e-07,
65
+ "loss": 1.1249,
66
+ "step": 4000
67
+ },
68
+ {
69
+ "epoch": 0.16,
70
+ "eval_accuracy": 0.3469061876247505,
71
+ "eval_loss": 1.1000434160232544,
72
+ "eval_runtime": 50.6414,
73
+ "eval_samples_per_second": 98.931,
74
+ "eval_steps_per_second": 12.381,
75
+ "step": 4000
76
+ },
77
+ {
78
+ "epoch": 0.2,
79
+ "learning_rate": 1.3880555555555558e-07,
80
+ "loss": 1.1206,
81
+ "step": 5000
82
+ },
83
+ {
84
+ "epoch": 0.2,
85
+ "eval_accuracy": 0.34171656686626745,
86
+ "eval_loss": 1.0996097326278687,
87
+ "eval_runtime": 50.6956,
88
+ "eval_samples_per_second": 98.825,
89
+ "eval_steps_per_second": 12.368,
90
+ "step": 5000
91
+ },
92
+ {
93
+ "epoch": 0.24,
94
+ "learning_rate": 1.6658333333333335e-07,
95
+ "loss": 1.1161,
96
+ "step": 6000
97
+ },
98
+ {
99
+ "epoch": 0.24,
100
+ "eval_accuracy": 0.3590818363273453,
101
+ "eval_loss": 1.096977710723877,
102
+ "eval_runtime": 50.5826,
103
+ "eval_samples_per_second": 99.046,
104
+ "eval_steps_per_second": 12.396,
105
+ "step": 6000
106
+ },
107
+ {
108
+ "epoch": 0.29,
109
+ "learning_rate": 1.9433333333333334e-07,
110
+ "loss": 1.1138,
111
+ "step": 7000
112
+ },
113
+ {
114
+ "epoch": 0.29,
115
+ "eval_accuracy": 0.3532934131736527,
116
+ "eval_loss": 1.0949815511703491,
117
+ "eval_runtime": 50.5604,
118
+ "eval_samples_per_second": 99.089,
119
+ "eval_steps_per_second": 12.401,
120
+ "step": 7000
121
+ },
122
+ {
123
+ "epoch": 0.33,
124
+ "learning_rate": 2.2211111111111114e-07,
125
+ "loss": 1.1135,
126
+ "step": 8000
127
+ },
128
+ {
129
+ "epoch": 0.33,
130
+ "eval_accuracy": 0.36327345309381237,
131
+ "eval_loss": 1.093558430671692,
132
+ "eval_runtime": 50.608,
133
+ "eval_samples_per_second": 98.996,
134
+ "eval_steps_per_second": 12.389,
135
+ "step": 8000
136
+ },
137
+ {
138
+ "epoch": 0.37,
139
+ "learning_rate": 2.4986111111111113e-07,
140
+ "loss": 1.1063,
141
+ "step": 9000
142
+ },
143
+ {
144
+ "epoch": 0.37,
145
+ "eval_accuracy": 0.35548902195608784,
146
+ "eval_loss": 1.0992107391357422,
147
+ "eval_runtime": 50.5434,
148
+ "eval_samples_per_second": 99.123,
149
+ "eval_steps_per_second": 12.405,
150
+ "step": 9000
151
+ },
152
+ {
153
+ "epoch": 0.41,
154
+ "learning_rate": 2.776388888888889e-07,
155
+ "loss": 1.1083,
156
+ "step": 10000
157
+ },
158
+ {
159
+ "epoch": 0.41,
160
+ "eval_accuracy": 0.3836327345309381,
161
+ "eval_loss": 1.0920206308364868,
162
+ "eval_runtime": 50.5559,
163
+ "eval_samples_per_second": 99.098,
164
+ "eval_steps_per_second": 12.402,
165
+ "step": 10000
166
+ },
167
+ {
168
+ "epoch": 0.45,
169
+ "learning_rate": 3.053888888888889e-07,
170
+ "loss": 1.1044,
171
+ "step": 11000
172
+ },
173
+ {
174
+ "epoch": 0.45,
175
+ "eval_accuracy": 0.38762475049900197,
176
+ "eval_loss": 1.0904459953308105,
177
+ "eval_runtime": 50.9725,
178
+ "eval_samples_per_second": 98.288,
179
+ "eval_steps_per_second": 12.301,
180
+ "step": 11000
181
+ },
182
+ {
183
+ "epoch": 0.49,
184
+ "learning_rate": 3.331666666666667e-07,
185
+ "loss": 1.1019,
186
+ "step": 12000
187
+ },
188
+ {
189
+ "epoch": 0.49,
190
+ "eval_accuracy": 0.3912175648702595,
191
+ "eval_loss": 1.088543176651001,
192
+ "eval_runtime": 50.8297,
193
+ "eval_samples_per_second": 98.564,
194
+ "eval_steps_per_second": 12.335,
195
+ "step": 12000
196
+ },
197
+ {
198
+ "epoch": 0.53,
199
+ "learning_rate": 3.609166666666667e-07,
200
+ "loss": 1.1032,
201
+ "step": 13000
202
+ },
203
+ {
204
+ "epoch": 0.53,
205
+ "eval_accuracy": 0.3435129740518962,
206
+ "eval_loss": 1.0940682888031006,
207
+ "eval_runtime": 51.0031,
208
+ "eval_samples_per_second": 98.229,
209
+ "eval_steps_per_second": 12.293,
210
+ "step": 13000
211
+ },
212
+ {
213
+ "epoch": 0.57,
214
+ "learning_rate": 3.886944444444445e-07,
215
+ "loss": 1.1006,
216
+ "step": 14000
217
+ },
218
+ {
219
+ "epoch": 0.57,
220
+ "eval_accuracy": 0.36826347305389223,
221
+ "eval_loss": 1.0924910306930542,
222
+ "eval_runtime": 51.2753,
223
+ "eval_samples_per_second": 97.708,
224
+ "eval_steps_per_second": 12.228,
225
+ "step": 14000
226
+ },
227
+ {
228
+ "epoch": 0.61,
229
+ "learning_rate": 4.164444444444445e-07,
230
+ "loss": 1.1,
231
+ "step": 15000
232
+ },
233
+ {
234
+ "epoch": 0.61,
235
+ "eval_accuracy": 0.38323353293413176,
236
+ "eval_loss": 1.0879395008087158,
237
+ "eval_runtime": 51.3783,
238
+ "eval_samples_per_second": 97.512,
239
+ "eval_steps_per_second": 12.204,
240
+ "step": 15000
241
+ },
242
+ {
243
+ "epoch": 0.65,
244
+ "learning_rate": 4.442222222222223e-07,
245
+ "loss": 1.0982,
246
+ "step": 16000
247
+ },
248
+ {
249
+ "epoch": 0.65,
250
+ "eval_accuracy": 0.3964071856287425,
251
+ "eval_loss": 1.0822893381118774,
252
+ "eval_runtime": 51.2346,
253
+ "eval_samples_per_second": 97.786,
254
+ "eval_steps_per_second": 12.238,
255
+ "step": 16000
256
+ },
257
+ {
258
+ "epoch": 0.69,
259
+ "learning_rate": 4.7197222222222224e-07,
260
+ "loss": 1.0953,
261
+ "step": 17000
262
+ },
263
+ {
264
+ "epoch": 0.69,
265
+ "eval_accuracy": 0.4121756487025948,
266
+ "eval_loss": 1.0795228481292725,
267
+ "eval_runtime": 51.31,
268
+ "eval_samples_per_second": 97.642,
269
+ "eval_steps_per_second": 12.22,
270
+ "step": 17000
271
+ },
272
+ {
273
+ "epoch": 0.73,
274
+ "learning_rate": 4.997500000000001e-07,
275
+ "loss": 1.0943,
276
+ "step": 18000
277
+ },
278
+ {
279
+ "epoch": 0.73,
280
+ "eval_accuracy": 0.42015968063872255,
281
+ "eval_loss": 1.079458236694336,
282
+ "eval_runtime": 51.3255,
283
+ "eval_samples_per_second": 97.612,
284
+ "eval_steps_per_second": 12.216,
285
+ "step": 18000
286
+ },
287
+ {
288
+ "epoch": 0.77,
289
+ "learning_rate": 5.275e-07,
290
+ "loss": 1.0885,
291
+ "step": 19000
292
+ },
293
+ {
294
+ "epoch": 0.77,
295
+ "eval_accuracy": 0.437125748502994,
296
+ "eval_loss": 1.0667964220046997,
297
+ "eval_runtime": 51.1737,
298
+ "eval_samples_per_second": 97.902,
299
+ "eval_steps_per_second": 12.252,
300
+ "step": 19000
301
+ },
302
+ {
303
+ "epoch": 0.81,
304
+ "learning_rate": 5.552777777777778e-07,
305
+ "loss": 1.0792,
306
+ "step": 20000
307
+ },
308
+ {
309
+ "epoch": 0.81,
310
+ "eval_accuracy": 0.4315369261477046,
311
+ "eval_loss": 1.0628777742385864,
312
+ "eval_runtime": 51.4162,
313
+ "eval_samples_per_second": 97.44,
314
+ "eval_steps_per_second": 12.195,
315
+ "step": 20000
316
+ },
317
+ {
318
+ "epoch": 0.86,
319
+ "learning_rate": 5.830277777777779e-07,
320
+ "loss": 1.0755,
321
+ "step": 21000
322
+ },
323
+ {
324
+ "epoch": 0.86,
325
+ "eval_accuracy": 0.4343313373253493,
326
+ "eval_loss": 1.0604455471038818,
327
+ "eval_runtime": 51.2629,
328
+ "eval_samples_per_second": 97.731,
329
+ "eval_steps_per_second": 12.231,
330
+ "step": 21000
331
+ },
332
+ {
333
+ "epoch": 0.9,
334
+ "learning_rate": 6.108055555555556e-07,
335
+ "loss": 1.073,
336
+ "step": 22000
337
+ },
338
+ {
339
+ "epoch": 0.9,
340
+ "eval_accuracy": 0.43353293413173655,
341
+ "eval_loss": 1.0552711486816406,
342
+ "eval_runtime": 51.2402,
343
+ "eval_samples_per_second": 97.775,
344
+ "eval_steps_per_second": 12.236,
345
+ "step": 22000
346
+ },
347
+ {
348
+ "epoch": 0.94,
349
+ "learning_rate": 6.385555555555556e-07,
350
+ "loss": 1.0699,
351
+ "step": 23000
352
+ },
353
+ {
354
+ "epoch": 0.94,
355
+ "eval_accuracy": 0.4313373253493014,
356
+ "eval_loss": 1.052398920059204,
357
+ "eval_runtime": 51.1758,
358
+ "eval_samples_per_second": 97.898,
359
+ "eval_steps_per_second": 12.252,
360
+ "step": 23000
361
+ },
362
+ {
363
+ "epoch": 0.98,
364
+ "learning_rate": 6.663333333333334e-07,
365
+ "loss": 1.0695,
366
+ "step": 24000
367
+ },
368
+ {
369
+ "epoch": 0.98,
370
+ "eval_accuracy": 0.4407185628742515,
371
+ "eval_loss": 1.047132968902588,
372
+ "eval_runtime": 51.3549,
373
+ "eval_samples_per_second": 97.556,
374
+ "eval_steps_per_second": 12.209,
375
+ "step": 24000
376
+ },
377
+ {
378
+ "epoch": 1.02,
379
+ "learning_rate": 6.940833333333334e-07,
380
+ "loss": 1.0656,
381
+ "step": 25000
382
+ },
383
+ {
384
+ "epoch": 1.02,
385
+ "eval_accuracy": 0.4439121756487026,
386
+ "eval_loss": 1.045853853225708,
387
+ "eval_runtime": 51.1548,
388
+ "eval_samples_per_second": 97.938,
389
+ "eval_steps_per_second": 12.257,
390
+ "step": 25000
391
+ },
392
+ {
393
+ "epoch": 1.06,
394
+ "learning_rate": 7.218611111111112e-07,
395
+ "loss": 1.0645,
396
+ "step": 26000
397
+ },
398
+ {
399
+ "epoch": 1.06,
400
+ "eval_accuracy": 0.43273453093812375,
401
+ "eval_loss": 1.0467982292175293,
402
+ "eval_runtime": 51.1774,
403
+ "eval_samples_per_second": 97.895,
404
+ "eval_steps_per_second": 12.252,
405
+ "step": 26000
406
+ },
407
+ {
408
+ "epoch": 1.1,
409
+ "learning_rate": 7.496111111111112e-07,
410
+ "loss": 1.0639,
411
+ "step": 27000
412
+ },
413
+ {
414
+ "epoch": 1.1,
415
+ "eval_accuracy": 0.43313373253493015,
416
+ "eval_loss": 1.0436949729919434,
417
+ "eval_runtime": 51.1313,
418
+ "eval_samples_per_second": 97.983,
419
+ "eval_steps_per_second": 12.263,
420
+ "step": 27000
421
+ },
422
+ {
423
+ "epoch": 1.14,
424
+ "learning_rate": 7.77388888888889e-07,
425
+ "loss": 1.0638,
426
+ "step": 28000
427
+ },
428
+ {
429
+ "epoch": 1.14,
430
+ "eval_accuracy": 0.436127744510978,
431
+ "eval_loss": 1.0424726009368896,
432
+ "eval_runtime": 51.2359,
433
+ "eval_samples_per_second": 97.783,
434
+ "eval_steps_per_second": 12.238,
435
+ "step": 28000
436
+ },
437
+ {
438
+ "epoch": 1.18,
439
+ "learning_rate": 8.051388888888889e-07,
440
+ "loss": 1.0591,
441
+ "step": 29000
442
+ },
443
+ {
444
+ "epoch": 1.18,
445
+ "eval_accuracy": 0.43852295409181635,
446
+ "eval_loss": 1.0379680395126343,
447
+ "eval_runtime": 51.0917,
448
+ "eval_samples_per_second": 98.059,
449
+ "eval_steps_per_second": 12.272,
450
+ "step": 29000
451
+ },
452
+ {
453
+ "epoch": 1.22,
454
+ "learning_rate": 8.329166666666667e-07,
455
+ "loss": 1.0571,
456
+ "step": 30000
457
+ },
458
+ {
459
+ "epoch": 1.22,
460
+ "eval_accuracy": 0.43932135728542915,
461
+ "eval_loss": 1.034009575843811,
462
+ "eval_runtime": 51.187,
463
+ "eval_samples_per_second": 97.876,
464
+ "eval_steps_per_second": 12.249,
465
+ "step": 30000
466
+ },
467
+ {
468
+ "epoch": 1.26,
469
+ "learning_rate": 8.606666666666668e-07,
470
+ "loss": 1.0609,
471
+ "step": 31000
472
+ },
473
+ {
474
+ "epoch": 1.26,
475
+ "eval_accuracy": 0.4489021956087824,
476
+ "eval_loss": 1.0380228757858276,
477
+ "eval_runtime": 51.2533,
478
+ "eval_samples_per_second": 97.75,
479
+ "eval_steps_per_second": 12.233,
480
+ "step": 31000
481
+ },
482
+ {
483
+ "epoch": 1.3,
484
+ "learning_rate": 8.884444444444445e-07,
485
+ "loss": 1.0536,
486
+ "step": 32000
487
+ },
488
+ {
489
+ "epoch": 1.3,
490
+ "eval_accuracy": 0.4347305389221557,
491
+ "eval_loss": 1.035749912261963,
492
+ "eval_runtime": 51.1708,
493
+ "eval_samples_per_second": 97.907,
494
+ "eval_steps_per_second": 12.253,
495
+ "step": 32000
496
+ },
497
+ {
498
+ "epoch": 1.34,
499
+ "learning_rate": 9.161944444444445e-07,
500
+ "loss": 1.0526,
501
+ "step": 33000
502
+ },
503
+ {
504
+ "epoch": 1.34,
505
+ "eval_accuracy": 0.4471057884231537,
506
+ "eval_loss": 1.0298662185668945,
507
+ "eval_runtime": 51.2742,
508
+ "eval_samples_per_second": 97.71,
509
+ "eval_steps_per_second": 12.228,
510
+ "step": 33000
511
+ },
512
+ {
513
+ "epoch": 1.39,
514
+ "learning_rate": 9.439722222222223e-07,
515
+ "loss": 1.0454,
516
+ "step": 34000
517
+ },
518
+ {
519
+ "epoch": 1.39,
520
+ "eval_accuracy": 0.449500998003992,
521
+ "eval_loss": 1.028323769569397,
522
+ "eval_runtime": 51.2636,
523
+ "eval_samples_per_second": 97.73,
524
+ "eval_steps_per_second": 12.231,
525
+ "step": 34000
526
+ },
527
+ {
528
+ "epoch": 1.43,
529
+ "learning_rate": 9.717222222222224e-07,
530
+ "loss": 1.0507,
531
+ "step": 35000
532
+ },
533
+ {
534
+ "epoch": 1.43,
535
+ "eval_accuracy": 0.4554890219560878,
536
+ "eval_loss": 1.0231916904449463,
537
+ "eval_runtime": 51.2328,
538
+ "eval_samples_per_second": 97.789,
539
+ "eval_steps_per_second": 12.238,
540
+ "step": 35000
541
+ },
542
+ {
543
+ "epoch": 1.47,
544
+ "learning_rate": 9.995000000000001e-07,
545
+ "loss": 1.0462,
546
+ "step": 36000
547
+ },
548
+ {
549
+ "epoch": 1.47,
550
+ "eval_accuracy": 0.4590818363273453,
551
+ "eval_loss": 1.02061927318573,
552
+ "eval_runtime": 50.6455,
553
+ "eval_samples_per_second": 98.923,
554
+ "eval_steps_per_second": 12.38,
555
+ "step": 36000
556
+ },
557
+ {
558
+ "epoch": 1.51,
559
+ "learning_rate": 1.02725e-06,
560
+ "loss": 1.0423,
561
+ "step": 37000
562
+ },
563
+ {
564
+ "epoch": 1.51,
565
+ "eval_accuracy": 0.4722554890219561,
566
+ "eval_loss": 1.0115904808044434,
567
+ "eval_runtime": 51.0173,
568
+ "eval_samples_per_second": 98.202,
569
+ "eval_steps_per_second": 12.29,
570
+ "step": 37000
571
+ },
572
+ {
573
+ "epoch": 1.55,
574
+ "learning_rate": 1.055027777777778e-06,
575
+ "loss": 1.041,
576
+ "step": 38000
577
+ },
578
+ {
579
+ "epoch": 1.55,
580
+ "eval_accuracy": 0.47365269461077847,
581
+ "eval_loss": 1.0123876333236694,
582
+ "eval_runtime": 51.1777,
583
+ "eval_samples_per_second": 97.894,
584
+ "eval_steps_per_second": 12.251,
585
+ "step": 38000
586
+ },
587
+ {
588
+ "epoch": 1.59,
589
+ "learning_rate": 1.0827777777777778e-06,
590
+ "loss": 1.0422,
591
+ "step": 39000
592
+ },
593
+ {
594
+ "epoch": 1.59,
595
+ "eval_accuracy": 0.46367265469061875,
596
+ "eval_loss": 1.019059181213379,
597
+ "eval_runtime": 51.1786,
598
+ "eval_samples_per_second": 97.892,
599
+ "eval_steps_per_second": 12.251,
600
+ "step": 39000
601
+ },
602
+ {
603
+ "epoch": 1.63,
604
+ "learning_rate": 1.1105555555555556e-06,
605
+ "loss": 1.034,
606
+ "step": 40000
607
+ },
608
+ {
609
+ "epoch": 1.63,
610
+ "eval_accuracy": 0.4782435129740519,
611
+ "eval_loss": 1.0045363903045654,
612
+ "eval_runtime": 51.2481,
613
+ "eval_samples_per_second": 97.76,
614
+ "eval_steps_per_second": 12.235,
615
+ "step": 40000
616
+ },
617
+ {
618
+ "epoch": 1.67,
619
+ "learning_rate": 1.1383055555555557e-06,
620
+ "loss": 1.0349,
621
+ "step": 41000
622
+ },
623
+ {
624
+ "epoch": 1.67,
625
+ "eval_accuracy": 0.4550898203592814,
626
+ "eval_loss": 1.0253337621688843,
627
+ "eval_runtime": 51.219,
628
+ "eval_samples_per_second": 97.815,
629
+ "eval_steps_per_second": 12.242,
630
+ "step": 41000
631
+ },
632
+ {
633
+ "epoch": 1.71,
634
+ "learning_rate": 1.1660833333333334e-06,
635
+ "loss": 1.0338,
636
+ "step": 42000
637
+ },
638
+ {
639
+ "epoch": 1.71,
640
+ "eval_accuracy": 0.480439121756487,
641
+ "eval_loss": 1.000636339187622,
642
+ "eval_runtime": 51.1853,
643
+ "eval_samples_per_second": 97.88,
644
+ "eval_steps_per_second": 12.25,
645
+ "step": 42000
646
+ },
647
+ {
648
+ "epoch": 1.75,
649
+ "learning_rate": 1.1938333333333335e-06,
650
+ "loss": 1.031,
651
+ "step": 43000
652
+ },
653
+ {
654
+ "epoch": 1.75,
655
+ "eval_accuracy": 0.45888223552894214,
656
+ "eval_loss": 1.0288445949554443,
657
+ "eval_runtime": 51.295,
658
+ "eval_samples_per_second": 97.67,
659
+ "eval_steps_per_second": 12.223,
660
+ "step": 43000
661
+ },
662
+ {
663
+ "epoch": 1.79,
664
+ "learning_rate": 1.2216111111111113e-06,
665
+ "loss": 1.0327,
666
+ "step": 44000
667
+ },
668
+ {
669
+ "epoch": 1.79,
670
+ "eval_accuracy": 0.493812375249501,
671
+ "eval_loss": 0.9989358186721802,
672
+ "eval_runtime": 51.1981,
673
+ "eval_samples_per_second": 97.855,
674
+ "eval_steps_per_second": 12.247,
675
+ "step": 44000
676
+ },
677
+ {
678
+ "epoch": 1.83,
679
+ "learning_rate": 1.2493611111111112e-06,
680
+ "loss": 1.0321,
681
+ "step": 45000
682
+ },
683
+ {
684
+ "epoch": 1.83,
685
+ "eval_accuracy": 0.4870259481037924,
686
+ "eval_loss": 1.003555417060852,
687
+ "eval_runtime": 50.8148,
688
+ "eval_samples_per_second": 98.593,
689
+ "eval_steps_per_second": 12.339,
690
+ "step": 45000
691
+ },
692
+ {
693
+ "epoch": 1.87,
694
+ "learning_rate": 1.277138888888889e-06,
695
+ "loss": 1.0326,
696
+ "step": 46000
697
+ },
698
+ {
699
+ "epoch": 1.87,
700
+ "eval_accuracy": 0.4810379241516966,
701
+ "eval_loss": 1.00687575340271,
702
+ "eval_runtime": 50.7043,
703
+ "eval_samples_per_second": 98.808,
704
+ "eval_steps_per_second": 12.366,
705
+ "step": 46000
706
+ },
707
+ {
708
+ "epoch": 1.91,
709
+ "learning_rate": 1.304888888888889e-06,
710
+ "loss": 1.0266,
711
+ "step": 47000
712
+ },
713
+ {
714
+ "epoch": 1.91,
715
+ "eval_accuracy": 0.49840319361277446,
716
+ "eval_loss": 0.9924121499061584,
717
+ "eval_runtime": 50.7174,
718
+ "eval_samples_per_second": 98.783,
719
+ "eval_steps_per_second": 12.363,
720
+ "step": 47000
721
+ },
722
+ {
723
+ "epoch": 1.96,
724
+ "learning_rate": 1.3326666666666668e-06,
725
+ "loss": 1.0285,
726
+ "step": 48000
727
+ },
728
+ {
729
+ "epoch": 1.96,
730
+ "eval_accuracy": 0.45788423153692615,
731
+ "eval_loss": 1.0277597904205322,
732
+ "eval_runtime": 50.7995,
733
+ "eval_samples_per_second": 98.623,
734
+ "eval_steps_per_second": 12.343,
735
+ "step": 48000
736
+ },
737
+ {
738
+ "epoch": 2.0,
739
+ "learning_rate": 1.3604166666666668e-06,
740
+ "loss": 1.0236,
741
+ "step": 49000
742
+ },
743
+ {
744
+ "epoch": 2.0,
745
+ "eval_accuracy": 0.5021956087824351,
746
+ "eval_loss": 0.9895390868186951,
747
+ "eval_runtime": 50.824,
748
+ "eval_samples_per_second": 98.576,
749
+ "eval_steps_per_second": 12.337,
750
+ "step": 49000
751
+ },
752
+ {
753
+ "epoch": 2.04,
754
+ "learning_rate": 1.3881944444444446e-06,
755
+ "loss": 1.0243,
756
+ "step": 50000
757
+ },
758
+ {
759
+ "epoch": 2.04,
760
+ "eval_accuracy": 0.47984031936127747,
761
+ "eval_loss": 1.0068507194519043,
762
+ "eval_runtime": 50.7665,
763
+ "eval_samples_per_second": 98.687,
764
+ "eval_steps_per_second": 12.351,
765
+ "step": 50000
766
+ },
767
+ {
768
+ "epoch": 2.08,
769
+ "learning_rate": 1.4159444444444445e-06,
770
+ "loss": 1.0237,
771
+ "step": 51000
772
+ },
773
+ {
774
+ "epoch": 2.08,
775
+ "eval_accuracy": 0.5053892215568863,
776
+ "eval_loss": 0.9906230568885803,
777
+ "eval_runtime": 50.9162,
778
+ "eval_samples_per_second": 98.397,
779
+ "eval_steps_per_second": 12.314,
780
+ "step": 51000
781
+ },
782
+ {
783
+ "epoch": 2.12,
784
+ "learning_rate": 1.4437222222222225e-06,
785
+ "loss": 1.0184,
786
+ "step": 52000
787
+ },
788
+ {
789
+ "epoch": 2.12,
790
+ "eval_accuracy": 0.47784431137724553,
791
+ "eval_loss": 1.015572428703308,
792
+ "eval_runtime": 51.1054,
793
+ "eval_samples_per_second": 98.033,
794
+ "eval_steps_per_second": 12.269,
795
+ "step": 52000
796
+ },
797
+ {
798
+ "epoch": 2.16,
799
+ "learning_rate": 1.4714722222222223e-06,
800
+ "loss": 1.0234,
801
+ "step": 53000
802
+ },
803
+ {
804
+ "epoch": 2.16,
805
+ "eval_accuracy": 0.5031936127744511,
806
+ "eval_loss": 0.9886481761932373,
807
+ "eval_runtime": 51.2847,
808
+ "eval_samples_per_second": 97.69,
809
+ "eval_steps_per_second": 12.226,
810
+ "step": 53000
811
+ },
812
+ {
813
+ "epoch": 2.2,
814
+ "learning_rate": 1.49925e-06,
815
+ "loss": 1.0195,
816
+ "step": 54000
817
+ },
818
+ {
819
+ "epoch": 2.2,
820
+ "eval_accuracy": 0.5041916167664671,
821
+ "eval_loss": 0.991216778755188,
822
+ "eval_runtime": 51.0131,
823
+ "eval_samples_per_second": 98.21,
824
+ "eval_steps_per_second": 12.291,
825
+ "step": 54000
826
+ },
827
+ {
828
+ "epoch": 2.24,
829
+ "learning_rate": 1.5270000000000002e-06,
830
+ "loss": 1.0195,
831
+ "step": 55000
832
+ },
833
+ {
834
+ "epoch": 2.24,
835
+ "eval_accuracy": 0.5053892215568863,
836
+ "eval_loss": 0.9881049394607544,
837
+ "eval_runtime": 51.1896,
838
+ "eval_samples_per_second": 97.871,
839
+ "eval_steps_per_second": 12.249,
840
+ "step": 55000
841
+ },
842
+ {
843
+ "epoch": 2.28,
844
+ "learning_rate": 1.554777777777778e-06,
845
+ "loss": 1.0131,
846
+ "step": 56000
847
+ },
848
+ {
849
+ "epoch": 2.28,
850
+ "eval_accuracy": 0.46447105788423154,
851
+ "eval_loss": 1.0177615880966187,
852
+ "eval_runtime": 51.2121,
853
+ "eval_samples_per_second": 97.828,
854
+ "eval_steps_per_second": 12.243,
855
+ "step": 56000
856
+ },
857
+ {
858
+ "epoch": 2.32,
859
+ "learning_rate": 1.582527777777778e-06,
860
+ "loss": 1.0132,
861
+ "step": 57000
862
+ },
863
+ {
864
+ "epoch": 2.32,
865
+ "eval_accuracy": 0.5025948103792415,
866
+ "eval_loss": 0.9875446557998657,
867
+ "eval_runtime": 51.1782,
868
+ "eval_samples_per_second": 97.893,
869
+ "eval_steps_per_second": 12.251,
870
+ "step": 57000
871
+ },
872
+ {
873
+ "epoch": 2.36,
874
+ "learning_rate": 1.6103055555555556e-06,
875
+ "loss": 1.0167,
876
+ "step": 58000
877
+ },
878
+ {
879
+ "epoch": 2.36,
880
+ "eval_accuracy": 0.49880239520958086,
881
+ "eval_loss": 0.9983100295066833,
882
+ "eval_runtime": 50.8713,
883
+ "eval_samples_per_second": 98.484,
884
+ "eval_steps_per_second": 12.325,
885
+ "step": 58000
886
+ },
887
+ {
888
+ "epoch": 2.4,
889
+ "learning_rate": 1.6380555555555554e-06,
890
+ "loss": 1.0126,
891
+ "step": 59000
892
+ },
893
+ {
894
+ "epoch": 2.4,
895
+ "eval_accuracy": 0.5005988023952096,
896
+ "eval_loss": 0.9835163950920105,
897
+ "eval_runtime": 50.8767,
898
+ "eval_samples_per_second": 98.473,
899
+ "eval_steps_per_second": 12.324,
900
+ "step": 59000
901
+ },
902
+ {
903
+ "epoch": 2.44,
904
+ "learning_rate": 1.6658333333333334e-06,
905
+ "loss": 1.0167,
906
+ "step": 60000
907
+ },
908
+ {
909
+ "epoch": 2.44,
910
+ "eval_accuracy": 0.5091816367265469,
911
+ "eval_loss": 0.9835397601127625,
912
+ "eval_runtime": 50.8449,
913
+ "eval_samples_per_second": 98.535,
914
+ "eval_steps_per_second": 12.332,
915
+ "step": 60000
916
+ },
917
+ {
918
+ "epoch": 2.49,
919
+ "learning_rate": 1.6935833333333335e-06,
920
+ "loss": 1.013,
921
+ "step": 61000
922
+ },
923
+ {
924
+ "epoch": 2.49,
925
+ "eval_accuracy": 0.5045908183632735,
926
+ "eval_loss": 0.982866108417511,
927
+ "eval_runtime": 50.742,
928
+ "eval_samples_per_second": 98.735,
929
+ "eval_steps_per_second": 12.357,
930
+ "step": 61000
931
+ },
932
+ {
933
+ "epoch": 2.53,
934
+ "learning_rate": 1.7213333333333336e-06,
935
+ "loss": 1.0165,
936
+ "step": 62000
937
+ },
938
+ {
939
+ "epoch": 2.53,
940
+ "eval_accuracy": 0.49261477045908186,
941
+ "eval_loss": 0.9988529682159424,
942
+ "eval_runtime": 51.2665,
943
+ "eval_samples_per_second": 97.725,
944
+ "eval_steps_per_second": 12.23,
945
+ "step": 62000
946
+ },
947
+ {
948
+ "epoch": 2.57,
949
+ "learning_rate": 1.7491111111111111e-06,
950
+ "loss": 1.0144,
951
+ "step": 63000
952
+ },
953
+ {
954
+ "epoch": 2.57,
955
+ "eval_accuracy": 0.5007984031936128,
956
+ "eval_loss": 0.9871107339859009,
957
+ "eval_runtime": 51.2665,
958
+ "eval_samples_per_second": 97.725,
959
+ "eval_steps_per_second": 12.23,
960
+ "step": 63000
961
+ },
962
+ {
963
+ "epoch": 2.61,
964
+ "learning_rate": 1.7768611111111112e-06,
965
+ "loss": 1.0107,
966
+ "step": 64000
967
+ },
968
+ {
969
+ "epoch": 2.61,
970
+ "eval_accuracy": 0.5141716566866268,
971
+ "eval_loss": 0.9800056219100952,
972
+ "eval_runtime": 50.933,
973
+ "eval_samples_per_second": 98.364,
974
+ "eval_steps_per_second": 12.31,
975
+ "step": 64000
976
+ },
977
+ {
978
+ "epoch": 2.65,
979
+ "learning_rate": 1.804638888888889e-06,
980
+ "loss": 1.0114,
981
+ "step": 65000
982
+ },
983
+ {
984
+ "epoch": 2.65,
985
+ "eval_accuracy": 0.5167664670658683,
986
+ "eval_loss": 0.9798646569252014,
987
+ "eval_runtime": 50.8211,
988
+ "eval_samples_per_second": 98.581,
989
+ "eval_steps_per_second": 12.337,
990
+ "step": 65000
991
+ },
992
+ {
993
+ "epoch": 2.69,
994
+ "learning_rate": 1.8324166666666667e-06,
995
+ "loss": 1.0132,
996
+ "step": 66000
997
+ },
998
+ {
999
+ "epoch": 2.69,
1000
+ "eval_accuracy": 0.5083832335329341,
1001
+ "eval_loss": 0.9781086444854736,
1002
+ "eval_runtime": 50.7983,
1003
+ "eval_samples_per_second": 98.625,
1004
+ "eval_steps_per_second": 12.343,
1005
+ "step": 66000
1006
+ },
1007
+ {
1008
+ "epoch": 2.73,
1009
+ "learning_rate": 1.8601944444444447e-06,
1010
+ "loss": 1.0091,
1011
+ "step": 67000
1012
+ },
1013
+ {
1014
+ "epoch": 2.73,
1015
+ "eval_accuracy": 0.48622754491017967,
1016
+ "eval_loss": 0.9882963299751282,
1017
+ "eval_runtime": 50.9393,
1018
+ "eval_samples_per_second": 98.352,
1019
+ "eval_steps_per_second": 12.309,
1020
+ "step": 67000
1021
+ },
1022
+ {
1023
+ "epoch": 2.77,
1024
+ "learning_rate": 1.8879444444444446e-06,
1025
+ "loss": 1.0094,
1026
+ "step": 68000
1027
+ },
1028
+ {
1029
+ "epoch": 2.77,
1030
+ "eval_accuracy": 0.5109780439121756,
1031
+ "eval_loss": 0.9793729186058044,
1032
+ "eval_runtime": 50.8821,
1033
+ "eval_samples_per_second": 98.463,
1034
+ "eval_steps_per_second": 12.323,
1035
+ "step": 68000
1036
+ },
1037
+ {
1038
+ "epoch": 2.81,
1039
+ "learning_rate": 1.9156944444444444e-06,
1040
+ "loss": 1.007,
1041
+ "step": 69000
1042
+ },
1043
+ {
1044
+ "epoch": 2.81,
1045
+ "eval_accuracy": 0.5103792415169661,
1046
+ "eval_loss": 0.9774999022483826,
1047
+ "eval_runtime": 50.848,
1048
+ "eval_samples_per_second": 98.529,
1049
+ "eval_steps_per_second": 12.331,
1050
+ "step": 69000
1051
+ },
1052
+ {
1053
+ "epoch": 2.85,
1054
+ "learning_rate": 1.9434722222222224e-06,
1055
+ "loss": 1.0062,
1056
+ "step": 70000
1057
+ },
1058
+ {
1059
+ "epoch": 2.85,
1060
+ "eval_accuracy": 0.5131736526946108,
1061
+ "eval_loss": 0.9792296290397644,
1062
+ "eval_runtime": 51.4864,
1063
+ "eval_samples_per_second": 97.307,
1064
+ "eval_steps_per_second": 12.178,
1065
+ "step": 70000
1066
+ },
1067
+ {
1068
+ "epoch": 2.89,
1069
+ "learning_rate": 1.97125e-06,
1070
+ "loss": 0.9999,
1071
+ "step": 71000
1072
+ },
1073
+ {
1074
+ "epoch": 2.89,
1075
+ "eval_accuracy": 0.5089820359281437,
1076
+ "eval_loss": 0.9837301969528198,
1077
+ "eval_runtime": 51.1226,
1078
+ "eval_samples_per_second": 98.0,
1079
+ "eval_steps_per_second": 12.265,
1080
+ "step": 71000
1081
+ },
1082
+ {
1083
+ "epoch": 2.93,
1084
+ "learning_rate": 1.9990000000000003e-06,
1085
+ "loss": 1.0066,
1086
+ "step": 72000
1087
+ },
1088
+ {
1089
+ "epoch": 2.93,
1090
+ "eval_accuracy": 0.5077844311377245,
1091
+ "eval_loss": 0.9925987720489502,
1092
+ "eval_runtime": 50.9371,
1093
+ "eval_samples_per_second": 98.357,
1094
+ "eval_steps_per_second": 12.309,
1095
+ "step": 72000
1096
+ },
1097
+ {
1098
+ "epoch": 2.97,
1099
+ "learning_rate": 2.026777777777778e-06,
1100
+ "loss": 1.0051,
1101
+ "step": 73000
1102
+ },
1103
+ {
1104
+ "epoch": 2.97,
1105
+ "eval_accuracy": 0.5125748502994012,
1106
+ "eval_loss": 0.9769598245620728,
1107
+ "eval_runtime": 51.2214,
1108
+ "eval_samples_per_second": 97.811,
1109
+ "eval_steps_per_second": 12.241,
1110
+ "step": 73000
1111
+ },
1112
+ {
1113
+ "epoch": 3.01,
1114
+ "learning_rate": 2.054555555555556e-06,
1115
+ "loss": 1.0025,
1116
+ "step": 74000
1117
+ },
1118
+ {
1119
+ "epoch": 3.01,
1120
+ "eval_accuracy": 0.5189620758483033,
1121
+ "eval_loss": 0.969079315662384,
1122
+ "eval_runtime": 51.0735,
1123
+ "eval_samples_per_second": 98.094,
1124
+ "eval_steps_per_second": 12.276,
1125
+ "step": 74000
1126
+ },
1127
+ {
1128
+ "epoch": 3.06,
1129
+ "learning_rate": 2.0823333333333334e-06,
1130
+ "loss": 0.9933,
1131
+ "step": 75000
1132
+ },
1133
+ {
1134
+ "epoch": 3.06,
1135
+ "eval_accuracy": 0.5043912175648703,
1136
+ "eval_loss": 0.9786885976791382,
1137
+ "eval_runtime": 51.2569,
1138
+ "eval_samples_per_second": 97.743,
1139
+ "eval_steps_per_second": 12.233,
1140
+ "step": 75000
1141
+ },
1142
+ {
1143
+ "epoch": 3.1,
1144
+ "learning_rate": 2.1100833333333332e-06,
1145
+ "loss": 0.9932,
1146
+ "step": 76000
1147
+ },
1148
+ {
1149
+ "epoch": 3.1,
1150
+ "eval_accuracy": 0.512375249500998,
1151
+ "eval_loss": 0.968884289264679,
1152
+ "eval_runtime": 51.0713,
1153
+ "eval_samples_per_second": 98.098,
1154
+ "eval_steps_per_second": 12.277,
1155
+ "step": 76000
1156
+ },
1157
+ {
1158
+ "epoch": 3.14,
1159
+ "learning_rate": 2.137861111111111e-06,
1160
+ "loss": 0.9964,
1161
+ "step": 77000
1162
+ },
1163
+ {
1164
+ "epoch": 3.14,
1165
+ "eval_accuracy": 0.49221556886227547,
1166
+ "eval_loss": 1.0001007318496704,
1167
+ "eval_runtime": 51.275,
1168
+ "eval_samples_per_second": 97.708,
1169
+ "eval_steps_per_second": 12.228,
1170
+ "step": 77000
1171
+ },
1172
+ {
1173
+ "epoch": 3.18,
1174
+ "learning_rate": 2.1656111111111115e-06,
1175
+ "loss": 0.9872,
1176
+ "step": 78000
1177
+ },
1178
+ {
1179
+ "epoch": 3.18,
1180
+ "eval_accuracy": 0.5251497005988024,
1181
+ "eval_loss": 0.9642478823661804,
1182
+ "eval_runtime": 51.1066,
1183
+ "eval_samples_per_second": 98.03,
1184
+ "eval_steps_per_second": 12.268,
1185
+ "step": 78000
1186
+ },
1187
+ {
1188
+ "epoch": 3.22,
1189
+ "learning_rate": 2.193388888888889e-06,
1190
+ "loss": 0.9933,
1191
+ "step": 79000
1192
+ },
1193
+ {
1194
+ "epoch": 3.22,
1195
+ "eval_accuracy": 0.5053892215568863,
1196
+ "eval_loss": 0.9904850721359253,
1197
+ "eval_runtime": 51.1661,
1198
+ "eval_samples_per_second": 97.916,
1199
+ "eval_steps_per_second": 12.254,
1200
+ "step": 79000
1201
+ }
1202
+ ],
1203
+ "logging_steps": 1000,
1204
+ "max_steps": 10000000,
1205
+ "num_train_epochs": 408,
1206
+ "save_steps": 1000,
1207
+ "total_flos": 1.651372714158981e+17,
1208
+ "trial_name": null,
1209
+ "trial_params": null
1210
+ }
checkpoint-79000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:933e5c21c45262af343f5d708824cb21f8d339a9e2662221f8b52a4157df35fa
3
+ size 4091
checkpoint-79000/vocab.json ADDED
The diff for this file is too large to render. See raw diff