yaanhaan commited on
Commit
0254c20
1 Parent(s): 5c2a6bd

Upload 9 files

Browse files
Files changed (9) hide show
  1. config.json +26 -0
  2. merges.txt +0 -0
  3. optimizer.pt +3 -0
  4. pytorch_model.bin +3 -0
  5. rng_state.pth +3 -0
  6. scheduler.pt +3 -0
  7. trainer_state.json +1294 -0
  8. training_args.bin +3 -0
  9. vocab.json +0 -0
config.json ADDED
@@ -0,0 +1,26 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "RobertaForMaskedLM"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 512,
16
+ "model_type": "roberta",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 12,
19
+ "pad_token_id": 1,
20
+ "position_embedding_type": "absolute",
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.17.0",
23
+ "type_vocab_size": 1,
24
+ "use_cache": true,
25
+ "vocab_size": 52000
26
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a38dec490d5d02b17617d5353d0b03a663ebb28af23671271dc33f00fafabac2
3
+ size 1008359813
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38c9f23bf1631afc264e1c39b3ae8e6a20d0d340207eec0447523f3bb343f7fa
3
+ size 504208825
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:88a9bcc908001537b9220e76687bc5bc72fc3ccf3d0dd606796ce524412df4f7
3
+ size 17577
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b58b3f11105c4300aff38ceeb3b077a9949fa173a36a71c55b17f928d181616
3
+ size 627
trainer_state.json ADDED
@@ -0,0 +1,1294 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 4.986592651832337,
5
+ "global_step": 106000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 4.999952956673096e-05,
13
+ "loss": 11.0435,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 4.9764783365479614e-05,
19
+ "loss": 6.9955,
20
+ "step": 500
21
+ },
22
+ {
23
+ "epoch": 0.05,
24
+ "learning_rate": 4.952956673095921e-05,
25
+ "loss": 6.0131,
26
+ "step": 1000
27
+ },
28
+ {
29
+ "epoch": 0.07,
30
+ "learning_rate": 4.9294350096438824e-05,
31
+ "loss": 5.5677,
32
+ "step": 1500
33
+ },
34
+ {
35
+ "epoch": 0.09,
36
+ "learning_rate": 4.905913346191843e-05,
37
+ "loss": 5.2936,
38
+ "step": 2000
39
+ },
40
+ {
41
+ "epoch": 0.12,
42
+ "learning_rate": 4.8823916827398035e-05,
43
+ "loss": 5.0561,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.14,
48
+ "learning_rate": 4.8588700192877647e-05,
49
+ "loss": 4.8787,
50
+ "step": 3000
51
+ },
52
+ {
53
+ "epoch": 0.16,
54
+ "learning_rate": 4.835348355835725e-05,
55
+ "loss": 4.7244,
56
+ "step": 3500
57
+ },
58
+ {
59
+ "epoch": 0.19,
60
+ "learning_rate": 4.811826692383686e-05,
61
+ "loss": 4.5982,
62
+ "step": 4000
63
+ },
64
+ {
65
+ "epoch": 0.21,
66
+ "learning_rate": 4.788305028931646e-05,
67
+ "loss": 4.4815,
68
+ "step": 4500
69
+ },
70
+ {
71
+ "epoch": 0.24,
72
+ "learning_rate": 4.764783365479607e-05,
73
+ "loss": 4.4,
74
+ "step": 5000
75
+ },
76
+ {
77
+ "epoch": 0.26,
78
+ "learning_rate": 4.741261702027567e-05,
79
+ "loss": 4.3032,
80
+ "step": 5500
81
+ },
82
+ {
83
+ "epoch": 0.28,
84
+ "learning_rate": 4.7177400385755284e-05,
85
+ "loss": 4.2385,
86
+ "step": 6000
87
+ },
88
+ {
89
+ "epoch": 0.31,
90
+ "learning_rate": 4.694218375123489e-05,
91
+ "loss": 4.1932,
92
+ "step": 6500
93
+ },
94
+ {
95
+ "epoch": 0.33,
96
+ "learning_rate": 4.6706967116714494e-05,
97
+ "loss": 4.1363,
98
+ "step": 7000
99
+ },
100
+ {
101
+ "epoch": 0.35,
102
+ "learning_rate": 4.6471750482194106e-05,
103
+ "loss": 4.0896,
104
+ "step": 7500
105
+ },
106
+ {
107
+ "epoch": 0.38,
108
+ "learning_rate": 4.6236533847673704e-05,
109
+ "loss": 4.0412,
110
+ "step": 8000
111
+ },
112
+ {
113
+ "epoch": 0.4,
114
+ "learning_rate": 4.6001317213153316e-05,
115
+ "loss": 3.9997,
116
+ "step": 8500
117
+ },
118
+ {
119
+ "epoch": 0.42,
120
+ "learning_rate": 4.576610057863293e-05,
121
+ "loss": 3.9699,
122
+ "step": 9000
123
+ },
124
+ {
125
+ "epoch": 0.45,
126
+ "learning_rate": 4.5530883944112527e-05,
127
+ "loss": 3.9251,
128
+ "step": 9500
129
+ },
130
+ {
131
+ "epoch": 0.47,
132
+ "learning_rate": 4.529566730959214e-05,
133
+ "loss": 3.9045,
134
+ "step": 10000
135
+ },
136
+ {
137
+ "epoch": 0.49,
138
+ "learning_rate": 4.5060450675071744e-05,
139
+ "loss": 3.8665,
140
+ "step": 10500
141
+ },
142
+ {
143
+ "epoch": 0.52,
144
+ "learning_rate": 4.482523404055135e-05,
145
+ "loss": 3.836,
146
+ "step": 11000
147
+ },
148
+ {
149
+ "epoch": 0.54,
150
+ "learning_rate": 4.4590017406030954e-05,
151
+ "loss": 3.8186,
152
+ "step": 11500
153
+ },
154
+ {
155
+ "epoch": 0.56,
156
+ "learning_rate": 4.4354800771510566e-05,
157
+ "loss": 3.8041,
158
+ "step": 12000
159
+ },
160
+ {
161
+ "epoch": 0.59,
162
+ "learning_rate": 4.411958413699017e-05,
163
+ "loss": 3.7728,
164
+ "step": 12500
165
+ },
166
+ {
167
+ "epoch": 0.61,
168
+ "learning_rate": 4.3884367502469776e-05,
169
+ "loss": 3.7502,
170
+ "step": 13000
171
+ },
172
+ {
173
+ "epoch": 0.64,
174
+ "learning_rate": 4.364915086794938e-05,
175
+ "loss": 3.7264,
176
+ "step": 13500
177
+ },
178
+ {
179
+ "epoch": 0.66,
180
+ "learning_rate": 4.3413934233428986e-05,
181
+ "loss": 3.7029,
182
+ "step": 14000
183
+ },
184
+ {
185
+ "epoch": 0.68,
186
+ "learning_rate": 4.31787175989086e-05,
187
+ "loss": 3.6765,
188
+ "step": 14500
189
+ },
190
+ {
191
+ "epoch": 0.71,
192
+ "learning_rate": 4.29435009643882e-05,
193
+ "loss": 3.6528,
194
+ "step": 15000
195
+ },
196
+ {
197
+ "epoch": 0.73,
198
+ "learning_rate": 4.270828432986781e-05,
199
+ "loss": 3.6563,
200
+ "step": 15500
201
+ },
202
+ {
203
+ "epoch": 0.75,
204
+ "learning_rate": 4.247306769534742e-05,
205
+ "loss": 3.6245,
206
+ "step": 16000
207
+ },
208
+ {
209
+ "epoch": 0.78,
210
+ "learning_rate": 4.223785106082702e-05,
211
+ "loss": 3.6056,
212
+ "step": 16500
213
+ },
214
+ {
215
+ "epoch": 0.8,
216
+ "learning_rate": 4.200263442630663e-05,
217
+ "loss": 3.5829,
218
+ "step": 17000
219
+ },
220
+ {
221
+ "epoch": 0.82,
222
+ "learning_rate": 4.176741779178624e-05,
223
+ "loss": 3.5855,
224
+ "step": 17500
225
+ },
226
+ {
227
+ "epoch": 0.85,
228
+ "learning_rate": 4.153220115726584e-05,
229
+ "loss": 3.5703,
230
+ "step": 18000
231
+ },
232
+ {
233
+ "epoch": 0.87,
234
+ "learning_rate": 4.129698452274545e-05,
235
+ "loss": 3.5545,
236
+ "step": 18500
237
+ },
238
+ {
239
+ "epoch": 0.89,
240
+ "learning_rate": 4.106176788822506e-05,
241
+ "loss": 3.5338,
242
+ "step": 19000
243
+ },
244
+ {
245
+ "epoch": 0.92,
246
+ "learning_rate": 4.082655125370466e-05,
247
+ "loss": 3.5202,
248
+ "step": 19500
249
+ },
250
+ {
251
+ "epoch": 0.94,
252
+ "learning_rate": 4.059133461918427e-05,
253
+ "loss": 3.5066,
254
+ "step": 20000
255
+ },
256
+ {
257
+ "epoch": 0.96,
258
+ "learning_rate": 4.035611798466388e-05,
259
+ "loss": 3.5027,
260
+ "step": 20500
261
+ },
262
+ {
263
+ "epoch": 0.99,
264
+ "learning_rate": 4.0120901350143485e-05,
265
+ "loss": 3.477,
266
+ "step": 21000
267
+ },
268
+ {
269
+ "epoch": 1.01,
270
+ "learning_rate": 3.988568471562309e-05,
271
+ "loss": 3.4658,
272
+ "step": 21500
273
+ },
274
+ {
275
+ "epoch": 1.03,
276
+ "learning_rate": 3.96504680811027e-05,
277
+ "loss": 3.4505,
278
+ "step": 22000
279
+ },
280
+ {
281
+ "epoch": 1.06,
282
+ "learning_rate": 3.94152514465823e-05,
283
+ "loss": 3.445,
284
+ "step": 22500
285
+ },
286
+ {
287
+ "epoch": 1.08,
288
+ "learning_rate": 3.918003481206191e-05,
289
+ "loss": 3.4314,
290
+ "step": 23000
291
+ },
292
+ {
293
+ "epoch": 1.11,
294
+ "learning_rate": 3.894481817754152e-05,
295
+ "loss": 3.4278,
296
+ "step": 23500
297
+ },
298
+ {
299
+ "epoch": 1.13,
300
+ "learning_rate": 3.870960154302112e-05,
301
+ "loss": 3.4179,
302
+ "step": 24000
303
+ },
304
+ {
305
+ "epoch": 1.15,
306
+ "learning_rate": 3.8474384908500734e-05,
307
+ "loss": 3.4051,
308
+ "step": 24500
309
+ },
310
+ {
311
+ "epoch": 1.18,
312
+ "learning_rate": 3.823916827398034e-05,
313
+ "loss": 3.3858,
314
+ "step": 25000
315
+ },
316
+ {
317
+ "epoch": 1.2,
318
+ "learning_rate": 3.8003951639459944e-05,
319
+ "loss": 3.3851,
320
+ "step": 25500
321
+ },
322
+ {
323
+ "epoch": 1.22,
324
+ "learning_rate": 3.7768735004939556e-05,
325
+ "loss": 3.3592,
326
+ "step": 26000
327
+ },
328
+ {
329
+ "epoch": 1.25,
330
+ "learning_rate": 3.7533518370419154e-05,
331
+ "loss": 3.3617,
332
+ "step": 26500
333
+ },
334
+ {
335
+ "epoch": 1.27,
336
+ "learning_rate": 3.7298301735898766e-05,
337
+ "loss": 3.351,
338
+ "step": 27000
339
+ },
340
+ {
341
+ "epoch": 1.29,
342
+ "learning_rate": 3.706308510137837e-05,
343
+ "loss": 3.3456,
344
+ "step": 27500
345
+ },
346
+ {
347
+ "epoch": 1.32,
348
+ "learning_rate": 3.6827868466857977e-05,
349
+ "loss": 3.3412,
350
+ "step": 28000
351
+ },
352
+ {
353
+ "epoch": 1.34,
354
+ "learning_rate": 3.659265183233758e-05,
355
+ "loss": 3.3275,
356
+ "step": 28500
357
+ },
358
+ {
359
+ "epoch": 1.36,
360
+ "learning_rate": 3.6357435197817194e-05,
361
+ "loss": 3.3048,
362
+ "step": 29000
363
+ },
364
+ {
365
+ "epoch": 1.39,
366
+ "learning_rate": 3.61222185632968e-05,
367
+ "loss": 3.3148,
368
+ "step": 29500
369
+ },
370
+ {
371
+ "epoch": 1.41,
372
+ "learning_rate": 3.5887001928776404e-05,
373
+ "loss": 3.3077,
374
+ "step": 30000
375
+ },
376
+ {
377
+ "epoch": 1.43,
378
+ "learning_rate": 3.5651785294256016e-05,
379
+ "loss": 3.2981,
380
+ "step": 30500
381
+ },
382
+ {
383
+ "epoch": 1.46,
384
+ "learning_rate": 3.5416568659735614e-05,
385
+ "loss": 3.2814,
386
+ "step": 31000
387
+ },
388
+ {
389
+ "epoch": 1.48,
390
+ "learning_rate": 3.5181352025215226e-05,
391
+ "loss": 3.2877,
392
+ "step": 31500
393
+ },
394
+ {
395
+ "epoch": 1.51,
396
+ "learning_rate": 3.494613539069483e-05,
397
+ "loss": 3.2722,
398
+ "step": 32000
399
+ },
400
+ {
401
+ "epoch": 1.53,
402
+ "learning_rate": 3.4710918756174436e-05,
403
+ "loss": 3.269,
404
+ "step": 32500
405
+ },
406
+ {
407
+ "epoch": 1.55,
408
+ "learning_rate": 3.447570212165405e-05,
409
+ "loss": 3.2491,
410
+ "step": 33000
411
+ },
412
+ {
413
+ "epoch": 1.58,
414
+ "learning_rate": 3.424048548713365e-05,
415
+ "loss": 3.2531,
416
+ "step": 33500
417
+ },
418
+ {
419
+ "epoch": 1.6,
420
+ "learning_rate": 3.400526885261326e-05,
421
+ "loss": 3.2399,
422
+ "step": 34000
423
+ },
424
+ {
425
+ "epoch": 1.62,
426
+ "learning_rate": 3.377005221809286e-05,
427
+ "loss": 3.24,
428
+ "step": 34500
429
+ },
430
+ {
431
+ "epoch": 1.65,
432
+ "learning_rate": 3.353483558357247e-05,
433
+ "loss": 3.2312,
434
+ "step": 35000
435
+ },
436
+ {
437
+ "epoch": 1.67,
438
+ "learning_rate": 3.329961894905208e-05,
439
+ "loss": 3.2293,
440
+ "step": 35500
441
+ },
442
+ {
443
+ "epoch": 1.69,
444
+ "learning_rate": 3.3064402314531685e-05,
445
+ "loss": 3.2214,
446
+ "step": 36000
447
+ },
448
+ {
449
+ "epoch": 1.72,
450
+ "learning_rate": 3.282918568001129e-05,
451
+ "loss": 3.2154,
452
+ "step": 36500
453
+ },
454
+ {
455
+ "epoch": 1.74,
456
+ "learning_rate": 3.2593969045490896e-05,
457
+ "loss": 3.2198,
458
+ "step": 37000
459
+ },
460
+ {
461
+ "epoch": 1.76,
462
+ "learning_rate": 3.235875241097051e-05,
463
+ "loss": 3.1968,
464
+ "step": 37500
465
+ },
466
+ {
467
+ "epoch": 1.79,
468
+ "learning_rate": 3.212353577645011e-05,
469
+ "loss": 3.1938,
470
+ "step": 38000
471
+ },
472
+ {
473
+ "epoch": 1.81,
474
+ "learning_rate": 3.188831914192972e-05,
475
+ "loss": 3.1861,
476
+ "step": 38500
477
+ },
478
+ {
479
+ "epoch": 1.83,
480
+ "learning_rate": 3.165310250740933e-05,
481
+ "loss": 3.1805,
482
+ "step": 39000
483
+ },
484
+ {
485
+ "epoch": 1.86,
486
+ "learning_rate": 3.141788587288893e-05,
487
+ "loss": 3.1778,
488
+ "step": 39500
489
+ },
490
+ {
491
+ "epoch": 1.88,
492
+ "learning_rate": 3.118266923836854e-05,
493
+ "loss": 3.1748,
494
+ "step": 40000
495
+ },
496
+ {
497
+ "epoch": 1.91,
498
+ "learning_rate": 3.0947452603848145e-05,
499
+ "loss": 3.1688,
500
+ "step": 40500
501
+ },
502
+ {
503
+ "epoch": 1.93,
504
+ "learning_rate": 3.071223596932775e-05,
505
+ "loss": 3.1617,
506
+ "step": 41000
507
+ },
508
+ {
509
+ "epoch": 1.95,
510
+ "learning_rate": 3.0477019334807362e-05,
511
+ "loss": 3.1537,
512
+ "step": 41500
513
+ },
514
+ {
515
+ "epoch": 1.98,
516
+ "learning_rate": 3.0241802700286964e-05,
517
+ "loss": 3.1581,
518
+ "step": 42000
519
+ },
520
+ {
521
+ "epoch": 2.0,
522
+ "learning_rate": 3.0006586065766572e-05,
523
+ "loss": 3.149,
524
+ "step": 42500
525
+ },
526
+ {
527
+ "epoch": 2.02,
528
+ "learning_rate": 2.9771369431246177e-05,
529
+ "loss": 3.1317,
530
+ "step": 43000
531
+ },
532
+ {
533
+ "epoch": 2.05,
534
+ "learning_rate": 2.9536152796725786e-05,
535
+ "loss": 3.1136,
536
+ "step": 43500
537
+ },
538
+ {
539
+ "epoch": 2.07,
540
+ "learning_rate": 2.9300936162205394e-05,
541
+ "loss": 3.1115,
542
+ "step": 44000
543
+ },
544
+ {
545
+ "epoch": 2.09,
546
+ "learning_rate": 2.9065719527685e-05,
547
+ "loss": 3.1165,
548
+ "step": 44500
549
+ },
550
+ {
551
+ "epoch": 2.12,
552
+ "learning_rate": 2.8830502893164608e-05,
553
+ "loss": 3.0959,
554
+ "step": 45000
555
+ },
556
+ {
557
+ "epoch": 2.14,
558
+ "learning_rate": 2.859528625864421e-05,
559
+ "loss": 3.1135,
560
+ "step": 45500
561
+ },
562
+ {
563
+ "epoch": 2.16,
564
+ "learning_rate": 2.8360069624123818e-05,
565
+ "loss": 3.0856,
566
+ "step": 46000
567
+ },
568
+ {
569
+ "epoch": 2.19,
570
+ "learning_rate": 2.812485298960343e-05,
571
+ "loss": 3.1023,
572
+ "step": 46500
573
+ },
574
+ {
575
+ "epoch": 2.21,
576
+ "learning_rate": 2.7889636355083032e-05,
577
+ "loss": 3.0848,
578
+ "step": 47000
579
+ },
580
+ {
581
+ "epoch": 2.23,
582
+ "learning_rate": 2.765441972056264e-05,
583
+ "loss": 3.0901,
584
+ "step": 47500
585
+ },
586
+ {
587
+ "epoch": 2.26,
588
+ "learning_rate": 2.7419203086042245e-05,
589
+ "loss": 3.0776,
590
+ "step": 48000
591
+ },
592
+ {
593
+ "epoch": 2.28,
594
+ "learning_rate": 2.7183986451521854e-05,
595
+ "loss": 3.0703,
596
+ "step": 48500
597
+ },
598
+ {
599
+ "epoch": 2.31,
600
+ "learning_rate": 2.6948769817001456e-05,
601
+ "loss": 3.0671,
602
+ "step": 49000
603
+ },
604
+ {
605
+ "epoch": 2.33,
606
+ "learning_rate": 2.6713553182481067e-05,
607
+ "loss": 3.0574,
608
+ "step": 49500
609
+ },
610
+ {
611
+ "epoch": 2.35,
612
+ "learning_rate": 2.6478336547960676e-05,
613
+ "loss": 3.0583,
614
+ "step": 50000
615
+ },
616
+ {
617
+ "epoch": 2.38,
618
+ "learning_rate": 2.6243119913440278e-05,
619
+ "loss": 3.0599,
620
+ "step": 50500
621
+ },
622
+ {
623
+ "epoch": 2.4,
624
+ "learning_rate": 2.6007903278919886e-05,
625
+ "loss": 3.0596,
626
+ "step": 51000
627
+ },
628
+ {
629
+ "epoch": 2.42,
630
+ "learning_rate": 2.577268664439949e-05,
631
+ "loss": 3.0522,
632
+ "step": 51500
633
+ },
634
+ {
635
+ "epoch": 2.45,
636
+ "learning_rate": 2.55374700098791e-05,
637
+ "loss": 3.0484,
638
+ "step": 52000
639
+ },
640
+ {
641
+ "epoch": 2.47,
642
+ "learning_rate": 2.5302253375358708e-05,
643
+ "loss": 3.0332,
644
+ "step": 52500
645
+ },
646
+ {
647
+ "epoch": 2.49,
648
+ "learning_rate": 2.5067036740838313e-05,
649
+ "loss": 3.0302,
650
+ "step": 53000
651
+ },
652
+ {
653
+ "epoch": 2.52,
654
+ "learning_rate": 2.4831820106317922e-05,
655
+ "loss": 3.0321,
656
+ "step": 53500
657
+ },
658
+ {
659
+ "epoch": 2.54,
660
+ "learning_rate": 2.4596603471797527e-05,
661
+ "loss": 3.0304,
662
+ "step": 54000
663
+ },
664
+ {
665
+ "epoch": 2.56,
666
+ "learning_rate": 2.4361386837277132e-05,
667
+ "loss": 3.0296,
668
+ "step": 54500
669
+ },
670
+ {
671
+ "epoch": 2.59,
672
+ "learning_rate": 2.412617020275674e-05,
673
+ "loss": 3.0164,
674
+ "step": 55000
675
+ },
676
+ {
677
+ "epoch": 2.61,
678
+ "learning_rate": 2.3890953568236346e-05,
679
+ "loss": 3.0118,
680
+ "step": 55500
681
+ },
682
+ {
683
+ "epoch": 2.63,
684
+ "learning_rate": 2.365573693371595e-05,
685
+ "loss": 3.0146,
686
+ "step": 56000
687
+ },
688
+ {
689
+ "epoch": 2.66,
690
+ "learning_rate": 2.3420520299195563e-05,
691
+ "loss": 3.0152,
692
+ "step": 56500
693
+ },
694
+ {
695
+ "epoch": 2.68,
696
+ "learning_rate": 2.3185303664675168e-05,
697
+ "loss": 3.0084,
698
+ "step": 57000
699
+ },
700
+ {
701
+ "epoch": 2.7,
702
+ "learning_rate": 2.2950087030154773e-05,
703
+ "loss": 2.9993,
704
+ "step": 57500
705
+ },
706
+ {
707
+ "epoch": 2.73,
708
+ "learning_rate": 2.271487039563438e-05,
709
+ "loss": 3.0059,
710
+ "step": 58000
711
+ },
712
+ {
713
+ "epoch": 2.75,
714
+ "learning_rate": 2.2479653761113986e-05,
715
+ "loss": 2.998,
716
+ "step": 58500
717
+ },
718
+ {
719
+ "epoch": 2.78,
720
+ "learning_rate": 2.224443712659359e-05,
721
+ "loss": 2.9874,
722
+ "step": 59000
723
+ },
724
+ {
725
+ "epoch": 2.8,
726
+ "learning_rate": 2.20092204920732e-05,
727
+ "loss": 2.9934,
728
+ "step": 59500
729
+ },
730
+ {
731
+ "epoch": 2.82,
732
+ "learning_rate": 2.177400385755281e-05,
733
+ "loss": 2.9885,
734
+ "step": 60000
735
+ },
736
+ {
737
+ "epoch": 2.85,
738
+ "learning_rate": 2.1538787223032414e-05,
739
+ "loss": 2.9806,
740
+ "step": 60500
741
+ },
742
+ {
743
+ "epoch": 2.87,
744
+ "learning_rate": 2.130357058851202e-05,
745
+ "loss": 2.9801,
746
+ "step": 61000
747
+ },
748
+ {
749
+ "epoch": 2.89,
750
+ "learning_rate": 2.1068353953991627e-05,
751
+ "loss": 2.9726,
752
+ "step": 61500
753
+ },
754
+ {
755
+ "epoch": 2.92,
756
+ "learning_rate": 2.0833137319471236e-05,
757
+ "loss": 2.975,
758
+ "step": 62000
759
+ },
760
+ {
761
+ "epoch": 2.94,
762
+ "learning_rate": 2.059792068495084e-05,
763
+ "loss": 2.9697,
764
+ "step": 62500
765
+ },
766
+ {
767
+ "epoch": 2.96,
768
+ "learning_rate": 2.036270405043045e-05,
769
+ "loss": 2.9604,
770
+ "step": 63000
771
+ },
772
+ {
773
+ "epoch": 2.99,
774
+ "learning_rate": 2.0127487415910055e-05,
775
+ "loss": 2.955,
776
+ "step": 63500
777
+ },
778
+ {
779
+ "epoch": 3.01,
780
+ "learning_rate": 1.989227078138966e-05,
781
+ "loss": 2.9571,
782
+ "step": 64000
783
+ },
784
+ {
785
+ "epoch": 3.03,
786
+ "learning_rate": 1.9657054146869268e-05,
787
+ "loss": 2.9388,
788
+ "step": 64500
789
+ },
790
+ {
791
+ "epoch": 3.06,
792
+ "learning_rate": 1.9421837512348877e-05,
793
+ "loss": 2.9447,
794
+ "step": 65000
795
+ },
796
+ {
797
+ "epoch": 3.08,
798
+ "learning_rate": 1.9186620877828482e-05,
799
+ "loss": 2.9359,
800
+ "step": 65500
801
+ },
802
+ {
803
+ "epoch": 3.1,
804
+ "learning_rate": 1.8951404243308087e-05,
805
+ "loss": 2.9429,
806
+ "step": 66000
807
+ },
808
+ {
809
+ "epoch": 3.13,
810
+ "learning_rate": 1.8716187608787695e-05,
811
+ "loss": 2.9315,
812
+ "step": 66500
813
+ },
814
+ {
815
+ "epoch": 3.15,
816
+ "learning_rate": 1.84809709742673e-05,
817
+ "loss": 2.9258,
818
+ "step": 67000
819
+ },
820
+ {
821
+ "epoch": 3.18,
822
+ "learning_rate": 1.8245754339746906e-05,
823
+ "loss": 2.9285,
824
+ "step": 67500
825
+ },
826
+ {
827
+ "epoch": 3.2,
828
+ "learning_rate": 1.8010537705226514e-05,
829
+ "loss": 2.9277,
830
+ "step": 68000
831
+ },
832
+ {
833
+ "epoch": 3.22,
834
+ "learning_rate": 1.7775321070706123e-05,
835
+ "loss": 2.9209,
836
+ "step": 68500
837
+ },
838
+ {
839
+ "epoch": 3.25,
840
+ "learning_rate": 1.7540104436185728e-05,
841
+ "loss": 2.925,
842
+ "step": 69000
843
+ },
844
+ {
845
+ "epoch": 3.27,
846
+ "learning_rate": 1.7304887801665333e-05,
847
+ "loss": 2.9119,
848
+ "step": 69500
849
+ },
850
+ {
851
+ "epoch": 3.29,
852
+ "learning_rate": 1.706967116714494e-05,
853
+ "loss": 2.9145,
854
+ "step": 70000
855
+ },
856
+ {
857
+ "epoch": 3.32,
858
+ "learning_rate": 1.6834454532624546e-05,
859
+ "loss": 2.9079,
860
+ "step": 70500
861
+ },
862
+ {
863
+ "epoch": 3.34,
864
+ "learning_rate": 1.6599237898104155e-05,
865
+ "loss": 2.9089,
866
+ "step": 71000
867
+ },
868
+ {
869
+ "epoch": 3.36,
870
+ "learning_rate": 1.6364021263583763e-05,
871
+ "loss": 2.9052,
872
+ "step": 71500
873
+ },
874
+ {
875
+ "epoch": 3.39,
876
+ "learning_rate": 1.612880462906337e-05,
877
+ "loss": 2.897,
878
+ "step": 72000
879
+ },
880
+ {
881
+ "epoch": 3.41,
882
+ "learning_rate": 1.5893587994542974e-05,
883
+ "loss": 2.9026,
884
+ "step": 72500
885
+ },
886
+ {
887
+ "epoch": 3.43,
888
+ "learning_rate": 1.5658371360022582e-05,
889
+ "loss": 2.9071,
890
+ "step": 73000
891
+ },
892
+ {
893
+ "epoch": 3.46,
894
+ "learning_rate": 1.5423154725502187e-05,
895
+ "loss": 2.9021,
896
+ "step": 73500
897
+ },
898
+ {
899
+ "epoch": 3.48,
900
+ "learning_rate": 1.5187938090981796e-05,
901
+ "loss": 2.8918,
902
+ "step": 74000
903
+ },
904
+ {
905
+ "epoch": 3.5,
906
+ "learning_rate": 1.4952721456461402e-05,
907
+ "loss": 2.8995,
908
+ "step": 74500
909
+ },
910
+ {
911
+ "epoch": 3.53,
912
+ "learning_rate": 1.4717504821941008e-05,
913
+ "loss": 2.8874,
914
+ "step": 75000
915
+ },
916
+ {
917
+ "epoch": 3.55,
918
+ "learning_rate": 1.4482288187420614e-05,
919
+ "loss": 2.8779,
920
+ "step": 75500
921
+ },
922
+ {
923
+ "epoch": 3.58,
924
+ "learning_rate": 1.4247071552900221e-05,
925
+ "loss": 2.8961,
926
+ "step": 76000
927
+ },
928
+ {
929
+ "epoch": 3.6,
930
+ "learning_rate": 1.401185491837983e-05,
931
+ "loss": 2.8852,
932
+ "step": 76500
933
+ },
934
+ {
935
+ "epoch": 3.62,
936
+ "learning_rate": 1.3776638283859436e-05,
937
+ "loss": 2.8786,
938
+ "step": 77000
939
+ },
940
+ {
941
+ "epoch": 3.65,
942
+ "learning_rate": 1.3541421649339042e-05,
943
+ "loss": 2.8713,
944
+ "step": 77500
945
+ },
946
+ {
947
+ "epoch": 3.67,
948
+ "learning_rate": 1.3306205014818648e-05,
949
+ "loss": 2.8909,
950
+ "step": 78000
951
+ },
952
+ {
953
+ "epoch": 3.69,
954
+ "learning_rate": 1.3070988380298255e-05,
955
+ "loss": 2.8746,
956
+ "step": 78500
957
+ },
958
+ {
959
+ "epoch": 3.72,
960
+ "learning_rate": 1.283577174577786e-05,
961
+ "loss": 2.8682,
962
+ "step": 79000
963
+ },
964
+ {
965
+ "epoch": 3.74,
966
+ "learning_rate": 1.260055511125747e-05,
967
+ "loss": 2.871,
968
+ "step": 79500
969
+ },
970
+ {
971
+ "epoch": 3.76,
972
+ "learning_rate": 1.2365338476737076e-05,
973
+ "loss": 2.8652,
974
+ "step": 80000
975
+ },
976
+ {
977
+ "epoch": 3.79,
978
+ "learning_rate": 1.2130121842216682e-05,
979
+ "loss": 2.8685,
980
+ "step": 80500
981
+ },
982
+ {
983
+ "epoch": 3.81,
984
+ "learning_rate": 1.189490520769629e-05,
985
+ "loss": 2.859,
986
+ "step": 81000
987
+ },
988
+ {
989
+ "epoch": 3.83,
990
+ "learning_rate": 1.1659688573175896e-05,
991
+ "loss": 2.8699,
992
+ "step": 81500
993
+ },
994
+ {
995
+ "epoch": 3.86,
996
+ "learning_rate": 1.1424471938655503e-05,
997
+ "loss": 2.8599,
998
+ "step": 82000
999
+ },
1000
+ {
1001
+ "epoch": 3.88,
1002
+ "learning_rate": 1.1189255304135108e-05,
1003
+ "loss": 2.8636,
1004
+ "step": 82500
1005
+ },
1006
+ {
1007
+ "epoch": 3.9,
1008
+ "learning_rate": 1.0954038669614716e-05,
1009
+ "loss": 2.8681,
1010
+ "step": 83000
1011
+ },
1012
+ {
1013
+ "epoch": 3.93,
1014
+ "learning_rate": 1.0718822035094323e-05,
1015
+ "loss": 2.8511,
1016
+ "step": 83500
1017
+ },
1018
+ {
1019
+ "epoch": 3.95,
1020
+ "learning_rate": 1.0483605400573928e-05,
1021
+ "loss": 2.8546,
1022
+ "step": 84000
1023
+ },
1024
+ {
1025
+ "epoch": 3.98,
1026
+ "learning_rate": 1.0248388766053537e-05,
1027
+ "loss": 2.8476,
1028
+ "step": 84500
1029
+ },
1030
+ {
1031
+ "epoch": 4.0,
1032
+ "learning_rate": 1.0013172131533142e-05,
1033
+ "loss": 2.8516,
1034
+ "step": 85000
1035
+ },
1036
+ {
1037
+ "epoch": 4.02,
1038
+ "learning_rate": 9.777955497012749e-06,
1039
+ "loss": 2.8439,
1040
+ "step": 85500
1041
+ },
1042
+ {
1043
+ "epoch": 4.05,
1044
+ "learning_rate": 9.542738862492356e-06,
1045
+ "loss": 2.8427,
1046
+ "step": 86000
1047
+ },
1048
+ {
1049
+ "epoch": 4.07,
1050
+ "learning_rate": 9.307522227971962e-06,
1051
+ "loss": 2.8392,
1052
+ "step": 86500
1053
+ },
1054
+ {
1055
+ "epoch": 4.09,
1056
+ "learning_rate": 9.07230559345157e-06,
1057
+ "loss": 2.8381,
1058
+ "step": 87000
1059
+ },
1060
+ {
1061
+ "epoch": 4.12,
1062
+ "learning_rate": 8.837088958931176e-06,
1063
+ "loss": 2.8347,
1064
+ "step": 87500
1065
+ },
1066
+ {
1067
+ "epoch": 4.14,
1068
+ "learning_rate": 8.601872324410783e-06,
1069
+ "loss": 2.8354,
1070
+ "step": 88000
1071
+ },
1072
+ {
1073
+ "epoch": 4.16,
1074
+ "learning_rate": 8.36665568989039e-06,
1075
+ "loss": 2.8351,
1076
+ "step": 88500
1077
+ },
1078
+ {
1079
+ "epoch": 4.19,
1080
+ "learning_rate": 8.131439055369996e-06,
1081
+ "loss": 2.8283,
1082
+ "step": 89000
1083
+ },
1084
+ {
1085
+ "epoch": 4.21,
1086
+ "learning_rate": 7.896222420849603e-06,
1087
+ "loss": 2.8277,
1088
+ "step": 89500
1089
+ },
1090
+ {
1091
+ "epoch": 4.23,
1092
+ "learning_rate": 7.661005786329208e-06,
1093
+ "loss": 2.83,
1094
+ "step": 90000
1095
+ },
1096
+ {
1097
+ "epoch": 4.26,
1098
+ "learning_rate": 7.425789151808817e-06,
1099
+ "loss": 2.8247,
1100
+ "step": 90500
1101
+ },
1102
+ {
1103
+ "epoch": 4.28,
1104
+ "learning_rate": 7.190572517288423e-06,
1105
+ "loss": 2.8238,
1106
+ "step": 91000
1107
+ },
1108
+ {
1109
+ "epoch": 4.3,
1110
+ "learning_rate": 6.955355882768029e-06,
1111
+ "loss": 2.8292,
1112
+ "step": 91500
1113
+ },
1114
+ {
1115
+ "epoch": 4.33,
1116
+ "learning_rate": 6.720139248247636e-06,
1117
+ "loss": 2.8368,
1118
+ "step": 92000
1119
+ },
1120
+ {
1121
+ "epoch": 4.35,
1122
+ "learning_rate": 6.484922613727243e-06,
1123
+ "loss": 2.8247,
1124
+ "step": 92500
1125
+ },
1126
+ {
1127
+ "epoch": 4.38,
1128
+ "learning_rate": 6.24970597920685e-06,
1129
+ "loss": 2.8258,
1130
+ "step": 93000
1131
+ },
1132
+ {
1133
+ "epoch": 4.4,
1134
+ "learning_rate": 6.014489344686457e-06,
1135
+ "loss": 2.8128,
1136
+ "step": 93500
1137
+ },
1138
+ {
1139
+ "epoch": 4.42,
1140
+ "learning_rate": 5.779272710166063e-06,
1141
+ "loss": 2.82,
1142
+ "step": 94000
1143
+ },
1144
+ {
1145
+ "epoch": 4.45,
1146
+ "learning_rate": 5.5440560756456695e-06,
1147
+ "loss": 2.8101,
1148
+ "step": 94500
1149
+ },
1150
+ {
1151
+ "epoch": 4.47,
1152
+ "learning_rate": 5.308839441125277e-06,
1153
+ "loss": 2.8218,
1154
+ "step": 95000
1155
+ },
1156
+ {
1157
+ "epoch": 4.49,
1158
+ "learning_rate": 5.073622806604884e-06,
1159
+ "loss": 2.8205,
1160
+ "step": 95500
1161
+ },
1162
+ {
1163
+ "epoch": 4.52,
1164
+ "learning_rate": 4.83840617208449e-06,
1165
+ "loss": 2.8134,
1166
+ "step": 96000
1167
+ },
1168
+ {
1169
+ "epoch": 4.54,
1170
+ "learning_rate": 4.603189537564097e-06,
1171
+ "loss": 2.8081,
1172
+ "step": 96500
1173
+ },
1174
+ {
1175
+ "epoch": 4.56,
1176
+ "learning_rate": 4.3679729030437035e-06,
1177
+ "loss": 2.8082,
1178
+ "step": 97000
1179
+ },
1180
+ {
1181
+ "epoch": 4.59,
1182
+ "learning_rate": 4.13275626852331e-06,
1183
+ "loss": 2.8099,
1184
+ "step": 97500
1185
+ },
1186
+ {
1187
+ "epoch": 4.61,
1188
+ "learning_rate": 3.897539634002917e-06,
1189
+ "loss": 2.8096,
1190
+ "step": 98000
1191
+ },
1192
+ {
1193
+ "epoch": 4.63,
1194
+ "learning_rate": 3.662322999482524e-06,
1195
+ "loss": 2.8078,
1196
+ "step": 98500
1197
+ },
1198
+ {
1199
+ "epoch": 4.66,
1200
+ "learning_rate": 3.42710636496213e-06,
1201
+ "loss": 2.8151,
1202
+ "step": 99000
1203
+ },
1204
+ {
1205
+ "epoch": 4.68,
1206
+ "learning_rate": 3.191889730441737e-06,
1207
+ "loss": 2.795,
1208
+ "step": 99500
1209
+ },
1210
+ {
1211
+ "epoch": 4.7,
1212
+ "learning_rate": 2.956673095921344e-06,
1213
+ "loss": 2.8095,
1214
+ "step": 100000
1215
+ },
1216
+ {
1217
+ "epoch": 4.73,
1218
+ "learning_rate": 2.7214564614009503e-06,
1219
+ "loss": 2.8146,
1220
+ "step": 100500
1221
+ },
1222
+ {
1223
+ "epoch": 4.75,
1224
+ "learning_rate": 2.486239826880557e-06,
1225
+ "loss": 2.8022,
1226
+ "step": 101000
1227
+ },
1228
+ {
1229
+ "epoch": 4.77,
1230
+ "learning_rate": 2.251023192360164e-06,
1231
+ "loss": 2.8,
1232
+ "step": 101500
1233
+ },
1234
+ {
1235
+ "epoch": 4.8,
1236
+ "learning_rate": 2.0158065578397707e-06,
1237
+ "loss": 2.7974,
1238
+ "step": 102000
1239
+ },
1240
+ {
1241
+ "epoch": 4.82,
1242
+ "learning_rate": 1.780589923319377e-06,
1243
+ "loss": 2.8006,
1244
+ "step": 102500
1245
+ },
1246
+ {
1247
+ "epoch": 4.85,
1248
+ "learning_rate": 1.545373288798984e-06,
1249
+ "loss": 2.795,
1250
+ "step": 103000
1251
+ },
1252
+ {
1253
+ "epoch": 4.87,
1254
+ "learning_rate": 1.3101566542785905e-06,
1255
+ "loss": 2.8012,
1256
+ "step": 103500
1257
+ },
1258
+ {
1259
+ "epoch": 4.89,
1260
+ "learning_rate": 1.0749400197581973e-06,
1261
+ "loss": 2.8025,
1262
+ "step": 104000
1263
+ },
1264
+ {
1265
+ "epoch": 4.92,
1266
+ "learning_rate": 8.397233852378041e-07,
1267
+ "loss": 2.8024,
1268
+ "step": 104500
1269
+ },
1270
+ {
1271
+ "epoch": 4.94,
1272
+ "learning_rate": 6.045067507174108e-07,
1273
+ "loss": 2.7982,
1274
+ "step": 105000
1275
+ },
1276
+ {
1277
+ "epoch": 4.96,
1278
+ "learning_rate": 3.6929011619701747e-07,
1279
+ "loss": 2.7953,
1280
+ "step": 105500
1281
+ },
1282
+ {
1283
+ "epoch": 4.99,
1284
+ "learning_rate": 1.340734816766242e-07,
1285
+ "loss": 2.8075,
1286
+ "step": 106000
1287
+ }
1288
+ ],
1289
+ "max_steps": 106285,
1290
+ "num_train_epochs": 5,
1291
+ "total_flos": 3.567911344090743e+18,
1292
+ "trial_name": null,
1293
+ "trial_params": null
1294
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:964183d647b297eb67c45e9921fa035afd6f61598344e546ceced898ea9b913d
3
+ size 3003
vocab.json ADDED
The diff for this file is too large to render. See raw diff