tadsatlawa-na commited on
Commit
3f43d0c
1 Parent(s): 7eebdd3

Add human_heavy_model

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 +719 -0
  8. training_args.bin +3 -0
  9. vocab.json +27 -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": 320,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "layer_norm_eps": 1e-12,
15
+ "max_position_embeddings": 202,
16
+ "model_type": "roberta",
17
+ "num_attention_heads": 20,
18
+ "num_hidden_layers": 6,
19
+ "pad_token_id": 1,
20
+ "position_embedding_type": "absolute",
21
+ "torch_dtype": "float32",
22
+ "transformers_version": "4.32.1",
23
+ "type_vocab_size": 1,
24
+ "use_cache": true,
25
+ "vocab_size": 25
26
+ }
merges.txt ADDED
File without changes
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7938a9001da6ef796f009200a9e1fb957b692fab1d3c11a932c84f3b00f14a8b
3
+ size 115822597
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:278807d75adc9f58a665030c4a6a0de5faaa07393efd436617522c105e8214d0
3
+ size 57903285
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:727a60f6a2c65d7708d64932ca144a518147a4c4548d32a95d84fc33aa759ee3
3
+ size 14511
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c23d4145a9ebb7344c7d6ca631d5dd3d8c34e9352fa0e03d8c5bf972a43fd7c6
3
+ size 627
trainer_state.json ADDED
@@ -0,0 +1,719 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.991040318566451,
5
+ "eval_steps": 10000,
6
+ "global_step": 500000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.04,
13
+ "learning_rate": 9.83937751004016e-06,
14
+ "loss": 1.3677,
15
+ "step": 10000
16
+ },
17
+ {
18
+ "epoch": 0.04,
19
+ "eval_loss": 0.8635309934616089,
20
+ "eval_runtime": 127.2968,
21
+ "eval_samples_per_second": 1932.492,
22
+ "eval_steps_per_second": 20.134,
23
+ "step": 10000
24
+ },
25
+ {
26
+ "epoch": 0.08,
27
+ "learning_rate": 9.638674698795182e-06,
28
+ "loss": 0.7723,
29
+ "step": 20000
30
+ },
31
+ {
32
+ "epoch": 0.08,
33
+ "eval_loss": 0.6709557175636292,
34
+ "eval_runtime": 127.3309,
35
+ "eval_samples_per_second": 1931.974,
36
+ "eval_steps_per_second": 20.129,
37
+ "step": 20000
38
+ },
39
+ {
40
+ "epoch": 0.12,
41
+ "learning_rate": 9.437951807228917e-06,
42
+ "loss": 0.6635,
43
+ "step": 30000
44
+ },
45
+ {
46
+ "epoch": 0.12,
47
+ "eval_loss": 0.6188660264015198,
48
+ "eval_runtime": 127.3716,
49
+ "eval_samples_per_second": 1931.357,
50
+ "eval_steps_per_second": 20.122,
51
+ "step": 30000
52
+ },
53
+ {
54
+ "epoch": 0.16,
55
+ "learning_rate": 9.237248995983937e-06,
56
+ "loss": 0.62,
57
+ "step": 40000
58
+ },
59
+ {
60
+ "epoch": 0.16,
61
+ "eval_loss": 0.5857706069946289,
62
+ "eval_runtime": 128.3689,
63
+ "eval_samples_per_second": 1916.353,
64
+ "eval_steps_per_second": 19.966,
65
+ "step": 40000
66
+ },
67
+ {
68
+ "epoch": 0.2,
69
+ "learning_rate": 9.036465863453816e-06,
70
+ "loss": 0.5933,
71
+ "step": 50000
72
+ },
73
+ {
74
+ "epoch": 0.2,
75
+ "eval_loss": 0.5672558546066284,
76
+ "eval_runtime": 128.5129,
77
+ "eval_samples_per_second": 1914.205,
78
+ "eval_steps_per_second": 19.944,
79
+ "step": 50000
80
+ },
81
+ {
82
+ "epoch": 0.24,
83
+ "learning_rate": 8.835763052208836e-06,
84
+ "loss": 0.5755,
85
+ "step": 60000
86
+ },
87
+ {
88
+ "epoch": 0.24,
89
+ "eval_loss": 0.5512102842330933,
90
+ "eval_runtime": 127.3167,
91
+ "eval_samples_per_second": 1932.19,
92
+ "eval_steps_per_second": 20.131,
93
+ "step": 60000
94
+ },
95
+ {
96
+ "epoch": 0.28,
97
+ "learning_rate": 8.63504016064257e-06,
98
+ "loss": 0.5622,
99
+ "step": 70000
100
+ },
101
+ {
102
+ "epoch": 0.28,
103
+ "eval_loss": 0.5416693687438965,
104
+ "eval_runtime": 127.5393,
105
+ "eval_samples_per_second": 1928.817,
106
+ "eval_steps_per_second": 20.096,
107
+ "step": 70000
108
+ },
109
+ {
110
+ "epoch": 0.32,
111
+ "learning_rate": 8.434337349397592e-06,
112
+ "loss": 0.5522,
113
+ "step": 80000
114
+ },
115
+ {
116
+ "epoch": 0.32,
117
+ "eval_loss": 0.5347551107406616,
118
+ "eval_runtime": 127.3132,
119
+ "eval_samples_per_second": 1932.243,
120
+ "eval_steps_per_second": 20.131,
121
+ "step": 80000
122
+ },
123
+ {
124
+ "epoch": 0.36,
125
+ "learning_rate": 8.23363453815261e-06,
126
+ "loss": 0.5437,
127
+ "step": 90000
128
+ },
129
+ {
130
+ "epoch": 0.36,
131
+ "eval_loss": 0.5291008353233337,
132
+ "eval_runtime": 127.3464,
133
+ "eval_samples_per_second": 1931.739,
134
+ "eval_steps_per_second": 20.126,
135
+ "step": 90000
136
+ },
137
+ {
138
+ "epoch": 0.4,
139
+ "learning_rate": 8.032931726907631e-06,
140
+ "loss": 0.5365,
141
+ "step": 100000
142
+ },
143
+ {
144
+ "epoch": 0.4,
145
+ "eval_loss": 0.5225369334220886,
146
+ "eval_runtime": 127.5532,
147
+ "eval_samples_per_second": 1928.608,
148
+ "eval_steps_per_second": 20.094,
149
+ "step": 100000
150
+ },
151
+ {
152
+ "epoch": 0.44,
153
+ "learning_rate": 7.832228915662651e-06,
154
+ "loss": 0.5309,
155
+ "step": 110000
156
+ },
157
+ {
158
+ "epoch": 0.44,
159
+ "eval_loss": 0.5173077583312988,
160
+ "eval_runtime": 127.5475,
161
+ "eval_samples_per_second": 1928.693,
162
+ "eval_steps_per_second": 20.094,
163
+ "step": 110000
164
+ },
165
+ {
166
+ "epoch": 0.48,
167
+ "learning_rate": 7.63152610441767e-06,
168
+ "loss": 0.5252,
169
+ "step": 120000
170
+ },
171
+ {
172
+ "epoch": 0.48,
173
+ "eval_loss": 0.513536274433136,
174
+ "eval_runtime": 130.7886,
175
+ "eval_samples_per_second": 1880.897,
176
+ "eval_steps_per_second": 19.597,
177
+ "step": 120000
178
+ },
179
+ {
180
+ "epoch": 0.52,
181
+ "learning_rate": 7.430823293172691e-06,
182
+ "loss": 0.5204,
183
+ "step": 130000
184
+ },
185
+ {
186
+ "epoch": 0.52,
187
+ "eval_loss": 0.5111202001571655,
188
+ "eval_runtime": 127.6816,
189
+ "eval_samples_per_second": 1926.668,
190
+ "eval_steps_per_second": 20.073,
191
+ "step": 130000
192
+ },
193
+ {
194
+ "epoch": 0.56,
195
+ "learning_rate": 7.230100401606426e-06,
196
+ "loss": 0.518,
197
+ "step": 140000
198
+ },
199
+ {
200
+ "epoch": 0.56,
201
+ "eval_loss": 0.5084987282752991,
202
+ "eval_runtime": 128.2814,
203
+ "eval_samples_per_second": 1917.659,
204
+ "eval_steps_per_second": 19.98,
205
+ "step": 140000
206
+ },
207
+ {
208
+ "epoch": 0.6,
209
+ "learning_rate": 7.029397590361447e-06,
210
+ "loss": 0.5135,
211
+ "step": 150000
212
+ },
213
+ {
214
+ "epoch": 0.6,
215
+ "eval_loss": 0.5029130578041077,
216
+ "eval_runtime": 130.8657,
217
+ "eval_samples_per_second": 1879.79,
218
+ "eval_steps_per_second": 19.585,
219
+ "step": 150000
220
+ },
221
+ {
222
+ "epoch": 0.64,
223
+ "learning_rate": 6.828694779116466e-06,
224
+ "loss": 0.5101,
225
+ "step": 160000
226
+ },
227
+ {
228
+ "epoch": 0.64,
229
+ "eval_loss": 0.5005983710289001,
230
+ "eval_runtime": 128.9427,
231
+ "eval_samples_per_second": 1907.824,
232
+ "eval_steps_per_second": 19.877,
233
+ "step": 160000
234
+ },
235
+ {
236
+ "epoch": 0.68,
237
+ "learning_rate": 6.627991967871487e-06,
238
+ "loss": 0.5065,
239
+ "step": 170000
240
+ },
241
+ {
242
+ "epoch": 0.68,
243
+ "eval_loss": 0.4987814128398895,
244
+ "eval_runtime": 127.8532,
245
+ "eval_samples_per_second": 1924.081,
246
+ "eval_steps_per_second": 20.046,
247
+ "step": 170000
248
+ },
249
+ {
250
+ "epoch": 0.72,
251
+ "learning_rate": 6.427289156626506e-06,
252
+ "loss": 0.5052,
253
+ "step": 180000
254
+ },
255
+ {
256
+ "epoch": 0.72,
257
+ "eval_loss": 0.49448052048683167,
258
+ "eval_runtime": 128.0108,
259
+ "eval_samples_per_second": 1921.713,
260
+ "eval_steps_per_second": 20.022,
261
+ "step": 180000
262
+ },
263
+ {
264
+ "epoch": 0.76,
265
+ "learning_rate": 6.226586345381527e-06,
266
+ "loss": 0.5025,
267
+ "step": 190000
268
+ },
269
+ {
270
+ "epoch": 0.76,
271
+ "eval_loss": 0.49206921458244324,
272
+ "eval_runtime": 127.1399,
273
+ "eval_samples_per_second": 1934.877,
274
+ "eval_steps_per_second": 20.159,
275
+ "step": 190000
276
+ },
277
+ {
278
+ "epoch": 0.8,
279
+ "learning_rate": 6.025883534136546e-06,
280
+ "loss": 0.4998,
281
+ "step": 200000
282
+ },
283
+ {
284
+ "epoch": 0.8,
285
+ "eval_loss": 0.4929586946964264,
286
+ "eval_runtime": 127.6742,
287
+ "eval_samples_per_second": 1926.779,
288
+ "eval_steps_per_second": 20.075,
289
+ "step": 200000
290
+ },
291
+ {
292
+ "epoch": 0.84,
293
+ "learning_rate": 5.825180722891567e-06,
294
+ "loss": 0.4982,
295
+ "step": 210000
296
+ },
297
+ {
298
+ "epoch": 0.84,
299
+ "eval_loss": 0.48860839009284973,
300
+ "eval_runtime": 127.3878,
301
+ "eval_samples_per_second": 1931.112,
302
+ "eval_steps_per_second": 20.12,
303
+ "step": 210000
304
+ },
305
+ {
306
+ "epoch": 0.88,
307
+ "learning_rate": 5.6244578313253014e-06,
308
+ "loss": 0.4969,
309
+ "step": 220000
310
+ },
311
+ {
312
+ "epoch": 0.88,
313
+ "eval_loss": 0.4888823628425598,
314
+ "eval_runtime": 127.2533,
315
+ "eval_samples_per_second": 1933.152,
316
+ "eval_steps_per_second": 20.141,
317
+ "step": 220000
318
+ },
319
+ {
320
+ "epoch": 0.92,
321
+ "learning_rate": 5.423755020080321e-06,
322
+ "loss": 0.495,
323
+ "step": 230000
324
+ },
325
+ {
326
+ "epoch": 0.92,
327
+ "eval_loss": 0.4841912090778351,
328
+ "eval_runtime": 126.968,
329
+ "eval_samples_per_second": 1937.496,
330
+ "eval_steps_per_second": 20.186,
331
+ "step": 230000
332
+ },
333
+ {
334
+ "epoch": 0.96,
335
+ "learning_rate": 5.223052208835342e-06,
336
+ "loss": 0.4927,
337
+ "step": 240000
338
+ },
339
+ {
340
+ "epoch": 0.96,
341
+ "eval_loss": 0.4853549897670746,
342
+ "eval_runtime": 127.4501,
343
+ "eval_samples_per_second": 1930.167,
344
+ "eval_steps_per_second": 20.11,
345
+ "step": 240000
346
+ },
347
+ {
348
+ "epoch": 1.0,
349
+ "learning_rate": 5.022349397590361e-06,
350
+ "loss": 0.4914,
351
+ "step": 250000
352
+ },
353
+ {
354
+ "epoch": 1.0,
355
+ "eval_loss": 0.4826248586177826,
356
+ "eval_runtime": 127.3161,
357
+ "eval_samples_per_second": 1932.198,
358
+ "eval_steps_per_second": 20.131,
359
+ "step": 250000
360
+ },
361
+ {
362
+ "epoch": 1.04,
363
+ "learning_rate": 4.821646586345382e-06,
364
+ "loss": 0.4902,
365
+ "step": 260000
366
+ },
367
+ {
368
+ "epoch": 1.04,
369
+ "eval_loss": 0.48145654797554016,
370
+ "eval_runtime": 127.4143,
371
+ "eval_samples_per_second": 1930.709,
372
+ "eval_steps_per_second": 20.115,
373
+ "step": 260000
374
+ },
375
+ {
376
+ "epoch": 1.08,
377
+ "learning_rate": 4.620943775100402e-06,
378
+ "loss": 0.4894,
379
+ "step": 270000
380
+ },
381
+ {
382
+ "epoch": 1.08,
383
+ "eval_loss": 0.47896286845207214,
384
+ "eval_runtime": 127.419,
385
+ "eval_samples_per_second": 1930.638,
386
+ "eval_steps_per_second": 20.115,
387
+ "step": 270000
388
+ },
389
+ {
390
+ "epoch": 1.11,
391
+ "learning_rate": 4.420240963855422e-06,
392
+ "loss": 0.4881,
393
+ "step": 280000
394
+ },
395
+ {
396
+ "epoch": 1.11,
397
+ "eval_loss": 0.48297473788261414,
398
+ "eval_runtime": 127.4472,
399
+ "eval_samples_per_second": 1930.211,
400
+ "eval_steps_per_second": 20.11,
401
+ "step": 280000
402
+ },
403
+ {
404
+ "epoch": 1.15,
405
+ "learning_rate": 4.219538152610443e-06,
406
+ "loss": 0.487,
407
+ "step": 290000
408
+ },
409
+ {
410
+ "epoch": 1.15,
411
+ "eval_loss": 0.47816893458366394,
412
+ "eval_runtime": 127.1599,
413
+ "eval_samples_per_second": 1934.573,
414
+ "eval_steps_per_second": 20.156,
415
+ "step": 290000
416
+ },
417
+ {
418
+ "epoch": 1.19,
419
+ "learning_rate": 4.018835341365462e-06,
420
+ "loss": 0.4859,
421
+ "step": 300000
422
+ },
423
+ {
424
+ "epoch": 1.19,
425
+ "eval_loss": 0.4779074192047119,
426
+ "eval_runtime": 127.96,
427
+ "eval_samples_per_second": 1922.476,
428
+ "eval_steps_per_second": 20.03,
429
+ "step": 300000
430
+ },
431
+ {
432
+ "epoch": 1.23,
433
+ "learning_rate": 3.818132530120483e-06,
434
+ "loss": 0.4845,
435
+ "step": 310000
436
+ },
437
+ {
438
+ "epoch": 1.23,
439
+ "eval_loss": 0.47683581709861755,
440
+ "eval_runtime": 127.9081,
441
+ "eval_samples_per_second": 1923.256,
442
+ "eval_steps_per_second": 20.038,
443
+ "step": 310000
444
+ },
445
+ {
446
+ "epoch": 1.27,
447
+ "learning_rate": 3.6174297188755025e-06,
448
+ "loss": 0.4835,
449
+ "step": 320000
450
+ },
451
+ {
452
+ "epoch": 1.27,
453
+ "eval_loss": 0.4755454957485199,
454
+ "eval_runtime": 127.7316,
455
+ "eval_samples_per_second": 1925.914,
456
+ "eval_steps_per_second": 20.066,
457
+ "step": 320000
458
+ },
459
+ {
460
+ "epoch": 1.31,
461
+ "learning_rate": 3.4167068273092375e-06,
462
+ "loss": 0.483,
463
+ "step": 330000
464
+ },
465
+ {
466
+ "epoch": 1.31,
467
+ "eval_loss": 0.4744352400302887,
468
+ "eval_runtime": 128.197,
469
+ "eval_samples_per_second": 1918.921,
470
+ "eval_steps_per_second": 19.993,
471
+ "step": 330000
472
+ },
473
+ {
474
+ "epoch": 1.35,
475
+ "learning_rate": 3.2160040160642576e-06,
476
+ "loss": 0.4819,
477
+ "step": 340000
478
+ },
479
+ {
480
+ "epoch": 1.35,
481
+ "eval_loss": 0.4745638966560364,
482
+ "eval_runtime": 128.0584,
483
+ "eval_samples_per_second": 1920.999,
484
+ "eval_steps_per_second": 20.014,
485
+ "step": 340000
486
+ },
487
+ {
488
+ "epoch": 1.39,
489
+ "learning_rate": 3.0153012048192777e-06,
490
+ "loss": 0.481,
491
+ "step": 350000
492
+ },
493
+ {
494
+ "epoch": 1.39,
495
+ "eval_loss": 0.4744107723236084,
496
+ "eval_runtime": 128.1739,
497
+ "eval_samples_per_second": 1919.268,
498
+ "eval_steps_per_second": 19.996,
499
+ "step": 350000
500
+ },
501
+ {
502
+ "epoch": 1.43,
503
+ "learning_rate": 2.8145983935742978e-06,
504
+ "loss": 0.481,
505
+ "step": 360000
506
+ },
507
+ {
508
+ "epoch": 1.43,
509
+ "eval_loss": 0.472385436296463,
510
+ "eval_runtime": 128.2659,
511
+ "eval_samples_per_second": 1917.89,
512
+ "eval_steps_per_second": 19.982,
513
+ "step": 360000
514
+ },
515
+ {
516
+ "epoch": 1.47,
517
+ "learning_rate": 2.613895582329318e-06,
518
+ "loss": 0.4799,
519
+ "step": 370000
520
+ },
521
+ {
522
+ "epoch": 1.47,
523
+ "eval_loss": 0.4733026921749115,
524
+ "eval_runtime": 127.698,
525
+ "eval_samples_per_second": 1926.419,
526
+ "eval_steps_per_second": 20.071,
527
+ "step": 370000
528
+ },
529
+ {
530
+ "epoch": 1.51,
531
+ "learning_rate": 2.4131927710843376e-06,
532
+ "loss": 0.4795,
533
+ "step": 380000
534
+ },
535
+ {
536
+ "epoch": 1.51,
537
+ "eval_loss": 0.4719351530075073,
538
+ "eval_runtime": 128.2337,
539
+ "eval_samples_per_second": 1918.372,
540
+ "eval_steps_per_second": 19.987,
541
+ "step": 380000
542
+ },
543
+ {
544
+ "epoch": 1.55,
545
+ "learning_rate": 2.2124899598393577e-06,
546
+ "loss": 0.4784,
547
+ "step": 390000
548
+ },
549
+ {
550
+ "epoch": 1.55,
551
+ "eval_loss": 0.4699419438838959,
552
+ "eval_runtime": 127.9847,
553
+ "eval_samples_per_second": 1922.105,
554
+ "eval_steps_per_second": 20.026,
555
+ "step": 390000
556
+ },
557
+ {
558
+ "epoch": 1.59,
559
+ "learning_rate": 2.0117871485943778e-06,
560
+ "loss": 0.4785,
561
+ "step": 400000
562
+ },
563
+ {
564
+ "epoch": 1.59,
565
+ "eval_loss": 0.4711839556694031,
566
+ "eval_runtime": 127.89,
567
+ "eval_samples_per_second": 1923.527,
568
+ "eval_steps_per_second": 20.041,
569
+ "step": 400000
570
+ },
571
+ {
572
+ "epoch": 1.63,
573
+ "learning_rate": 1.8110843373493979e-06,
574
+ "loss": 0.4777,
575
+ "step": 410000
576
+ },
577
+ {
578
+ "epoch": 1.63,
579
+ "eval_loss": 0.46987083554267883,
580
+ "eval_runtime": 128.85,
581
+ "eval_samples_per_second": 1909.197,
582
+ "eval_steps_per_second": 19.891,
583
+ "step": 410000
584
+ },
585
+ {
586
+ "epoch": 1.67,
587
+ "learning_rate": 1.6103614457831327e-06,
588
+ "loss": 0.477,
589
+ "step": 420000
590
+ },
591
+ {
592
+ "epoch": 1.67,
593
+ "eval_loss": 0.46960577368736267,
594
+ "eval_runtime": 130.3922,
595
+ "eval_samples_per_second": 1886.616,
596
+ "eval_steps_per_second": 19.656,
597
+ "step": 420000
598
+ },
599
+ {
600
+ "epoch": 1.71,
601
+ "learning_rate": 1.4096586345381528e-06,
602
+ "loss": 0.4771,
603
+ "step": 430000
604
+ },
605
+ {
606
+ "epoch": 1.71,
607
+ "eval_loss": 0.47003933787345886,
608
+ "eval_runtime": 129.4605,
609
+ "eval_samples_per_second": 1900.193,
610
+ "eval_steps_per_second": 19.798,
611
+ "step": 430000
612
+ },
613
+ {
614
+ "epoch": 1.75,
615
+ "learning_rate": 1.2089558232931729e-06,
616
+ "loss": 0.4766,
617
+ "step": 440000
618
+ },
619
+ {
620
+ "epoch": 1.75,
621
+ "eval_loss": 0.47017282247543335,
622
+ "eval_runtime": 129.4902,
623
+ "eval_samples_per_second": 1899.758,
624
+ "eval_steps_per_second": 19.793,
625
+ "step": 440000
626
+ },
627
+ {
628
+ "epoch": 1.79,
629
+ "learning_rate": 1.008253012048193e-06,
630
+ "loss": 0.476,
631
+ "step": 450000
632
+ },
633
+ {
634
+ "epoch": 1.79,
635
+ "eval_loss": 0.46954795718193054,
636
+ "eval_runtime": 129.5407,
637
+ "eval_samples_per_second": 1899.017,
638
+ "eval_steps_per_second": 19.785,
639
+ "step": 450000
640
+ },
641
+ {
642
+ "epoch": 1.83,
643
+ "learning_rate": 8.07550200803213e-07,
644
+ "loss": 0.4757,
645
+ "step": 460000
646
+ },
647
+ {
648
+ "epoch": 1.83,
649
+ "eval_loss": 0.4694086015224457,
650
+ "eval_runtime": 129.2469,
651
+ "eval_samples_per_second": 1903.334,
652
+ "eval_steps_per_second": 19.83,
653
+ "step": 460000
654
+ },
655
+ {
656
+ "epoch": 1.87,
657
+ "learning_rate": 6.068273092369479e-07,
658
+ "loss": 0.4758,
659
+ "step": 470000
660
+ },
661
+ {
662
+ "epoch": 1.87,
663
+ "eval_loss": 0.4685874581336975,
664
+ "eval_runtime": 129.0023,
665
+ "eval_samples_per_second": 1906.943,
666
+ "eval_steps_per_second": 19.868,
667
+ "step": 470000
668
+ },
669
+ {
670
+ "epoch": 1.91,
671
+ "learning_rate": 4.061244979919679e-07,
672
+ "loss": 0.4754,
673
+ "step": 480000
674
+ },
675
+ {
676
+ "epoch": 1.91,
677
+ "eval_loss": 0.46817249059677124,
678
+ "eval_runtime": 130.8406,
679
+ "eval_samples_per_second": 1880.15,
680
+ "eval_steps_per_second": 19.589,
681
+ "step": 480000
682
+ },
683
+ {
684
+ "epoch": 1.95,
685
+ "learning_rate": 2.0542168674698798e-07,
686
+ "loss": 0.475,
687
+ "step": 490000
688
+ },
689
+ {
690
+ "epoch": 1.95,
691
+ "eval_loss": 0.4691283404827118,
692
+ "eval_runtime": 129.0096,
693
+ "eval_samples_per_second": 1906.836,
694
+ "eval_steps_per_second": 19.867,
695
+ "step": 490000
696
+ },
697
+ {
698
+ "epoch": 1.99,
699
+ "learning_rate": 4.718875502008032e-09,
700
+ "loss": 0.4756,
701
+ "step": 500000
702
+ },
703
+ {
704
+ "epoch": 1.99,
705
+ "eval_loss": 0.46795061230659485,
706
+ "eval_runtime": 129.0896,
707
+ "eval_samples_per_second": 1905.653,
708
+ "eval_steps_per_second": 19.854,
709
+ "step": 500000
710
+ }
711
+ ],
712
+ "logging_steps": 10000,
713
+ "max_steps": 500000,
714
+ "num_train_epochs": 2,
715
+ "save_steps": 10000,
716
+ "total_flos": 8.290835482935528e+17,
717
+ "trial_name": null,
718
+ "trial_params": null
719
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b42a4e020ab10a141d461fb57e52cea112c30887d6fb4aef002f23c2065856e
3
+ size 4155
vocab.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "<s>": 0,
3
+ "</s>": 2,
4
+ "<pad>": 1,
5
+ "<unk>": 3,
6
+ "<mask>": 4,
7
+ "A": 5,
8
+ "C": 6,
9
+ "D": 7,
10
+ "E": 8,
11
+ "F": 9,
12
+ "G": 10,
13
+ "H": 11,
14
+ "I": 12,
15
+ "K": 13,
16
+ "L": 14,
17
+ "M": 15,
18
+ "N": 16,
19
+ "P": 17,
20
+ "Q": 18,
21
+ "R": 19,
22
+ "S": 20,
23
+ "T": 21,
24
+ "V": 22,
25
+ "W": 23,
26
+ "Y": 24
27
+ }