nbtpj commited on
Commit
9273cef
1 Parent(s): 55321c0

End of training

Browse files
last-checkpoint/config.json DELETED
@@ -1,75 +0,0 @@
1
- {
2
- "_name_or_path": "facebook/bart-base",
3
- "activation_dropout": 0.1,
4
- "activation_function": "gelu",
5
- "add_bias_logits": false,
6
- "add_final_layer_norm": false,
7
- "architectures": [
8
- "BartForConditionalGeneration"
9
- ],
10
- "attention_dropout": 0.1,
11
- "bos_token_id": 0,
12
- "classif_dropout": 0.1,
13
- "classifier_dropout": 0.0,
14
- "d_model": 768,
15
- "decoder_attention_heads": 12,
16
- "decoder_ffn_dim": 3072,
17
- "decoder_layerdrop": 0.0,
18
- "decoder_layers": 6,
19
- "decoder_start_token_id": 2,
20
- "dropout": 0.1,
21
- "early_stopping": true,
22
- "encoder_attention_heads": 12,
23
- "encoder_ffn_dim": 3072,
24
- "encoder_layerdrop": 0.0,
25
- "encoder_layers": 6,
26
- "eos_token_id": 2,
27
- "forced_bos_token_id": 0,
28
- "forced_eos_token_id": 2,
29
- "gradient_checkpointing": false,
30
- "id2label": {
31
- "0": "LABEL_0",
32
- "1": "LABEL_1",
33
- "2": "LABEL_2"
34
- },
35
- "init_std": 0.02,
36
- "is_encoder_decoder": true,
37
- "label2id": {
38
- "LABEL_0": 0,
39
- "LABEL_1": 1,
40
- "LABEL_2": 2
41
- },
42
- "max_position_embeddings": 1024,
43
- "model_type": "bart",
44
- "no_repeat_ngram_size": 3,
45
- "normalize_before": false,
46
- "normalize_embedding": true,
47
- "num_beams": 4,
48
- "num_hidden_layers": 6,
49
- "pad_token_id": 1,
50
- "scale_embedding": false,
51
- "task_specific_params": {
52
- "summarization": {
53
- "length_penalty": 1.0,
54
- "max_length": 128,
55
- "min_length": 12,
56
- "num_beams": 4
57
- },
58
- "summarization_cnn": {
59
- "length_penalty": 2.0,
60
- "max_length": 142,
61
- "min_length": 56,
62
- "num_beams": 4
63
- },
64
- "summarization_xsum": {
65
- "length_penalty": 1.0,
66
- "max_length": 62,
67
- "min_length": 11,
68
- "num_beams": 6
69
- }
70
- },
71
- "torch_dtype": "float32",
72
- "transformers_version": "4.20.1",
73
- "use_cache": true,
74
- "vocab_size": 50265
75
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
last-checkpoint/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:38bbc7cd108d4066f57dece94d18666a0619186fb83ead8f178ba5374f85c4d4
3
- size 1115513717
 
 
 
 
last-checkpoint/pytorch_model.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:7b3390a170051de185cec86c8c4131be12aeb79afd4f2697385bcd3c5fe72132
3
- size 557969145
 
 
 
 
last-checkpoint/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:4d8425ca758719b66e03746ec5ac1a9a95c444f7ea14cfc4567c5f50fa51c740
3
- size 15523
 
 
 
 
last-checkpoint/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:03778c2e3b15a96003ceff9d06ee5b9ecdadc81dcfa36477cad7912647bf1ff9
3
- size 623
 
 
 
 
last-checkpoint/trainer_state.json DELETED
@@ -1,916 +0,0 @@
1
- {
2
- "best_metric": null,
3
- "best_model_checkpoint": null,
4
- "epoch": 1.7227518088893994,
5
- "global_step": 15000,
6
- "is_hyper_param_search": false,
7
- "is_local_process_zero": true,
8
- "is_world_process_zero": true,
9
- "log_history": [
10
- {
11
- "epoch": 0.01,
12
- "learning_rate": 4.968098002934984e-05,
13
- "loss": 0.8326,
14
- "step": 100
15
- },
16
- {
17
- "epoch": 0.02,
18
- "learning_rate": 4.936196005869967e-05,
19
- "loss": 0.8028,
20
- "step": 200
21
- },
22
- {
23
- "epoch": 0.03,
24
- "learning_rate": 4.9042940088049516e-05,
25
- "loss": 0.7492,
26
- "step": 300
27
- },
28
- {
29
- "epoch": 0.05,
30
- "learning_rate": 4.8723920117399354e-05,
31
- "loss": 0.7413,
32
- "step": 400
33
- },
34
- {
35
- "epoch": 0.06,
36
- "learning_rate": 4.840490014674919e-05,
37
- "loss": 0.7788,
38
- "step": 500
39
- },
40
- {
41
- "epoch": 0.07,
42
- "learning_rate": 4.808588017609902e-05,
43
- "loss": 0.6938,
44
- "step": 600
45
- },
46
- {
47
- "epoch": 0.08,
48
- "learning_rate": 4.776686020544887e-05,
49
- "loss": 0.7157,
50
- "step": 700
51
- },
52
- {
53
- "epoch": 0.09,
54
- "learning_rate": 4.7447840234798705e-05,
55
- "loss": 0.7606,
56
- "step": 800
57
- },
58
- {
59
- "epoch": 0.1,
60
- "learning_rate": 4.7128820264148536e-05,
61
- "loss": 0.7222,
62
- "step": 900
63
- },
64
- {
65
- "epoch": 0.11,
66
- "learning_rate": 4.6809800293498373e-05,
67
- "loss": 0.6986,
68
- "step": 1000
69
- },
70
- {
71
- "epoch": 0.13,
72
- "learning_rate": 4.649078032284822e-05,
73
- "loss": 0.7202,
74
- "step": 1100
75
- },
76
- {
77
- "epoch": 0.14,
78
- "learning_rate": 4.617176035219805e-05,
79
- "loss": 0.7456,
80
- "step": 1200
81
- },
82
- {
83
- "epoch": 0.15,
84
- "learning_rate": 4.585274038154789e-05,
85
- "loss": 0.6753,
86
- "step": 1300
87
- },
88
- {
89
- "epoch": 0.16,
90
- "learning_rate": 4.5533720410897725e-05,
91
- "loss": 0.6983,
92
- "step": 1400
93
- },
94
- {
95
- "epoch": 0.17,
96
- "learning_rate": 4.521470044024756e-05,
97
- "loss": 0.7113,
98
- "step": 1500
99
- },
100
- {
101
- "epoch": 0.18,
102
- "learning_rate": 4.48956804695974e-05,
103
- "loss": 0.7059,
104
- "step": 1600
105
- },
106
- {
107
- "epoch": 0.2,
108
- "learning_rate": 4.457666049894724e-05,
109
- "loss": 0.691,
110
- "step": 1700
111
- },
112
- {
113
- "epoch": 0.21,
114
- "learning_rate": 4.4257640528297076e-05,
115
- "loss": 0.752,
116
- "step": 1800
117
- },
118
- {
119
- "epoch": 0.22,
120
- "learning_rate": 4.393862055764691e-05,
121
- "loss": 0.681,
122
- "step": 1900
123
- },
124
- {
125
- "epoch": 0.23,
126
- "learning_rate": 4.361960058699675e-05,
127
- "loss": 0.7501,
128
- "step": 2000
129
- },
130
- {
131
- "epoch": 0.24,
132
- "learning_rate": 4.330058061634659e-05,
133
- "loss": 0.6956,
134
- "step": 2100
135
- },
136
- {
137
- "epoch": 0.25,
138
- "learning_rate": 4.298156064569642e-05,
139
- "loss": 0.6996,
140
- "step": 2200
141
- },
142
- {
143
- "epoch": 0.26,
144
- "learning_rate": 4.266254067504626e-05,
145
- "loss": 0.6891,
146
- "step": 2300
147
- },
148
- {
149
- "epoch": 0.28,
150
- "learning_rate": 4.2343520704396096e-05,
151
- "loss": 0.7419,
152
- "step": 2400
153
- },
154
- {
155
- "epoch": 0.29,
156
- "learning_rate": 4.202450073374593e-05,
157
- "loss": 0.6979,
158
- "step": 2500
159
- },
160
- {
161
- "epoch": 0.3,
162
- "learning_rate": 4.170548076309577e-05,
163
- "loss": 0.6425,
164
- "step": 2600
165
- },
166
- {
167
- "epoch": 0.31,
168
- "learning_rate": 4.138646079244561e-05,
169
- "loss": 0.7243,
170
- "step": 2700
171
- },
172
- {
173
- "epoch": 0.32,
174
- "learning_rate": 4.106744082179545e-05,
175
- "loss": 0.7348,
176
- "step": 2800
177
- },
178
- {
179
- "epoch": 0.33,
180
- "learning_rate": 4.0748420851145285e-05,
181
- "loss": 0.6352,
182
- "step": 2900
183
- },
184
- {
185
- "epoch": 0.34,
186
- "learning_rate": 4.042940088049512e-05,
187
- "loss": 0.6469,
188
- "step": 3000
189
- },
190
- {
191
- "epoch": 0.36,
192
- "learning_rate": 4.011038090984496e-05,
193
- "loss": 0.718,
194
- "step": 3100
195
- },
196
- {
197
- "epoch": 0.37,
198
- "learning_rate": 3.979136093919479e-05,
199
- "loss": 0.7129,
200
- "step": 3200
201
- },
202
- {
203
- "epoch": 0.38,
204
- "learning_rate": 3.9472340968544636e-05,
205
- "loss": 0.672,
206
- "step": 3300
207
- },
208
- {
209
- "epoch": 0.39,
210
- "learning_rate": 3.9153320997894473e-05,
211
- "loss": 0.6616,
212
- "step": 3400
213
- },
214
- {
215
- "epoch": 0.4,
216
- "learning_rate": 3.8834301027244304e-05,
217
- "loss": 0.7105,
218
- "step": 3500
219
- },
220
- {
221
- "epoch": 0.41,
222
- "learning_rate": 3.851528105659414e-05,
223
- "loss": 0.7299,
224
- "step": 3600
225
- },
226
- {
227
- "epoch": 0.42,
228
- "learning_rate": 3.819626108594398e-05,
229
- "loss": 0.7022,
230
- "step": 3700
231
- },
232
- {
233
- "epoch": 0.44,
234
- "learning_rate": 3.787724111529382e-05,
235
- "loss": 0.6225,
236
- "step": 3800
237
- },
238
- {
239
- "epoch": 0.45,
240
- "learning_rate": 3.7558221144643656e-05,
241
- "loss": 0.6461,
242
- "step": 3900
243
- },
244
- {
245
- "epoch": 0.46,
246
- "learning_rate": 3.723920117399349e-05,
247
- "loss": 0.6935,
248
- "step": 4000
249
- },
250
- {
251
- "epoch": 0.47,
252
- "learning_rate": 3.692018120334333e-05,
253
- "loss": 0.671,
254
- "step": 4100
255
- },
256
- {
257
- "epoch": 0.48,
258
- "learning_rate": 3.660116123269317e-05,
259
- "loss": 0.6818,
260
- "step": 4200
261
- },
262
- {
263
- "epoch": 0.49,
264
- "learning_rate": 3.628214126204301e-05,
265
- "loss": 0.6343,
266
- "step": 4300
267
- },
268
- {
269
- "epoch": 0.51,
270
- "learning_rate": 3.5963121291392844e-05,
271
- "loss": 0.6623,
272
- "step": 4400
273
- },
274
- {
275
- "epoch": 0.52,
276
- "learning_rate": 3.5644101320742675e-05,
277
- "loss": 0.6896,
278
- "step": 4500
279
- },
280
- {
281
- "epoch": 0.53,
282
- "learning_rate": 3.532508135009251e-05,
283
- "loss": 0.6735,
284
- "step": 4600
285
- },
286
- {
287
- "epoch": 0.54,
288
- "learning_rate": 3.500606137944236e-05,
289
- "loss": 0.6723,
290
- "step": 4700
291
- },
292
- {
293
- "epoch": 0.55,
294
- "learning_rate": 3.4687041408792196e-05,
295
- "loss": 0.6224,
296
- "step": 4800
297
- },
298
- {
299
- "epoch": 0.56,
300
- "learning_rate": 3.4368021438142027e-05,
301
- "loss": 0.6902,
302
- "step": 4900
303
- },
304
- {
305
- "epoch": 0.57,
306
- "learning_rate": 3.4049001467491864e-05,
307
- "loss": 0.6264,
308
- "step": 5000
309
- },
310
- {
311
- "epoch": 0.59,
312
- "learning_rate": 3.372998149684171e-05,
313
- "loss": 0.6278,
314
- "step": 5100
315
- },
316
- {
317
- "epoch": 0.6,
318
- "learning_rate": 3.341096152619154e-05,
319
- "loss": 0.653,
320
- "step": 5200
321
- },
322
- {
323
- "epoch": 0.61,
324
- "learning_rate": 3.309194155554138e-05,
325
- "loss": 0.6709,
326
- "step": 5300
327
- },
328
- {
329
- "epoch": 0.62,
330
- "learning_rate": 3.2772921584891215e-05,
331
- "loss": 0.7207,
332
- "step": 5400
333
- },
334
- {
335
- "epoch": 0.63,
336
- "learning_rate": 3.245390161424105e-05,
337
- "loss": 0.671,
338
- "step": 5500
339
- },
340
- {
341
- "epoch": 0.64,
342
- "learning_rate": 3.213488164359089e-05,
343
- "loss": 0.7015,
344
- "step": 5600
345
- },
346
- {
347
- "epoch": 0.65,
348
- "learning_rate": 3.181586167294073e-05,
349
- "loss": 0.6896,
350
- "step": 5700
351
- },
352
- {
353
- "epoch": 0.67,
354
- "learning_rate": 3.1496841702290567e-05,
355
- "loss": 0.6674,
356
- "step": 5800
357
- },
358
- {
359
- "epoch": 0.68,
360
- "learning_rate": 3.11778217316404e-05,
361
- "loss": 0.6784,
362
- "step": 5900
363
- },
364
- {
365
- "epoch": 0.69,
366
- "learning_rate": 3.085880176099024e-05,
367
- "loss": 0.7048,
368
- "step": 6000
369
- },
370
- {
371
- "epoch": 0.7,
372
- "learning_rate": 3.053978179034008e-05,
373
- "loss": 0.6401,
374
- "step": 6100
375
- },
376
- {
377
- "epoch": 0.71,
378
- "learning_rate": 3.0220761819689914e-05,
379
- "loss": 0.7202,
380
- "step": 6200
381
- },
382
- {
383
- "epoch": 0.72,
384
- "learning_rate": 2.990174184903975e-05,
385
- "loss": 0.6472,
386
- "step": 6300
387
- },
388
- {
389
- "epoch": 0.74,
390
- "learning_rate": 2.958272187838959e-05,
391
- "loss": 0.6601,
392
- "step": 6400
393
- },
394
- {
395
- "epoch": 0.75,
396
- "learning_rate": 2.9263701907739428e-05,
397
- "loss": 0.6788,
398
- "step": 6500
399
- },
400
- {
401
- "epoch": 0.76,
402
- "learning_rate": 2.8944681937089262e-05,
403
- "loss": 0.745,
404
- "step": 6600
405
- },
406
- {
407
- "epoch": 0.77,
408
- "learning_rate": 2.86256619664391e-05,
409
- "loss": 0.6654,
410
- "step": 6700
411
- },
412
- {
413
- "epoch": 0.78,
414
- "learning_rate": 2.8306641995788934e-05,
415
- "loss": 0.6988,
416
- "step": 6800
417
- },
418
- {
419
- "epoch": 0.79,
420
- "learning_rate": 2.7987622025138775e-05,
421
- "loss": 0.6084,
422
- "step": 6900
423
- },
424
- {
425
- "epoch": 0.8,
426
- "learning_rate": 2.7668602054488613e-05,
427
- "loss": 0.6551,
428
- "step": 7000
429
- },
430
- {
431
- "epoch": 0.82,
432
- "learning_rate": 2.7349582083838448e-05,
433
- "loss": 0.6264,
434
- "step": 7100
435
- },
436
- {
437
- "epoch": 0.83,
438
- "learning_rate": 2.7030562113188285e-05,
439
- "loss": 0.6456,
440
- "step": 7200
441
- },
442
- {
443
- "epoch": 0.84,
444
- "learning_rate": 2.6711542142538127e-05,
445
- "loss": 0.7103,
446
- "step": 7300
447
- },
448
- {
449
- "epoch": 0.85,
450
- "learning_rate": 2.639252217188796e-05,
451
- "loss": 0.7238,
452
- "step": 7400
453
- },
454
- {
455
- "epoch": 0.86,
456
- "learning_rate": 2.60735022012378e-05,
457
- "loss": 0.6286,
458
- "step": 7500
459
- },
460
- {
461
- "epoch": 0.87,
462
- "learning_rate": 2.5754482230587633e-05,
463
- "loss": 0.6284,
464
- "step": 7600
465
- },
466
- {
467
- "epoch": 0.88,
468
- "learning_rate": 2.5435462259937474e-05,
469
- "loss": 0.6169,
470
- "step": 7700
471
- },
472
- {
473
- "epoch": 0.9,
474
- "learning_rate": 2.5116442289287312e-05,
475
- "loss": 0.6334,
476
- "step": 7800
477
- },
478
- {
479
- "epoch": 0.91,
480
- "learning_rate": 2.4797422318637146e-05,
481
- "loss": 0.6573,
482
- "step": 7900
483
- },
484
- {
485
- "epoch": 0.92,
486
- "learning_rate": 2.4478402347986984e-05,
487
- "loss": 0.6026,
488
- "step": 8000
489
- },
490
- {
491
- "epoch": 0.93,
492
- "learning_rate": 2.4159382377336822e-05,
493
- "loss": 0.6836,
494
- "step": 8100
495
- },
496
- {
497
- "epoch": 0.94,
498
- "learning_rate": 2.384036240668666e-05,
499
- "loss": 0.6093,
500
- "step": 8200
501
- },
502
- {
503
- "epoch": 0.95,
504
- "learning_rate": 2.3521342436036498e-05,
505
- "loss": 0.6603,
506
- "step": 8300
507
- },
508
- {
509
- "epoch": 0.96,
510
- "learning_rate": 2.3202322465386332e-05,
511
- "loss": 0.6312,
512
- "step": 8400
513
- },
514
- {
515
- "epoch": 0.98,
516
- "learning_rate": 2.2883302494736173e-05,
517
- "loss": 0.6312,
518
- "step": 8500
519
- },
520
- {
521
- "epoch": 0.99,
522
- "learning_rate": 2.2564282524086008e-05,
523
- "loss": 0.6278,
524
- "step": 8600
525
- },
526
- {
527
- "epoch": 1.0,
528
- "learning_rate": 2.224526255343585e-05,
529
- "loss": 0.6115,
530
- "step": 8700
531
- },
532
- {
533
- "epoch": 1.01,
534
- "learning_rate": 2.1926242582785683e-05,
535
- "loss": 0.6666,
536
- "step": 8800
537
- },
538
- {
539
- "epoch": 1.02,
540
- "learning_rate": 2.160722261213552e-05,
541
- "loss": 0.6047,
542
- "step": 8900
543
- },
544
- {
545
- "epoch": 1.03,
546
- "learning_rate": 2.128820264148536e-05,
547
- "loss": 0.6174,
548
- "step": 9000
549
- },
550
- {
551
- "epoch": 1.05,
552
- "learning_rate": 2.0969182670835193e-05,
553
- "loss": 0.5977,
554
- "step": 9100
555
- },
556
- {
557
- "epoch": 1.06,
558
- "learning_rate": 2.0650162700185034e-05,
559
- "loss": 0.553,
560
- "step": 9200
561
- },
562
- {
563
- "epoch": 1.07,
564
- "learning_rate": 2.033114272953487e-05,
565
- "loss": 0.6447,
566
- "step": 9300
567
- },
568
- {
569
- "epoch": 1.08,
570
- "learning_rate": 2.0012122758884706e-05,
571
- "loss": 0.5979,
572
- "step": 9400
573
- },
574
- {
575
- "epoch": 1.09,
576
- "learning_rate": 1.9693102788234544e-05,
577
- "loss": 0.6234,
578
- "step": 9500
579
- },
580
- {
581
- "epoch": 1.1,
582
- "learning_rate": 1.9374082817584382e-05,
583
- "loss": 0.5939,
584
- "step": 9600
585
- },
586
- {
587
- "epoch": 1.11,
588
- "learning_rate": 1.905506284693422e-05,
589
- "loss": 0.6481,
590
- "step": 9700
591
- },
592
- {
593
- "epoch": 1.13,
594
- "learning_rate": 1.8736042876284058e-05,
595
- "loss": 0.592,
596
- "step": 9800
597
- },
598
- {
599
- "epoch": 1.14,
600
- "learning_rate": 1.8417022905633892e-05,
601
- "loss": 0.6491,
602
- "step": 9900
603
- },
604
- {
605
- "epoch": 1.15,
606
- "learning_rate": 1.8098002934983733e-05,
607
- "loss": 0.5978,
608
- "step": 10000
609
- },
610
- {
611
- "epoch": 1.16,
612
- "learning_rate": 1.7778982964333567e-05,
613
- "loss": 0.6034,
614
- "step": 10100
615
- },
616
- {
617
- "epoch": 1.17,
618
- "learning_rate": 1.7459962993683405e-05,
619
- "loss": 0.6037,
620
- "step": 10200
621
- },
622
- {
623
- "epoch": 1.18,
624
- "learning_rate": 1.7140943023033243e-05,
625
- "loss": 0.6251,
626
- "step": 10300
627
- },
628
- {
629
- "epoch": 1.19,
630
- "learning_rate": 1.6821923052383077e-05,
631
- "loss": 0.581,
632
- "step": 10400
633
- },
634
- {
635
- "epoch": 1.21,
636
- "learning_rate": 1.650290308173292e-05,
637
- "loss": 0.5919,
638
- "step": 10500
639
- },
640
- {
641
- "epoch": 1.22,
642
- "learning_rate": 1.6183883111082753e-05,
643
- "loss": 0.6073,
644
- "step": 10600
645
- },
646
- {
647
- "epoch": 1.23,
648
- "learning_rate": 1.5864863140432594e-05,
649
- "loss": 0.6273,
650
- "step": 10700
651
- },
652
- {
653
- "epoch": 1.24,
654
- "learning_rate": 1.554584316978243e-05,
655
- "loss": 0.6198,
656
- "step": 10800
657
- },
658
- {
659
- "epoch": 1.25,
660
- "learning_rate": 1.5226823199132268e-05,
661
- "loss": 0.5993,
662
- "step": 10900
663
- },
664
- {
665
- "epoch": 1.26,
666
- "learning_rate": 1.4907803228482104e-05,
667
- "loss": 0.5519,
668
- "step": 11000
669
- },
670
- {
671
- "epoch": 1.27,
672
- "learning_rate": 1.4588783257831942e-05,
673
- "loss": 0.6152,
674
- "step": 11100
675
- },
676
- {
677
- "epoch": 1.29,
678
- "learning_rate": 1.4269763287181778e-05,
679
- "loss": 0.6251,
680
- "step": 11200
681
- },
682
- {
683
- "epoch": 1.3,
684
- "learning_rate": 1.3950743316531614e-05,
685
- "loss": 0.6747,
686
- "step": 11300
687
- },
688
- {
689
- "epoch": 1.31,
690
- "learning_rate": 1.3631723345881454e-05,
691
- "loss": 0.6253,
692
- "step": 11400
693
- },
694
- {
695
- "epoch": 1.32,
696
- "learning_rate": 1.331270337523129e-05,
697
- "loss": 0.6009,
698
- "step": 11500
699
- },
700
- {
701
- "epoch": 1.33,
702
- "learning_rate": 1.2993683404581127e-05,
703
- "loss": 0.6019,
704
- "step": 11600
705
- },
706
- {
707
- "epoch": 1.34,
708
- "learning_rate": 1.2674663433930964e-05,
709
- "loss": 0.6286,
710
- "step": 11700
711
- },
712
- {
713
- "epoch": 1.36,
714
- "learning_rate": 1.2355643463280801e-05,
715
- "loss": 0.5796,
716
- "step": 11800
717
- },
718
- {
719
- "epoch": 1.37,
720
- "learning_rate": 1.2036623492630639e-05,
721
- "loss": 0.6077,
722
- "step": 11900
723
- },
724
- {
725
- "epoch": 1.38,
726
- "learning_rate": 1.1717603521980477e-05,
727
- "loss": 0.6186,
728
- "step": 12000
729
- },
730
- {
731
- "epoch": 1.39,
732
- "learning_rate": 1.1398583551330313e-05,
733
- "loss": 0.6632,
734
- "step": 12100
735
- },
736
- {
737
- "epoch": 1.4,
738
- "learning_rate": 1.107956358068015e-05,
739
- "loss": 0.5796,
740
- "step": 12200
741
- },
742
- {
743
- "epoch": 1.41,
744
- "learning_rate": 1.0760543610029989e-05,
745
- "loss": 0.5951,
746
- "step": 12300
747
- },
748
- {
749
- "epoch": 1.42,
750
- "learning_rate": 1.0441523639379826e-05,
751
- "loss": 0.5743,
752
- "step": 12400
753
- },
754
- {
755
- "epoch": 1.44,
756
- "learning_rate": 1.0122503668729662e-05,
757
- "loss": 0.6705,
758
- "step": 12500
759
- },
760
- {
761
- "epoch": 1.45,
762
- "learning_rate": 9.8034836980795e-06,
763
- "loss": 0.5761,
764
- "step": 12600
765
- },
766
- {
767
- "epoch": 1.46,
768
- "learning_rate": 9.484463727429338e-06,
769
- "loss": 0.5709,
770
- "step": 12700
771
- },
772
- {
773
- "epoch": 1.47,
774
- "learning_rate": 9.165443756779176e-06,
775
- "loss": 0.562,
776
- "step": 12800
777
- },
778
- {
779
- "epoch": 1.48,
780
- "learning_rate": 8.846423786129014e-06,
781
- "loss": 0.6113,
782
- "step": 12900
783
- },
784
- {
785
- "epoch": 1.49,
786
- "learning_rate": 8.52740381547885e-06,
787
- "loss": 0.6137,
788
- "step": 13000
789
- },
790
- {
791
- "epoch": 1.5,
792
- "learning_rate": 8.208383844828686e-06,
793
- "loss": 0.5669,
794
- "step": 13100
795
- },
796
- {
797
- "epoch": 1.52,
798
- "learning_rate": 7.889363874178523e-06,
799
- "loss": 0.6175,
800
- "step": 13200
801
- },
802
- {
803
- "epoch": 1.53,
804
- "learning_rate": 7.570343903528361e-06,
805
- "loss": 0.6069,
806
- "step": 13300
807
- },
808
- {
809
- "epoch": 1.54,
810
- "learning_rate": 7.251323932878198e-06,
811
- "loss": 0.5949,
812
- "step": 13400
813
- },
814
- {
815
- "epoch": 1.55,
816
- "learning_rate": 6.932303962228036e-06,
817
- "loss": 0.6207,
818
- "step": 13500
819
- },
820
- {
821
- "epoch": 1.56,
822
- "learning_rate": 6.613283991577873e-06,
823
- "loss": 0.596,
824
- "step": 13600
825
- },
826
- {
827
- "epoch": 1.57,
828
- "learning_rate": 6.294264020927711e-06,
829
- "loss": 0.5972,
830
- "step": 13700
831
- },
832
- {
833
- "epoch": 1.58,
834
- "learning_rate": 5.975244050277548e-06,
835
- "loss": 0.6127,
836
- "step": 13800
837
- },
838
- {
839
- "epoch": 1.6,
840
- "learning_rate": 5.6562240796273845e-06,
841
- "loss": 0.6214,
842
- "step": 13900
843
- },
844
- {
845
- "epoch": 1.61,
846
- "learning_rate": 5.337204108977222e-06,
847
- "loss": 0.6009,
848
- "step": 14000
849
- },
850
- {
851
- "epoch": 1.62,
852
- "learning_rate": 5.018184138327059e-06,
853
- "loss": 0.628,
854
- "step": 14100
855
- },
856
- {
857
- "epoch": 1.63,
858
- "learning_rate": 4.699164167676897e-06,
859
- "loss": 0.6096,
860
- "step": 14200
861
- },
862
- {
863
- "epoch": 1.64,
864
- "learning_rate": 4.380144197026735e-06,
865
- "loss": 0.5396,
866
- "step": 14300
867
- },
868
- {
869
- "epoch": 1.65,
870
- "learning_rate": 4.061124226376571e-06,
871
- "loss": 0.554,
872
- "step": 14400
873
- },
874
- {
875
- "epoch": 1.67,
876
- "learning_rate": 3.7421042557264087e-06,
877
- "loss": 0.5867,
878
- "step": 14500
879
- },
880
- {
881
- "epoch": 1.68,
882
- "learning_rate": 3.423084285076246e-06,
883
- "loss": 0.584,
884
- "step": 14600
885
- },
886
- {
887
- "epoch": 1.69,
888
- "learning_rate": 3.1040643144260834e-06,
889
- "loss": 0.587,
890
- "step": 14700
891
- },
892
- {
893
- "epoch": 1.7,
894
- "learning_rate": 2.7850443437759203e-06,
895
- "loss": 0.605,
896
- "step": 14800
897
- },
898
- {
899
- "epoch": 1.71,
900
- "learning_rate": 2.4660243731257577e-06,
901
- "loss": 0.5905,
902
- "step": 14900
903
- },
904
- {
905
- "epoch": 1.72,
906
- "learning_rate": 2.147004402475595e-06,
907
- "loss": 0.6208,
908
- "step": 15000
909
- }
910
- ],
911
- "max_steps": 15673,
912
- "num_train_epochs": 2,
913
- "total_flos": 5.929193491393536e+16,
914
- "trial_name": null,
915
- "trial_params": null
916
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
last-checkpoint/training_args.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:fab8dd6db5a19553c3856081555c13cd9dabcada806f1bf66328d790e5db97f1
3
- size 3439
 
 
 
 
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7b3390a170051de185cec86c8c4131be12aeb79afd4f2697385bcd3c5fe72132
3
  size 557969145
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d559a092bb0a487233ab7cbe18198aabe17bcd9e577027d411f6d587b21a314
3
  size 557969145
runs/Jan26_03-13-45_b3489f7155a5/events.out.tfevents.1674703022.b3489f7155a5.24.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:42156792ecc5fb337e2c5b5d1ed6fdc0fe4bf653273b9d8e8c643e6dabea8931
3
- size 28385
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:606e98b595db58997a7a6e392e3068441e771425cecc7313eb162435373ebe2d
3
+ size 29681