HealthTeam commited on
Commit
6ffd7f3
1 Parent(s): ea0d69b

End of training

Browse files
last-checkpoint/generation_config.json → generation_config.json RENAMED
File without changes
last-checkpoint/added_tokens.json DELETED
@@ -1,6 +0,0 @@
1
- {
2
- "translate Arabic to English:": 250101,
3
- "translate Arabic to Thai:": 250100,
4
- "translate Thai to Arabic:": 250102,
5
- "translate Thai to English:": 250103
6
- }
 
 
 
 
 
 
 
last-checkpoint/config.json DELETED
@@ -1,32 +0,0 @@
1
- {
2
- "_name_or_path": "Shularp/mt5-small-finetuned-MultiHead-230207-finetuned-MultiHead-230210",
3
- "architectures": [
4
- "MT5ForConditionalGeneration"
5
- ],
6
- "d_ff": 1024,
7
- "d_kv": 64,
8
- "d_model": 512,
9
- "decoder_start_token_id": 0,
10
- "dense_act_fn": "gelu_new",
11
- "dropout_rate": 0.1,
12
- "eos_token_id": 1,
13
- "feed_forward_proj": "gated-gelu",
14
- "initializer_factor": 1.0,
15
- "is_encoder_decoder": true,
16
- "is_gated_act": true,
17
- "layer_norm_epsilon": 1e-06,
18
- "max_length": 1024,
19
- "model_type": "mt5",
20
- "num_decoder_layers": 8,
21
- "num_heads": 6,
22
- "num_layers": 8,
23
- "pad_token_id": 0,
24
- "relative_attention_max_distance": 128,
25
- "relative_attention_num_buckets": 32,
26
- "tie_word_embeddings": false,
27
- "tokenizer_class": "T5Tokenizer",
28
- "torch_dtype": "float32",
29
- "transformers_version": "4.26.1",
30
- "use_cache": true,
31
- "vocab_size": 250104
32
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
last-checkpoint/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:e99160596deaf1a941f401564fa9f387da61ce4fcbd3c8be41a5796b4d9c8fcb
3
- size 2401461637
 
 
 
 
last-checkpoint/pytorch_model.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:95a868f38e06c1d0679daaf4d5510e7ff02a3e67b90ee3a4b468a09643d77131
3
- size 1200739717
 
 
 
 
last-checkpoint/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:049a8b27323a9ab88596ec47eb567e57317d171edf3ddee0d967d13cf176058a
3
- size 14575
 
 
 
 
last-checkpoint/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:0f57334800401c750c6751652d2fefd75968842550a273b6662abb88233eab71
3
- size 627
 
 
 
 
last-checkpoint/special_tokens_map.json DELETED
@@ -1,11 +0,0 @@
1
- {
2
- "additional_special_tokens": [
3
- "translate Arabic to Thai:",
4
- "translate Arabic to English:",
5
- "translate Thai to Arabic:",
6
- "translate Thai to English:"
7
- ],
8
- "eos_token": "</s>",
9
- "pad_token": "<pad>",
10
- "unk_token": "<unk>"
11
- }
 
 
 
 
 
 
 
 
 
 
 
 
last-checkpoint/spiece.model DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:ef78f86560d809067d12bac6c09f19a462cb3af3f54d2b8acbba26e1433125d6
3
- size 4309802
 
 
 
 
last-checkpoint/tokenizer_config.json DELETED
@@ -1,12 +0,0 @@
1
- {
2
- "additional_special_tokens": null,
3
- "eos_token": "</s>",
4
- "extra_ids": 0,
5
- "model_max_length": 1000000000000000019884624838656,
6
- "name_or_path": "Shularp/mt5-small-finetuned-MultiHead-230207-finetuned-MultiHead-230210",
7
- "pad_token": "<pad>",
8
- "sp_model_kwargs": {},
9
- "special_tokens_map_file": "/root/.cache/huggingface/hub/models--google--mt5-small/snapshots/38f23af8ec210eb6c376d40e9c56bd25a80f195d/special_tokens_map.json",
10
- "tokenizer_class": "T5Tokenizer",
11
- "unk_token": "<unk>"
12
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
last-checkpoint/trainer_state.json DELETED
@@ -1,1222 +0,0 @@
1
- {
2
- "best_metric": null,
3
- "best_model_checkpoint": null,
4
- "epoch": 2.9583172175775787,
5
- "global_step": 99432,
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": 1.9900826118433453e-05,
13
- "loss": 2.5088,
14
- "step": 500
15
- },
16
- {
17
- "epoch": 0.03,
18
- "learning_rate": 1.9801652236866898e-05,
19
- "loss": 2.4955,
20
- "step": 1000
21
- },
22
- {
23
- "epoch": 0.04,
24
- "learning_rate": 1.970247835530035e-05,
25
- "loss": 2.5035,
26
- "step": 1500
27
- },
28
- {
29
- "epoch": 0.06,
30
- "learning_rate": 1.96033044737338e-05,
31
- "loss": 2.4969,
32
- "step": 2000
33
- },
34
- {
35
- "epoch": 0.07,
36
- "learning_rate": 1.9504130592167246e-05,
37
- "loss": 2.4963,
38
- "step": 2500
39
- },
40
- {
41
- "epoch": 0.09,
42
- "learning_rate": 1.9404956710600698e-05,
43
- "loss": 2.5023,
44
- "step": 3000
45
- },
46
- {
47
- "epoch": 0.1,
48
- "learning_rate": 1.930578282903415e-05,
49
- "loss": 2.5007,
50
- "step": 3500
51
- },
52
- {
53
- "epoch": 0.12,
54
- "learning_rate": 1.9206608947467594e-05,
55
- "loss": 2.4953,
56
- "step": 4000
57
- },
58
- {
59
- "epoch": 0.13,
60
- "learning_rate": 1.9107435065901046e-05,
61
- "loss": 2.499,
62
- "step": 4500
63
- },
64
- {
65
- "epoch": 0.15,
66
- "learning_rate": 1.9008261184334497e-05,
67
- "loss": 2.4919,
68
- "step": 5000
69
- },
70
- {
71
- "epoch": 0.16,
72
- "learning_rate": 1.8909087302767945e-05,
73
- "loss": 2.4988,
74
- "step": 5500
75
- },
76
- {
77
- "epoch": 0.18,
78
- "learning_rate": 1.8809913421201393e-05,
79
- "loss": 2.5076,
80
- "step": 6000
81
- },
82
- {
83
- "epoch": 0.19,
84
- "learning_rate": 1.8710739539634845e-05,
85
- "loss": 2.4915,
86
- "step": 6500
87
- },
88
- {
89
- "epoch": 0.21,
90
- "learning_rate": 1.8611565658068293e-05,
91
- "loss": 2.4823,
92
- "step": 7000
93
- },
94
- {
95
- "epoch": 0.22,
96
- "learning_rate": 1.851239177650174e-05,
97
- "loss": 2.4859,
98
- "step": 7500
99
- },
100
- {
101
- "epoch": 0.24,
102
- "learning_rate": 1.8413217894935193e-05,
103
- "loss": 2.4951,
104
- "step": 8000
105
- },
106
- {
107
- "epoch": 0.25,
108
- "learning_rate": 1.831404401336864e-05,
109
- "loss": 2.4816,
110
- "step": 8500
111
- },
112
- {
113
- "epoch": 0.27,
114
- "learning_rate": 1.821487013180209e-05,
115
- "loss": 2.485,
116
- "step": 9000
117
- },
118
- {
119
- "epoch": 0.28,
120
- "learning_rate": 1.811569625023554e-05,
121
- "loss": 2.4889,
122
- "step": 9500
123
- },
124
- {
125
- "epoch": 0.3,
126
- "learning_rate": 1.801652236866899e-05,
127
- "loss": 2.4812,
128
- "step": 10000
129
- },
130
- {
131
- "epoch": 0.31,
132
- "learning_rate": 1.7917348487102437e-05,
133
- "loss": 2.4774,
134
- "step": 10500
135
- },
136
- {
137
- "epoch": 0.33,
138
- "learning_rate": 1.781817460553589e-05,
139
- "loss": 2.4796,
140
- "step": 11000
141
- },
142
- {
143
- "epoch": 0.34,
144
- "learning_rate": 1.7719000723969337e-05,
145
- "loss": 2.4706,
146
- "step": 11500
147
- },
148
- {
149
- "epoch": 0.36,
150
- "learning_rate": 1.7619826842402785e-05,
151
- "loss": 2.4703,
152
- "step": 12000
153
- },
154
- {
155
- "epoch": 0.37,
156
- "learning_rate": 1.7520652960836234e-05,
157
- "loss": 2.4717,
158
- "step": 12500
159
- },
160
- {
161
- "epoch": 0.39,
162
- "learning_rate": 1.7421479079269685e-05,
163
- "loss": 2.4818,
164
- "step": 13000
165
- },
166
- {
167
- "epoch": 0.4,
168
- "learning_rate": 1.7322305197703137e-05,
169
- "loss": 2.4719,
170
- "step": 13500
171
- },
172
- {
173
- "epoch": 0.42,
174
- "learning_rate": 1.722313131613658e-05,
175
- "loss": 2.4671,
176
- "step": 14000
177
- },
178
- {
179
- "epoch": 0.43,
180
- "learning_rate": 1.7123957434570033e-05,
181
- "loss": 2.4853,
182
- "step": 14500
183
- },
184
- {
185
- "epoch": 0.45,
186
- "learning_rate": 1.7024783553003485e-05,
187
- "loss": 2.4823,
188
- "step": 15000
189
- },
190
- {
191
- "epoch": 0.46,
192
- "learning_rate": 1.692560967143693e-05,
193
- "loss": 2.4886,
194
- "step": 15500
195
- },
196
- {
197
- "epoch": 0.48,
198
- "learning_rate": 1.682643578987038e-05,
199
- "loss": 2.4839,
200
- "step": 16000
201
- },
202
- {
203
- "epoch": 0.49,
204
- "learning_rate": 1.6727261908303833e-05,
205
- "loss": 2.4844,
206
- "step": 16500
207
- },
208
- {
209
- "epoch": 0.51,
210
- "learning_rate": 1.6628088026737277e-05,
211
- "loss": 2.4591,
212
- "step": 17000
213
- },
214
- {
215
- "epoch": 0.52,
216
- "learning_rate": 1.652891414517073e-05,
217
- "loss": 2.4776,
218
- "step": 17500
219
- },
220
- {
221
- "epoch": 0.54,
222
- "learning_rate": 1.642974026360418e-05,
223
- "loss": 2.472,
224
- "step": 18000
225
- },
226
- {
227
- "epoch": 0.55,
228
- "learning_rate": 1.6330566382037625e-05,
229
- "loss": 2.4665,
230
- "step": 18500
231
- },
232
- {
233
- "epoch": 0.57,
234
- "learning_rate": 1.6231392500471077e-05,
235
- "loss": 2.4709,
236
- "step": 19000
237
- },
238
- {
239
- "epoch": 0.58,
240
- "learning_rate": 1.613221861890453e-05,
241
- "loss": 2.4676,
242
- "step": 19500
243
- },
244
- {
245
- "epoch": 0.6,
246
- "learning_rate": 1.6033044737337973e-05,
247
- "loss": 2.472,
248
- "step": 20000
249
- },
250
- {
251
- "epoch": 0.61,
252
- "learning_rate": 1.5933870855771425e-05,
253
- "loss": 2.4703,
254
- "step": 20500
255
- },
256
- {
257
- "epoch": 0.62,
258
- "learning_rate": 1.5834696974204877e-05,
259
- "loss": 2.4775,
260
- "step": 21000
261
- },
262
- {
263
- "epoch": 0.64,
264
- "learning_rate": 1.5735523092638325e-05,
265
- "loss": 2.4544,
266
- "step": 21500
267
- },
268
- {
269
- "epoch": 0.65,
270
- "learning_rate": 1.5636349211071773e-05,
271
- "loss": 2.4677,
272
- "step": 22000
273
- },
274
- {
275
- "epoch": 0.67,
276
- "learning_rate": 1.5537175329505225e-05,
277
- "loss": 2.4533,
278
- "step": 22500
279
- },
280
- {
281
- "epoch": 0.68,
282
- "learning_rate": 1.5438001447938673e-05,
283
- "loss": 2.4545,
284
- "step": 23000
285
- },
286
- {
287
- "epoch": 0.7,
288
- "learning_rate": 1.533882756637212e-05,
289
- "loss": 2.472,
290
- "step": 23500
291
- },
292
- {
293
- "epoch": 0.71,
294
- "learning_rate": 1.523965368480557e-05,
295
- "loss": 2.471,
296
- "step": 24000
297
- },
298
- {
299
- "epoch": 0.73,
300
- "learning_rate": 1.514047980323902e-05,
301
- "loss": 2.4615,
302
- "step": 24500
303
- },
304
- {
305
- "epoch": 0.74,
306
- "learning_rate": 1.5041305921672469e-05,
307
- "loss": 2.4607,
308
- "step": 25000
309
- },
310
- {
311
- "epoch": 0.76,
312
- "learning_rate": 1.4942132040105919e-05,
313
- "loss": 2.4587,
314
- "step": 25500
315
- },
316
- {
317
- "epoch": 0.77,
318
- "learning_rate": 1.4842958158539369e-05,
319
- "loss": 2.4554,
320
- "step": 26000
321
- },
322
- {
323
- "epoch": 0.79,
324
- "learning_rate": 1.4743784276972817e-05,
325
- "loss": 2.4506,
326
- "step": 26500
327
- },
328
- {
329
- "epoch": 0.8,
330
- "learning_rate": 1.4644610395406267e-05,
331
- "loss": 2.4608,
332
- "step": 27000
333
- },
334
- {
335
- "epoch": 0.82,
336
- "learning_rate": 1.4545436513839717e-05,
337
- "loss": 2.4417,
338
- "step": 27500
339
- },
340
- {
341
- "epoch": 0.83,
342
- "learning_rate": 1.4446262632273165e-05,
343
- "loss": 2.446,
344
- "step": 28000
345
- },
346
- {
347
- "epoch": 0.85,
348
- "learning_rate": 1.4347088750706615e-05,
349
- "loss": 2.445,
350
- "step": 28500
351
- },
352
- {
353
- "epoch": 0.86,
354
- "learning_rate": 1.4247914869140065e-05,
355
- "loss": 2.4568,
356
- "step": 29000
357
- },
358
- {
359
- "epoch": 0.88,
360
- "learning_rate": 1.4148740987573514e-05,
361
- "loss": 2.4488,
362
- "step": 29500
363
- },
364
- {
365
- "epoch": 0.89,
366
- "learning_rate": 1.4049567106006963e-05,
367
- "loss": 2.428,
368
- "step": 30000
369
- },
370
- {
371
- "epoch": 0.91,
372
- "learning_rate": 1.3950393224440413e-05,
373
- "loss": 2.4268,
374
- "step": 30500
375
- },
376
- {
377
- "epoch": 0.92,
378
- "learning_rate": 1.3851219342873862e-05,
379
- "loss": 2.4473,
380
- "step": 31000
381
- },
382
- {
383
- "epoch": 0.94,
384
- "learning_rate": 1.375204546130731e-05,
385
- "loss": 2.4441,
386
- "step": 31500
387
- },
388
- {
389
- "epoch": 0.95,
390
- "learning_rate": 1.365287157974076e-05,
391
- "loss": 2.4408,
392
- "step": 32000
393
- },
394
- {
395
- "epoch": 0.97,
396
- "learning_rate": 1.355369769817421e-05,
397
- "loss": 2.4462,
398
- "step": 32500
399
- },
400
- {
401
- "epoch": 0.98,
402
- "learning_rate": 1.3454523816607659e-05,
403
- "loss": 2.4452,
404
- "step": 33000
405
- },
406
- {
407
- "epoch": 1.0,
408
- "learning_rate": 1.3355349935041108e-05,
409
- "loss": 2.4438,
410
- "step": 33500
411
- },
412
- {
413
- "epoch": 1.0,
414
- "eval_bleu": 16.655484084216944,
415
- "eval_loss": 2.04453706741333,
416
- "eval_runtime": 4558.2525,
417
- "eval_samples_per_second": 8.87,
418
- "eval_steps_per_second": 0.277,
419
- "step": 33611
420
- },
421
- {
422
- "epoch": 1.01,
423
- "learning_rate": 1.3256176053474558e-05,
424
- "loss": 2.44,
425
- "step": 34000
426
- },
427
- {
428
- "epoch": 1.03,
429
- "learning_rate": 1.3157002171908007e-05,
430
- "loss": 2.4188,
431
- "step": 34500
432
- },
433
- {
434
- "epoch": 1.04,
435
- "learning_rate": 1.3057828290341456e-05,
436
- "loss": 2.4264,
437
- "step": 35000
438
- },
439
- {
440
- "epoch": 1.06,
441
- "learning_rate": 1.2958654408774906e-05,
442
- "loss": 2.4337,
443
- "step": 35500
444
- },
445
- {
446
- "epoch": 1.07,
447
- "learning_rate": 1.2859480527208354e-05,
448
- "loss": 2.433,
449
- "step": 36000
450
- },
451
- {
452
- "epoch": 1.09,
453
- "learning_rate": 1.2760306645641804e-05,
454
- "loss": 2.4277,
455
- "step": 36500
456
- },
457
- {
458
- "epoch": 1.1,
459
- "learning_rate": 1.2661132764075254e-05,
460
- "loss": 2.43,
461
- "step": 37000
462
- },
463
- {
464
- "epoch": 1.12,
465
- "learning_rate": 1.2561958882508702e-05,
466
- "loss": 2.4343,
467
- "step": 37500
468
- },
469
- {
470
- "epoch": 1.13,
471
- "learning_rate": 1.2462785000942152e-05,
472
- "loss": 2.4278,
473
- "step": 38000
474
- },
475
- {
476
- "epoch": 1.15,
477
- "learning_rate": 1.2363611119375602e-05,
478
- "loss": 2.4462,
479
- "step": 38500
480
- },
481
- {
482
- "epoch": 1.16,
483
- "learning_rate": 1.2264437237809052e-05,
484
- "loss": 2.4115,
485
- "step": 39000
486
- },
487
- {
488
- "epoch": 1.18,
489
- "learning_rate": 1.21652633562425e-05,
490
- "loss": 2.426,
491
- "step": 39500
492
- },
493
- {
494
- "epoch": 1.19,
495
- "learning_rate": 1.206608947467595e-05,
496
- "loss": 2.4384,
497
- "step": 40000
498
- },
499
- {
500
- "epoch": 1.2,
501
- "learning_rate": 1.19669155931094e-05,
502
- "loss": 2.418,
503
- "step": 40500
504
- },
505
- {
506
- "epoch": 1.22,
507
- "learning_rate": 1.1867741711542848e-05,
508
- "loss": 2.4409,
509
- "step": 41000
510
- },
511
- {
512
- "epoch": 1.23,
513
- "learning_rate": 1.1768567829976298e-05,
514
- "loss": 2.4147,
515
- "step": 41500
516
- },
517
- {
518
- "epoch": 1.25,
519
- "learning_rate": 1.1669393948409748e-05,
520
- "loss": 2.4325,
521
- "step": 42000
522
- },
523
- {
524
- "epoch": 1.26,
525
- "learning_rate": 1.1570220066843196e-05,
526
- "loss": 2.4317,
527
- "step": 42500
528
- },
529
- {
530
- "epoch": 1.28,
531
- "learning_rate": 1.1471046185276646e-05,
532
- "loss": 2.4253,
533
- "step": 43000
534
- },
535
- {
536
- "epoch": 1.29,
537
- "learning_rate": 1.1371872303710096e-05,
538
- "loss": 2.4249,
539
- "step": 43500
540
- },
541
- {
542
- "epoch": 1.31,
543
- "learning_rate": 1.1272698422143544e-05,
544
- "loss": 2.4322,
545
- "step": 44000
546
- },
547
- {
548
- "epoch": 1.32,
549
- "learning_rate": 1.1173524540576994e-05,
550
- "loss": 2.4176,
551
- "step": 44500
552
- },
553
- {
554
- "epoch": 1.34,
555
- "learning_rate": 1.1074350659010444e-05,
556
- "loss": 2.4333,
557
- "step": 45000
558
- },
559
- {
560
- "epoch": 1.35,
561
- "learning_rate": 1.0975176777443892e-05,
562
- "loss": 2.4222,
563
- "step": 45500
564
- },
565
- {
566
- "epoch": 1.37,
567
- "learning_rate": 1.0876002895877342e-05,
568
- "loss": 2.4313,
569
- "step": 46000
570
- },
571
- {
572
- "epoch": 1.38,
573
- "learning_rate": 1.0776829014310792e-05,
574
- "loss": 2.4179,
575
- "step": 46500
576
- },
577
- {
578
- "epoch": 1.4,
579
- "learning_rate": 1.0677655132744242e-05,
580
- "loss": 2.4222,
581
- "step": 47000
582
- },
583
- {
584
- "epoch": 1.41,
585
- "learning_rate": 1.057848125117769e-05,
586
- "loss": 2.4182,
587
- "step": 47500
588
- },
589
- {
590
- "epoch": 1.43,
591
- "learning_rate": 1.047930736961114e-05,
592
- "loss": 2.4215,
593
- "step": 48000
594
- },
595
- {
596
- "epoch": 1.44,
597
- "learning_rate": 1.038013348804459e-05,
598
- "loss": 2.4029,
599
- "step": 48500
600
- },
601
- {
602
- "epoch": 1.46,
603
- "learning_rate": 1.0280959606478038e-05,
604
- "loss": 2.4089,
605
- "step": 49000
606
- },
607
- {
608
- "epoch": 1.47,
609
- "learning_rate": 1.0181785724911488e-05,
610
- "loss": 2.4046,
611
- "step": 49500
612
- },
613
- {
614
- "epoch": 1.49,
615
- "learning_rate": 1.0082611843344938e-05,
616
- "loss": 2.4158,
617
- "step": 50000
618
- },
619
- {
620
- "epoch": 1.5,
621
- "learning_rate": 9.983437961778388e-06,
622
- "loss": 2.4139,
623
- "step": 50500
624
- },
625
- {
626
- "epoch": 1.52,
627
- "learning_rate": 9.884264080211836e-06,
628
- "loss": 2.4127,
629
- "step": 51000
630
- },
631
- {
632
- "epoch": 1.53,
633
- "learning_rate": 9.785090198645286e-06,
634
- "loss": 2.4178,
635
- "step": 51500
636
- },
637
- {
638
- "epoch": 1.55,
639
- "learning_rate": 9.685916317078736e-06,
640
- "loss": 2.4217,
641
- "step": 52000
642
- },
643
- {
644
- "epoch": 1.56,
645
- "learning_rate": 9.586742435512184e-06,
646
- "loss": 2.4053,
647
- "step": 52500
648
- },
649
- {
650
- "epoch": 1.58,
651
- "learning_rate": 9.487568553945634e-06,
652
- "loss": 2.3942,
653
- "step": 53000
654
- },
655
- {
656
- "epoch": 1.59,
657
- "learning_rate": 9.388394672379084e-06,
658
- "loss": 2.4107,
659
- "step": 53500
660
- },
661
- {
662
- "epoch": 1.61,
663
- "learning_rate": 9.289220790812532e-06,
664
- "loss": 2.4023,
665
- "step": 54000
666
- },
667
- {
668
- "epoch": 1.62,
669
- "learning_rate": 9.190046909245982e-06,
670
- "loss": 2.4252,
671
- "step": 54500
672
- },
673
- {
674
- "epoch": 1.64,
675
- "learning_rate": 9.090873027679432e-06,
676
- "loss": 2.4244,
677
- "step": 55000
678
- },
679
- {
680
- "epoch": 1.65,
681
- "learning_rate": 8.99169914611288e-06,
682
- "loss": 2.42,
683
- "step": 55500
684
- },
685
- {
686
- "epoch": 1.67,
687
- "learning_rate": 8.89252526454633e-06,
688
- "loss": 2.4224,
689
- "step": 56000
690
- },
691
- {
692
- "epoch": 1.68,
693
- "learning_rate": 8.79335138297978e-06,
694
- "loss": 2.4077,
695
- "step": 56500
696
- },
697
- {
698
- "epoch": 1.7,
699
- "learning_rate": 8.69417750141323e-06,
700
- "loss": 2.4192,
701
- "step": 57000
702
- },
703
- {
704
- "epoch": 1.71,
705
- "learning_rate": 8.595003619846678e-06,
706
- "loss": 2.4083,
707
- "step": 57500
708
- },
709
- {
710
- "epoch": 1.73,
711
- "learning_rate": 8.495829738280127e-06,
712
- "loss": 2.4172,
713
- "step": 58000
714
- },
715
- {
716
- "epoch": 1.74,
717
- "learning_rate": 8.396655856713577e-06,
718
- "loss": 2.4149,
719
- "step": 58500
720
- },
721
- {
722
- "epoch": 1.76,
723
- "learning_rate": 8.297481975147026e-06,
724
- "loss": 2.4268,
725
- "step": 59000
726
- },
727
- {
728
- "epoch": 1.77,
729
- "learning_rate": 8.198308093580475e-06,
730
- "loss": 2.4092,
731
- "step": 59500
732
- },
733
- {
734
- "epoch": 1.79,
735
- "learning_rate": 8.099134212013925e-06,
736
- "loss": 2.4079,
737
- "step": 60000
738
- },
739
- {
740
- "epoch": 1.8,
741
- "learning_rate": 7.999960330447374e-06,
742
- "loss": 2.4135,
743
- "step": 60500
744
- },
745
- {
746
- "epoch": 1.81,
747
- "learning_rate": 7.900786448880823e-06,
748
- "loss": 2.3943,
749
- "step": 61000
750
- },
751
- {
752
- "epoch": 1.83,
753
- "learning_rate": 7.801612567314273e-06,
754
- "loss": 2.4081,
755
- "step": 61500
756
- },
757
- {
758
- "epoch": 1.84,
759
- "learning_rate": 7.702438685747721e-06,
760
- "loss": 2.4058,
761
- "step": 62000
762
- },
763
- {
764
- "epoch": 1.86,
765
- "learning_rate": 7.603264804181172e-06,
766
- "loss": 2.4079,
767
- "step": 62500
768
- },
769
- {
770
- "epoch": 1.87,
771
- "learning_rate": 7.50409092261462e-06,
772
- "loss": 2.4085,
773
- "step": 63000
774
- },
775
- {
776
- "epoch": 1.89,
777
- "learning_rate": 7.4049170410480695e-06,
778
- "loss": 2.4206,
779
- "step": 63500
780
- },
781
- {
782
- "epoch": 1.9,
783
- "learning_rate": 7.30574315948152e-06,
784
- "loss": 2.4053,
785
- "step": 64000
786
- },
787
- {
788
- "epoch": 1.92,
789
- "learning_rate": 7.206569277914968e-06,
790
- "loss": 2.4024,
791
- "step": 64500
792
- },
793
- {
794
- "epoch": 1.93,
795
- "learning_rate": 7.107395396348419e-06,
796
- "loss": 2.4064,
797
- "step": 65000
798
- },
799
- {
800
- "epoch": 1.95,
801
- "learning_rate": 7.008221514781868e-06,
802
- "loss": 2.4093,
803
- "step": 65500
804
- },
805
- {
806
- "epoch": 1.96,
807
- "learning_rate": 6.909047633215316e-06,
808
- "loss": 2.3942,
809
- "step": 66000
810
- },
811
- {
812
- "epoch": 1.98,
813
- "learning_rate": 6.809873751648767e-06,
814
- "loss": 2.4201,
815
- "step": 66500
816
- },
817
- {
818
- "epoch": 1.99,
819
- "learning_rate": 6.710699870082215e-06,
820
- "loss": 2.3878,
821
- "step": 67000
822
- },
823
- {
824
- "epoch": 2.0,
825
- "eval_bleu": 17.38844508561531,
826
- "eval_loss": 2.0142550468444824,
827
- "eval_runtime": 4302.3863,
828
- "eval_samples_per_second": 9.398,
829
- "eval_steps_per_second": 0.294,
830
- "step": 67222
831
- },
832
- {
833
- "epoch": 2.01,
834
- "learning_rate": 6.611525988515664e-06,
835
- "loss": 2.4155,
836
- "step": 67500
837
- },
838
- {
839
- "epoch": 2.02,
840
- "learning_rate": 6.512352106949115e-06,
841
- "loss": 2.3969,
842
- "step": 68000
843
- },
844
- {
845
- "epoch": 2.04,
846
- "learning_rate": 6.413178225382563e-06,
847
- "loss": 2.3964,
848
- "step": 68500
849
- },
850
- {
851
- "epoch": 2.05,
852
- "learning_rate": 6.314004343816014e-06,
853
- "loss": 2.3973,
854
- "step": 69000
855
- },
856
- {
857
- "epoch": 2.07,
858
- "learning_rate": 6.214830462249463e-06,
859
- "loss": 2.3925,
860
- "step": 69500
861
- },
862
- {
863
- "epoch": 2.08,
864
- "learning_rate": 6.115656580682911e-06,
865
- "loss": 2.3968,
866
- "step": 70000
867
- },
868
- {
869
- "epoch": 2.1,
870
- "learning_rate": 6.016482699116362e-06,
871
- "loss": 2.3967,
872
- "step": 70500
873
- },
874
- {
875
- "epoch": 2.11,
876
- "learning_rate": 5.917308817549811e-06,
877
- "loss": 2.3827,
878
- "step": 71000
879
- },
880
- {
881
- "epoch": 2.13,
882
- "learning_rate": 5.818134935983259e-06,
883
- "loss": 2.4147,
884
- "step": 71500
885
- },
886
- {
887
- "epoch": 2.14,
888
- "learning_rate": 5.71896105441671e-06,
889
- "loss": 2.4108,
890
- "step": 72000
891
- },
892
- {
893
- "epoch": 2.16,
894
- "learning_rate": 5.619787172850158e-06,
895
- "loss": 2.4031,
896
- "step": 72500
897
- },
898
- {
899
- "epoch": 2.17,
900
- "learning_rate": 5.520613291283607e-06,
901
- "loss": 2.3952,
902
- "step": 73000
903
- },
904
- {
905
- "epoch": 2.19,
906
- "learning_rate": 5.421439409717058e-06,
907
- "loss": 2.4134,
908
- "step": 73500
909
- },
910
- {
911
- "epoch": 2.2,
912
- "learning_rate": 5.322265528150506e-06,
913
- "loss": 2.3976,
914
- "step": 74000
915
- },
916
- {
917
- "epoch": 2.22,
918
- "learning_rate": 5.223091646583957e-06,
919
- "loss": 2.3929,
920
- "step": 74500
921
- },
922
- {
923
- "epoch": 2.23,
924
- "learning_rate": 5.123917765017406e-06,
925
- "loss": 2.3933,
926
- "step": 75000
927
- },
928
- {
929
- "epoch": 2.25,
930
- "learning_rate": 5.024743883450854e-06,
931
- "loss": 2.3927,
932
- "step": 75500
933
- },
934
- {
935
- "epoch": 2.26,
936
- "learning_rate": 4.925570001884304e-06,
937
- "loss": 2.3998,
938
- "step": 76000
939
- },
940
- {
941
- "epoch": 2.28,
942
- "learning_rate": 4.826396120317754e-06,
943
- "loss": 2.3989,
944
- "step": 76500
945
- },
946
- {
947
- "epoch": 2.29,
948
- "learning_rate": 4.727222238751203e-06,
949
- "loss": 2.4074,
950
- "step": 77000
951
- },
952
- {
953
- "epoch": 2.31,
954
- "learning_rate": 4.628048357184653e-06,
955
- "loss": 2.3954,
956
- "step": 77500
957
- },
958
- {
959
- "epoch": 2.32,
960
- "learning_rate": 4.528874475618102e-06,
961
- "loss": 2.4112,
962
- "step": 78000
963
- },
964
- {
965
- "epoch": 2.34,
966
- "learning_rate": 4.429700594051551e-06,
967
- "loss": 2.3899,
968
- "step": 78500
969
- },
970
- {
971
- "epoch": 2.35,
972
- "learning_rate": 4.330526712485001e-06,
973
- "loss": 2.3889,
974
- "step": 79000
975
- },
976
- {
977
- "epoch": 2.37,
978
- "learning_rate": 4.23135283091845e-06,
979
- "loss": 2.3987,
980
- "step": 79500
981
- },
982
- {
983
- "epoch": 2.38,
984
- "learning_rate": 4.132178949351899e-06,
985
- "loss": 2.3962,
986
- "step": 80000
987
- },
988
- {
989
- "epoch": 2.4,
990
- "learning_rate": 4.033005067785349e-06,
991
- "loss": 2.4047,
992
- "step": 80500
993
- },
994
- {
995
- "epoch": 2.41,
996
- "learning_rate": 3.933831186218798e-06,
997
- "loss": 2.4066,
998
- "step": 81000
999
- },
1000
- {
1001
- "epoch": 2.42,
1002
- "learning_rate": 3.834657304652248e-06,
1003
- "loss": 2.4025,
1004
- "step": 81500
1005
- },
1006
- {
1007
- "epoch": 2.44,
1008
- "learning_rate": 3.735483423085696e-06,
1009
- "loss": 2.3984,
1010
- "step": 82000
1011
- },
1012
- {
1013
- "epoch": 2.45,
1014
- "learning_rate": 3.6363095415191457e-06,
1015
- "loss": 2.4053,
1016
- "step": 82500
1017
- },
1018
- {
1019
- "epoch": 2.47,
1020
- "learning_rate": 3.537135659952595e-06,
1021
- "loss": 2.3999,
1022
- "step": 83000
1023
- },
1024
- {
1025
- "epoch": 2.48,
1026
- "learning_rate": 3.4379617783860446e-06,
1027
- "loss": 2.399,
1028
- "step": 83500
1029
- },
1030
- {
1031
- "epoch": 2.5,
1032
- "learning_rate": 3.3387878968194936e-06,
1033
- "loss": 2.398,
1034
- "step": 84000
1035
- },
1036
- {
1037
- "epoch": 2.51,
1038
- "learning_rate": 3.239614015252943e-06,
1039
- "loss": 2.3872,
1040
- "step": 84500
1041
- },
1042
- {
1043
- "epoch": 2.53,
1044
- "learning_rate": 3.1404401336863925e-06,
1045
- "loss": 2.3963,
1046
- "step": 85000
1047
- },
1048
- {
1049
- "epoch": 2.54,
1050
- "learning_rate": 3.041266252119842e-06,
1051
- "loss": 2.3929,
1052
- "step": 85500
1053
- },
1054
- {
1055
- "epoch": 2.56,
1056
- "learning_rate": 2.942092370553291e-06,
1057
- "loss": 2.3752,
1058
- "step": 86000
1059
- },
1060
- {
1061
- "epoch": 2.57,
1062
- "learning_rate": 2.8429184889867405e-06,
1063
- "loss": 2.3801,
1064
- "step": 86500
1065
- },
1066
- {
1067
- "epoch": 2.59,
1068
- "learning_rate": 2.74374460742019e-06,
1069
- "loss": 2.3842,
1070
- "step": 87000
1071
- },
1072
- {
1073
- "epoch": 2.6,
1074
- "learning_rate": 2.6445707258536394e-06,
1075
- "loss": 2.4046,
1076
- "step": 87500
1077
- },
1078
- {
1079
- "epoch": 2.62,
1080
- "learning_rate": 2.5453968442870885e-06,
1081
- "loss": 2.3949,
1082
- "step": 88000
1083
- },
1084
- {
1085
- "epoch": 2.63,
1086
- "learning_rate": 2.446222962720538e-06,
1087
- "loss": 2.3923,
1088
- "step": 88500
1089
- },
1090
- {
1091
- "epoch": 2.65,
1092
- "learning_rate": 2.3470490811539874e-06,
1093
- "loss": 2.3844,
1094
- "step": 89000
1095
- },
1096
- {
1097
- "epoch": 2.66,
1098
- "learning_rate": 2.247875199587437e-06,
1099
- "loss": 2.3928,
1100
- "step": 89500
1101
- },
1102
- {
1103
- "epoch": 2.68,
1104
- "learning_rate": 2.1487013180208863e-06,
1105
- "loss": 2.3907,
1106
- "step": 90000
1107
- },
1108
- {
1109
- "epoch": 2.69,
1110
- "learning_rate": 2.0495274364543354e-06,
1111
- "loss": 2.3999,
1112
- "step": 90500
1113
- },
1114
- {
1115
- "epoch": 2.71,
1116
- "learning_rate": 1.950353554887785e-06,
1117
- "loss": 2.4117,
1118
- "step": 91000
1119
- },
1120
- {
1121
- "epoch": 2.72,
1122
- "learning_rate": 1.8511796733212343e-06,
1123
- "loss": 2.4146,
1124
- "step": 91500
1125
- },
1126
- {
1127
- "epoch": 2.74,
1128
- "learning_rate": 1.7520057917546838e-06,
1129
- "loss": 2.3863,
1130
- "step": 92000
1131
- },
1132
- {
1133
- "epoch": 2.75,
1134
- "learning_rate": 1.652831910188133e-06,
1135
- "loss": 2.3857,
1136
- "step": 92500
1137
- },
1138
- {
1139
- "epoch": 2.77,
1140
- "learning_rate": 1.5536580286215825e-06,
1141
- "loss": 2.4035,
1142
- "step": 93000
1143
- },
1144
- {
1145
- "epoch": 2.78,
1146
- "learning_rate": 1.4544841470550317e-06,
1147
- "loss": 2.39,
1148
- "step": 93500
1149
- },
1150
- {
1151
- "epoch": 2.8,
1152
- "learning_rate": 1.3553102654884812e-06,
1153
- "loss": 2.3929,
1154
- "step": 94000
1155
- },
1156
- {
1157
- "epoch": 2.81,
1158
- "learning_rate": 1.2561363839219304e-06,
1159
- "loss": 2.3998,
1160
- "step": 94500
1161
- },
1162
- {
1163
- "epoch": 2.83,
1164
- "learning_rate": 1.1569625023553797e-06,
1165
- "loss": 2.3968,
1166
- "step": 95000
1167
- },
1168
- {
1169
- "epoch": 2.84,
1170
- "learning_rate": 1.0577886207888292e-06,
1171
- "loss": 2.3752,
1172
- "step": 95500
1173
- },
1174
- {
1175
- "epoch": 2.86,
1176
- "learning_rate": 9.586147392222784e-07,
1177
- "loss": 2.3943,
1178
- "step": 96000
1179
- },
1180
- {
1181
- "epoch": 2.87,
1182
- "learning_rate": 8.594408576557279e-07,
1183
- "loss": 2.3826,
1184
- "step": 96500
1185
- },
1186
- {
1187
- "epoch": 2.89,
1188
- "learning_rate": 7.602669760891772e-07,
1189
- "loss": 2.395,
1190
- "step": 97000
1191
- },
1192
- {
1193
- "epoch": 2.9,
1194
- "learning_rate": 6.610930945226266e-07,
1195
- "loss": 2.403,
1196
- "step": 97500
1197
- },
1198
- {
1199
- "epoch": 2.92,
1200
- "learning_rate": 5.619192129560759e-07,
1201
- "loss": 2.4108,
1202
- "step": 98000
1203
- },
1204
- {
1205
- "epoch": 2.93,
1206
- "learning_rate": 4.6274533138952524e-07,
1207
- "loss": 2.3714,
1208
- "step": 98500
1209
- },
1210
- {
1211
- "epoch": 2.95,
1212
- "learning_rate": 3.6357144982297465e-07,
1213
- "loss": 2.4135,
1214
- "step": 99000
1215
- }
1216
- ],
1217
- "max_steps": 100833,
1218
- "num_train_epochs": 3,
1219
- "total_flos": 2.677110317041582e+17,
1220
- "trial_name": null,
1221
- "trial_params": null
1222
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
last-checkpoint/training_args.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:25d27f125b86e4c017aefda34e143225bc4d6c87bc4f71d560b204a01dacffa1
3
- size 3771
 
 
 
 
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:95a868f38e06c1d0679daaf4d5510e7ff02a3e67b90ee3a4b468a09643d77131
3
  size 1200739717
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d86fd483d214231052b2a478d73464f09487abc2fdc7b300c3237619d004e0c3
3
  size 1200739717
runs/Feb14_18-29-07_39730b194efc/events.out.tfevents.1676399391.39730b194efc.3664.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:30c1c47592e796d4ffdc458b36558eef785bbf2aa92130eff24c37516d6638f2
3
- size 36597
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c25f3603579878cff325e1a606f53b136cde2aa48d4fe512d4b45ac5c96fbf81
3
+ size 37762