TinyPixel commited on
Commit
6064792
·
1 Parent(s): 4638c64

Upload folder using huggingface_hub

Browse files
adapter_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:59d432a48d9bc716a9372fa34a1f9be98c5a566d3590de58b1e8c167486670d2
3
  size 134264397
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:728b71db6b78b273f9498bef25c21d78d54318d754e4a46bd46fee1e1f1cf745
3
  size 134264397
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1157f3b85f292b2cbc7f9535bde31800d02bf73fbb6419bcbbe9b173453c1cb0
3
  size 268514437
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a5c89fab8c9d839a29de40ce795d1bb735eafc0d436b9fd5c6e622dfb6200cce
3
  size 268514437
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8e11e097e8cfe839d01dd21bf715ab74e8ac4dfcd8f5680dd229ae8d57f91c42
3
  size 14575
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:98e26949a92355d875e793b20e3109e63a5abef66fa917564e9566db9d56b401
3
  size 14575
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0bd7b80ec246d44d8b81124b8ba36f2bf12bf8f35c030e50fad44a064bd194e8
3
  size 627
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aaf30ebf0bcd9ed5383b022bb911f7c2039ea918b316aeff6dd531cc9d7be661
3
  size 627
special_tokens_map.json CHANGED
@@ -1,10 +1,13 @@
1
  {
2
  "additional_special_tokens": [
3
- "PROMPT:",
4
- "ANSWER:"
 
 
 
5
  ],
6
- "bos_token": "<|im_start|>",
7
- "eos_token": "<|im_end|>",
8
- "pad_token": "<|endoftext|>",
9
  "unk_token": "<|endoftext|>"
10
  }
 
1
  {
2
  "additional_special_tokens": [
3
+ "<|im_end|>",
4
+ "<|im_start|>",
5
+ "<|end_of_turn|>",
6
+ "USER:",
7
+ "ASSISTANT:"
8
  ],
9
+ "bos_token": "<|endoftext|>",
10
+ "eos_token": "<|endoftext|>",
11
+ "pad_token": "[PAD]",
12
  "unk_token": "<|endoftext|>"
13
  }
tokenizer.json CHANGED
@@ -235,7 +235,7 @@
235
  },
236
  {
237
  "id": 50277,
238
- "content": "<|im_start|>",
239
  "single_word": false,
240
  "lstrip": false,
241
  "rstrip": false,
@@ -244,7 +244,7 @@
244
  },
245
  {
246
  "id": 50278,
247
- "content": "<|im_end|>",
248
  "single_word": false,
249
  "lstrip": false,
250
  "rstrip": false,
@@ -253,7 +253,7 @@
253
  },
254
  {
255
  "id": 50279,
256
- "content": "PROMPT:",
257
  "single_word": false,
258
  "lstrip": false,
259
  "rstrip": false,
@@ -262,7 +262,43 @@
262
  },
263
  {
264
  "id": 50280,
265
- "content": "ANSWER:",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
266
  "single_word": false,
267
  "lstrip": false,
268
  "rstrip": false,
 
235
  },
236
  {
237
  "id": 50277,
238
+ "content": "PROMPT:",
239
  "single_word": false,
240
  "lstrip": false,
241
  "rstrip": false,
 
244
  },
245
  {
246
  "id": 50278,
247
+ "content": "ANSWER:",
248
  "single_word": false,
249
  "lstrip": false,
250
  "rstrip": false,
 
253
  },
254
  {
255
  "id": 50279,
256
+ "content": "<|im_start|>",
257
  "single_word": false,
258
  "lstrip": false,
259
  "rstrip": false,
 
262
  },
263
  {
264
  "id": 50280,
265
+ "content": "<|im_end|>",
266
+ "single_word": false,
267
+ "lstrip": false,
268
+ "rstrip": false,
269
+ "normalized": false,
270
+ "special": true
271
+ },
272
+ {
273
+ "id": 50281,
274
+ "content": "<|end_of_turn|>",
275
+ "single_word": false,
276
+ "lstrip": false,
277
+ "rstrip": false,
278
+ "normalized": false,
279
+ "special": true
280
+ },
281
+ {
282
+ "id": 50282,
283
+ "content": "[PAD]",
284
+ "single_word": false,
285
+ "lstrip": false,
286
+ "rstrip": false,
287
+ "normalized": false,
288
+ "special": true
289
+ },
290
+ {
291
+ "id": 50283,
292
+ "content": "USER:",
293
+ "single_word": false,
294
+ "lstrip": false,
295
+ "rstrip": false,
296
+ "normalized": false,
297
+ "special": true
298
+ },
299
+ {
300
+ "id": 50284,
301
+ "content": "ASSISTANT:",
302
  "single_word": false,
303
  "lstrip": false,
304
  "rstrip": false,
trainer_state.json CHANGED
@@ -1,595 +1,211 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 2.9825242718446603,
5
  "eval_steps": 500,
6
- "global_step": 192,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "epoch": 0.03,
13
  "learning_rate": 2e-05,
14
- "loss": 2.2907,
15
  "step": 2
16
  },
17
  {
18
- "epoch": 0.06,
19
  "learning_rate": 2e-05,
20
- "loss": 2.5081,
21
  "step": 4
22
  },
23
  {
24
- "epoch": 0.09,
25
  "learning_rate": 2e-05,
26
- "loss": 2.3899,
27
  "step": 6
28
  },
29
  {
30
- "epoch": 0.12,
31
  "learning_rate": 2e-05,
32
- "loss": 2.6247,
33
  "step": 8
34
  },
35
  {
36
- "epoch": 0.16,
37
  "learning_rate": 2e-05,
38
- "loss": 2.6502,
39
  "step": 10
40
  },
41
  {
42
- "epoch": 0.19,
43
  "learning_rate": 2e-05,
44
- "loss": 2.8024,
45
  "step": 12
46
  },
47
  {
48
- "epoch": 0.22,
49
  "learning_rate": 2e-05,
50
- "loss": 2.8058,
51
  "step": 14
52
  },
53
  {
54
- "epoch": 0.25,
55
  "learning_rate": 2e-05,
56
- "loss": 3.4057,
57
  "step": 16
58
  },
59
  {
60
- "epoch": 0.28,
61
  "learning_rate": 2e-05,
62
- "loss": 2.2647,
63
  "step": 18
64
  },
65
  {
66
- "epoch": 0.31,
67
  "learning_rate": 2e-05,
68
- "loss": 2.3353,
69
  "step": 20
70
  },
71
  {
72
- "epoch": 0.34,
73
  "learning_rate": 2e-05,
74
- "loss": 2.3558,
75
  "step": 22
76
  },
77
  {
78
- "epoch": 0.37,
79
  "learning_rate": 2e-05,
80
- "loss": 2.646,
81
  "step": 24
82
  },
83
  {
84
- "epoch": 0.4,
85
  "learning_rate": 2e-05,
86
- "loss": 2.6764,
87
  "step": 26
88
  },
89
  {
90
- "epoch": 0.43,
91
  "learning_rate": 2e-05,
92
- "loss": 2.843,
93
  "step": 28
94
  },
95
  {
96
- "epoch": 0.47,
97
  "learning_rate": 2e-05,
98
- "loss": 2.8074,
99
  "step": 30
100
  },
101
  {
102
- "epoch": 0.5,
103
  "learning_rate": 2e-05,
104
- "loss": 3.5751,
105
  "step": 32
106
  },
107
  {
108
- "epoch": 0.53,
109
  "learning_rate": 2e-05,
110
- "loss": 2.2423,
111
  "step": 34
112
  },
113
  {
114
- "epoch": 0.56,
115
  "learning_rate": 2e-05,
116
- "loss": 2.4882,
117
  "step": 36
118
  },
119
  {
120
- "epoch": 0.59,
121
  "learning_rate": 2e-05,
122
- "loss": 2.5508,
123
  "step": 38
124
  },
125
  {
126
- "epoch": 0.62,
127
  "learning_rate": 2e-05,
128
- "loss": 2.6119,
129
  "step": 40
130
  },
131
  {
132
- "epoch": 0.65,
133
  "learning_rate": 2e-05,
134
- "loss": 2.7165,
135
  "step": 42
136
  },
137
  {
138
- "epoch": 0.68,
139
  "learning_rate": 2e-05,
140
- "loss": 2.6439,
141
  "step": 44
142
  },
143
  {
144
- "epoch": 0.71,
145
  "learning_rate": 2e-05,
146
- "loss": 2.6984,
147
  "step": 46
148
  },
149
  {
150
- "epoch": 0.75,
151
  "learning_rate": 2e-05,
152
- "loss": 3.1777,
153
  "step": 48
154
  },
155
  {
156
- "epoch": 0.78,
157
  "learning_rate": 2e-05,
158
- "loss": 2.3067,
159
  "step": 50
160
  },
161
  {
162
- "epoch": 0.81,
163
  "learning_rate": 2e-05,
164
- "loss": 2.41,
165
  "step": 52
166
  },
167
  {
168
- "epoch": 0.84,
169
  "learning_rate": 2e-05,
170
- "loss": 2.5112,
171
  "step": 54
172
  },
173
  {
174
- "epoch": 0.87,
175
  "learning_rate": 2e-05,
176
- "loss": 2.4753,
177
  "step": 56
178
  },
179
  {
180
- "epoch": 0.9,
181
  "learning_rate": 2e-05,
182
- "loss": 2.6149,
183
  "step": 58
184
  },
185
  {
186
- "epoch": 0.93,
187
  "learning_rate": 2e-05,
188
- "loss": 2.618,
189
  "step": 60
190
  },
191
  {
192
- "epoch": 0.96,
193
  "learning_rate": 2e-05,
194
- "loss": 2.8237,
195
  "step": 62
196
  },
197
  {
198
- "epoch": 0.99,
199
  "learning_rate": 2e-05,
200
- "loss": 3.0225,
201
  "step": 64
202
- },
203
- {
204
- "epoch": 1.03,
205
- "learning_rate": 2e-05,
206
- "loss": 2.3223,
207
- "step": 66
208
- },
209
- {
210
- "epoch": 1.06,
211
- "learning_rate": 2e-05,
212
- "loss": 2.3016,
213
- "step": 68
214
- },
215
- {
216
- "epoch": 1.09,
217
- "learning_rate": 2e-05,
218
- "loss": 2.3603,
219
- "step": 70
220
- },
221
- {
222
- "epoch": 1.12,
223
- "learning_rate": 2e-05,
224
- "loss": 2.4986,
225
- "step": 72
226
- },
227
- {
228
- "epoch": 1.15,
229
- "learning_rate": 2e-05,
230
- "loss": 2.6638,
231
- "step": 74
232
- },
233
- {
234
- "epoch": 1.18,
235
- "learning_rate": 2e-05,
236
- "loss": 2.5076,
237
- "step": 76
238
- },
239
- {
240
- "epoch": 1.21,
241
- "learning_rate": 2e-05,
242
- "loss": 2.7194,
243
- "step": 78
244
- },
245
- {
246
- "epoch": 1.24,
247
- "learning_rate": 2e-05,
248
- "loss": 2.7815,
249
- "step": 80
250
- },
251
- {
252
- "epoch": 1.27,
253
- "learning_rate": 2e-05,
254
- "loss": 2.385,
255
- "step": 82
256
- },
257
- {
258
- "epoch": 1.3,
259
- "learning_rate": 2e-05,
260
- "loss": 2.2657,
261
- "step": 84
262
- },
263
- {
264
- "epoch": 1.34,
265
- "learning_rate": 2e-05,
266
- "loss": 2.4204,
267
- "step": 86
268
- },
269
- {
270
- "epoch": 1.37,
271
- "learning_rate": 2e-05,
272
- "loss": 2.4338,
273
- "step": 88
274
- },
275
- {
276
- "epoch": 1.4,
277
- "learning_rate": 2e-05,
278
- "loss": 2.5382,
279
- "step": 90
280
- },
281
- {
282
- "epoch": 1.43,
283
- "learning_rate": 2e-05,
284
- "loss": 2.5632,
285
- "step": 92
286
- },
287
- {
288
- "epoch": 1.46,
289
- "learning_rate": 2e-05,
290
- "loss": 2.5847,
291
- "step": 94
292
- },
293
- {
294
- "epoch": 1.49,
295
- "learning_rate": 2e-05,
296
- "loss": 2.8775,
297
- "step": 96
298
- },
299
- {
300
- "epoch": 1.52,
301
- "learning_rate": 2e-05,
302
- "loss": 2.5724,
303
- "step": 98
304
- },
305
- {
306
- "epoch": 1.55,
307
- "learning_rate": 2e-05,
308
- "loss": 2.2229,
309
- "step": 100
310
- },
311
- {
312
- "epoch": 1.58,
313
- "learning_rate": 2e-05,
314
- "loss": 2.409,
315
- "step": 102
316
- },
317
- {
318
- "epoch": 1.62,
319
- "learning_rate": 2e-05,
320
- "loss": 2.4577,
321
- "step": 104
322
- },
323
- {
324
- "epoch": 1.65,
325
- "learning_rate": 2e-05,
326
- "loss": 2.5779,
327
- "step": 106
328
- },
329
- {
330
- "epoch": 1.68,
331
- "learning_rate": 2e-05,
332
- "loss": 2.6785,
333
- "step": 108
334
- },
335
- {
336
- "epoch": 1.71,
337
- "learning_rate": 2e-05,
338
- "loss": 2.6454,
339
- "step": 110
340
- },
341
- {
342
- "epoch": 1.74,
343
- "learning_rate": 2e-05,
344
- "loss": 2.7884,
345
- "step": 112
346
- },
347
- {
348
- "epoch": 1.77,
349
- "learning_rate": 2e-05,
350
- "loss": 2.523,
351
- "step": 114
352
- },
353
- {
354
- "epoch": 1.8,
355
- "learning_rate": 2e-05,
356
- "loss": 2.3859,
357
- "step": 116
358
- },
359
- {
360
- "epoch": 1.83,
361
- "learning_rate": 2e-05,
362
- "loss": 2.4611,
363
- "step": 118
364
- },
365
- {
366
- "epoch": 1.86,
367
- "learning_rate": 2e-05,
368
- "loss": 2.5436,
369
- "step": 120
370
- },
371
- {
372
- "epoch": 1.9,
373
- "learning_rate": 2e-05,
374
- "loss": 2.495,
375
- "step": 122
376
- },
377
- {
378
- "epoch": 1.93,
379
- "learning_rate": 2e-05,
380
- "loss": 2.537,
381
- "step": 124
382
- },
383
- {
384
- "epoch": 1.96,
385
- "learning_rate": 2e-05,
386
- "loss": 2.6619,
387
- "step": 126
388
- },
389
- {
390
- "epoch": 1.99,
391
- "learning_rate": 2e-05,
392
- "loss": 2.6684,
393
- "step": 128
394
- },
395
- {
396
- "epoch": 2.02,
397
- "learning_rate": 2e-05,
398
- "loss": 2.344,
399
- "step": 130
400
- },
401
- {
402
- "epoch": 2.05,
403
- "learning_rate": 2e-05,
404
- "loss": 2.3504,
405
- "step": 132
406
- },
407
- {
408
- "epoch": 2.08,
409
- "learning_rate": 2e-05,
410
- "loss": 2.4477,
411
- "step": 134
412
- },
413
- {
414
- "epoch": 2.11,
415
- "learning_rate": 2e-05,
416
- "loss": 2.4448,
417
- "step": 136
418
- },
419
- {
420
- "epoch": 2.14,
421
- "learning_rate": 2e-05,
422
- "loss": 2.5899,
423
- "step": 138
424
- },
425
- {
426
- "epoch": 2.17,
427
- "learning_rate": 2e-05,
428
- "loss": 2.5766,
429
- "step": 140
430
- },
431
- {
432
- "epoch": 2.21,
433
- "learning_rate": 2e-05,
434
- "loss": 2.6561,
435
- "step": 142
436
- },
437
- {
438
- "epoch": 2.24,
439
- "learning_rate": 2e-05,
440
- "loss": 2.6808,
441
- "step": 144
442
- },
443
- {
444
- "epoch": 2.27,
445
- "learning_rate": 2e-05,
446
- "loss": 2.3978,
447
- "step": 146
448
- },
449
- {
450
- "epoch": 2.3,
451
- "learning_rate": 2e-05,
452
- "loss": 2.2291,
453
- "step": 148
454
- },
455
- {
456
- "epoch": 2.33,
457
- "learning_rate": 2e-05,
458
- "loss": 2.4869,
459
- "step": 150
460
- },
461
- {
462
- "epoch": 2.36,
463
- "learning_rate": 2e-05,
464
- "loss": 2.3935,
465
- "step": 152
466
- },
467
- {
468
- "epoch": 2.39,
469
- "learning_rate": 2e-05,
470
- "loss": 2.3914,
471
- "step": 154
472
- },
473
- {
474
- "epoch": 2.42,
475
- "learning_rate": 2e-05,
476
- "loss": 2.4832,
477
- "step": 156
478
- },
479
- {
480
- "epoch": 2.45,
481
- "learning_rate": 2e-05,
482
- "loss": 2.6037,
483
- "step": 158
484
- },
485
- {
486
- "epoch": 2.49,
487
- "learning_rate": 2e-05,
488
- "loss": 2.6918,
489
- "step": 160
490
- },
491
- {
492
- "epoch": 2.52,
493
- "learning_rate": 2e-05,
494
- "loss": 2.4658,
495
- "step": 162
496
- },
497
- {
498
- "epoch": 2.55,
499
- "learning_rate": 2e-05,
500
- "loss": 2.3468,
501
- "step": 164
502
- },
503
- {
504
- "epoch": 2.58,
505
- "learning_rate": 2e-05,
506
- "loss": 2.4063,
507
- "step": 166
508
- },
509
- {
510
- "epoch": 2.61,
511
- "learning_rate": 2e-05,
512
- "loss": 2.2269,
513
- "step": 168
514
- },
515
- {
516
- "epoch": 2.64,
517
- "learning_rate": 2e-05,
518
- "loss": 2.6863,
519
- "step": 170
520
- },
521
- {
522
- "epoch": 2.67,
523
- "learning_rate": 2e-05,
524
- "loss": 2.4768,
525
- "step": 172
526
- },
527
- {
528
- "epoch": 2.7,
529
- "learning_rate": 2e-05,
530
- "loss": 2.5682,
531
- "step": 174
532
- },
533
- {
534
- "epoch": 2.73,
535
- "learning_rate": 2e-05,
536
- "loss": 2.6645,
537
- "step": 176
538
- },
539
- {
540
- "epoch": 2.77,
541
- "learning_rate": 2e-05,
542
- "loss": 2.4437,
543
- "step": 178
544
- },
545
- {
546
- "epoch": 2.8,
547
- "learning_rate": 2e-05,
548
- "loss": 2.2549,
549
- "step": 180
550
- },
551
- {
552
- "epoch": 2.83,
553
- "learning_rate": 2e-05,
554
- "loss": 2.2415,
555
- "step": 182
556
- },
557
- {
558
- "epoch": 2.86,
559
- "learning_rate": 2e-05,
560
- "loss": 2.3616,
561
- "step": 184
562
- },
563
- {
564
- "epoch": 2.89,
565
- "learning_rate": 2e-05,
566
- "loss": 2.6455,
567
- "step": 186
568
- },
569
- {
570
- "epoch": 2.92,
571
- "learning_rate": 2e-05,
572
- "loss": 2.4196,
573
- "step": 188
574
- },
575
- {
576
- "epoch": 2.95,
577
- "learning_rate": 2e-05,
578
- "loss": 2.3695,
579
- "step": 190
580
- },
581
- {
582
- "epoch": 2.98,
583
- "learning_rate": 2e-05,
584
- "loss": 2.6522,
585
- "step": 192
586
  }
587
  ],
588
  "logging_steps": 2,
589
- "max_steps": 192,
590
  "num_train_epochs": 3,
591
  "save_steps": 64,
592
- "total_flos": 5247227070922752.0,
593
  "trial_name": null,
594
  "trial_params": null
595
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 0.029938018945152613,
5
  "eval_steps": 500,
6
+ "global_step": 64,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "epoch": 0.0,
13
  "learning_rate": 2e-05,
14
+ "loss": 1.2083,
15
  "step": 2
16
  },
17
  {
18
+ "epoch": 0.0,
19
  "learning_rate": 2e-05,
20
+ "loss": 1.3515,
21
  "step": 4
22
  },
23
  {
24
+ "epoch": 0.0,
25
  "learning_rate": 2e-05,
26
+ "loss": 1.3634,
27
  "step": 6
28
  },
29
  {
30
+ "epoch": 0.0,
31
  "learning_rate": 2e-05,
32
+ "loss": 1.5072,
33
  "step": 8
34
  },
35
  {
36
+ "epoch": 0.0,
37
  "learning_rate": 2e-05,
38
+ "loss": 1.3421,
39
  "step": 10
40
  },
41
  {
42
+ "epoch": 0.01,
43
  "learning_rate": 2e-05,
44
+ "loss": 1.4479,
45
  "step": 12
46
  },
47
  {
48
+ "epoch": 0.01,
49
  "learning_rate": 2e-05,
50
+ "loss": 1.4508,
51
  "step": 14
52
  },
53
  {
54
+ "epoch": 0.01,
55
  "learning_rate": 2e-05,
56
+ "loss": 1.6295,
57
  "step": 16
58
  },
59
  {
60
+ "epoch": 0.01,
61
  "learning_rate": 2e-05,
62
+ "loss": 1.7485,
63
  "step": 18
64
  },
65
  {
66
+ "epoch": 0.01,
67
  "learning_rate": 2e-05,
68
+ "loss": 1.7204,
69
  "step": 20
70
  },
71
  {
72
+ "epoch": 0.01,
73
  "learning_rate": 2e-05,
74
+ "loss": 1.6516,
75
  "step": 22
76
  },
77
  {
78
+ "epoch": 0.01,
79
  "learning_rate": 2e-05,
80
+ "loss": 1.8028,
81
  "step": 24
82
  },
83
  {
84
+ "epoch": 0.01,
85
  "learning_rate": 2e-05,
86
+ "loss": 1.8158,
87
  "step": 26
88
  },
89
  {
90
+ "epoch": 0.01,
91
  "learning_rate": 2e-05,
92
+ "loss": 1.7422,
93
  "step": 28
94
  },
95
  {
96
+ "epoch": 0.01,
97
  "learning_rate": 2e-05,
98
+ "loss": 1.9044,
99
  "step": 30
100
  },
101
  {
102
+ "epoch": 0.01,
103
  "learning_rate": 2e-05,
104
+ "loss": 1.9552,
105
  "step": 32
106
  },
107
  {
108
+ "epoch": 0.02,
109
  "learning_rate": 2e-05,
110
+ "loss": 1.9569,
111
  "step": 34
112
  },
113
  {
114
+ "epoch": 0.02,
115
  "learning_rate": 2e-05,
116
+ "loss": 1.8889,
117
  "step": 36
118
  },
119
  {
120
+ "epoch": 0.02,
121
  "learning_rate": 2e-05,
122
+ "loss": 2.0743,
123
  "step": 38
124
  },
125
  {
126
+ "epoch": 0.02,
127
  "learning_rate": 2e-05,
128
+ "loss": 2.2455,
129
  "step": 40
130
  },
131
  {
132
+ "epoch": 0.02,
133
  "learning_rate": 2e-05,
134
+ "loss": 2.4897,
135
  "step": 42
136
  },
137
  {
138
+ "epoch": 0.02,
139
  "learning_rate": 2e-05,
140
+ "loss": 3.4686,
141
  "step": 44
142
  },
143
  {
144
+ "epoch": 0.02,
145
  "learning_rate": 2e-05,
146
+ "loss": 4.6284,
147
  "step": 46
148
  },
149
  {
150
+ "epoch": 0.02,
151
  "learning_rate": 2e-05,
152
+ "loss": 5.1497,
153
  "step": 48
154
  },
155
  {
156
+ "epoch": 0.02,
157
  "learning_rate": 2e-05,
158
+ "loss": 5.8505,
159
  "step": 50
160
  },
161
  {
162
+ "epoch": 0.02,
163
  "learning_rate": 2e-05,
164
+ "loss": 1.2356,
165
  "step": 52
166
  },
167
  {
168
+ "epoch": 0.03,
169
  "learning_rate": 2e-05,
170
+ "loss": 1.0891,
171
  "step": 54
172
  },
173
  {
174
+ "epoch": 0.03,
175
  "learning_rate": 2e-05,
176
+ "loss": 1.3606,
177
  "step": 56
178
  },
179
  {
180
+ "epoch": 0.03,
181
  "learning_rate": 2e-05,
182
+ "loss": 1.2032,
183
  "step": 58
184
  },
185
  {
186
+ "epoch": 0.03,
187
  "learning_rate": 2e-05,
188
+ "loss": 1.5731,
189
  "step": 60
190
  },
191
  {
192
+ "epoch": 0.03,
193
  "learning_rate": 2e-05,
194
+ "loss": 1.3972,
195
  "step": 62
196
  },
197
  {
198
+ "epoch": 0.03,
199
  "learning_rate": 2e-05,
200
+ "loss": 1.2838,
201
  "step": 64
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
202
  }
203
  ],
204
  "logging_steps": 2,
205
+ "max_steps": 6411,
206
  "num_train_epochs": 3,
207
  "save_steps": 64,
208
+ "total_flos": 1773658295746560.0,
209
  "trial_name": null,
210
  "trial_params": null
211
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f63f9a244ea9fa2da2b8f04aed940ca2e2687b0a7e2cc08aebb2592b5b348d42
3
  size 4027
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a4cab28c541135f31db77ef933651a034c665fbcad5b7aef448b8f45a0ce5092
3
  size 4027