AlekseyKorshuk commited on
Commit
1790f07
·
1 Parent(s): 3b4951b

huggingartists

Browse files
README.md CHANGED
@@ -45,15 +45,15 @@ from datasets import load_dataset
45
  dataset = load_dataset("huggingartists/queen")
46
  ```
47
 
48
- [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/33biqcks/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
49
 
50
  ## Training procedure
51
 
52
  The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Queen's lyrics.
53
 
54
- Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/1k3htq7p) for full transparency and reproducibility.
55
 
56
- At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/1k3htq7p/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
 
45
  dataset = load_dataset("huggingartists/queen")
46
  ```
47
 
48
+ [Explore the data](https://wandb.ai/huggingartists/huggingartists/runs/29edwcv0/artifacts), which is tracked with [W&B artifacts](https://docs.wandb.com/artifacts) at every step of the pipeline.
49
 
50
  ## Training procedure
51
 
52
  The model is based on a pre-trained [GPT-2](https://huggingface.co/gpt2) which is fine-tuned on Queen's lyrics.
53
 
54
+ Hyperparameters and metrics are recorded in the [W&B training run](https://wandb.ai/huggingartists/huggingartists/runs/135qvcfh) for full transparency and reproducibility.
55
 
56
+ At the end of training, [the final model](https://wandb.ai/huggingartists/huggingartists/runs/135qvcfh/artifacts) is logged and versioned.
57
 
58
  ## How to use
59
 
evaluation.txt CHANGED
@@ -1 +1 @@
1
- {"eval_loss": 0.9721857309341431, "eval_runtime": 1.902, "eval_samples_per_second": 44.689, "eval_steps_per_second": 5.783, "epoch": 11.0}
 
1
+ {"eval_loss": 0.9767317771911621, "eval_runtime": 2.1966, "eval_samples_per_second": 43.704, "eval_steps_per_second": 5.463, "epoch": 12.0}
flax_model.msgpack CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:800bffdd4c52f619bf17474c0d9f571b39795cdad6155abd76ab3d0e828edbee
3
  size 497764120
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d59d4092b6fbaa70e9f950aa8fcb06e8531e931e125851c1d4493cca00a1d6d3
3
  size 497764120
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d44e3f3b76fe08f524c40d07198c2501597e28b7eaa180a40a79abc32258dba5
3
  size 995604017
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ee8cef8141948570d21835c08d084a9b314fbdef3ba8f904e6e203013456a0d
3
  size 995604017
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:56180aae9f0dfc24b3d1cdbb07dc2eb5acaab5a5b5e60892d48dca7feaf72ef9
3
  size 510396521
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3308c600cab57fc29528fed25a3b6bbe8e7e16d4e20d2e4b9967b339a59a76b
3
  size 510396521
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f441152599d4ecfcd40f2f229153cf77289fd381959e35ff164226f90d545b2b
3
  size 14503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:afadc6c82078abf8ebc3154d7c31c5702ef7fa29db799979801c5d8393338a19
3
  size 14503
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ae81c5f3c8df017bd46e14a116a6befbfbd6eda859addceb4a11946640558415
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d825f1ad452557c55ae3db5704cfc8ff0c2aef068797ec1bf426331ec8788e52
3
  size 623
trainer_state.json CHANGED
@@ -1,1026 +1,102 @@
1
  {
2
- "best_metric": 0.9721857309341431,
3
- "best_model_checkpoint": "output/queen/checkpoint-759",
4
- "epoch": 11.0,
5
- "global_step": 759,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
  "epoch": 0.07,
12
- "learning_rate": 0.00013537785052914354,
13
- "loss": 3.396,
14
  "step": 5
15
  },
16
  {
17
  "epoch": 0.15,
18
  "learning_rate": 0.0001300082017869573,
19
- "loss": 3.2987,
20
  "step": 10
21
  },
22
  {
23
  "epoch": 0.22,
24
- "learning_rate": 0.0001213763104094255,
25
- "loss": 2.8576,
26
  "step": 15
27
  },
28
  {
29
  "epoch": 0.29,
30
- "learning_rate": 0.000109940736055617,
31
- "loss": 2.9968,
32
  "step": 20
33
  },
34
  {
35
  "epoch": 0.37,
36
- "learning_rate": 9.630898093421199e-05,
37
- "loss": 2.9324,
38
  "step": 25
39
  },
40
  {
41
  "epoch": 0.44,
42
- "learning_rate": 8.120521692221673e-05,
43
- "loss": 2.9931,
44
  "step": 30
45
  },
46
  {
47
  "epoch": 0.51,
48
  "learning_rate": 6.543181473690228e-05,
49
- "loss": 2.8919,
50
  "step": 35
51
  },
52
  {
53
  "epoch": 0.59,
54
- "learning_rate": 4.982671888105512e-05,
55
- "loss": 2.7977,
56
  "step": 40
57
  },
58
  {
59
  "epoch": 0.66,
60
- "learning_rate": 3.5218932770313667e-05,
61
- "loss": 2.9689,
62
  "step": 45
63
  },
64
  {
65
  "epoch": 0.74,
66
- "learning_rate": 2.2384478845846175e-05,
67
- "loss": 2.756,
68
  "step": 50
69
  },
70
  {
71
  "epoch": 0.81,
72
- "learning_rate": 1.200517324255647e-05,
73
- "loss": 2.8716,
74
  "step": 55
75
  },
76
  {
77
  "epoch": 0.88,
78
- "learning_rate": 4.6324050628611986e-06,
79
- "loss": 2.9236,
80
  "step": 60
81
  },
82
  {
83
  "epoch": 0.96,
84
- "learning_rate": 6.578444419609313e-07,
85
- "loss": 2.8184,
86
  "step": 65
87
  },
88
  {
89
  "epoch": 1.0,
90
- "eval_loss": 2.78603196144104,
91
- "eval_runtime": 4.3884,
92
- "eval_samples_per_second": 21.192,
93
- "eval_steps_per_second": 2.734,
94
  "step": 68
95
- },
96
- {
97
- "epoch": 1.03,
98
- "learning_rate": 2.926355061606279e-07,
99
- "loss": 2.8201,
100
- "step": 70
101
- },
102
- {
103
- "epoch": 1.1,
104
- "learning_rate": 3.55617957525975e-06,
105
- "loss": 2.7876,
106
- "step": 75
107
- },
108
- {
109
- "epoch": 1.18,
110
- "learning_rate": 1.0275104488948473e-05,
111
- "loss": 2.8738,
112
- "step": 80
113
- },
114
- {
115
- "epoch": 1.25,
116
- "learning_rate": 2.009247481060283e-05,
117
- "loss": 2.6439,
118
- "step": 85
119
- },
120
- {
121
- "epoch": 1.32,
122
- "learning_rate": 3.2486753626613365e-05,
123
- "loss": 2.5993,
124
- "step": 90
125
- },
126
- {
127
- "epoch": 1.4,
128
- "learning_rate": 4.679950861668151e-05,
129
- "loss": 2.7072,
130
- "step": 95
131
- },
132
- {
133
- "epoch": 1.47,
134
- "learning_rate": 6.227039054081748e-05,
135
- "loss": 2.7465,
136
- "step": 100
137
- },
138
- {
139
- "epoch": 1.54,
140
- "learning_rate": 7.807752594969911e-05,
141
- "loss": 2.5926,
142
- "step": 105
143
- },
144
- {
145
- "epoch": 1.62,
146
- "learning_rate": 9.338117830043867e-05,
147
- "loss": 2.7178,
148
- "step": 110
149
- },
150
- {
151
- "epoch": 1.69,
152
- "learning_rate": 0.00010736835802938978,
153
- "loss": 2.7027,
154
- "step": 115
155
- },
156
- {
157
- "epoch": 1.76,
158
- "learning_rate": 0.00011929601172133719,
159
- "loss": 2.4457,
160
- "step": 120
161
- },
162
- {
163
- "epoch": 1.84,
164
- "learning_rate": 0.00012853049599834097,
165
- "loss": 2.6806,
166
- "step": 125
167
- },
168
- {
169
- "epoch": 1.91,
170
- "learning_rate": 0.00013458123912165538,
171
- "loss": 2.4508,
172
- "step": 130
173
- },
174
- {
175
- "epoch": 1.99,
176
- "learning_rate": 0.00013712680207144277,
177
- "loss": 2.6492,
178
- "step": 135
179
- },
180
- {
181
- "epoch": 2.0,
182
- "eval_loss": 2.3717236518859863,
183
- "eval_runtime": 4.3106,
184
- "eval_samples_per_second": 22.039,
185
- "eval_steps_per_second": 2.784,
186
- "step": 136
187
- },
188
- {
189
- "epoch": 2.0,
190
- "eval_loss": 2.222139835357666,
191
- "eval_runtime": 3.7161,
192
- "eval_samples_per_second": 23.412,
193
- "eval_steps_per_second": 2.96,
194
- "step": 138
195
- },
196
- {
197
- "epoch": 2.03,
198
- "learning_rate": 0.00013691577939766304,
199
- "loss": 2.3223,
200
- "step": 140
201
- },
202
- {
203
- "epoch": 2.1,
204
- "learning_rate": 0.00013374528334456193,
205
- "loss": 2.4227,
206
- "step": 145
207
- },
208
- {
209
- "epoch": 2.17,
210
- "learning_rate": 0.00012721317115188912,
211
- "loss": 2.3956,
212
- "step": 150
213
- },
214
- {
215
- "epoch": 2.25,
216
- "learning_rate": 0.00011765651179294606,
217
- "loss": 2.3273,
218
- "step": 155
219
- },
220
- {
221
- "epoch": 2.32,
222
- "learning_rate": 0.0001055684464238085,
223
- "loss": 2.4048,
224
- "step": 160
225
- },
226
- {
227
- "epoch": 2.39,
228
- "learning_rate": 9.157274139492967e-05,
229
- "loss": 2.4168,
230
- "step": 165
231
- },
232
- {
233
- "epoch": 2.46,
234
- "learning_rate": 7.639160076306266e-05,
235
- "loss": 2.4496,
236
- "step": 170
237
- },
238
- {
239
- "epoch": 2.54,
240
- "learning_rate": 6.0808399236937385e-05,
241
- "loss": 2.0669,
242
- "step": 175
243
- },
244
- {
245
- "epoch": 2.61,
246
- "learning_rate": 4.5627258605070364e-05,
247
- "loss": 2.2692,
248
- "step": 180
249
- },
250
- {
251
- "epoch": 2.68,
252
- "learning_rate": 3.1631553576191535e-05,
253
- "loss": 2.1242,
254
- "step": 185
255
- },
256
- {
257
- "epoch": 2.75,
258
- "learning_rate": 1.954348820705391e-05,
259
- "loss": 2.3064,
260
- "step": 190
261
- },
262
- {
263
- "epoch": 2.83,
264
- "learning_rate": 9.986828848110892e-06,
265
- "loss": 2.146,
266
- "step": 195
267
- },
268
- {
269
- "epoch": 2.9,
270
- "learning_rate": 3.454716655438096e-06,
271
- "loss": 2.3787,
272
- "step": 200
273
- },
274
- {
275
- "epoch": 2.97,
276
- "learning_rate": 2.8422060233695394e-07,
277
- "loss": 2.2533,
278
- "step": 205
279
- },
280
- {
281
- "epoch": 3.0,
282
- "eval_loss": 2.043508768081665,
283
- "eval_runtime": 3.8679,
284
- "eval_samples_per_second": 22.493,
285
- "eval_steps_per_second": 2.844,
286
- "step": 207
287
- },
288
- {
289
- "epoch": 3.04,
290
- "learning_rate": 6.389441019077102e-07,
291
- "loss": 2.1995,
292
- "step": 210
293
- },
294
- {
295
- "epoch": 3.12,
296
- "learning_rate": 4.500582770777967e-06,
297
- "loss": 2.2239,
298
- "step": 215
299
- },
300
- {
301
- "epoch": 3.19,
302
- "learning_rate": 1.166986896886688e-05,
303
- "loss": 2.1124,
304
- "step": 220
305
- },
306
- {
307
- "epoch": 3.26,
308
- "learning_rate": 2.177685437520025e-05,
309
- "loss": 2.2442,
310
- "step": 225
311
- },
312
- {
313
- "epoch": 3.33,
314
- "learning_rate": 3.4300000000000014e-05,
315
- "loss": 2.0765,
316
- "step": 230
317
- },
318
- {
319
- "epoch": 3.41,
320
- "learning_rate": 4.859308855339902e-05,
321
- "loss": 2.042,
322
- "step": 235
323
- },
324
- {
325
- "epoch": 3.48,
326
- "learning_rate": 6.391857044318346e-05,
327
- "loss": 2.0462,
328
- "step": 240
329
- },
330
- {
331
- "epoch": 3.55,
332
- "learning_rate": 7.948562268689874e-05,
333
- "loss": 1.9776,
334
- "step": 245
335
- },
336
- {
337
- "epoch": 3.62,
338
- "learning_rate": 9.449095682862937e-05,
339
- "loss": 2.1824,
340
- "step": 250
341
- },
342
- {
343
- "epoch": 3.7,
344
- "learning_rate": 0.0001081602700970798,
345
- "loss": 1.996,
346
- "step": 255
347
- },
348
- {
349
- "epoch": 3.77,
350
- "learning_rate": 0.00011978820084915117,
351
- "loss": 2.1329,
352
- "step": 260
353
- },
354
- {
355
- "epoch": 3.84,
356
- "learning_rate": 0.00012877472652481797,
357
- "loss": 2.1854,
358
- "step": 265
359
- },
360
- {
361
- "epoch": 3.91,
362
- "learning_rate": 0.00013465612591205902,
363
- "loss": 1.8201,
364
- "step": 270
365
- },
366
- {
367
- "epoch": 3.99,
368
- "learning_rate": 0.00013712890801216552,
369
- "loss": 1.8969,
370
- "step": 275
371
- },
372
- {
373
- "epoch": 4.0,
374
- "eval_loss": 1.9531660079956055,
375
- "eval_runtime": 3.8382,
376
- "eval_samples_per_second": 22.667,
377
- "eval_steps_per_second": 2.866,
378
- "step": 276
379
- },
380
- {
381
- "epoch": 4.06,
382
- "learning_rate": 0.0001360654727320747,
383
- "loss": 1.817,
384
- "step": 280
385
- },
386
- {
387
- "epoch": 4.13,
388
- "learning_rate": 0.0001315206952832741,
389
- "loss": 1.6319,
390
- "step": 285
391
- },
392
- {
393
- "epoch": 4.2,
394
- "learning_rate": 0.00012372909452021153,
395
- "loss": 1.7738,
396
- "step": 290
397
- },
398
- {
399
- "epoch": 4.28,
400
- "learning_rate": 0.00011309273133714503,
401
- "loss": 1.7757,
402
- "step": 295
403
- },
404
- {
405
- "epoch": 4.35,
406
- "learning_rate": 0.0001001604615883571,
407
- "loss": 1.9699,
408
- "step": 300
409
- },
410
- {
411
- "epoch": 4.42,
412
- "learning_rate": 8.5599614119113e-05,
413
- "loss": 1.9681,
414
- "step": 305
415
- },
416
- {
417
- "epoch": 4.49,
418
- "learning_rate": 7.016155537292879e-05,
419
- "loss": 1.9245,
420
- "step": 310
421
- },
422
- {
423
- "epoch": 4.57,
424
- "learning_rate": 5.464291750458938e-05,
425
- "loss": 1.7497,
426
- "step": 315
427
- },
428
- {
429
- "epoch": 4.64,
430
- "learning_rate": 3.98444906994169e-05,
431
- "loss": 1.7975,
432
- "step": 320
433
- },
434
- {
435
- "epoch": 4.71,
436
- "learning_rate": 2.6529900930537204e-05,
437
- "loss": 1.6675,
438
- "step": 325
439
- },
440
- {
441
- "epoch": 4.78,
442
- "learning_rate": 1.5386205457676833e-05,
443
- "loss": 1.9876,
444
- "step": 330
445
- },
446
- {
447
- "epoch": 4.86,
448
- "learning_rate": 6.988439412050469e-06,
449
- "loss": 1.691,
450
- "step": 335
451
- },
452
- {
453
- "epoch": 4.93,
454
- "learning_rate": 1.7699429285993718e-06,
455
- "loss": 1.8689,
456
- "step": 340
457
- },
458
- {
459
- "epoch": 5.0,
460
- "learning_rate": 0.0,
461
- "loss": 1.9327,
462
- "step": 345
463
- },
464
- {
465
- "epoch": 5.0,
466
- "eval_loss": 1.8510653972625732,
467
- "eval_runtime": 3.8275,
468
- "eval_samples_per_second": 22.73,
469
- "eval_steps_per_second": 2.874,
470
- "step": 345
471
- },
472
- {
473
- "epoch": 5.07,
474
- "learning_rate": 1.7699429285993566e-06,
475
- "loss": 1.5991,
476
- "step": 350
477
- },
478
- {
479
- "epoch": 5.14,
480
- "learning_rate": 6.988439412050431e-06,
481
- "loss": 1.6155,
482
- "step": 355
483
- },
484
- {
485
- "epoch": 5.22,
486
- "learning_rate": 1.538620545767678e-05,
487
- "loss": 1.6707,
488
- "step": 360
489
- },
490
- {
491
- "epoch": 5.29,
492
- "learning_rate": 2.6529900930537035e-05,
493
- "loss": 1.7262,
494
- "step": 365
495
- },
496
- {
497
- "epoch": 5.36,
498
- "learning_rate": 3.984449069941682e-05,
499
- "loss": 1.7087,
500
- "step": 370
501
- },
502
- {
503
- "epoch": 5.43,
504
- "learning_rate": 5.4642917504589295e-05,
505
- "loss": 1.7766,
506
- "step": 375
507
- },
508
- {
509
- "epoch": 5.51,
510
- "learning_rate": 7.016155537292883e-05,
511
- "loss": 1.5992,
512
- "step": 380
513
- },
514
- {
515
- "epoch": 5.58,
516
- "learning_rate": 8.559961411911291e-05,
517
- "loss": 1.6561,
518
- "step": 385
519
- },
520
- {
521
- "epoch": 5.65,
522
- "learning_rate": 0.00010016046158835702,
523
- "loss": 1.6108,
524
- "step": 390
525
- },
526
- {
527
- "epoch": 5.72,
528
- "learning_rate": 0.00011309273133714504,
529
- "loss": 1.6236,
530
- "step": 395
531
- },
532
- {
533
- "epoch": 5.8,
534
- "learning_rate": 0.0001237290945202115,
535
- "loss": 1.6319,
536
- "step": 400
537
- },
538
- {
539
- "epoch": 5.87,
540
- "learning_rate": 0.00013152069528327408,
541
- "loss": 1.4268,
542
- "step": 405
543
- },
544
- {
545
- "epoch": 5.94,
546
- "learning_rate": 0.00013606547273207472,
547
- "loss": 1.7278,
548
- "step": 410
549
- },
550
- {
551
- "epoch": 6.0,
552
- "eval_loss": 1.8254233598709106,
553
- "eval_runtime": 3.854,
554
- "eval_samples_per_second": 22.574,
555
- "eval_steps_per_second": 2.854,
556
- "step": 414
557
- },
558
- {
559
- "epoch": 6.01,
560
- "learning_rate": 0.00013712890801216552,
561
- "loss": 1.8762,
562
- "step": 415
563
- },
564
- {
565
- "epoch": 6.09,
566
- "learning_rate": 0.00013465612591205902,
567
- "loss": 1.4495,
568
- "step": 420
569
- },
570
- {
571
- "epoch": 6.16,
572
- "learning_rate": 0.00012877472652481797,
573
- "loss": 1.5224,
574
- "step": 425
575
- },
576
- {
577
- "epoch": 6.23,
578
- "learning_rate": 0.00011978820084915123,
579
- "loss": 1.5038,
580
- "step": 430
581
- },
582
- {
583
- "epoch": 6.3,
584
- "learning_rate": 0.00010816027009708009,
585
- "loss": 1.4097,
586
- "step": 435
587
- },
588
- {
589
- "epoch": 6.38,
590
- "learning_rate": 9.449095682862935e-05,
591
- "loss": 1.3203,
592
- "step": 440
593
- },
594
- {
595
- "epoch": 6.45,
596
- "learning_rate": 7.948562268689883e-05,
597
- "loss": 1.3634,
598
- "step": 445
599
- },
600
- {
601
- "epoch": 6.52,
602
- "learning_rate": 6.39185704431838e-05,
603
- "loss": 1.4181,
604
- "step": 450
605
- },
606
- {
607
- "epoch": 6.59,
608
- "learning_rate": 4.859308855339899e-05,
609
- "loss": 1.5025,
610
- "step": 455
611
- },
612
- {
613
- "epoch": 6.67,
614
- "learning_rate": 3.429999999999998e-05,
615
- "loss": 1.5881,
616
- "step": 460
617
- },
618
- {
619
- "epoch": 6.74,
620
- "learning_rate": 2.1776854375200487e-05,
621
- "loss": 1.5042,
622
- "step": 465
623
- },
624
- {
625
- "epoch": 6.81,
626
- "learning_rate": 1.1669868968866859e-05,
627
- "loss": 1.4229,
628
- "step": 470
629
- },
630
- {
631
- "epoch": 6.88,
632
- "learning_rate": 4.500582770777952e-06,
633
- "loss": 1.5551,
634
- "step": 475
635
- },
636
- {
637
- "epoch": 6.96,
638
- "learning_rate": 6.389441019077407e-07,
639
- "loss": 1.6253,
640
- "step": 480
641
- },
642
- {
643
- "epoch": 7.0,
644
- "eval_loss": 1.787222146987915,
645
- "eval_runtime": 3.8357,
646
- "eval_samples_per_second": 22.681,
647
- "eval_steps_per_second": 2.868,
648
- "step": 483
649
- },
650
- {
651
- "epoch": 7.03,
652
- "learning_rate": 2.8422060233696156e-07,
653
- "loss": 1.5683,
654
- "step": 485
655
- },
656
- {
657
- "epoch": 7.1,
658
- "learning_rate": 3.4547166554380356e-06,
659
- "loss": 1.1749,
660
- "step": 490
661
- },
662
- {
663
- "epoch": 7.17,
664
- "learning_rate": 9.986828848110785e-06,
665
- "loss": 1.2655,
666
- "step": 495
667
- },
668
- {
669
- "epoch": 7.25,
670
- "learning_rate": 1.9543488207053935e-05,
671
- "loss": 1.3456,
672
- "step": 500
673
- },
674
- {
675
- "epoch": 7.32,
676
- "learning_rate": 3.163155357619136e-05,
677
- "loss": 1.2074,
678
- "step": 505
679
- },
680
- {
681
- "epoch": 7.39,
682
- "learning_rate": 4.5627258605070174e-05,
683
- "loss": 1.3068,
684
- "step": 510
685
- },
686
- {
687
- "epoch": 7.46,
688
- "learning_rate": 6.080839923693748e-05,
689
- "loss": 1.1119,
690
- "step": 515
691
- },
692
- {
693
- "epoch": 7.54,
694
- "learning_rate": 7.639160076306239e-05,
695
- "loss": 1.3026,
696
- "step": 520
697
- },
698
- {
699
- "epoch": 7.61,
700
- "learning_rate": 9.15727413949297e-05,
701
- "loss": 1.4164,
702
- "step": 525
703
- },
704
- {
705
- "epoch": 7.68,
706
- "learning_rate": 0.00010556844642380854,
707
- "loss": 1.6188,
708
- "step": 530
709
- },
710
- {
711
- "epoch": 7.75,
712
- "learning_rate": 0.00011765651179294598,
713
- "loss": 1.438,
714
- "step": 535
715
- },
716
- {
717
- "epoch": 7.83,
718
- "learning_rate": 0.00012721317115188917,
719
- "loss": 1.267,
720
- "step": 540
721
- },
722
- {
723
- "epoch": 7.9,
724
- "learning_rate": 0.00013374528334456193,
725
- "loss": 1.2976,
726
- "step": 545
727
- },
728
- {
729
- "epoch": 7.97,
730
- "learning_rate": 0.00013691577939766304,
731
- "loss": 1.4411,
732
- "step": 550
733
- },
734
- {
735
- "epoch": 8.0,
736
- "eval_loss": 1.795061707496643,
737
- "eval_runtime": 3.8467,
738
- "eval_samples_per_second": 22.617,
739
- "eval_steps_per_second": 2.86,
740
- "step": 552
741
- },
742
- {
743
- "epoch": 8.04,
744
- "learning_rate": 0.0001365610558980923,
745
- "loss": 1.266,
746
- "step": 555
747
- },
748
- {
749
- "epoch": 8.12,
750
- "learning_rate": 0.000132699417229222,
751
- "loss": 1.2792,
752
- "step": 560
753
- },
754
- {
755
- "epoch": 8.19,
756
- "learning_rate": 0.0001255301310311332,
757
- "loss": 1.2561,
758
- "step": 565
759
- },
760
- {
761
- "epoch": 8.26,
762
- "learning_rate": 0.00011542314562479977,
763
- "loss": 1.2462,
764
- "step": 570
765
- },
766
- {
767
- "epoch": 8.33,
768
- "learning_rate": 0.00010289999999999993,
769
- "loss": 1.2013,
770
- "step": 575
771
- },
772
- {
773
- "epoch": 8.41,
774
- "learning_rate": 8.860691144660113e-05,
775
- "loss": 1.2052,
776
- "step": 580
777
- },
778
- {
779
- "epoch": 8.48,
780
- "learning_rate": 7.328142955681657e-05,
781
- "loss": 1.3885,
782
- "step": 585
783
- },
784
- {
785
- "epoch": 8.55,
786
- "learning_rate": 5.771437731310106e-05,
787
- "loss": 1.0822,
788
- "step": 590
789
- },
790
- {
791
- "epoch": 8.62,
792
- "learning_rate": 4.270904317137078e-05,
793
- "loss": 1.3064,
794
- "step": 595
795
- },
796
- {
797
- "epoch": 8.7,
798
- "learning_rate": 2.9039729902920224e-05,
799
- "loss": 1.2009,
800
- "step": 600
801
- },
802
- {
803
- "epoch": 8.77,
804
- "learning_rate": 1.7411799150848703e-05,
805
- "loss": 1.2057,
806
- "step": 605
807
- },
808
- {
809
- "epoch": 8.84,
810
- "learning_rate": 8.425273475182102e-06,
811
- "loss": 1.1498,
812
- "step": 610
813
- },
814
- {
815
- "epoch": 8.91,
816
- "learning_rate": 2.5438740879410024e-06,
817
- "loss": 1.0197,
818
- "step": 615
819
- },
820
- {
821
- "epoch": 8.99,
822
- "learning_rate": 7.109198783448199e-08,
823
- "loss": 1.1622,
824
- "step": 620
825
- },
826
- {
827
- "epoch": 9.0,
828
- "eval_loss": 1.775342345237732,
829
- "eval_runtime": 3.8327,
830
- "eval_samples_per_second": 22.699,
831
- "eval_steps_per_second": 2.87,
832
- "step": 621
833
- },
834
- {
835
- "epoch": 9.19,
836
- "learning_rate": 1.2005173242556402e-05,
837
- "loss": 1.3691,
838
- "step": 625
839
- },
840
- {
841
- "epoch": 9.26,
842
- "learning_rate": 2.2384478845846314e-05,
843
- "loss": 1.5005,
844
- "step": 630
845
- },
846
- {
847
- "epoch": 9.34,
848
- "learning_rate": 3.5218932770313436e-05,
849
- "loss": 1.4065,
850
- "step": 635
851
- },
852
- {
853
- "epoch": 9.41,
854
- "learning_rate": 4.982671888105515e-05,
855
- "loss": 1.3225,
856
- "step": 640
857
- },
858
- {
859
- "epoch": 9.49,
860
- "learning_rate": 6.543181473690211e-05,
861
- "loss": 1.0367,
862
- "step": 645
863
- },
864
- {
865
- "epoch": 9.56,
866
- "learning_rate": 8.120521692221662e-05,
867
- "loss": 1.1784,
868
- "step": 650
869
- },
870
- {
871
- "epoch": 9.63,
872
- "learning_rate": 9.630898093421192e-05,
873
- "loss": 1.2778,
874
- "step": 655
875
- },
876
- {
877
- "epoch": 9.71,
878
- "learning_rate": 0.00010994073605561679,
879
- "loss": 1.3754,
880
- "step": 660
881
- },
882
- {
883
- "epoch": 9.78,
884
- "learning_rate": 0.00012137631040942535,
885
- "loss": 1.1983,
886
- "step": 665
887
- },
888
- {
889
- "epoch": 9.85,
890
- "learning_rate": 0.00013000820178695732,
891
- "loss": 1.3077,
892
- "step": 670
893
- },
894
- {
895
- "epoch": 9.93,
896
- "learning_rate": 0.0001353778505291435,
897
- "loss": 1.4969,
898
- "step": 675
899
- },
900
- {
901
- "epoch": 10.0,
902
- "learning_rate": 0.0001372,
903
- "loss": 1.3263,
904
- "step": 680
905
- },
906
- {
907
- "epoch": 10.0,
908
- "eval_loss": 1.0593422651290894,
909
- "eval_runtime": 1.2161,
910
- "eval_samples_per_second": 75.649,
911
- "eval_steps_per_second": 9.867,
912
- "step": 680
913
- },
914
- {
915
- "epoch": 9.93,
916
- "learning_rate": 0.00013543005707140072,
917
- "loss": 1.2301,
918
- "step": 685
919
- },
920
- {
921
- "epoch": 10.0,
922
- "learning_rate": 0.0001372,
923
- "loss": 1.2039,
924
- "step": 690
925
- },
926
- {
927
- "epoch": 10.0,
928
- "eval_loss": 0.9929937720298767,
929
- "eval_runtime": 1.7672,
930
- "eval_samples_per_second": 48.098,
931
- "eval_steps_per_second": 6.224,
932
- "step": 690
933
- },
934
- {
935
- "epoch": 10.07,
936
- "learning_rate": 0.00013543005707140074,
937
- "loss": 1.1894,
938
- "step": 695
939
- },
940
- {
941
- "epoch": 10.14,
942
- "learning_rate": 0.0001302115605879495,
943
- "loss": 1.2106,
944
- "step": 700
945
- },
946
- {
947
- "epoch": 10.22,
948
- "learning_rate": 0.00012181379454232323,
949
- "loss": 1.144,
950
- "step": 705
951
- },
952
- {
953
- "epoch": 10.29,
954
- "learning_rate": 0.00011067009906946301,
955
- "loss": 1.3006,
956
- "step": 710
957
- },
958
- {
959
- "epoch": 10.36,
960
- "learning_rate": 9.735550930058323e-05,
961
- "loss": 1.2147,
962
- "step": 715
963
- },
964
- {
965
- "epoch": 10.43,
966
- "learning_rate": 8.255708249541075e-05,
967
- "loss": 1.1931,
968
- "step": 720
969
- },
970
- {
971
- "epoch": 10.51,
972
- "learning_rate": 6.703844462707122e-05,
973
- "loss": 1.2735,
974
- "step": 725
975
- },
976
- {
977
- "epoch": 10.58,
978
- "learning_rate": 5.160038588088712e-05,
979
- "loss": 1.1695,
980
- "step": 730
981
- },
982
- {
983
- "epoch": 10.65,
984
- "learning_rate": 3.703953841164302e-05,
985
- "loss": 1.2219,
986
- "step": 735
987
- },
988
- {
989
- "epoch": 10.72,
990
- "learning_rate": 2.4107268662854985e-05,
991
- "loss": 1.1271,
992
- "step": 740
993
- },
994
- {
995
- "epoch": 10.8,
996
- "learning_rate": 1.3470905479788523e-05,
997
- "loss": 1.2963,
998
- "step": 745
999
- },
1000
- {
1001
- "epoch": 10.87,
1002
- "learning_rate": 5.679304716725944e-06,
1003
- "loss": 1.0029,
1004
- "step": 750
1005
- },
1006
- {
1007
- "epoch": 10.94,
1008
- "learning_rate": 1.1345272679252971e-06,
1009
- "loss": 1.2789,
1010
- "step": 755
1011
- },
1012
- {
1013
- "epoch": 11.0,
1014
- "eval_loss": 0.9721857309341431,
1015
- "eval_runtime": 1.8495,
1016
- "eval_samples_per_second": 45.957,
1017
- "eval_steps_per_second": 5.947,
1018
- "step": 759
1019
  }
1020
  ],
1021
- "max_steps": 759,
1022
- "num_train_epochs": 11,
1023
- "total_flos": 785966432256000.0,
1024
  "trial_name": null,
1025
  "trial_params": null
1026
  }
 
1
  {
2
+ "best_metric": 0.9767317771911621,
3
+ "best_model_checkpoint": "output/queen/checkpoint-68",
4
+ "epoch": 1.0,
5
+ "global_step": 68,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
  "epoch": 0.07,
12
+ "learning_rate": 0.00013537785052914348,
13
+ "loss": 0.8946,
14
  "step": 5
15
  },
16
  {
17
  "epoch": 0.15,
18
  "learning_rate": 0.0001300082017869573,
19
+ "loss": 1.1571,
20
  "step": 10
21
  },
22
  {
23
  "epoch": 0.22,
24
+ "learning_rate": 0.00012137631040942562,
25
+ "loss": 1.0891,
26
  "step": 15
27
  },
28
  {
29
  "epoch": 0.29,
30
+ "learning_rate": 0.00010994073605561692,
31
+ "loss": 1.0026,
32
  "step": 20
33
  },
34
  {
35
  "epoch": 0.37,
36
+ "learning_rate": 9.63089809342121e-05,
37
+ "loss": 0.8956,
38
  "step": 25
39
  },
40
  {
41
  "epoch": 0.44,
42
+ "learning_rate": 8.120521692221655e-05,
43
+ "loss": 1.0037,
44
  "step": 30
45
  },
46
  {
47
  "epoch": 0.51,
48
  "learning_rate": 6.543181473690228e-05,
49
+ "loss": 0.9259,
50
  "step": 35
51
  },
52
  {
53
  "epoch": 0.59,
54
+ "learning_rate": 4.982671888105533e-05,
55
+ "loss": 0.9391,
56
  "step": 40
57
  },
58
  {
59
  "epoch": 0.66,
60
+ "learning_rate": 3.52189327703136e-05,
61
+ "loss": 0.897,
62
  "step": 45
63
  },
64
  {
65
  "epoch": 0.74,
66
+ "learning_rate": 2.2384478845846266e-05,
67
+ "loss": 0.969,
68
  "step": 50
69
  },
70
  {
71
  "epoch": 0.81,
72
+ "learning_rate": 1.2005173242556645e-05,
73
+ "loss": 0.9795,
74
  "step": 55
75
  },
76
  {
77
  "epoch": 0.88,
78
+ "learning_rate": 4.632405062861206e-06,
79
+ "loss": 1.0966,
80
  "step": 60
81
  },
82
  {
83
  "epoch": 0.96,
84
+ "learning_rate": 6.578444419609618e-07,
85
+ "loss": 0.9097,
86
  "step": 65
87
  },
88
  {
89
  "epoch": 1.0,
90
+ "eval_loss": 0.9767317771911621,
91
+ "eval_runtime": 2.0747,
92
+ "eval_samples_per_second": 46.272,
93
+ "eval_steps_per_second": 5.784,
94
  "step": 68
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
95
  }
96
  ],
97
+ "max_steps": 816,
98
+ "num_train_epochs": 12,
99
+ "total_flos": 70156910592000.0,
100
  "trial_name": null,
101
  "trial_params": null
102
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:04747b0b5177cd34b6949ed0755471acbad69a771c0daf9bdbba80e2464114ac
3
  size 3311
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de2b73e108df809dd60f215a7871329eac5738c939f3a5b41fa7a79d90523567
3
  size 3311