Hugging Face's logo

ccore
/
Llama-2-8k-2m-rethink

Text Generation
Transformers
PyTorch
llama
Generated from Trainer
text-generation-inference
Inference Endpoints
Model card Files Community
ccore commited on
Commit
93f1edb
•
1 Parent(s): e126367

Upload folder using huggingface_hub

Browse files
Files changed (14) hide show
  1. README.md +5 -5
  2. all_results.json +5 -5
  3. checkpoint-1000/config.json +5 -5
  4. checkpoint-1000/optimizer.pt +2 -2
  5. checkpoint-1000/pytorch_model.bin +2 -2
  6. checkpoint-1000/tokenizer_config.json +1 -1
  7. checkpoint-1000/trainer_state.json +203 -203
  8. checkpoint-1000/training_args.bin +1 -1
  9. config.json +5 -5
  10. pytorch_model.bin +2 -2
  11. tokenizer_config.json +1 -1
  12. train_results.json +5 -5
  13. trainer_state.json +0 -0
  14. training_args.bin +1 -1
README.md CHANGED
@@ -1,18 +1,18 @@
1
  ---
2
- base_model: ./output5
3
  tags:
4
  - generated_from_trainer
5
  model-index:
6
- - name: output5
7
  results: []
8
  ---
9
 
10
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
  should probably proofread and complete it, then remove this comment. -->
12
 
13
- # output5
14
 
15
- This model is a fine-tuned version of [./output5](https://huggingface.co/./output5) on an unknown dataset.
16
 
17
  ## Model description
18
 
@@ -37,7 +37,7 @@ The following hyperparameters were used during training:
37
  - seed: 42
38
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
39
  - lr_scheduler_type: constant
40
- - num_epochs: 5.0
41
 
42
  ### Training results
43
 
 
1
  ---
2
+ base_model: ./model
3
  tags:
4
  - generated_from_trainer
5
  model-index:
6
+ - name: output6
7
  results: []
8
  ---
9
 
10
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
11
  should probably proofread and complete it, then remove this comment. -->
12
 
13
+ # output6
14
 
15
+ This model is a fine-tuned version of [./model](https://huggingface.co/./model) on an unknown dataset.
16
 
17
  ## Model description
18
 
 
37
  - seed: 42
38
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
39
  - lr_scheduler_type: constant
40
+ - num_epochs: 1.0
41
 
42
  ### Training results
43
 
all_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 5.0,
3
- "train_loss": 1.999870525792007,
4
- "train_runtime": 268.6642,
5
  "train_samples": 1528,
6
- "train_samples_per_second": 28.437,
7
- "train_steps_per_second": 28.437
8
  }
 
1
  {
2
+ "epoch": 1.0,
3
+ "train_loss": 4.918852043401508,
4
+ "train_runtime": 510.3139,
5
  "train_samples": 1528,
6
+ "train_samples_per_second": 2.994,
7
+ "train_steps_per_second": 2.994
8
  }
checkpoint-1000/config.json CHANGED
@@ -1,18 +1,18 @@
1
  {
2
- "_name_or_path": "./output5",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
6
  "bos_token_id": 1,
7
  "eos_token_id": 2,
8
  "hidden_act": "silu",
9
- "hidden_size": 256,
10
  "initializer_range": 0.02,
11
- "intermediate_size": 512,
12
- "max_position_embeddings": 8192,
13
  "model_type": "llama",
14
  "num_attention_heads": 4,
15
- "num_hidden_layers": 2,
16
  "num_key_value_heads": 4,
17
  "pretraining_tp": 2,
18
  "rms_norm_eps": 1e-05,
 
1
  {
2
+ "_name_or_path": "./model",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
6
  "bos_token_id": 1,
7
  "eos_token_id": 2,
8
  "hidden_act": "silu",
9
+ "hidden_size": 64,
10
  "initializer_range": 0.02,
11
+ "intermediate_size": 256,
12
+ "max_position_embeddings": 2048,
13
  "model_type": "llama",
14
  "num_attention_heads": 4,
15
+ "num_hidden_layers": 64,
16
  "num_key_value_heads": 4,
17
  "pretraining_tp": 2,
18
  "rms_norm_eps": 1e-05,
checkpoint-1000/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:75d9e2af855228dddbe0658b3aca8fceb49da183db44df0f8f6e98b8b81f5e76
3
- size 141585770
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c7cda44b230bbcb85da0d7236da1eb35bb9377fa2bb6fd30f7b3d1b4b84598c
3
+ size 66872806
checkpoint-1000/pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fb42b85976a37ecf6745788717d8c16a2c224e77a5574c54a93294058563d9de
3
- size 70790896
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d3c73bd6ab978ac50d3a23630d3368f4cfa257b0358bb46d412c42ebbd5fa3f
3
+ size 33387545
checkpoint-1000/tokenizer_config.json CHANGED
@@ -29,7 +29,7 @@
29
  "bos_token": "<s>",
30
  "clean_up_tokenization_spaces": false,
31
  "eos_token": "</s>",
32
- "model_max_length": 8192,
33
  "pad_token": null,
34
  "sp_model_kwargs": {},
35
  "tokenizer_class": "LlamaTokenizer",
 
29
  "bos_token": "<s>",
30
  "clean_up_tokenization_spaces": false,
31
  "eos_token": "</s>",
32
+ "model_max_length": 2048,
33
  "pad_token": null,
34
  "sp_model_kwargs": {},
35
  "tokenizer_class": "LlamaTokenizer",
checkpoint-1000/trainer_state.json CHANGED
@@ -11,1209 +11,1209 @@
11
  {
12
  "epoch": 0.0,
13
  "learning_rate": 0.001,
14
- "loss": 1.3845,
15
  "step": 5
16
  },
17
  {
18
  "epoch": 0.01,
19
  "learning_rate": 0.001,
20
- "loss": 1.6284,
21
  "step": 10
22
  },
23
  {
24
  "epoch": 0.01,
25
  "learning_rate": 0.001,
26
- "loss": 1.3152,
27
  "step": 15
28
  },
29
  {
30
  "epoch": 0.01,
31
  "learning_rate": 0.001,
32
- "loss": 1.6192,
33
  "step": 20
34
  },
35
  {
36
  "epoch": 0.02,
37
  "learning_rate": 0.001,
38
- "loss": 1.7394,
39
  "step": 25
40
  },
41
  {
42
  "epoch": 0.02,
43
  "learning_rate": 0.001,
44
- "loss": 1.9925,
45
  "step": 30
46
  },
47
  {
48
  "epoch": 0.02,
49
  "learning_rate": 0.001,
50
- "loss": 1.902,
51
  "step": 35
52
  },
53
  {
54
  "epoch": 0.03,
55
  "learning_rate": 0.001,
56
- "loss": 1.822,
57
  "step": 40
58
  },
59
  {
60
  "epoch": 0.03,
61
  "learning_rate": 0.001,
62
- "loss": 2.137,
63
  "step": 45
64
  },
65
  {
66
  "epoch": 0.03,
67
  "learning_rate": 0.001,
68
- "loss": 2.2556,
69
  "step": 50
70
  },
71
  {
72
  "epoch": 0.04,
73
  "learning_rate": 0.001,
74
- "loss": 2.2468,
75
  "step": 55
76
  },
77
  {
78
  "epoch": 0.04,
79
  "learning_rate": 0.001,
80
- "loss": 2.0746,
81
  "step": 60
82
  },
83
  {
84
  "epoch": 0.04,
85
  "learning_rate": 0.001,
86
- "loss": 2.1092,
87
  "step": 65
88
  },
89
  {
90
  "epoch": 0.05,
91
  "learning_rate": 0.001,
92
- "loss": 1.9553,
93
  "step": 70
94
  },
95
  {
96
  "epoch": 0.05,
97
  "learning_rate": 0.001,
98
- "loss": 2.121,
99
  "step": 75
100
  },
101
  {
102
  "epoch": 0.05,
103
  "learning_rate": 0.001,
104
- "loss": 2.1722,
105
  "step": 80
106
  },
107
  {
108
  "epoch": 0.06,
109
  "learning_rate": 0.001,
110
- "loss": 2.1552,
111
  "step": 85
112
  },
113
  {
114
  "epoch": 0.06,
115
  "learning_rate": 0.001,
116
- "loss": 2.5971,
117
  "step": 90
118
  },
119
  {
120
  "epoch": 0.06,
121
  "learning_rate": 0.001,
122
- "loss": 2.4684,
123
  "step": 95
124
  },
125
  {
126
  "epoch": 0.07,
127
  "learning_rate": 0.001,
128
- "loss": 2.0248,
129
  "step": 100
130
  },
131
  {
132
  "epoch": 0.07,
133
  "learning_rate": 0.001,
134
- "loss": 2.5156,
135
  "step": 105
136
  },
137
  {
138
  "epoch": 0.07,
139
  "learning_rate": 0.001,
140
- "loss": 2.3322,
141
  "step": 110
142
  },
143
  {
144
  "epoch": 0.08,
145
  "learning_rate": 0.001,
146
- "loss": 1.7698,
147
  "step": 115
148
  },
149
  {
150
  "epoch": 0.08,
151
  "learning_rate": 0.001,
152
- "loss": 2.3363,
153
  "step": 120
154
  },
155
  {
156
  "epoch": 0.08,
157
  "learning_rate": 0.001,
158
- "loss": 2.1079,
159
  "step": 125
160
  },
161
  {
162
  "epoch": 0.09,
163
  "learning_rate": 0.001,
164
- "loss": 2.0998,
165
  "step": 130
166
  },
167
  {
168
  "epoch": 0.09,
169
  "learning_rate": 0.001,
170
- "loss": 2.8265,
171
  "step": 135
172
  },
173
  {
174
  "epoch": 0.09,
175
  "learning_rate": 0.001,
176
- "loss": 2.6611,
177
  "step": 140
178
  },
179
  {
180
  "epoch": 0.09,
181
  "learning_rate": 0.001,
182
- "loss": 2.2187,
183
  "step": 145
184
  },
185
  {
186
  "epoch": 0.1,
187
  "learning_rate": 0.001,
188
- "loss": 2.33,
189
  "step": 150
190
  },
191
  {
192
  "epoch": 0.1,
193
  "learning_rate": 0.001,
194
- "loss": 2.6119,
195
  "step": 155
196
  },
197
  {
198
  "epoch": 0.1,
199
  "learning_rate": 0.001,
200
- "loss": 2.2203,
201
  "step": 160
202
  },
203
  {
204
  "epoch": 0.11,
205
  "learning_rate": 0.001,
206
- "loss": 2.5474,
207
  "step": 165
208
  },
209
  {
210
  "epoch": 0.11,
211
  "learning_rate": 0.001,
212
- "loss": 2.4763,
213
  "step": 170
214
  },
215
  {
216
  "epoch": 0.11,
217
  "learning_rate": 0.001,
218
- "loss": 2.6068,
219
  "step": 175
220
  },
221
  {
222
  "epoch": 0.12,
223
  "learning_rate": 0.001,
224
- "loss": 2.1221,
225
  "step": 180
226
  },
227
  {
228
  "epoch": 0.12,
229
  "learning_rate": 0.001,
230
- "loss": 2.1185,
231
  "step": 185
232
  },
233
  {
234
  "epoch": 0.12,
235
  "learning_rate": 0.001,
236
- "loss": 1.9481,
237
  "step": 190
238
  },
239
  {
240
  "epoch": 0.13,
241
  "learning_rate": 0.001,
242
- "loss": 2.823,
243
  "step": 195
244
  },
245
  {
246
  "epoch": 0.13,
247
  "learning_rate": 0.001,
248
- "loss": 2.4702,
249
  "step": 200
250
  },
251
  {
252
  "epoch": 0.13,
253
  "learning_rate": 0.001,
254
- "loss": 2.3721,
255
  "step": 205
256
  },
257
  {
258
  "epoch": 0.14,
259
  "learning_rate": 0.001,
260
- "loss": 2.4932,
261
  "step": 210
262
  },
263
  {
264
  "epoch": 0.14,
265
  "learning_rate": 0.001,
266
- "loss": 2.3977,
267
  "step": 215
268
  },
269
  {
270
  "epoch": 0.14,
271
  "learning_rate": 0.001,
272
- "loss": 2.5299,
273
  "step": 220
274
  },
275
  {
276
  "epoch": 0.15,
277
  "learning_rate": 0.001,
278
- "loss": 2.4818,
279
  "step": 225
280
  },
281
  {
282
  "epoch": 0.15,
283
  "learning_rate": 0.001,
284
- "loss": 2.3404,
285
  "step": 230
286
  },
287
  {
288
  "epoch": 0.15,
289
  "learning_rate": 0.001,
290
- "loss": 2.3614,
291
  "step": 235
292
  },
293
  {
294
  "epoch": 0.16,
295
  "learning_rate": 0.001,
296
- "loss": 2.1509,
297
  "step": 240
298
  },
299
  {
300
  "epoch": 0.16,
301
  "learning_rate": 0.001,
302
- "loss": 2.5301,
303
  "step": 245
304
  },
305
  {
306
  "epoch": 0.16,
307
  "learning_rate": 0.001,
308
- "loss": 2.3328,
309
  "step": 250
310
  },
311
  {
312
  "epoch": 0.17,
313
  "learning_rate": 0.001,
314
- "loss": 2.412,
315
  "step": 255
316
  },
317
  {
318
  "epoch": 0.17,
319
  "learning_rate": 0.001,
320
- "loss": 2.5923,
321
  "step": 260
322
  },
323
  {
324
  "epoch": 0.17,
325
  "learning_rate": 0.001,
326
- "loss": 2.2655,
327
  "step": 265
328
  },
329
  {
330
  "epoch": 0.18,
331
  "learning_rate": 0.001,
332
- "loss": 2.7524,
333
  "step": 270
334
  },
335
  {
336
  "epoch": 0.18,
337
  "learning_rate": 0.001,
338
- "loss": 2.6456,
339
  "step": 275
340
  },
341
  {
342
  "epoch": 0.18,
343
  "learning_rate": 0.001,
344
- "loss": 2.3814,
345
  "step": 280
346
  },
347
  {
348
  "epoch": 0.19,
349
  "learning_rate": 0.001,
350
- "loss": 2.3805,
351
  "step": 285
352
  },
353
  {
354
  "epoch": 0.19,
355
  "learning_rate": 0.001,
356
- "loss": 2.6727,
357
  "step": 290
358
  },
359
  {
360
  "epoch": 0.19,
361
  "learning_rate": 0.001,
362
- "loss": 2.3931,
363
  "step": 295
364
  },
365
  {
366
  "epoch": 0.2,
367
  "learning_rate": 0.001,
368
- "loss": 1.9526,
369
  "step": 300
370
  },
371
  {
372
  "epoch": 0.2,
373
  "learning_rate": 0.001,
374
- "loss": 2.7219,
375
  "step": 305
376
  },
377
  {
378
  "epoch": 0.2,
379
  "learning_rate": 0.001,
380
- "loss": 2.4869,
381
  "step": 310
382
  },
383
  {
384
  "epoch": 0.21,
385
  "learning_rate": 0.001,
386
- "loss": 2.4303,
387
  "step": 315
388
  },
389
  {
390
  "epoch": 0.21,
391
  "learning_rate": 0.001,
392
- "loss": 2.5916,
393
  "step": 320
394
  },
395
  {
396
  "epoch": 0.21,
397
  "learning_rate": 0.001,
398
- "loss": 2.3917,
399
  "step": 325
400
  },
401
  {
402
  "epoch": 0.22,
403
  "learning_rate": 0.001,
404
- "loss": 2.4881,
405
  "step": 330
406
  },
407
  {
408
  "epoch": 0.22,
409
  "learning_rate": 0.001,
410
- "loss": 2.3369,
411
  "step": 335
412
  },
413
  {
414
  "epoch": 0.22,
415
  "learning_rate": 0.001,
416
- "loss": 2.45,
417
  "step": 340
418
  },
419
  {
420
  "epoch": 0.23,
421
  "learning_rate": 0.001,
422
- "loss": 2.1916,
423
  "step": 345
424
  },
425
  {
426
  "epoch": 0.23,
427
  "learning_rate": 0.001,
428
- "loss": 2.2606,
429
  "step": 350
430
  },
431
  {
432
  "epoch": 0.23,
433
  "learning_rate": 0.001,
434
- "loss": 2.5533,
435
  "step": 355
436
  },
437
  {
438
  "epoch": 0.24,
439
  "learning_rate": 0.001,
440
- "loss": 2.1873,
441
  "step": 360
442
  },
443
  {
444
  "epoch": 0.24,
445
  "learning_rate": 0.001,
446
- "loss": 2.6057,
447
  "step": 365
448
  },
449
  {
450
  "epoch": 0.24,
451
  "learning_rate": 0.001,
452
- "loss": 2.4611,
453
  "step": 370
454
  },
455
  {
456
  "epoch": 0.25,
457
  "learning_rate": 0.001,
458
- "loss": 2.606,
459
  "step": 375
460
  },
461
  {
462
  "epoch": 0.25,
463
  "learning_rate": 0.001,
464
- "loss": 2.5645,
465
  "step": 380
466
  },
467
  {
468
  "epoch": 0.25,
469
  "learning_rate": 0.001,
470
- "loss": 2.7591,
471
  "step": 385
472
  },
473
  {
474
  "epoch": 0.26,
475
  "learning_rate": 0.001,
476
- "loss": 2.3399,
477
  "step": 390
478
  },
479
  {
480
  "epoch": 0.26,
481
  "learning_rate": 0.001,
482
- "loss": 2.3675,
483
  "step": 395
484
  },
485
  {
486
  "epoch": 0.26,
487
  "learning_rate": 0.001,
488
- "loss": 2.5141,
489
  "step": 400
490
  },
491
  {
492
  "epoch": 0.27,
493
  "learning_rate": 0.001,
494
- "loss": 2.2407,
495
  "step": 405
496
  },
497
  {
498
  "epoch": 0.27,
499
  "learning_rate": 0.001,
500
- "loss": 2.6321,
501
  "step": 410
502
  },
503
  {
504
  "epoch": 0.27,
505
  "learning_rate": 0.001,
506
- "loss": 2.6431,
507
  "step": 415
508
  },
509
  {
510
  "epoch": 0.27,
511
  "learning_rate": 0.001,
512
- "loss": 2.4707,
513
  "step": 420
514
  },
515
  {
516
  "epoch": 0.28,
517
  "learning_rate": 0.001,
518
- "loss": 2.3744,
519
  "step": 425
520
  },
521
  {
522
  "epoch": 0.28,
523
  "learning_rate": 0.001,
524
- "loss": 2.5658,
525
  "step": 430
526
  },
527
  {
528
  "epoch": 0.28,
529
  "learning_rate": 0.001,
530
- "loss": 1.9922,
531
  "step": 435
532
  },
533
  {
534
  "epoch": 0.29,
535
  "learning_rate": 0.001,
536
- "loss": 2.5948,
537
  "step": 440
538
  },
539
  {
540
  "epoch": 0.29,
541
  "learning_rate": 0.001,
542
- "loss": 2.2519,
543
  "step": 445
544
  },
545
  {
546
  "epoch": 0.29,
547
  "learning_rate": 0.001,
548
- "loss": 2.6248,
549
  "step": 450
550
  },
551
  {
552
  "epoch": 0.3,
553
  "learning_rate": 0.001,
554
- "loss": 2.5877,
555
  "step": 455
556
  },
557
  {
558
  "epoch": 0.3,
559
  "learning_rate": 0.001,
560
- "loss": 2.1311,
561
  "step": 460
562
  },
563
  {
564
  "epoch": 0.3,
565
  "learning_rate": 0.001,
566
- "loss": 2.6091,
567
  "step": 465
568
  },
569
  {
570
  "epoch": 0.31,
571
  "learning_rate": 0.001,
572
- "loss": 2.5488,
573
  "step": 470
574
  },
575
  {
576
  "epoch": 0.31,
577
  "learning_rate": 0.001,
578
- "loss": 2.499,
579
  "step": 475
580
  },
581
  {
582
  "epoch": 0.31,
583
  "learning_rate": 0.001,
584
- "loss": 2.5031,
585
  "step": 480
586
  },
587
  {
588
  "epoch": 0.32,
589
  "learning_rate": 0.001,
590
- "loss": 2.5783,
591
  "step": 485
592
  },
593
  {
594
  "epoch": 0.32,
595
  "learning_rate": 0.001,
596
- "loss": 2.6303,
597
  "step": 490
598
  },
599
  {
600
  "epoch": 0.32,
601
  "learning_rate": 0.001,
602
- "loss": 2.6344,
603
  "step": 495
604
  },
605
  {
606
  "epoch": 0.33,
607
  "learning_rate": 0.001,
608
- "loss": 2.7031,
609
  "step": 500
610
  },
611
  {
612
  "epoch": 0.33,
613
  "learning_rate": 0.001,
614
- "loss": 2.6677,
615
  "step": 505
616
  },
617
  {
618
  "epoch": 0.33,
619
  "learning_rate": 0.001,
620
- "loss": 2.7617,
621
  "step": 510
622
  },
623
  {
624
  "epoch": 0.34,
625
  "learning_rate": 0.001,
626
- "loss": 2.3122,
627
  "step": 515
628
  },
629
  {
630
  "epoch": 0.34,
631
  "learning_rate": 0.001,
632
- "loss": 2.3887,
633
  "step": 520
634
  },
635
  {
636
  "epoch": 0.34,
637
  "learning_rate": 0.001,
638
- "loss": 2.5569,
639
  "step": 525
640
  },
641
  {
642
  "epoch": 0.35,
643
  "learning_rate": 0.001,
644
- "loss": 2.4038,
645
  "step": 530
646
  },
647
  {
648
  "epoch": 0.35,
649
  "learning_rate": 0.001,
650
- "loss": 2.7413,
651
  "step": 535
652
  },
653
  {
654
  "epoch": 0.35,
655
  "learning_rate": 0.001,
656
- "loss": 2.5264,
657
  "step": 540
658
  },
659
  {
660
  "epoch": 0.36,
661
  "learning_rate": 0.001,
662
- "loss": 2.335,
663
  "step": 545
664
  },
665
  {
666
  "epoch": 0.36,
667
  "learning_rate": 0.001,
668
- "loss": 2.7574,
669
  "step": 550
670
  },
671
  {
672
  "epoch": 0.36,
673
  "learning_rate": 0.001,
674
- "loss": 2.4542,
675
  "step": 555
676
  },
677
  {
678
  "epoch": 0.37,
679
  "learning_rate": 0.001,
680
- "loss": 2.0885,
681
  "step": 560
682
  },
683
  {
684
  "epoch": 0.37,
685
  "learning_rate": 0.001,
686
- "loss": 2.6057,
687
  "step": 565
688
  },
689
  {
690
  "epoch": 0.37,
691
  "learning_rate": 0.001,
692
- "loss": 2.4971,
693
  "step": 570
694
  },
695
  {
696
  "epoch": 0.38,
697
  "learning_rate": 0.001,
698
- "loss": 2.6251,
699
  "step": 575
700
  },
701
  {
702
  "epoch": 0.38,
703
  "learning_rate": 0.001,
704
- "loss": 2.5784,
705
  "step": 580
706
  },
707
  {
708
  "epoch": 0.38,
709
  "learning_rate": 0.001,
710
- "loss": 2.3878,
711
  "step": 585
712
  },
713
  {
714
  "epoch": 0.39,
715
  "learning_rate": 0.001,
716
- "loss": 2.4271,
717
  "step": 590
718
  },
719
  {
720
  "epoch": 0.39,
721
  "learning_rate": 0.001,
722
- "loss": 2.6751,
723
  "step": 595
724
  },
725
  {
726
  "epoch": 0.39,
727
  "learning_rate": 0.001,
728
- "loss": 2.17,
729
  "step": 600
730
  },
731
  {
732
  "epoch": 0.4,
733
  "learning_rate": 0.001,
734
- "loss": 2.0794,
735
  "step": 605
736
  },
737
  {
738
  "epoch": 0.4,
739
  "learning_rate": 0.001,
740
- "loss": 2.6073,
741
  "step": 610
742
  },
743
  {
744
  "epoch": 0.4,
745
  "learning_rate": 0.001,
746
- "loss": 2.6717,
747
  "step": 615
748
  },
749
  {
750
  "epoch": 0.41,
751
  "learning_rate": 0.001,
752
- "loss": 2.8236,
753
  "step": 620
754
  },
755
  {
756
  "epoch": 0.41,
757
  "learning_rate": 0.001,
758
- "loss": 2.4193,
759
  "step": 625
760
  },
761
  {
762
  "epoch": 0.41,
763
  "learning_rate": 0.001,
764
- "loss": 2.7181,
765
  "step": 630
766
  },
767
  {
768
  "epoch": 0.42,
769
  "learning_rate": 0.001,
770
- "loss": 2.6567,
771
  "step": 635
772
  },
773
  {
774
  "epoch": 0.42,
775
  "learning_rate": 0.001,
776
- "loss": 2.1919,
777
  "step": 640
778
  },
779
  {
780
  "epoch": 0.42,
781
  "learning_rate": 0.001,
782
- "loss": 2.5645,
783
  "step": 645
784
  },
785
  {
786
  "epoch": 0.43,
787
  "learning_rate": 0.001,
788
- "loss": 2.3313,
789
  "step": 650
790
  },
791
  {
792
  "epoch": 0.43,
793
  "learning_rate": 0.001,
794
- "loss": 2.533,
795
  "step": 655
796
  },
797
  {
798
  "epoch": 0.43,
799
  "learning_rate": 0.001,
800
- "loss": 2.621,
801
  "step": 660
802
  },
803
  {
804
  "epoch": 0.44,
805
  "learning_rate": 0.001,
806
- "loss": 2.5693,
807
  "step": 665
808
  },
809
  {
810
  "epoch": 0.44,
811
  "learning_rate": 0.001,
812
- "loss": 2.6943,
813
  "step": 670
814
  },
815
  {
816
  "epoch": 0.44,
817
  "learning_rate": 0.001,
818
- "loss": 2.2715,
819
  "step": 675
820
  },
821
  {
822
  "epoch": 0.45,
823
  "learning_rate": 0.001,
824
- "loss": 2.5614,
825
  "step": 680
826
  },
827
  {
828
  "epoch": 0.45,
829
  "learning_rate": 0.001,
830
- "loss": 2.5784,
831
  "step": 685
832
  },
833
  {
834
  "epoch": 0.45,
835
  "learning_rate": 0.001,
836
- "loss": 2.0613,
837
  "step": 690
838
  },
839
  {
840
  "epoch": 0.45,
841
  "learning_rate": 0.001,
842
- "loss": 2.582,
843
  "step": 695
844
  },
845
  {
846
  "epoch": 0.46,
847
  "learning_rate": 0.001,
848
- "loss": 2.5178,
849
  "step": 700
850
  },
851
  {
852
  "epoch": 0.46,
853
  "learning_rate": 0.001,
854
- "loss": 2.432,
855
  "step": 705
856
  },
857
  {
858
  "epoch": 0.46,
859
  "learning_rate": 0.001,
860
- "loss": 2.1979,
861
  "step": 710
862
  },
863
  {
864
  "epoch": 0.47,
865
  "learning_rate": 0.001,
866
- "loss": 2.4317,
867
  "step": 715
868
  },
869
  {
870
  "epoch": 0.47,
871
  "learning_rate": 0.001,
872
- "loss": 2.285,
873
  "step": 720
874
  },
875
  {
876
  "epoch": 0.47,
877
  "learning_rate": 0.001,
878
- "loss": 2.6127,
879
  "step": 725
880
  },
881
  {
882
  "epoch": 0.48,
883
  "learning_rate": 0.001,
884
- "loss": 2.6221,
885
  "step": 730
886
  },
887
  {
888
  "epoch": 0.48,
889
  "learning_rate": 0.001,
890
- "loss": 2.7519,
891
  "step": 735
892
  },
893
  {
894
  "epoch": 0.48,
895
  "learning_rate": 0.001,
896
- "loss": 2.7351,
897
  "step": 740
898
  },
899
  {
900
  "epoch": 0.49,
901
  "learning_rate": 0.001,
902
- "loss": 2.5224,
903
  "step": 745
904
  },
905
  {
906
  "epoch": 0.49,
907
  "learning_rate": 0.001,
908
- "loss": 2.5145,
909
  "step": 750
910
  },
911
  {
912
  "epoch": 0.49,
913
  "learning_rate": 0.001,
914
- "loss": 2.5122,
915
  "step": 755
916
  },
917
  {
918
  "epoch": 0.5,
919
  "learning_rate": 0.001,
920
- "loss": 2.2971,
921
  "step": 760
922
  },
923
  {
924
  "epoch": 0.5,
925
  "learning_rate": 0.001,
926
- "loss": 2.5628,
927
  "step": 765
928
  },
929
  {
930
  "epoch": 0.5,
931
  "learning_rate": 0.001,
932
- "loss": 2.5803,
933
  "step": 770
934
  },
935
  {
936
  "epoch": 0.51,
937
  "learning_rate": 0.001,
938
- "loss": 2.5209,
939
  "step": 775
940
  },
941
  {
942
  "epoch": 0.51,
943
  "learning_rate": 0.001,
944
- "loss": 2.5051,
945
  "step": 780
946
  },
947
  {
948
  "epoch": 0.51,
949
  "learning_rate": 0.001,
950
- "loss": 2.5275,
951
  "step": 785
952
  },
953
  {
954
  "epoch": 0.52,
955
  "learning_rate": 0.001,
956
- "loss": 2.5019,
957
  "step": 790
958
  },
959
  {
960
  "epoch": 0.52,
961
  "learning_rate": 0.001,
962
- "loss": 2.1714,
963
  "step": 795
964
  },
965
  {
966
  "epoch": 0.52,
967
  "learning_rate": 0.001,
968
- "loss": 2.87,
969
  "step": 800
970
  },
971
  {
972
  "epoch": 0.53,
973
  "learning_rate": 0.001,
974
- "loss": 2.6468,
975
  "step": 805
976
  },
977
  {
978
  "epoch": 0.53,
979
  "learning_rate": 0.001,
980
- "loss": 2.4115,
981
  "step": 810
982
  },
983
  {
984
  "epoch": 0.53,
985
  "learning_rate": 0.001,
986
- "loss": 2.5377,
987
  "step": 815
988
  },
989
  {
990
  "epoch": 0.54,
991
  "learning_rate": 0.001,
992
- "loss": 2.4986,
993
  "step": 820
994
  },
995
  {
996
  "epoch": 0.54,
997
  "learning_rate": 0.001,
998
- "loss": 2.6485,
999
  "step": 825
1000
  },
1001
  {
1002
  "epoch": 0.54,
1003
  "learning_rate": 0.001,
1004
- "loss": 2.6314,
1005
  "step": 830
1006
  },
1007
  {
1008
  "epoch": 0.55,
1009
  "learning_rate": 0.001,
1010
- "loss": 2.3516,
1011
  "step": 835
1012
  },
1013
  {
1014
  "epoch": 0.55,
1015
  "learning_rate": 0.001,
1016
- "loss": 2.39,
1017
  "step": 840
1018
  },
1019
  {
1020
  "epoch": 0.55,
1021
  "learning_rate": 0.001,
1022
- "loss": 2.3759,
1023
  "step": 845
1024
  },
1025
  {
1026
  "epoch": 0.56,
1027
  "learning_rate": 0.001,
1028
- "loss": 2.5348,
1029
  "step": 850
1030
  },
1031
  {
1032
  "epoch": 0.56,
1033
  "learning_rate": 0.001,
1034
- "loss": 2.6843,
1035
  "step": 855
1036
  },
1037
  {
1038
  "epoch": 0.56,
1039
  "learning_rate": 0.001,
1040
- "loss": 2.6289,
1041
  "step": 860
1042
  },
1043
  {
1044
  "epoch": 0.57,
1045
  "learning_rate": 0.001,
1046
- "loss": 2.4023,
1047
  "step": 865
1048
  },
1049
  {
1050
  "epoch": 0.57,
1051
  "learning_rate": 0.001,
1052
- "loss": 2.5979,
1053
  "step": 870
1054
  },
1055
  {
1056
  "epoch": 0.57,
1057
  "learning_rate": 0.001,
1058
- "loss": 2.4683,
1059
  "step": 875
1060
  },
1061
  {
1062
  "epoch": 0.58,
1063
  "learning_rate": 0.001,
1064
- "loss": 2.7099,
1065
  "step": 880
1066
  },
1067
  {
1068
  "epoch": 0.58,
1069
  "learning_rate": 0.001,
1070
- "loss": 2.3129,
1071
  "step": 885
1072
  },
1073
  {
1074
  "epoch": 0.58,
1075
  "learning_rate": 0.001,
1076
- "loss": 2.4508,
1077
  "step": 890
1078
  },
1079
  {
1080
  "epoch": 0.59,
1081
  "learning_rate": 0.001,
1082
- "loss": 2.6484,
1083
  "step": 895
1084
  },
1085
  {
1086
  "epoch": 0.59,
1087
  "learning_rate": 0.001,
1088
- "loss": 2.4541,
1089
  "step": 900
1090
  },
1091
  {
1092
  "epoch": 0.59,
1093
  "learning_rate": 0.001,
1094
- "loss": 2.5101,
1095
  "step": 905
1096
  },
1097
  {
1098
  "epoch": 0.6,
1099
  "learning_rate": 0.001,
1100
- "loss": 2.6002,
1101
  "step": 910
1102
  },
1103
  {
1104
  "epoch": 0.6,
1105
  "learning_rate": 0.001,
1106
- "loss": 2.742,
1107
  "step": 915
1108
  },
1109
  {
1110
  "epoch": 0.6,
1111
  "learning_rate": 0.001,
1112
- "loss": 2.4772,
1113
  "step": 920
1114
  },
1115
  {
1116
  "epoch": 0.61,
1117
  "learning_rate": 0.001,
1118
- "loss": 2.4328,
1119
  "step": 925
1120
  },
1121
  {
1122
  "epoch": 0.61,
1123
  "learning_rate": 0.001,
1124
- "loss": 2.7404,
1125
  "step": 930
1126
  },
1127
  {
1128
  "epoch": 0.61,
1129
  "learning_rate": 0.001,
1130
- "loss": 2.5969,
1131
  "step": 935
1132
  },
1133
  {
1134
  "epoch": 0.62,
1135
  "learning_rate": 0.001,
1136
- "loss": 2.1367,
1137
  "step": 940
1138
  },
1139
  {
1140
  "epoch": 0.62,
1141
  "learning_rate": 0.001,
1142
- "loss": 2.703,
1143
  "step": 945
1144
  },
1145
  {
1146
  "epoch": 0.62,
1147
  "learning_rate": 0.001,
1148
- "loss": 2.5638,
1149
  "step": 950
1150
  },
1151
  {
1152
  "epoch": 0.62,
1153
  "learning_rate": 0.001,
1154
- "loss": 1.8322,
1155
  "step": 955
1156
  },
1157
  {
1158
  "epoch": 0.63,
1159
  "learning_rate": 0.001,
1160
- "loss": 2.5764,
1161
  "step": 960
1162
  },
1163
  {
1164
  "epoch": 0.63,
1165
  "learning_rate": 0.001,
1166
- "loss": 2.4679,
1167
  "step": 965
1168
  },
1169
  {
1170
  "epoch": 0.63,
1171
  "learning_rate": 0.001,
1172
- "loss": 2.1691,
1173
  "step": 970
1174
  },
1175
  {
1176
  "epoch": 0.64,
1177
  "learning_rate": 0.001,
1178
- "loss": 2.3252,
1179
  "step": 975
1180
  },
1181
  {
1182
  "epoch": 0.64,
1183
  "learning_rate": 0.001,
1184
- "loss": 2.3988,
1185
  "step": 980
1186
  },
1187
  {
1188
  "epoch": 0.64,
1189
  "learning_rate": 0.001,
1190
- "loss": 2.221,
1191
  "step": 985
1192
  },
1193
  {
1194
  "epoch": 0.65,
1195
  "learning_rate": 0.001,
1196
- "loss": 2.4715,
1197
  "step": 990
1198
  },
1199
  {
1200
  "epoch": 0.65,
1201
  "learning_rate": 0.001,
1202
- "loss": 2.2479,
1203
  "step": 995
1204
  },
1205
  {
1206
  "epoch": 0.65,
1207
  "learning_rate": 0.001,
1208
- "loss": 2.3859,
1209
  "step": 1000
1210
  }
1211
  ],
1212
  "logging_steps": 5,
1213
- "max_steps": 7640,
1214
- "num_train_epochs": 5,
1215
  "save_steps": 1000,
1216
- "total_flos": 116785152000000.0,
1217
  "trial_name": null,
1218
  "trial_params": null
1219
  }
 
11
  {
12
  "epoch": 0.0,
13
  "learning_rate": 0.001,
14
+ "loss": 10.3529,
15
  "step": 5
16
  },
17
  {
18
  "epoch": 0.01,
19
  "learning_rate": 0.001,
20
+ "loss": 10.0747,
21
  "step": 10
22
  },
23
  {
24
  "epoch": 0.01,
25
  "learning_rate": 0.001,
26
+ "loss": 9.7304,
27
  "step": 15
28
  },
29
  {
30
  "epoch": 0.01,
31
  "learning_rate": 0.001,
32
+ "loss": 9.4139,
33
  "step": 20
34
  },
35
  {
36
  "epoch": 0.02,
37
  "learning_rate": 0.001,
38
+ "loss": 9.0418,
39
  "step": 25
40
  },
41
  {
42
  "epoch": 0.02,
43
  "learning_rate": 0.001,
44
+ "loss": 8.7655,
45
  "step": 30
46
  },
47
  {
48
  "epoch": 0.02,
49
  "learning_rate": 0.001,
50
+ "loss": 8.3599,
51
  "step": 35
52
  },
53
  {
54
  "epoch": 0.03,
55
  "learning_rate": 0.001,
56
+ "loss": 8.0603,
57
  "step": 40
58
  },
59
  {
60
  "epoch": 0.03,
61
  "learning_rate": 0.001,
62
+ "loss": 7.7776,
63
  "step": 45
64
  },
65
  {
66
  "epoch": 0.03,
67
  "learning_rate": 0.001,
68
+ "loss": 7.5678,
69
  "step": 50
70
  },
71
  {
72
  "epoch": 0.04,
73
  "learning_rate": 0.001,
74
+ "loss": 7.5338,
75
  "step": 55
76
  },
77
  {
78
  "epoch": 0.04,
79
  "learning_rate": 0.001,
80
+ "loss": 7.2249,
81
  "step": 60
82
  },
83
  {
84
  "epoch": 0.04,
85
  "learning_rate": 0.001,
86
+ "loss": 7.0286,
87
  "step": 65
88
  },
89
  {
90
  "epoch": 0.05,
91
  "learning_rate": 0.001,
92
+ "loss": 6.9518,
93
  "step": 70
94
  },
95
  {
96
  "epoch": 0.05,
97
  "learning_rate": 0.001,
98
+ "loss": 6.7623,
99
  "step": 75
100
  },
101
  {
102
  "epoch": 0.05,
103
  "learning_rate": 0.001,
104
+ "loss": 6.7361,
105
  "step": 80
106
  },
107
  {
108
  "epoch": 0.06,
109
  "learning_rate": 0.001,
110
+ "loss": 6.9201,
111
  "step": 85
112
  },
113
  {
114
  "epoch": 0.06,
115
  "learning_rate": 0.001,
116
+ "loss": 6.8491,
117
  "step": 90
118
  },
119
  {
120
  "epoch": 0.06,
121
  "learning_rate": 0.001,
122
+ "loss": 6.7565,
123
  "step": 95
124
  },
125
  {
126
  "epoch": 0.07,
127
  "learning_rate": 0.001,
128
+ "loss": 6.6537,
129
  "step": 100
130
  },
131
  {
132
  "epoch": 0.07,
133
  "learning_rate": 0.001,
134
+ "loss": 6.8522,
135
  "step": 105
136
  },
137
  {
138
  "epoch": 0.07,
139
  "learning_rate": 0.001,
140
+ "loss": 6.6052,
141
  "step": 110
142
  },
143
  {
144
  "epoch": 0.08,
145
  "learning_rate": 0.001,
146
+ "loss": 6.4885,
147
  "step": 115
148
  },
149
  {
150
  "epoch": 0.08,
151
  "learning_rate": 0.001,
152
+ "loss": 6.5162,
153
  "step": 120
154
  },
155
  {
156
  "epoch": 0.08,
157
  "learning_rate": 0.001,
158
+ "loss": 6.4759,
159
  "step": 125
160
  },
161
  {
162
  "epoch": 0.09,
163
  "learning_rate": 0.001,
164
+ "loss": 6.3678,
165
  "step": 130
166
  },
167
  {
168
  "epoch": 0.09,
169
  "learning_rate": 0.001,
170
+ "loss": 6.6435,
171
  "step": 135
172
  },
173
  {
174
  "epoch": 0.09,
175
  "learning_rate": 0.001,
176
+ "loss": 6.4821,
177
  "step": 140
178
  },
179
  {
180
  "epoch": 0.09,
181
  "learning_rate": 0.001,
182
+ "loss": 6.3594,
183
  "step": 145
184
  },
185
  {
186
  "epoch": 0.1,
187
  "learning_rate": 0.001,
188
+ "loss": 6.1634,
189
  "step": 150
190
  },
191
  {
192
  "epoch": 0.1,
193
  "learning_rate": 0.001,
194
+ "loss": 6.377,
195
  "step": 155
196
  },
197
  {
198
  "epoch": 0.1,
199
  "learning_rate": 0.001,
200
+ "loss": 6.368,
201
  "step": 160
202
  },
203
  {
204
  "epoch": 0.11,
205
  "learning_rate": 0.001,
206
+ "loss": 6.1571,
207
  "step": 165
208
  },
209
  {
210
  "epoch": 0.11,
211
  "learning_rate": 0.001,
212
+ "loss": 6.122,
213
  "step": 170
214
  },
215
  {
216
  "epoch": 0.11,
217
  "learning_rate": 0.001,
218
+ "loss": 6.2277,
219
  "step": 175
220
  },
221
  {
222
  "epoch": 0.12,
223
  "learning_rate": 0.001,
224
+ "loss": 5.9614,
225
  "step": 180
226
  },
227
  {
228
  "epoch": 0.12,
229
  "learning_rate": 0.001,
230
+ "loss": 5.9394,
231
  "step": 185
232
  },
233
  {
234
  "epoch": 0.12,
235
  "learning_rate": 0.001,
236
+ "loss": 5.8261,
237
  "step": 190
238
  },
239
  {
240
  "epoch": 0.13,
241
  "learning_rate": 0.001,
242
+ "loss": 6.2232,
243
  "step": 195
244
  },
245
  {
246
  "epoch": 0.13,
247
  "learning_rate": 0.001,
248
+ "loss": 6.0747,
249
  "step": 200
250
  },
251
  {
252
  "epoch": 0.13,
253
  "learning_rate": 0.001,
254
+ "loss": 5.9842,
255
  "step": 205
256
  },
257
  {
258
  "epoch": 0.14,
259
  "learning_rate": 0.001,
260
+ "loss": 6.2139,
261
  "step": 210
262
  },
263
  {
264
  "epoch": 0.14,
265
  "learning_rate": 0.001,
266
+ "loss": 6.0711,
267
  "step": 215
268
  },
269
  {
270
  "epoch": 0.14,
271
  "learning_rate": 0.001,
272
+ "loss": 5.9545,
273
  "step": 220
274
  },
275
  {
276
  "epoch": 0.15,
277
  "learning_rate": 0.001,
278
+ "loss": 5.7681,
279
  "step": 225
280
  },
281
  {
282
  "epoch": 0.15,
283
  "learning_rate": 0.001,
284
+ "loss": 5.8673,
285
  "step": 230
286
  },
287
  {
288
  "epoch": 0.15,
289
  "learning_rate": 0.001,
290
+ "loss": 5.6984,
291
  "step": 235
292
  },
293
  {
294
  "epoch": 0.16,
295
  "learning_rate": 0.001,
296
+ "loss": 5.6021,
297
  "step": 240
298
  },
299
  {
300
  "epoch": 0.16,
301
  "learning_rate": 0.001,
302
+ "loss": 5.8764,
303
  "step": 245
304
  },
305
  {
306
  "epoch": 0.16,
307
  "learning_rate": 0.001,
308
+ "loss": 5.6526,
309
  "step": 250
310
  },
311
  {
312
  "epoch": 0.17,
313
  "learning_rate": 0.001,
314
+ "loss": 5.7407,
315
  "step": 255
316
  },
317
  {
318
  "epoch": 0.17,
319
  "learning_rate": 0.001,
320
+ "loss": 5.6711,
321
  "step": 260
322
  },
323
  {
324
  "epoch": 0.17,
325
  "learning_rate": 0.001,
326
+ "loss": 5.6802,
327
  "step": 265
328
  },
329
  {
330
  "epoch": 0.18,
331
  "learning_rate": 0.001,
332
+ "loss": 5.7385,
333
  "step": 270
334
  },
335
  {
336
  "epoch": 0.18,
337
  "learning_rate": 0.001,
338
+ "loss": 5.8663,
339
  "step": 275
340
  },
341
  {
342
  "epoch": 0.18,
343
  "learning_rate": 0.001,
344
+ "loss": 5.5553,
345
  "step": 280
346
  },
347
  {
348
  "epoch": 0.19,
349
  "learning_rate": 0.001,
350
+ "loss": 5.6528,
351
  "step": 285
352
  },
353
  {
354
  "epoch": 0.19,
355
  "learning_rate": 0.001,
356
+ "loss": 5.8758,
357
  "step": 290
358
  },
359
  {
360
  "epoch": 0.19,
361
  "learning_rate": 0.001,
362
+ "loss": 5.513,
363
  "step": 295
364
  },
365
  {
366
  "epoch": 0.2,
367
  "learning_rate": 0.001,
368
+ "loss": 5.3608,
369
  "step": 300
370
  },
371
  {
372
  "epoch": 0.2,
373
  "learning_rate": 0.001,
374
+ "loss": 5.7339,
375
  "step": 305
376
  },
377
  {
378
  "epoch": 0.2,
379
  "learning_rate": 0.001,
380
+ "loss": 5.5918,
381
  "step": 310
382
  },
383
  {
384
  "epoch": 0.21,
385
  "learning_rate": 0.001,
386
+ "loss": 5.7015,
387
  "step": 315
388
  },
389
  {
390
  "epoch": 0.21,
391
  "learning_rate": 0.001,
392
+ "loss": 5.5825,
393
  "step": 320
394
  },
395
  {
396
  "epoch": 0.21,
397
  "learning_rate": 0.001,
398
+ "loss": 5.3423,
399
  "step": 325
400
  },
401
  {
402
  "epoch": 0.22,
403
  "learning_rate": 0.001,
404
+ "loss": 5.2627,
405
  "step": 330
406
  },
407
  {
408
  "epoch": 0.22,
409
  "learning_rate": 0.001,
410
+ "loss": 5.4692,
411
  "step": 335
412
  },
413
  {
414
  "epoch": 0.22,
415
  "learning_rate": 0.001,
416
+ "loss": 5.474,
417
  "step": 340
418
  },
419
  {
420
  "epoch": 0.23,
421
  "learning_rate": 0.001,
422
+ "loss": 5.3762,
423
  "step": 345
424
  },
425
  {
426
  "epoch": 0.23,
427
  "learning_rate": 0.001,
428
+ "loss": 5.2908,
429
  "step": 350
430
  },
431
  {
432
  "epoch": 0.23,
433
  "learning_rate": 0.001,
434
+ "loss": 5.4183,
435
  "step": 355
436
  },
437
  {
438
  "epoch": 0.24,
439
  "learning_rate": 0.001,
440
+ "loss": 5.2073,
441
  "step": 360
442
  },
443
  {
444
  "epoch": 0.24,
445
  "learning_rate": 0.001,
446
+ "loss": 5.4598,
447
  "step": 365
448
  },
449
  {
450
  "epoch": 0.24,
451
  "learning_rate": 0.001,
452
+ "loss": 5.4487,
453
  "step": 370
454
  },
455
  {
456
  "epoch": 0.25,
457
  "learning_rate": 0.001,
458
+ "loss": 5.2604,
459
  "step": 375
460
  },
461
  {
462
  "epoch": 0.25,
463
  "learning_rate": 0.001,
464
+ "loss": 5.1575,
465
  "step": 380
466
  },
467
  {
468
  "epoch": 0.25,
469
  "learning_rate": 0.001,
470
+ "loss": 5.2809,
471
  "step": 385
472
  },
473
  {
474
  "epoch": 0.26,
475
  "learning_rate": 0.001,
476
+ "loss": 5.1923,
477
  "step": 390
478
  },
479
  {
480
  "epoch": 0.26,
481
  "learning_rate": 0.001,
482
+ "loss": 5.3494,
483
  "step": 395
484
  },
485
  {
486
  "epoch": 0.26,
487
  "learning_rate": 0.001,
488
+ "loss": 5.2338,
489
  "step": 400
490
  },
491
  {
492
  "epoch": 0.27,
493
  "learning_rate": 0.001,
494
+ "loss": 5.2143,
495
  "step": 405
496
  },
497
  {
498
  "epoch": 0.27,
499
  "learning_rate": 0.001,
500
+ "loss": 5.3615,
501
  "step": 410
502
  },
503
  {
504
  "epoch": 0.27,
505
  "learning_rate": 0.001,
506
+ "loss": 5.4357,
507
  "step": 415
508
  },
509
  {
510
  "epoch": 0.27,
511
  "learning_rate": 0.001,
512
+ "loss": 5.1693,
513
  "step": 420
514
  },
515
  {
516
  "epoch": 0.28,
517
  "learning_rate": 0.001,
518
+ "loss": 5.1072,
519
  "step": 425
520
  },
521
  {
522
  "epoch": 0.28,
523
  "learning_rate": 0.001,
524
+ "loss": 5.2298,
525
  "step": 430
526
  },
527
  {
528
  "epoch": 0.28,
529
  "learning_rate": 0.001,
530
+ "loss": 4.9306,
531
  "step": 435
532
  },
533
  {
534
  "epoch": 0.29,
535
  "learning_rate": 0.001,
536
+ "loss": 5.321,
537
  "step": 440
538
  },
539
  {
540
  "epoch": 0.29,
541
  "learning_rate": 0.001,
542
+ "loss": 4.9341,
543
  "step": 445
544
  },
545
  {
546
  "epoch": 0.29,
547
  "learning_rate": 0.001,
548
+ "loss": 5.0898,
549
  "step": 450
550
  },
551
  {
552
  "epoch": 0.3,
553
  "learning_rate": 0.001,
554
+ "loss": 5.1687,
555
  "step": 455
556
  },
557
  {
558
  "epoch": 0.3,
559
  "learning_rate": 0.001,
560
+ "loss": 4.8764,
561
  "step": 460
562
  },
563
  {
564
  "epoch": 0.3,
565
  "learning_rate": 0.001,
566
+ "loss": 5.1905,
567
  "step": 465
568
  },
569
  {
570
  "epoch": 0.31,
571
  "learning_rate": 0.001,
572
+ "loss": 5.0816,
573
  "step": 470
574
  },
575
  {
576
  "epoch": 0.31,
577
  "learning_rate": 0.001,
578
+ "loss": 5.2568,
579
  "step": 475
580
  },
581
  {
582
  "epoch": 0.31,
583
  "learning_rate": 0.001,
584
+ "loss": 5.2742,
585
  "step": 480
586
  },
587
  {
588
  "epoch": 0.32,
589
  "learning_rate": 0.001,
590
+ "loss": 5.0711,
591
  "step": 485
592
  },
593
  {
594
  "epoch": 0.32,
595
  "learning_rate": 0.001,
596
+ "loss": 5.0671,
597
  "step": 490
598
  },
599
  {
600
  "epoch": 0.32,
601
  "learning_rate": 0.001,
602
+ "loss": 5.1016,
603
  "step": 495
604
  },
605
  {
606
  "epoch": 0.33,
607
  "learning_rate": 0.001,
608
+ "loss": 5.3311,
609
  "step": 500
610
  },
611
  {
612
  "epoch": 0.33,
613
  "learning_rate": 0.001,
614
+ "loss": 5.2327,
615
  "step": 505
616
  },
617
  {
618
  "epoch": 0.33,
619
  "learning_rate": 0.001,
620
+ "loss": 5.3222,
621
  "step": 510
622
  },
623
  {
624
  "epoch": 0.34,
625
  "learning_rate": 0.001,
626
+ "loss": 4.9495,
627
  "step": 515
628
  },
629
  {
630
  "epoch": 0.34,
631
  "learning_rate": 0.001,
632
+ "loss": 4.957,
633
  "step": 520
634
  },
635
  {
636
  "epoch": 0.34,
637
  "learning_rate": 0.001,
638
+ "loss": 5.3251,
639
  "step": 525
640
  },
641
  {
642
  "epoch": 0.35,
643
  "learning_rate": 0.001,
644
+ "loss": 4.6873,
645
  "step": 530
646
  },
647
  {
648
  "epoch": 0.35,
649
  "learning_rate": 0.001,
650
+ "loss": 5.289,
651
  "step": 535
652
  },
653
  {
654
  "epoch": 0.35,
655
  "learning_rate": 0.001,
656
+ "loss": 5.0009,
657
  "step": 540
658
  },
659
  {
660
  "epoch": 0.36,
661
  "learning_rate": 0.001,
662
+ "loss": 4.8415,
663
  "step": 545
664
  },
665
  {
666
  "epoch": 0.36,
667
  "learning_rate": 0.001,
668
+ "loss": 5.2063,
669
  "step": 550
670
  },
671
  {
672
  "epoch": 0.36,
673
  "learning_rate": 0.001,
674
+ "loss": 4.8384,
675
  "step": 555
676
  },
677
  {
678
  "epoch": 0.37,
679
  "learning_rate": 0.001,
680
+ "loss": 4.3815,
681
  "step": 560
682
  },
683
  {
684
  "epoch": 0.37,
685
  "learning_rate": 0.001,
686
+ "loss": 4.9327,
687
  "step": 565
688
  },
689
  {
690
  "epoch": 0.37,
691
  "learning_rate": 0.001,
692
+ "loss": 4.8049,
693
  "step": 570
694
  },
695
  {
696
  "epoch": 0.38,
697
  "learning_rate": 0.001,
698
+ "loss": 4.9983,
699
  "step": 575
700
  },
701
  {
702
  "epoch": 0.38,
703
  "learning_rate": 0.001,
704
+ "loss": 4.9048,
705
  "step": 580
706
  },
707
  {
708
  "epoch": 0.38,
709
  "learning_rate": 0.001,
710
+ "loss": 4.8615,
711
  "step": 585
712
  },
713
  {
714
  "epoch": 0.39,
715
  "learning_rate": 0.001,
716
+ "loss": 4.8885,
717
  "step": 590
718
  },
719
  {
720
  "epoch": 0.39,
721
  "learning_rate": 0.001,
722
+ "loss": 4.9189,
723
  "step": 595
724
  },
725
  {
726
  "epoch": 0.39,
727
  "learning_rate": 0.001,
728
+ "loss": 4.537,
729
  "step": 600
730
  },
731
  {
732
  "epoch": 0.4,
733
  "learning_rate": 0.001,
734
+ "loss": 4.5421,
735
  "step": 605
736
  },
737
  {
738
  "epoch": 0.4,
739
  "learning_rate": 0.001,
740
+ "loss": 4.8229,
741
  "step": 610
742
  },
743
  {
744
  "epoch": 0.4,
745
  "learning_rate": 0.001,
746
+ "loss": 4.9116,
747
  "step": 615
748
  },
749
  {
750
  "epoch": 0.41,
751
  "learning_rate": 0.001,
752
+ "loss": 4.9846,
753
  "step": 620
754
  },
755
  {
756
  "epoch": 0.41,
757
  "learning_rate": 0.001,
758
+ "loss": 4.7512,
759
  "step": 625
760
  },
761
  {
762
  "epoch": 0.41,
763
  "learning_rate": 0.001,
764
+ "loss": 4.9718,
765
  "step": 630
766
  },
767
  {
768
  "epoch": 0.42,
769
  "learning_rate": 0.001,
770
+ "loss": 4.9162,
771
  "step": 635
772
  },
773
  {
774
  "epoch": 0.42,
775
  "learning_rate": 0.001,
776
+ "loss": 4.5091,
777
  "step": 640
778
  },
779
  {
780
  "epoch": 0.42,
781
  "learning_rate": 0.001,
782
+ "loss": 4.8638,
783
  "step": 645
784
  },
785
  {
786
  "epoch": 0.43,
787
  "learning_rate": 0.001,
788
+ "loss": 4.5735,
789
  "step": 650
790
  },
791
  {
792
  "epoch": 0.43,
793
  "learning_rate": 0.001,
794
+ "loss": 4.8966,
795
  "step": 655
796
  },
797
  {
798
  "epoch": 0.43,
799
  "learning_rate": 0.001,
800
+ "loss": 4.9169,
801
  "step": 660
802
  },
803
  {
804
  "epoch": 0.44,
805
  "learning_rate": 0.001,
806
+ "loss": 4.8778,
807
  "step": 665
808
  },
809
  {
810
  "epoch": 0.44,
811
  "learning_rate": 0.001,
812
+ "loss": 4.8185,
813
  "step": 670
814
  },
815
  {
816
  "epoch": 0.44,
817
  "learning_rate": 0.001,
818
+ "loss": 4.6256,
819
  "step": 675
820
  },
821
  {
822
  "epoch": 0.45,
823
  "learning_rate": 0.001,
824
+ "loss": 4.7773,
825
  "step": 680
826
  },
827
  {
828
  "epoch": 0.45,
829
  "learning_rate": 0.001,
830
+ "loss": 4.6482,
831
  "step": 685
832
  },
833
  {
834
  "epoch": 0.45,
835
  "learning_rate": 0.001,
836
+ "loss": 4.3789,
837
  "step": 690
838
  },
839
  {
840
  "epoch": 0.45,
841
  "learning_rate": 0.001,
842
+ "loss": 4.849,
843
  "step": 695
844
  },
845
  {
846
  "epoch": 0.46,
847
  "learning_rate": 0.001,
848
+ "loss": 4.7127,
849
  "step": 700
850
  },
851
  {
852
  "epoch": 0.46,
853
  "learning_rate": 0.001,
854
+ "loss": 4.6414,
855
  "step": 705
856
  },
857
  {
858
  "epoch": 0.46,
859
  "learning_rate": 0.001,
860
+ "loss": 4.512,
861
  "step": 710
862
  },
863
  {
864
  "epoch": 0.47,
865
  "learning_rate": 0.001,
866
+ "loss": 4.5855,
867
  "step": 715
868
  },
869
  {
870
  "epoch": 0.47,
871
  "learning_rate": 0.001,
872
+ "loss": 4.5004,
873
  "step": 720
874
  },
875
  {
876
  "epoch": 0.47,
877
  "learning_rate": 0.001,
878
+ "loss": 4.8402,
879
  "step": 725
880
  },
881
  {
882
  "epoch": 0.48,
883
  "learning_rate": 0.001,
884
+ "loss": 4.7665,
885
  "step": 730
886
  },
887
  {
888
  "epoch": 0.48,
889
  "learning_rate": 0.001,
890
+ "loss": 4.7962,
891
  "step": 735
892
  },
893
  {
894
  "epoch": 0.48,
895
  "learning_rate": 0.001,
896
+ "loss": 4.7931,
897
  "step": 740
898
  },
899
  {
900
  "epoch": 0.49,
901
  "learning_rate": 0.001,
902
+ "loss": 4.7293,
903
  "step": 745
904
  },
905
  {
906
  "epoch": 0.49,
907
  "learning_rate": 0.001,
908
+ "loss": 4.6017,
909
  "step": 750
910
  },
911
  {
912
  "epoch": 0.49,
913
  "learning_rate": 0.001,
914
+ "loss": 4.7313,
915
  "step": 755
916
  },
917
  {
918
  "epoch": 0.5,
919
  "learning_rate": 0.001,
920
+ "loss": 4.58,
921
  "step": 760
922
  },
923
  {
924
  "epoch": 0.5,
925
  "learning_rate": 0.001,
926
+ "loss": 4.6188,
927
  "step": 765
928
  },
929
  {
930
  "epoch": 0.5,
931
  "learning_rate": 0.001,
932
+ "loss": 4.5284,
933
  "step": 770
934
  },
935
  {
936
  "epoch": 0.51,
937
  "learning_rate": 0.001,
938
+ "loss": 4.5003,
939
  "step": 775
940
  },
941
  {
942
  "epoch": 0.51,
943
  "learning_rate": 0.001,
944
+ "loss": 4.5868,
945
  "step": 780
946
  },
947
  {
948
  "epoch": 0.51,
949
  "learning_rate": 0.001,
950
+ "loss": 4.5261,
951
  "step": 785
952
  },
953
  {
954
  "epoch": 0.52,
955
  "learning_rate": 0.001,
956
+ "loss": 4.6488,
957
  "step": 790
958
  },
959
  {
960
  "epoch": 0.52,
961
  "learning_rate": 0.001,
962
+ "loss": 4.2441,
963
  "step": 795
964
  },
965
  {
966
  "epoch": 0.52,
967
  "learning_rate": 0.001,
968
+ "loss": 4.9201,
969
  "step": 800
970
  },
971
  {
972
  "epoch": 0.53,
973
  "learning_rate": 0.001,
974
+ "loss": 4.5616,
975
  "step": 805
976
  },
977
  {
978
  "epoch": 0.53,
979
  "learning_rate": 0.001,
980
+ "loss": 4.4575,
981
  "step": 810
982
  },
983
  {
984
  "epoch": 0.53,
985
  "learning_rate": 0.001,
986
+ "loss": 4.4801,
987
  "step": 815
988
  },
989
  {
990
  "epoch": 0.54,
991
  "learning_rate": 0.001,
992
+ "loss": 4.4761,
993
  "step": 820
994
  },
995
  {
996
  "epoch": 0.54,
997
  "learning_rate": 0.001,
998
+ "loss": 4.7145,
999
  "step": 825
1000
  },
1001
  {
1002
  "epoch": 0.54,
1003
  "learning_rate": 0.001,
1004
+ "loss": 4.6759,
1005
  "step": 830
1006
  },
1007
  {
1008
  "epoch": 0.55,
1009
  "learning_rate": 0.001,
1010
+ "loss": 4.3177,
1011
  "step": 835
1012
  },
1013
  {
1014
  "epoch": 0.55,
1015
  "learning_rate": 0.001,
1016
+ "loss": 4.4118,
1017
  "step": 840
1018
  },
1019
  {
1020
  "epoch": 0.55,
1021
  "learning_rate": 0.001,
1022
+ "loss": 4.4915,
1023
  "step": 845
1024
  },
1025
  {
1026
  "epoch": 0.56,
1027
  "learning_rate": 0.001,
1028
+ "loss": 4.6066,
1029
  "step": 850
1030
  },
1031
  {
1032
  "epoch": 0.56,
1033
  "learning_rate": 0.001,
1034
+ "loss": 4.6189,
1035
  "step": 855
1036
  },
1037
  {
1038
  "epoch": 0.56,
1039
  "learning_rate": 0.001,
1040
+ "loss": 4.4977,
1041
  "step": 860
1042
  },
1043
  {
1044
  "epoch": 0.57,
1045
  "learning_rate": 0.001,
1046
+ "loss": 4.2084,
1047
  "step": 865
1048
  },
1049
  {
1050
  "epoch": 0.57,
1051
  "learning_rate": 0.001,
1052
+ "loss": 4.4593,
1053
  "step": 870
1054
  },
1055
  {
1056
  "epoch": 0.57,
1057
  "learning_rate": 0.001,
1058
+ "loss": 4.4324,
1059
  "step": 875
1060
  },
1061
  {
1062
  "epoch": 0.58,
1063
  "learning_rate": 0.001,
1064
+ "loss": 4.5696,
1065
  "step": 880
1066
  },
1067
  {
1068
  "epoch": 0.58,
1069
  "learning_rate": 0.001,
1070
+ "loss": 4.2641,
1071
  "step": 885
1072
  },
1073
  {
1074
  "epoch": 0.58,
1075
  "learning_rate": 0.001,
1076
+ "loss": 4.2248,
1077
  "step": 890
1078
  },
1079
  {
1080
  "epoch": 0.59,
1081
  "learning_rate": 0.001,
1082
+ "loss": 4.577,
1083
  "step": 895
1084
  },
1085
  {
1086
  "epoch": 0.59,
1087
  "learning_rate": 0.001,
1088
+ "loss": 4.3748,
1089
  "step": 900
1090
  },
1091
  {
1092
  "epoch": 0.59,
1093
  "learning_rate": 0.001,
1094
+ "loss": 4.4647,
1095
  "step": 905
1096
  },
1097
  {
1098
  "epoch": 0.6,
1099
  "learning_rate": 0.001,
1100
+ "loss": 4.3836,
1101
  "step": 910
1102
  },
1103
  {
1104
  "epoch": 0.6,
1105
  "learning_rate": 0.001,
1106
+ "loss": 4.7384,
1107
  "step": 915
1108
  },
1109
  {
1110
  "epoch": 0.6,
1111
  "learning_rate": 0.001,
1112
+ "loss": 4.3023,
1113
  "step": 920
1114
  },
1115
  {
1116
  "epoch": 0.61,
1117
  "learning_rate": 0.001,
1118
+ "loss": 4.2472,
1119
  "step": 925
1120
  },
1121
  {
1122
  "epoch": 0.61,
1123
  "learning_rate": 0.001,
1124
+ "loss": 4.6622,
1125
  "step": 930
1126
  },
1127
  {
1128
  "epoch": 0.61,
1129
  "learning_rate": 0.001,
1130
+ "loss": 4.5409,
1131
  "step": 935
1132
  },
1133
  {
1134
  "epoch": 0.62,
1135
  "learning_rate": 0.001,
1136
+ "loss": 3.9355,
1137
  "step": 940
1138
  },
1139
  {
1140
  "epoch": 0.62,
1141
  "learning_rate": 0.001,
1142
+ "loss": 4.4457,
1143
  "step": 945
1144
  },
1145
  {
1146
  "epoch": 0.62,
1147
  "learning_rate": 0.001,
1148
+ "loss": 4.4636,
1149
  "step": 950
1150
  },
1151
  {
1152
  "epoch": 0.62,
1153
  "learning_rate": 0.001,
1154
+ "loss": 3.7582,
1155
  "step": 955
1156
  },
1157
  {
1158
  "epoch": 0.63,
1159
  "learning_rate": 0.001,
1160
+ "loss": 4.4637,
1161
  "step": 960
1162
  },
1163
  {
1164
  "epoch": 0.63,
1165
  "learning_rate": 0.001,
1166
+ "loss": 4.4201,
1167
  "step": 965
1168
  },
1169
  {
1170
  "epoch": 0.63,
1171
  "learning_rate": 0.001,
1172
+ "loss": 3.9001,
1173
  "step": 970
1174
  },
1175
  {
1176
  "epoch": 0.64,
1177
  "learning_rate": 0.001,
1178
+ "loss": 4.1943,
1179
  "step": 975
1180
  },
1181
  {
1182
  "epoch": 0.64,
1183
  "learning_rate": 0.001,
1184
+ "loss": 4.1346,
1185
  "step": 980
1186
  },
1187
  {
1188
  "epoch": 0.64,
1189
  "learning_rate": 0.001,
1190
+ "loss": 4.057,
1191
  "step": 985
1192
  },
1193
  {
1194
  "epoch": 0.65,
1195
  "learning_rate": 0.001,
1196
+ "loss": 4.1786,
1197
  "step": 990
1198
  },
1199
  {
1200
  "epoch": 0.65,
1201
  "learning_rate": 0.001,
1202
+ "loss": 4.0282,
1203
  "step": 995
1204
  },
1205
  {
1206
  "epoch": 0.65,
1207
  "learning_rate": 0.001,
1208
+ "loss": 4.2299,
1209
  "step": 1000
1210
  }
1211
  ],
1212
  "logging_steps": 5,
1213
+ "max_steps": 1528,
1214
+ "num_train_epochs": 1,
1215
  "save_steps": 1000,
1216
+ "total_flos": 76806881280000.0,
1217
  "trial_name": null,
1218
  "trial_params": null
1219
  }
checkpoint-1000/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:026a40a22d79b976af4f05a529a919434f1a1af96981147d87e1c857264d40d7
3
  size 4027
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b462423a1a93a6e821517c0dc21398bf6ebcf0fa763dc09d72a23319257dcb4
3
  size 4027
config.json CHANGED
@@ -1,18 +1,18 @@
1
  {
2
- "_name_or_path": "./output5",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
6
  "bos_token_id": 1,
7
  "eos_token_id": 2,
8
  "hidden_act": "silu",
9
- "hidden_size": 256,
10
  "initializer_range": 0.02,
11
- "intermediate_size": 512,
12
- "max_position_embeddings": 8192,
13
  "model_type": "llama",
14
  "num_attention_heads": 4,
15
- "num_hidden_layers": 2,
16
  "num_key_value_heads": 4,
17
  "pretraining_tp": 2,
18
  "rms_norm_eps": 1e-05,
 
1
  {
2
+ "_name_or_path": "./model",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
6
  "bos_token_id": 1,
7
  "eos_token_id": 2,
8
  "hidden_act": "silu",
9
+ "hidden_size": 64,
10
  "initializer_range": 0.02,
11
+ "intermediate_size": 256,
12
+ "max_position_embeddings": 2048,
13
  "model_type": "llama",
14
  "num_attention_heads": 4,
15
+ "num_hidden_layers": 64,
16
  "num_key_value_heads": 4,
17
  "pretraining_tp": 2,
18
  "rms_norm_eps": 1e-05,
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4813035bd07556e93f50e8d269519bf9e2b59a15f2265677f99ab8809de94dc0
3
- size 70790896
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2e602f32a52b66f8373757839c5af315585c930fd659e3b4a9c59aa05decfc3
3
+ size 33387545
tokenizer_config.json CHANGED
@@ -29,7 +29,7 @@
29
  "bos_token": "<s>",
30
  "clean_up_tokenization_spaces": false,
31
  "eos_token": "</s>",
32
- "model_max_length": 8192,
33
  "pad_token": null,
34
  "sp_model_kwargs": {},
35
  "tokenizer_class": "LlamaTokenizer",
 
29
  "bos_token": "<s>",
30
  "clean_up_tokenization_spaces": false,
31
  "eos_token": "</s>",
32
+ "model_max_length": 2048,
33
  "pad_token": null,
34
  "sp_model_kwargs": {},
35
  "tokenizer_class": "LlamaTokenizer",
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 5.0,
3
- "train_loss": 1.999870525792007,
4
- "train_runtime": 268.6642,
5
  "train_samples": 1528,
6
- "train_samples_per_second": 28.437,
7
- "train_steps_per_second": 28.437
8
  }
 
1
  {
2
+ "epoch": 1.0,
3
+ "train_loss": 4.918852043401508,
4
+ "train_runtime": 510.3139,
5
  "train_samples": 1528,
6
+ "train_samples_per_second": 2.994,
7
+ "train_steps_per_second": 2.994
8
  }
trainer_state.json CHANGED
The diff for this file is too large to render. See raw diff
 
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:026a40a22d79b976af4f05a529a919434f1a1af96981147d87e1c857264d40d7
3
  size 4027
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b462423a1a93a6e821517c0dc21398bf6ebcf0fa763dc09d72a23319257dcb4
3
  size 4027