TinyPixel commited on
Commit
91ca22d
·
1 Parent(s): 2e25354

Upload folder using huggingface_hub

Browse files
adapter_config.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "auto_mapping": null,
3
- "base_model_name_or_path": "TinyPixel/redpajama-3b",
4
  "bias": "none",
5
  "fan_in_fan_out": false,
6
  "inference_mode": true,
@@ -14,10 +14,13 @@
14
  "r": 64,
15
  "revision": null,
16
  "target_modules": [
17
- "query_key_value",
18
- "dense",
19
- "dense_h_to_4h",
20
- "dense_4h_to_h"
 
 
 
21
  ],
22
  "task_type": "CAUSAL_LM"
23
  }
 
1
  {
2
  "auto_mapping": null,
3
+ "base_model_name_or_path": "TinyPixel/xgen-7b-4k-base-bf16-sharded",
4
  "bias": "none",
5
  "fan_in_fan_out": false,
6
  "inference_mode": true,
 
14
  "r": 64,
15
  "revision": null,
16
  "target_modules": [
17
+ "q_proj",
18
+ "up_proj",
19
+ "o_proj",
20
+ "k_proj",
21
+ "down_proj",
22
+ "gate_proj",
23
+ "v_proj"
24
  ],
25
  "task_type": "CAUSAL_LM"
26
  }
adapter_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7094f02097200fdc54c9dbe3dfa1fadd638d96096c07c065371fe6060c663b50
3
- size 335638221
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:727660330ea89b6697e56f2d72b1e9847bd9f3bc6e0d1a21279f5e86c28ef427
3
+ size 639792909
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:dca4086515bce75dcc8de82803510594c581f52868340b2fca4e5ccc9b1162ba
3
- size 671246853
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76f5ede7082c5897a50c985061c950b3a3c129d28b6e7add97b0969ab86eb352
3
+ size 1279539525
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:24c53aa6670010ef242c58d65904b8436c3809be37fa51341d65d63083dd6ddd
3
  size 14575
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:370bc9f640f0de057c4e91d360be8438630526997dd2b34dc8255e5aaeb210f7
3
  size 14575
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:aa8119b2a07925a797c9177dfb4076165e404eab5a0d14d455cd6c97c5af167c
3
  size 627
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f31903bd4cc4e6b3475b7fd5bead33ff7dd0e4da616a2713beabd1ca422d2d06
3
  size 627
special_tokens_map.json CHANGED
@@ -1,6 +1,10 @@
1
  {
2
- "bos_token": "<|endoftext|>",
3
- "eos_token": "<|endoftext|>",
4
- "pad_token": "<|endoftext|>",
5
- "unk_token": "<|endoftext|>"
 
 
 
 
6
  }
 
1
  {
2
+ "eos_token": {
3
+ "content": "<|endoftext|>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "pad_token": "<|endoftext|>"
10
  }
tokenizer_config.json CHANGED
@@ -1,9 +1,22 @@
1
  {
2
- "add_prefix_space": false,
3
- "bos_token": "<|endoftext|>",
 
 
 
 
 
 
4
  "clean_up_tokenization_spaces": true,
5
- "eos_token": "<|endoftext|>",
6
- "model_max_length": 2048,
7
- "tokenizer_class": "GPTNeoXTokenizer",
8
- "unk_token": "<|endoftext|>"
 
 
 
 
 
 
 
9
  }
 
1
  {
2
+ "add_eos_token": false,
3
+ "add_special_tokens": true,
4
+ "auto_map": {
5
+ "AutoTokenizer": [
6
+ "Salesforce/xgen-7b-4k-base--tokenization_xgen.XgenTokenizer",
7
+ null
8
+ ]
9
+ },
10
  "clean_up_tokenization_spaces": true,
11
+ "eos_token": {
12
+ "__type": "AddedToken",
13
+ "content": "<|endoftext|>",
14
+ "lstrip": false,
15
+ "normalized": true,
16
+ "rstrip": false,
17
+ "single_word": false
18
+ },
19
+ "model_max_length": 4096,
20
+ "pad_token": null,
21
+ "tokenizer_class": "XgenTokenizer"
22
  }
trainer_state.json CHANGED
@@ -1,1816 +1,166 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 0.07471805608524085,
5
- "global_step": 600,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
  "epoch": 0.0,
12
- "learning_rate": 5.532503457814662e-08,
13
- "loss": 1.305,
14
  "step": 2
15
  },
16
  {
17
- "epoch": 0.0,
18
- "learning_rate": 1.1065006915629324e-07,
19
- "loss": 1.2619,
20
  "step": 4
21
  },
22
  {
23
- "epoch": 0.0,
24
- "learning_rate": 1.6597510373443985e-07,
25
- "loss": 1.4923,
26
  "step": 6
27
  },
28
  {
29
- "epoch": 0.0,
30
- "learning_rate": 2.2130013831258647e-07,
31
- "loss": 1.5682,
32
  "step": 8
33
  },
34
  {
35
- "epoch": 0.0,
36
- "learning_rate": 2.766251728907331e-07,
37
- "loss": 1.6158,
38
  "step": 10
39
  },
40
  {
41
- "epoch": 0.0,
42
- "learning_rate": 3.319502074688797e-07,
43
- "loss": 1.6714,
44
  "step": 12
45
  },
46
  {
47
- "epoch": 0.0,
48
- "learning_rate": 3.872752420470263e-07,
49
- "loss": 1.6915,
50
  "step": 14
51
  },
52
  {
53
- "epoch": 0.0,
54
- "learning_rate": 4.4260027662517294e-07,
55
- "loss": 1.7026,
56
  "step": 16
57
  },
58
  {
59
- "epoch": 0.0,
60
- "learning_rate": 4.979253112033195e-07,
61
- "loss": 1.6197,
62
  "step": 18
63
  },
64
  {
65
- "epoch": 0.0,
66
- "learning_rate": 5.532503457814662e-07,
67
- "loss": 1.7253,
68
  "step": 20
69
  },
70
  {
71
- "epoch": 0.0,
72
- "learning_rate": 6.085753803596127e-07,
73
- "loss": 1.7767,
74
  "step": 22
75
  },
76
  {
77
- "epoch": 0.0,
78
- "learning_rate": 6.639004149377594e-07,
79
- "loss": 1.7851,
80
  "step": 24
81
  },
82
  {
83
- "epoch": 0.0,
84
- "learning_rate": 7.19225449515906e-07,
85
- "loss": 1.8175,
86
  "step": 26
87
  },
88
  {
89
- "epoch": 0.0,
90
- "learning_rate": 7.745504840940527e-07,
91
- "loss": 1.9565,
92
  "step": 28
93
  },
94
  {
95
- "epoch": 0.0,
96
- "learning_rate": 8.298755186721992e-07,
97
- "loss": 1.9234,
98
  "step": 30
99
  },
100
  {
101
- "epoch": 0.0,
102
- "learning_rate": 8.852005532503459e-07,
103
- "loss": 2.0439,
104
  "step": 32
105
  },
106
  {
107
- "epoch": 0.0,
108
- "learning_rate": 9.405255878284925e-07,
109
- "loss": 2.0397,
110
  "step": 34
111
  },
112
  {
113
- "epoch": 0.0,
114
- "learning_rate": 9.95850622406639e-07,
115
- "loss": 2.2069,
116
  "step": 36
117
  },
118
  {
119
- "epoch": 0.0,
120
- "learning_rate": 1.0511756569847857e-06,
121
- "loss": 2.2489,
122
  "step": 38
123
  },
124
  {
125
- "epoch": 0.0,
126
- "learning_rate": 1.1065006915629324e-06,
127
- "loss": 2.3121,
128
  "step": 40
129
  },
130
  {
131
- "epoch": 0.01,
132
- "learning_rate": 1.161825726141079e-06,
133
- "loss": 2.5232,
134
  "step": 42
135
  },
136
  {
137
- "epoch": 0.01,
138
- "learning_rate": 1.2171507607192255e-06,
139
- "loss": 2.725,
140
  "step": 44
141
  },
142
  {
143
- "epoch": 0.01,
144
- "learning_rate": 1.2724757952973722e-06,
145
- "loss": 3.0845,
146
  "step": 46
147
  },
148
  {
149
- "epoch": 0.01,
150
- "learning_rate": 1.3278008298755188e-06,
151
- "loss": 3.2097,
152
  "step": 48
153
  },
154
  {
155
- "epoch": 0.01,
156
- "learning_rate": 1.3831258644536655e-06,
157
- "loss": 3.3337,
158
  "step": 50
159
- },
160
- {
161
- "epoch": 0.01,
162
- "learning_rate": 1.438450899031812e-06,
163
- "loss": 1.3678,
164
- "step": 52
165
- },
166
- {
167
- "epoch": 0.01,
168
- "learning_rate": 1.4937759336099586e-06,
169
- "loss": 1.3633,
170
- "step": 54
171
- },
172
- {
173
- "epoch": 0.01,
174
- "learning_rate": 1.5491009681881053e-06,
175
- "loss": 1.4763,
176
- "step": 56
177
- },
178
- {
179
- "epoch": 0.01,
180
- "learning_rate": 1.604426002766252e-06,
181
- "loss": 1.6015,
182
- "step": 58
183
- },
184
- {
185
- "epoch": 0.01,
186
- "learning_rate": 1.6597510373443984e-06,
187
- "loss": 1.5423,
188
- "step": 60
189
- },
190
- {
191
- "epoch": 0.01,
192
- "learning_rate": 1.715076071922545e-06,
193
- "loss": 1.5975,
194
- "step": 62
195
- },
196
- {
197
- "epoch": 0.01,
198
- "learning_rate": 1.7704011065006918e-06,
199
- "loss": 1.7265,
200
- "step": 64
201
- },
202
- {
203
- "epoch": 0.01,
204
- "learning_rate": 1.8257261410788382e-06,
205
- "loss": 1.6917,
206
- "step": 66
207
- },
208
- {
209
- "epoch": 0.01,
210
- "learning_rate": 1.881051175656985e-06,
211
- "loss": 1.6335,
212
- "step": 68
213
- },
214
- {
215
- "epoch": 0.01,
216
- "learning_rate": 1.936376210235132e-06,
217
- "loss": 1.6513,
218
- "step": 70
219
- },
220
- {
221
- "epoch": 0.01,
222
- "learning_rate": 1.991701244813278e-06,
223
- "loss": 1.6958,
224
- "step": 72
225
- },
226
- {
227
- "epoch": 0.01,
228
- "learning_rate": 2.0470262793914247e-06,
229
- "loss": 1.7838,
230
- "step": 74
231
- },
232
- {
233
- "epoch": 0.01,
234
- "learning_rate": 2.1023513139695714e-06,
235
- "loss": 1.8354,
236
- "step": 76
237
- },
238
- {
239
- "epoch": 0.01,
240
- "learning_rate": 2.157676348547718e-06,
241
- "loss": 1.8646,
242
- "step": 78
243
- },
244
- {
245
- "epoch": 0.01,
246
- "learning_rate": 2.2130013831258647e-06,
247
- "loss": 1.9077,
248
- "step": 80
249
- },
250
- {
251
- "epoch": 0.01,
252
- "learning_rate": 2.2683264177040114e-06,
253
- "loss": 1.887,
254
- "step": 82
255
- },
256
- {
257
- "epoch": 0.01,
258
- "learning_rate": 2.323651452282158e-06,
259
- "loss": 2.129,
260
- "step": 84
261
- },
262
- {
263
- "epoch": 0.01,
264
- "learning_rate": 2.3789764868603043e-06,
265
- "loss": 2.1029,
266
- "step": 86
267
- },
268
- {
269
- "epoch": 0.01,
270
- "learning_rate": 2.434301521438451e-06,
271
- "loss": 2.2457,
272
- "step": 88
273
- },
274
- {
275
- "epoch": 0.01,
276
- "learning_rate": 2.4896265560165977e-06,
277
- "loss": 2.327,
278
- "step": 90
279
- },
280
- {
281
- "epoch": 0.01,
282
- "learning_rate": 2.5449515905947443e-06,
283
- "loss": 2.4578,
284
- "step": 92
285
- },
286
- {
287
- "epoch": 0.01,
288
- "learning_rate": 2.600276625172891e-06,
289
- "loss": 2.6359,
290
- "step": 94
291
- },
292
- {
293
- "epoch": 0.01,
294
- "learning_rate": 2.6556016597510377e-06,
295
- "loss": 2.9305,
296
- "step": 96
297
- },
298
- {
299
- "epoch": 0.01,
300
- "learning_rate": 2.710926694329184e-06,
301
- "loss": 3.1006,
302
- "step": 98
303
- },
304
- {
305
- "epoch": 0.01,
306
- "learning_rate": 2.766251728907331e-06,
307
- "loss": 3.2059,
308
- "step": 100
309
- },
310
- {
311
- "epoch": 0.01,
312
- "learning_rate": 2.8215767634854773e-06,
313
- "loss": 1.3058,
314
- "step": 102
315
- },
316
- {
317
- "epoch": 0.01,
318
- "learning_rate": 2.876901798063624e-06,
319
- "loss": 1.3116,
320
- "step": 104
321
- },
322
- {
323
- "epoch": 0.01,
324
- "learning_rate": 2.9322268326417706e-06,
325
- "loss": 1.3985,
326
- "step": 106
327
- },
328
- {
329
- "epoch": 0.01,
330
- "learning_rate": 2.9875518672199173e-06,
331
- "loss": 1.5591,
332
- "step": 108
333
- },
334
- {
335
- "epoch": 0.01,
336
- "learning_rate": 3.0428769017980635e-06,
337
- "loss": 1.621,
338
- "step": 110
339
- },
340
- {
341
- "epoch": 0.01,
342
- "learning_rate": 3.0982019363762106e-06,
343
- "loss": 1.4749,
344
- "step": 112
345
- },
346
- {
347
- "epoch": 0.01,
348
- "learning_rate": 3.153526970954357e-06,
349
- "loss": 1.6038,
350
- "step": 114
351
- },
352
- {
353
- "epoch": 0.01,
354
- "learning_rate": 3.208852005532504e-06,
355
- "loss": 1.6227,
356
- "step": 116
357
- },
358
- {
359
- "epoch": 0.01,
360
- "learning_rate": 3.26417704011065e-06,
361
- "loss": 1.5667,
362
- "step": 118
363
- },
364
- {
365
- "epoch": 0.01,
366
- "learning_rate": 3.319502074688797e-06,
367
- "loss": 1.7112,
368
- "step": 120
369
- },
370
- {
371
- "epoch": 0.02,
372
- "learning_rate": 3.374827109266944e-06,
373
- "loss": 1.7435,
374
- "step": 122
375
- },
376
- {
377
- "epoch": 0.02,
378
- "learning_rate": 3.43015214384509e-06,
379
- "loss": 1.7608,
380
- "step": 124
381
- },
382
- {
383
- "epoch": 0.02,
384
- "learning_rate": 3.4854771784232365e-06,
385
- "loss": 1.7508,
386
- "step": 126
387
- },
388
- {
389
- "epoch": 0.02,
390
- "learning_rate": 3.5408022130013836e-06,
391
- "loss": 1.7105,
392
- "step": 128
393
- },
394
- {
395
- "epoch": 0.02,
396
- "learning_rate": 3.5961272475795302e-06,
397
- "loss": 1.8219,
398
- "step": 130
399
- },
400
- {
401
- "epoch": 0.02,
402
- "learning_rate": 3.6514522821576765e-06,
403
- "loss": 1.9559,
404
- "step": 132
405
- },
406
- {
407
- "epoch": 0.02,
408
- "learning_rate": 3.7067773167358236e-06,
409
- "loss": 1.9084,
410
- "step": 134
411
- },
412
- {
413
- "epoch": 0.02,
414
- "learning_rate": 3.76210235131397e-06,
415
- "loss": 2.1159,
416
- "step": 136
417
- },
418
- {
419
- "epoch": 0.02,
420
- "learning_rate": 3.817427385892116e-06,
421
- "loss": 2.0861,
422
- "step": 138
423
- },
424
- {
425
- "epoch": 0.02,
426
- "learning_rate": 3.872752420470264e-06,
427
- "loss": 2.1838,
428
- "step": 140
429
- },
430
- {
431
- "epoch": 0.02,
432
- "learning_rate": 3.928077455048409e-06,
433
- "loss": 2.3962,
434
- "step": 142
435
- },
436
- {
437
- "epoch": 0.02,
438
- "learning_rate": 3.983402489626556e-06,
439
- "loss": 2.6263,
440
- "step": 144
441
- },
442
- {
443
- "epoch": 0.02,
444
- "learning_rate": 4.038727524204703e-06,
445
- "loss": 2.6236,
446
- "step": 146
447
- },
448
- {
449
- "epoch": 0.02,
450
- "learning_rate": 4.094052558782849e-06,
451
- "loss": 2.8199,
452
- "step": 148
453
- },
454
- {
455
- "epoch": 0.02,
456
- "learning_rate": 4.149377593360996e-06,
457
- "loss": 3.0848,
458
- "step": 150
459
- },
460
- {
461
- "epoch": 0.02,
462
- "learning_rate": 4.204702627939143e-06,
463
- "loss": 1.1621,
464
- "step": 152
465
- },
466
- {
467
- "epoch": 0.02,
468
- "learning_rate": 4.2600276625172894e-06,
469
- "loss": 1.3043,
470
- "step": 154
471
- },
472
- {
473
- "epoch": 0.02,
474
- "learning_rate": 4.315352697095436e-06,
475
- "loss": 1.4692,
476
- "step": 156
477
- },
478
- {
479
- "epoch": 0.02,
480
- "learning_rate": 4.370677731673583e-06,
481
- "loss": 1.4176,
482
- "step": 158
483
- },
484
- {
485
- "epoch": 0.02,
486
- "learning_rate": 4.4260027662517294e-06,
487
- "loss": 1.4805,
488
- "step": 160
489
- },
490
- {
491
- "epoch": 0.02,
492
- "learning_rate": 4.481327800829876e-06,
493
- "loss": 1.4791,
494
- "step": 162
495
- },
496
- {
497
- "epoch": 0.02,
498
- "learning_rate": 4.536652835408023e-06,
499
- "loss": 1.4865,
500
- "step": 164
501
- },
502
- {
503
- "epoch": 0.02,
504
- "learning_rate": 4.591977869986169e-06,
505
- "loss": 1.5246,
506
- "step": 166
507
- },
508
- {
509
- "epoch": 0.02,
510
- "learning_rate": 4.647302904564316e-06,
511
- "loss": 1.5192,
512
- "step": 168
513
- },
514
- {
515
- "epoch": 0.02,
516
- "learning_rate": 4.702627939142462e-06,
517
- "loss": 1.6564,
518
- "step": 170
519
- },
520
- {
521
- "epoch": 0.02,
522
- "learning_rate": 4.757952973720609e-06,
523
- "loss": 1.6141,
524
- "step": 172
525
- },
526
- {
527
- "epoch": 0.02,
528
- "learning_rate": 4.813278008298755e-06,
529
- "loss": 1.6751,
530
- "step": 174
531
- },
532
- {
533
- "epoch": 0.02,
534
- "learning_rate": 4.868603042876902e-06,
535
- "loss": 1.5923,
536
- "step": 176
537
- },
538
- {
539
- "epoch": 0.02,
540
- "learning_rate": 4.9239280774550495e-06,
541
- "loss": 1.6443,
542
- "step": 178
543
- },
544
- {
545
- "epoch": 0.02,
546
- "learning_rate": 4.979253112033195e-06,
547
- "loss": 1.6464,
548
- "step": 180
549
- },
550
- {
551
- "epoch": 0.02,
552
- "learning_rate": 5.034578146611342e-06,
553
- "loss": 1.6753,
554
- "step": 182
555
- },
556
- {
557
- "epoch": 0.02,
558
- "learning_rate": 5.089903181189489e-06,
559
- "loss": 1.6779,
560
- "step": 184
561
- },
562
- {
563
- "epoch": 0.02,
564
- "learning_rate": 5.1452282157676345e-06,
565
- "loss": 1.8974,
566
- "step": 186
567
- },
568
- {
569
- "epoch": 0.02,
570
- "learning_rate": 5.200553250345782e-06,
571
- "loss": 1.8848,
572
- "step": 188
573
- },
574
- {
575
- "epoch": 0.02,
576
- "learning_rate": 5.255878284923929e-06,
577
- "loss": 2.0583,
578
- "step": 190
579
- },
580
- {
581
- "epoch": 0.02,
582
- "learning_rate": 5.311203319502075e-06,
583
- "loss": 1.9503,
584
- "step": 192
585
- },
586
- {
587
- "epoch": 0.02,
588
- "learning_rate": 5.366528354080221e-06,
589
- "loss": 2.2817,
590
- "step": 194
591
- },
592
- {
593
- "epoch": 0.02,
594
- "learning_rate": 5.421853388658368e-06,
595
- "loss": 2.242,
596
- "step": 196
597
- },
598
- {
599
- "epoch": 0.02,
600
- "learning_rate": 5.477178423236515e-06,
601
- "loss": 2.2948,
602
- "step": 198
603
- },
604
- {
605
- "epoch": 0.02,
606
- "learning_rate": 5.532503457814662e-06,
607
- "loss": 2.593,
608
- "step": 200
609
- },
610
- {
611
- "epoch": 0.03,
612
- "learning_rate": 5.587828492392809e-06,
613
- "loss": 1.126,
614
- "step": 202
615
- },
616
- {
617
- "epoch": 0.03,
618
- "learning_rate": 5.6431535269709545e-06,
619
- "loss": 1.3656,
620
- "step": 204
621
- },
622
- {
623
- "epoch": 0.03,
624
- "learning_rate": 5.698478561549101e-06,
625
- "loss": 1.3054,
626
- "step": 206
627
- },
628
- {
629
- "epoch": 0.03,
630
- "learning_rate": 5.753803596127248e-06,
631
- "loss": 1.2756,
632
- "step": 208
633
- },
634
- {
635
- "epoch": 0.03,
636
- "learning_rate": 5.809128630705395e-06,
637
- "loss": 1.4594,
638
- "step": 210
639
- },
640
- {
641
- "epoch": 0.03,
642
- "learning_rate": 5.864453665283541e-06,
643
- "loss": 1.4164,
644
- "step": 212
645
- },
646
- {
647
- "epoch": 0.03,
648
- "learning_rate": 5.919778699861688e-06,
649
- "loss": 1.6161,
650
- "step": 214
651
- },
652
- {
653
- "epoch": 0.03,
654
- "learning_rate": 5.9751037344398345e-06,
655
- "loss": 1.5366,
656
- "step": 216
657
- },
658
- {
659
- "epoch": 0.03,
660
- "learning_rate": 6.030428769017981e-06,
661
- "loss": 1.3777,
662
- "step": 218
663
- },
664
- {
665
- "epoch": 0.03,
666
- "learning_rate": 6.085753803596127e-06,
667
- "loss": 1.5156,
668
- "step": 220
669
- },
670
- {
671
- "epoch": 0.03,
672
- "learning_rate": 6.1410788381742745e-06,
673
- "loss": 1.4669,
674
- "step": 222
675
- },
676
- {
677
- "epoch": 0.03,
678
- "learning_rate": 6.196403872752421e-06,
679
- "loss": 1.4233,
680
- "step": 224
681
- },
682
- {
683
- "epoch": 0.03,
684
- "learning_rate": 6.251728907330568e-06,
685
- "loss": 1.4206,
686
- "step": 226
687
- },
688
- {
689
- "epoch": 0.03,
690
- "learning_rate": 6.307053941908714e-06,
691
- "loss": 1.4126,
692
- "step": 228
693
- },
694
- {
695
- "epoch": 0.03,
696
- "learning_rate": 6.36237897648686e-06,
697
- "loss": 1.531,
698
- "step": 230
699
- },
700
- {
701
- "epoch": 0.03,
702
- "learning_rate": 6.417704011065008e-06,
703
- "loss": 1.4993,
704
- "step": 232
705
- },
706
- {
707
- "epoch": 0.03,
708
- "learning_rate": 6.4730290456431546e-06,
709
- "loss": 1.5724,
710
- "step": 234
711
- },
712
- {
713
- "epoch": 0.03,
714
- "learning_rate": 6.5283540802213e-06,
715
- "loss": 1.4972,
716
- "step": 236
717
- },
718
- {
719
- "epoch": 0.03,
720
- "learning_rate": 6.583679114799447e-06,
721
- "loss": 1.566,
722
- "step": 238
723
- },
724
- {
725
- "epoch": 0.03,
726
- "learning_rate": 6.639004149377594e-06,
727
- "loss": 1.649,
728
- "step": 240
729
- },
730
- {
731
- "epoch": 0.03,
732
- "learning_rate": 6.69432918395574e-06,
733
- "loss": 1.7134,
734
- "step": 242
735
- },
736
- {
737
- "epoch": 0.03,
738
- "learning_rate": 6.749654218533888e-06,
739
- "loss": 1.7004,
740
- "step": 244
741
- },
742
- {
743
- "epoch": 0.03,
744
- "learning_rate": 6.804979253112034e-06,
745
- "loss": 1.6889,
746
- "step": 246
747
- },
748
- {
749
- "epoch": 0.03,
750
- "learning_rate": 6.86030428769018e-06,
751
- "loss": 1.5918,
752
- "step": 248
753
- },
754
- {
755
- "epoch": 0.03,
756
- "learning_rate": 6.915629322268327e-06,
757
- "loss": 2.1735,
758
- "step": 250
759
- },
760
- {
761
- "epoch": 0.03,
762
- "learning_rate": 6.970954356846473e-06,
763
- "loss": 1.1666,
764
- "step": 252
765
- },
766
- {
767
- "epoch": 0.03,
768
- "learning_rate": 7.02627939142462e-06,
769
- "loss": 1.0931,
770
- "step": 254
771
- },
772
- {
773
- "epoch": 0.03,
774
- "learning_rate": 7.081604426002767e-06,
775
- "loss": 1.2902,
776
- "step": 256
777
- },
778
- {
779
- "epoch": 0.03,
780
- "learning_rate": 7.136929460580914e-06,
781
- "loss": 1.2619,
782
- "step": 258
783
- },
784
- {
785
- "epoch": 0.03,
786
- "learning_rate": 7.1922544951590604e-06,
787
- "loss": 1.3443,
788
- "step": 260
789
- },
790
- {
791
- "epoch": 0.03,
792
- "learning_rate": 7.247579529737206e-06,
793
- "loss": 1.2994,
794
- "step": 262
795
- },
796
- {
797
- "epoch": 0.03,
798
- "learning_rate": 7.302904564315353e-06,
799
- "loss": 1.3426,
800
- "step": 264
801
- },
802
- {
803
- "epoch": 0.03,
804
- "learning_rate": 7.3582295988935005e-06,
805
- "loss": 1.364,
806
- "step": 266
807
- },
808
- {
809
- "epoch": 0.03,
810
- "learning_rate": 7.413554633471647e-06,
811
- "loss": 1.3707,
812
- "step": 268
813
- },
814
- {
815
- "epoch": 0.03,
816
- "learning_rate": 7.468879668049793e-06,
817
- "loss": 1.3551,
818
- "step": 270
819
- },
820
- {
821
- "epoch": 0.03,
822
- "learning_rate": 7.52420470262794e-06,
823
- "loss": 1.3057,
824
- "step": 272
825
- },
826
- {
827
- "epoch": 0.03,
828
- "learning_rate": 7.579529737206086e-06,
829
- "loss": 1.2717,
830
- "step": 274
831
- },
832
- {
833
- "epoch": 0.03,
834
- "learning_rate": 7.634854771784232e-06,
835
- "loss": 1.4112,
836
- "step": 276
837
- },
838
- {
839
- "epoch": 0.03,
840
- "learning_rate": 7.69017980636238e-06,
841
- "loss": 1.3604,
842
- "step": 278
843
- },
844
- {
845
- "epoch": 0.03,
846
- "learning_rate": 7.745504840940527e-06,
847
- "loss": 1.4114,
848
- "step": 280
849
- },
850
- {
851
- "epoch": 0.04,
852
- "learning_rate": 7.800829875518673e-06,
853
- "loss": 1.3892,
854
- "step": 282
855
- },
856
- {
857
- "epoch": 0.04,
858
- "learning_rate": 7.856154910096819e-06,
859
- "loss": 1.4112,
860
- "step": 284
861
- },
862
- {
863
- "epoch": 0.04,
864
- "learning_rate": 7.911479944674966e-06,
865
- "loss": 1.5537,
866
- "step": 286
867
- },
868
- {
869
- "epoch": 0.04,
870
- "learning_rate": 7.966804979253112e-06,
871
- "loss": 1.4472,
872
- "step": 288
873
- },
874
- {
875
- "epoch": 0.04,
876
- "learning_rate": 8.02213001383126e-06,
877
- "loss": 1.5369,
878
- "step": 290
879
- },
880
- {
881
- "epoch": 0.04,
882
- "learning_rate": 8.077455048409405e-06,
883
- "loss": 1.512,
884
- "step": 292
885
- },
886
- {
887
- "epoch": 0.04,
888
- "learning_rate": 8.132780082987553e-06,
889
- "loss": 1.4981,
890
- "step": 294
891
- },
892
- {
893
- "epoch": 0.04,
894
- "learning_rate": 8.188105117565699e-06,
895
- "loss": 1.4316,
896
- "step": 296
897
- },
898
- {
899
- "epoch": 0.04,
900
- "learning_rate": 8.243430152143845e-06,
901
- "loss": 1.6334,
902
- "step": 298
903
- },
904
- {
905
- "epoch": 0.04,
906
- "learning_rate": 8.298755186721992e-06,
907
- "loss": 2.4904,
908
- "step": 300
909
- },
910
- {
911
- "epoch": 0.04,
912
- "learning_rate": 8.35408022130014e-06,
913
- "loss": 1.0349,
914
- "step": 302
915
- },
916
- {
917
- "epoch": 0.04,
918
- "learning_rate": 8.409405255878286e-06,
919
- "loss": 1.1243,
920
- "step": 304
921
- },
922
- {
923
- "epoch": 0.04,
924
- "learning_rate": 8.464730290456431e-06,
925
- "loss": 1.2724,
926
- "step": 306
927
- },
928
- {
929
- "epoch": 0.04,
930
- "learning_rate": 8.520055325034579e-06,
931
- "loss": 1.2668,
932
- "step": 308
933
- },
934
- {
935
- "epoch": 0.04,
936
- "learning_rate": 8.575380359612725e-06,
937
- "loss": 1.4017,
938
- "step": 310
939
- },
940
- {
941
- "epoch": 0.04,
942
- "learning_rate": 8.630705394190872e-06,
943
- "loss": 1.2285,
944
- "step": 312
945
- },
946
- {
947
- "epoch": 0.04,
948
- "learning_rate": 8.68603042876902e-06,
949
- "loss": 1.3107,
950
- "step": 314
951
- },
952
- {
953
- "epoch": 0.04,
954
- "learning_rate": 8.741355463347166e-06,
955
- "loss": 1.2918,
956
- "step": 316
957
- },
958
- {
959
- "epoch": 0.04,
960
- "learning_rate": 8.796680497925311e-06,
961
- "loss": 1.3214,
962
- "step": 318
963
- },
964
- {
965
- "epoch": 0.04,
966
- "learning_rate": 8.852005532503459e-06,
967
- "loss": 1.2502,
968
- "step": 320
969
- },
970
- {
971
- "epoch": 0.04,
972
- "learning_rate": 8.907330567081606e-06,
973
- "loss": 1.2628,
974
- "step": 322
975
- },
976
- {
977
- "epoch": 0.04,
978
- "learning_rate": 8.962655601659752e-06,
979
- "loss": 1.1691,
980
- "step": 324
981
- },
982
- {
983
- "epoch": 0.04,
984
- "learning_rate": 9.017980636237898e-06,
985
- "loss": 1.1576,
986
- "step": 326
987
- },
988
- {
989
- "epoch": 0.04,
990
- "learning_rate": 9.073305670816046e-06,
991
- "loss": 1.3607,
992
- "step": 328
993
- },
994
- {
995
- "epoch": 0.04,
996
- "learning_rate": 9.128630705394191e-06,
997
- "loss": 1.2192,
998
- "step": 330
999
- },
1000
- {
1001
- "epoch": 0.04,
1002
- "learning_rate": 9.183955739972337e-06,
1003
- "loss": 1.3484,
1004
- "step": 332
1005
- },
1006
- {
1007
- "epoch": 0.04,
1008
- "learning_rate": 9.239280774550485e-06,
1009
- "loss": 1.3634,
1010
- "step": 334
1011
- },
1012
- {
1013
- "epoch": 0.04,
1014
- "learning_rate": 9.294605809128632e-06,
1015
- "loss": 1.3289,
1016
- "step": 336
1017
- },
1018
- {
1019
- "epoch": 0.04,
1020
- "learning_rate": 9.349930843706778e-06,
1021
- "loss": 1.4129,
1022
- "step": 338
1023
- },
1024
- {
1025
- "epoch": 0.04,
1026
- "learning_rate": 9.405255878284924e-06,
1027
- "loss": 1.1766,
1028
- "step": 340
1029
- },
1030
- {
1031
- "epoch": 0.04,
1032
- "learning_rate": 9.460580912863071e-06,
1033
- "loss": 1.4086,
1034
- "step": 342
1035
- },
1036
- {
1037
- "epoch": 0.04,
1038
- "learning_rate": 9.515905947441217e-06,
1039
- "loss": 1.2979,
1040
- "step": 344
1041
- },
1042
- {
1043
- "epoch": 0.04,
1044
- "learning_rate": 9.571230982019365e-06,
1045
- "loss": 1.1013,
1046
- "step": 346
1047
- },
1048
- {
1049
- "epoch": 0.04,
1050
- "learning_rate": 9.62655601659751e-06,
1051
- "loss": 1.2951,
1052
- "step": 348
1053
- },
1054
- {
1055
- "epoch": 0.04,
1056
- "learning_rate": 9.681881051175658e-06,
1057
- "loss": 1.9857,
1058
- "step": 350
1059
- },
1060
- {
1061
- "epoch": 0.04,
1062
- "learning_rate": 9.737206085753804e-06,
1063
- "loss": 0.9824,
1064
- "step": 352
1065
- },
1066
- {
1067
- "epoch": 0.04,
1068
- "learning_rate": 9.792531120331951e-06,
1069
- "loss": 1.0624,
1070
- "step": 354
1071
- },
1072
- {
1073
- "epoch": 0.04,
1074
- "learning_rate": 9.847856154910099e-06,
1075
- "loss": 1.1545,
1076
- "step": 356
1077
- },
1078
- {
1079
- "epoch": 0.04,
1080
- "learning_rate": 9.903181189488245e-06,
1081
- "loss": 1.2113,
1082
- "step": 358
1083
- },
1084
- {
1085
- "epoch": 0.04,
1086
- "learning_rate": 9.95850622406639e-06,
1087
- "loss": 1.2999,
1088
- "step": 360
1089
- },
1090
- {
1091
- "epoch": 0.05,
1092
- "learning_rate": 1.0013831258644538e-05,
1093
- "loss": 1.4667,
1094
- "step": 362
1095
- },
1096
- {
1097
- "epoch": 0.05,
1098
- "learning_rate": 1.0069156293222684e-05,
1099
- "loss": 1.3276,
1100
- "step": 364
1101
- },
1102
- {
1103
- "epoch": 0.05,
1104
- "learning_rate": 1.012448132780083e-05,
1105
- "loss": 1.2589,
1106
- "step": 366
1107
- },
1108
- {
1109
- "epoch": 0.05,
1110
- "learning_rate": 1.0179806362378977e-05,
1111
- "loss": 1.252,
1112
- "step": 368
1113
- },
1114
- {
1115
- "epoch": 0.05,
1116
- "learning_rate": 1.0235131396957123e-05,
1117
- "loss": 1.1554,
1118
- "step": 370
1119
- },
1120
- {
1121
- "epoch": 0.05,
1122
- "learning_rate": 1.0290456431535269e-05,
1123
- "loss": 1.1559,
1124
- "step": 372
1125
- },
1126
- {
1127
- "epoch": 0.05,
1128
- "learning_rate": 1.0345781466113418e-05,
1129
- "loss": 1.2338,
1130
- "step": 374
1131
- },
1132
- {
1133
- "epoch": 0.05,
1134
- "learning_rate": 1.0401106500691564e-05,
1135
- "loss": 1.1119,
1136
- "step": 376
1137
- },
1138
- {
1139
- "epoch": 0.05,
1140
- "learning_rate": 1.0456431535269711e-05,
1141
- "loss": 1.1909,
1142
- "step": 378
1143
- },
1144
- {
1145
- "epoch": 0.05,
1146
- "learning_rate": 1.0511756569847857e-05,
1147
- "loss": 1.1185,
1148
- "step": 380
1149
- },
1150
- {
1151
- "epoch": 0.05,
1152
- "learning_rate": 1.0567081604426003e-05,
1153
- "loss": 1.2531,
1154
- "step": 382
1155
- },
1156
- {
1157
- "epoch": 0.05,
1158
- "learning_rate": 1.062240663900415e-05,
1159
- "loss": 1.2135,
1160
- "step": 384
1161
- },
1162
- {
1163
- "epoch": 0.05,
1164
- "learning_rate": 1.0677731673582296e-05,
1165
- "loss": 1.2216,
1166
- "step": 386
1167
- },
1168
- {
1169
- "epoch": 0.05,
1170
- "learning_rate": 1.0733056708160442e-05,
1171
- "loss": 1.2253,
1172
- "step": 388
1173
- },
1174
- {
1175
- "epoch": 0.05,
1176
- "learning_rate": 1.078838174273859e-05,
1177
- "loss": 1.2166,
1178
- "step": 390
1179
- },
1180
- {
1181
- "epoch": 0.05,
1182
- "learning_rate": 1.0843706777316736e-05,
1183
- "loss": 1.1889,
1184
- "step": 392
1185
- },
1186
- {
1187
- "epoch": 0.05,
1188
- "learning_rate": 1.0899031811894883e-05,
1189
- "loss": 0.9758,
1190
- "step": 394
1191
- },
1192
- {
1193
- "epoch": 0.05,
1194
- "learning_rate": 1.095435684647303e-05,
1195
- "loss": 0.9412,
1196
- "step": 396
1197
- },
1198
- {
1199
- "epoch": 0.05,
1200
- "learning_rate": 1.1009681881051178e-05,
1201
- "loss": 1.1546,
1202
- "step": 398
1203
- },
1204
- {
1205
- "epoch": 0.05,
1206
- "learning_rate": 1.1065006915629324e-05,
1207
- "loss": 1.7417,
1208
- "step": 400
1209
- },
1210
- {
1211
- "epoch": 0.05,
1212
- "learning_rate": 1.112033195020747e-05,
1213
- "loss": 1.0201,
1214
- "step": 402
1215
- },
1216
- {
1217
- "epoch": 0.05,
1218
- "learning_rate": 1.1175656984785617e-05,
1219
- "loss": 0.986,
1220
- "step": 404
1221
- },
1222
- {
1223
- "epoch": 0.05,
1224
- "learning_rate": 1.1230982019363763e-05,
1225
- "loss": 1.0251,
1226
- "step": 406
1227
- },
1228
- {
1229
- "epoch": 0.05,
1230
- "learning_rate": 1.1286307053941909e-05,
1231
- "loss": 1.1938,
1232
- "step": 408
1233
- },
1234
- {
1235
- "epoch": 0.05,
1236
- "learning_rate": 1.1341632088520057e-05,
1237
- "loss": 1.2092,
1238
- "step": 410
1239
- },
1240
- {
1241
- "epoch": 0.05,
1242
- "learning_rate": 1.1396957123098202e-05,
1243
- "loss": 1.2018,
1244
- "step": 412
1245
- },
1246
- {
1247
- "epoch": 0.05,
1248
- "learning_rate": 1.1452282157676348e-05,
1249
- "loss": 1.2467,
1250
- "step": 414
1251
- },
1252
- {
1253
- "epoch": 0.05,
1254
- "learning_rate": 1.1507607192254496e-05,
1255
- "loss": 1.2304,
1256
- "step": 416
1257
- },
1258
- {
1259
- "epoch": 0.05,
1260
- "learning_rate": 1.1562932226832643e-05,
1261
- "loss": 1.128,
1262
- "step": 418
1263
- },
1264
- {
1265
- "epoch": 0.05,
1266
- "learning_rate": 1.161825726141079e-05,
1267
- "loss": 1.1305,
1268
- "step": 420
1269
- },
1270
- {
1271
- "epoch": 0.05,
1272
- "learning_rate": 1.1673582295988937e-05,
1273
- "loss": 1.0733,
1274
- "step": 422
1275
- },
1276
- {
1277
- "epoch": 0.05,
1278
- "learning_rate": 1.1728907330567082e-05,
1279
- "loss": 1.1259,
1280
- "step": 424
1281
- },
1282
- {
1283
- "epoch": 0.05,
1284
- "learning_rate": 1.178423236514523e-05,
1285
- "loss": 1.0519,
1286
- "step": 426
1287
- },
1288
- {
1289
- "epoch": 0.05,
1290
- "learning_rate": 1.1839557399723376e-05,
1291
- "loss": 1.1726,
1292
- "step": 428
1293
- },
1294
- {
1295
- "epoch": 0.05,
1296
- "learning_rate": 1.1894882434301522e-05,
1297
- "loss": 1.0113,
1298
- "step": 430
1299
- },
1300
- {
1301
- "epoch": 0.05,
1302
- "learning_rate": 1.1950207468879669e-05,
1303
- "loss": 1.1983,
1304
- "step": 432
1305
- },
1306
- {
1307
- "epoch": 0.05,
1308
- "learning_rate": 1.2005532503457815e-05,
1309
- "loss": 1.0579,
1310
- "step": 434
1311
- },
1312
- {
1313
- "epoch": 0.05,
1314
- "learning_rate": 1.2060857538035962e-05,
1315
- "loss": 1.1345,
1316
- "step": 436
1317
- },
1318
- {
1319
- "epoch": 0.05,
1320
- "learning_rate": 1.2116182572614108e-05,
1321
- "loss": 1.0332,
1322
- "step": 438
1323
- },
1324
- {
1325
- "epoch": 0.05,
1326
- "learning_rate": 1.2171507607192254e-05,
1327
- "loss": 1.0807,
1328
- "step": 440
1329
- },
1330
- {
1331
- "epoch": 0.06,
1332
- "learning_rate": 1.2226832641770403e-05,
1333
- "loss": 1.0151,
1334
- "step": 442
1335
- },
1336
- {
1337
- "epoch": 0.06,
1338
- "learning_rate": 1.2282157676348549e-05,
1339
- "loss": 1.1693,
1340
- "step": 444
1341
- },
1342
- {
1343
- "epoch": 0.06,
1344
- "learning_rate": 1.2337482710926697e-05,
1345
- "loss": 0.9797,
1346
- "step": 446
1347
- },
1348
- {
1349
- "epoch": 0.06,
1350
- "learning_rate": 1.2392807745504842e-05,
1351
- "loss": 1.3063,
1352
- "step": 448
1353
- },
1354
- {
1355
- "epoch": 0.06,
1356
- "learning_rate": 1.2448132780082988e-05,
1357
- "loss": 1.6451,
1358
- "step": 450
1359
- },
1360
- {
1361
- "epoch": 0.06,
1362
- "learning_rate": 1.2503457814661136e-05,
1363
- "loss": 1.0108,
1364
- "step": 452
1365
- },
1366
- {
1367
- "epoch": 0.06,
1368
- "learning_rate": 1.2558782849239282e-05,
1369
- "loss": 1.0353,
1370
- "step": 454
1371
- },
1372
- {
1373
- "epoch": 0.06,
1374
- "learning_rate": 1.2614107883817427e-05,
1375
- "loss": 1.146,
1376
- "step": 456
1377
- },
1378
- {
1379
- "epoch": 0.06,
1380
- "learning_rate": 1.2669432918395575e-05,
1381
- "loss": 1.1282,
1382
- "step": 458
1383
- },
1384
- {
1385
- "epoch": 0.06,
1386
- "learning_rate": 1.272475795297372e-05,
1387
- "loss": 1.1537,
1388
- "step": 460
1389
- },
1390
- {
1391
- "epoch": 0.06,
1392
- "learning_rate": 1.2780082987551867e-05,
1393
- "loss": 1.1784,
1394
- "step": 462
1395
- },
1396
- {
1397
- "epoch": 0.06,
1398
- "learning_rate": 1.2835408022130016e-05,
1399
- "loss": 1.3489,
1400
- "step": 464
1401
- },
1402
- {
1403
- "epoch": 0.06,
1404
- "learning_rate": 1.2890733056708162e-05,
1405
- "loss": 1.1302,
1406
- "step": 466
1407
- },
1408
- {
1409
- "epoch": 0.06,
1410
- "learning_rate": 1.2946058091286309e-05,
1411
- "loss": 1.239,
1412
- "step": 468
1413
- },
1414
- {
1415
- "epoch": 0.06,
1416
- "learning_rate": 1.3001383125864455e-05,
1417
- "loss": 1.147,
1418
- "step": 470
1419
- },
1420
- {
1421
- "epoch": 0.06,
1422
- "learning_rate": 1.30567081604426e-05,
1423
- "loss": 1.2085,
1424
- "step": 472
1425
- },
1426
- {
1427
- "epoch": 0.06,
1428
- "learning_rate": 1.3112033195020748e-05,
1429
- "loss": 1.088,
1430
- "step": 474
1431
- },
1432
- {
1433
- "epoch": 0.06,
1434
- "learning_rate": 1.3167358229598894e-05,
1435
- "loss": 1.1087,
1436
- "step": 476
1437
- },
1438
- {
1439
- "epoch": 0.06,
1440
- "learning_rate": 1.3222683264177042e-05,
1441
- "loss": 1.1567,
1442
- "step": 478
1443
- },
1444
- {
1445
- "epoch": 0.06,
1446
- "learning_rate": 1.3278008298755187e-05,
1447
- "loss": 1.1561,
1448
- "step": 480
1449
- },
1450
- {
1451
- "epoch": 0.06,
1452
- "learning_rate": 1.3333333333333333e-05,
1453
- "loss": 1.0608,
1454
- "step": 482
1455
- },
1456
- {
1457
- "epoch": 0.06,
1458
- "learning_rate": 1.338865836791148e-05,
1459
- "loss": 1.0994,
1460
- "step": 484
1461
- },
1462
- {
1463
- "epoch": 0.06,
1464
- "learning_rate": 1.3443983402489628e-05,
1465
- "loss": 1.1067,
1466
- "step": 486
1467
- },
1468
- {
1469
- "epoch": 0.06,
1470
- "learning_rate": 1.3499308437067776e-05,
1471
- "loss": 1.2688,
1472
- "step": 488
1473
- },
1474
- {
1475
- "epoch": 0.06,
1476
- "learning_rate": 1.3554633471645922e-05,
1477
- "loss": 1.215,
1478
- "step": 490
1479
- },
1480
- {
1481
- "epoch": 0.06,
1482
- "learning_rate": 1.3609958506224067e-05,
1483
- "loss": 1.253,
1484
- "step": 492
1485
- },
1486
- {
1487
- "epoch": 0.06,
1488
- "learning_rate": 1.3665283540802215e-05,
1489
- "loss": 1.0932,
1490
- "step": 494
1491
- },
1492
- {
1493
- "epoch": 0.06,
1494
- "learning_rate": 1.372060857538036e-05,
1495
- "loss": 1.1944,
1496
- "step": 496
1497
- },
1498
- {
1499
- "epoch": 0.06,
1500
- "learning_rate": 1.3775933609958507e-05,
1501
- "loss": 1.0259,
1502
- "step": 498
1503
- },
1504
- {
1505
- "epoch": 0.06,
1506
- "learning_rate": 1.3831258644536654e-05,
1507
- "loss": 1.512,
1508
- "step": 500
1509
- },
1510
- {
1511
- "epoch": 0.06,
1512
- "learning_rate": 1.38865836791148e-05,
1513
- "loss": 1.0774,
1514
- "step": 502
1515
- },
1516
- {
1517
- "epoch": 0.06,
1518
- "learning_rate": 1.3941908713692946e-05,
1519
- "loss": 1.0722,
1520
- "step": 504
1521
- },
1522
- {
1523
- "epoch": 0.06,
1524
- "learning_rate": 1.3997233748271093e-05,
1525
- "loss": 1.125,
1526
- "step": 506
1527
- },
1528
- {
1529
- "epoch": 0.06,
1530
- "learning_rate": 1.405255878284924e-05,
1531
- "loss": 1.0927,
1532
- "step": 508
1533
- },
1534
- {
1535
- "epoch": 0.06,
1536
- "learning_rate": 1.4107883817427388e-05,
1537
- "loss": 1.1557,
1538
- "step": 510
1539
- },
1540
- {
1541
- "epoch": 0.06,
1542
- "learning_rate": 1.4163208852005534e-05,
1543
- "loss": 1.2796,
1544
- "step": 512
1545
- },
1546
- {
1547
- "epoch": 0.06,
1548
- "learning_rate": 1.421853388658368e-05,
1549
- "loss": 1.3694,
1550
- "step": 514
1551
- },
1552
- {
1553
- "epoch": 0.06,
1554
- "learning_rate": 1.4273858921161828e-05,
1555
- "loss": 1.1705,
1556
- "step": 516
1557
- },
1558
- {
1559
- "epoch": 0.06,
1560
- "learning_rate": 1.4329183955739973e-05,
1561
- "loss": 1.15,
1562
- "step": 518
1563
- },
1564
- {
1565
- "epoch": 0.06,
1566
- "learning_rate": 1.4384508990318121e-05,
1567
- "loss": 1.2782,
1568
- "step": 520
1569
- },
1570
- {
1571
- "epoch": 0.07,
1572
- "learning_rate": 1.4439834024896267e-05,
1573
- "loss": 1.1961,
1574
- "step": 522
1575
- },
1576
- {
1577
- "epoch": 0.07,
1578
- "learning_rate": 1.4495159059474413e-05,
1579
- "loss": 0.9539,
1580
- "step": 524
1581
- },
1582
- {
1583
- "epoch": 0.07,
1584
- "learning_rate": 1.455048409405256e-05,
1585
- "loss": 1.0402,
1586
- "step": 526
1587
- },
1588
- {
1589
- "epoch": 0.07,
1590
- "learning_rate": 1.4605809128630706e-05,
1591
- "loss": 1.0121,
1592
- "step": 528
1593
- },
1594
- {
1595
- "epoch": 0.07,
1596
- "learning_rate": 1.4661134163208852e-05,
1597
- "loss": 1.0516,
1598
- "step": 530
1599
- },
1600
- {
1601
- "epoch": 0.07,
1602
- "learning_rate": 1.4716459197787001e-05,
1603
- "loss": 1.1038,
1604
- "step": 532
1605
- },
1606
- {
1607
- "epoch": 0.07,
1608
- "learning_rate": 1.4771784232365147e-05,
1609
- "loss": 1.0192,
1610
- "step": 534
1611
- },
1612
- {
1613
- "epoch": 0.07,
1614
- "learning_rate": 1.4827109266943294e-05,
1615
- "loss": 1.1145,
1616
- "step": 536
1617
- },
1618
- {
1619
- "epoch": 0.07,
1620
- "learning_rate": 1.488243430152144e-05,
1621
- "loss": 1.0995,
1622
- "step": 538
1623
- },
1624
- {
1625
- "epoch": 0.07,
1626
- "learning_rate": 1.4937759336099586e-05,
1627
- "loss": 1.0171,
1628
- "step": 540
1629
- },
1630
- {
1631
- "epoch": 0.07,
1632
- "learning_rate": 1.4993084370677733e-05,
1633
- "loss": 1.1557,
1634
- "step": 542
1635
- },
1636
- {
1637
- "epoch": 0.07,
1638
- "learning_rate": 1.504840940525588e-05,
1639
- "loss": 1.0628,
1640
- "step": 544
1641
- },
1642
- {
1643
- "epoch": 0.07,
1644
- "learning_rate": 1.5103734439834025e-05,
1645
- "loss": 0.987,
1646
- "step": 546
1647
- },
1648
- {
1649
- "epoch": 0.07,
1650
- "learning_rate": 1.5159059474412173e-05,
1651
- "loss": 0.9558,
1652
- "step": 548
1653
- },
1654
- {
1655
- "epoch": 0.07,
1656
- "learning_rate": 1.5214384508990318e-05,
1657
- "loss": 1.4141,
1658
- "step": 550
1659
- },
1660
- {
1661
- "epoch": 0.07,
1662
- "learning_rate": 1.5269709543568464e-05,
1663
- "loss": 0.9219,
1664
- "step": 552
1665
- },
1666
- {
1667
- "epoch": 0.07,
1668
- "learning_rate": 1.5325034578146613e-05,
1669
- "loss": 1.0382,
1670
- "step": 554
1671
- },
1672
- {
1673
- "epoch": 0.07,
1674
- "learning_rate": 1.538035961272476e-05,
1675
- "loss": 1.1114,
1676
- "step": 556
1677
- },
1678
- {
1679
- "epoch": 0.07,
1680
- "learning_rate": 1.5435684647302905e-05,
1681
- "loss": 1.1823,
1682
- "step": 558
1683
- },
1684
- {
1685
- "epoch": 0.07,
1686
- "learning_rate": 1.5491009681881054e-05,
1687
- "loss": 1.1055,
1688
- "step": 560
1689
- },
1690
- {
1691
- "epoch": 0.07,
1692
- "learning_rate": 1.55463347164592e-05,
1693
- "loss": 1.1221,
1694
- "step": 562
1695
- },
1696
- {
1697
- "epoch": 0.07,
1698
- "learning_rate": 1.5601659751037346e-05,
1699
- "loss": 1.1758,
1700
- "step": 564
1701
- },
1702
- {
1703
- "epoch": 0.07,
1704
- "learning_rate": 1.5656984785615492e-05,
1705
- "loss": 1.0823,
1706
- "step": 566
1707
- },
1708
- {
1709
- "epoch": 0.07,
1710
- "learning_rate": 1.5712309820193638e-05,
1711
- "loss": 1.1184,
1712
- "step": 568
1713
- },
1714
- {
1715
- "epoch": 0.07,
1716
- "learning_rate": 1.5767634854771783e-05,
1717
- "loss": 1.1239,
1718
- "step": 570
1719
- },
1720
- {
1721
- "epoch": 0.07,
1722
- "learning_rate": 1.5822959889349933e-05,
1723
- "loss": 1.1376,
1724
- "step": 572
1725
- },
1726
- {
1727
- "epoch": 0.07,
1728
- "learning_rate": 1.587828492392808e-05,
1729
- "loss": 0.9875,
1730
- "step": 574
1731
- },
1732
- {
1733
- "epoch": 0.07,
1734
- "learning_rate": 1.5933609958506224e-05,
1735
- "loss": 1.0247,
1736
- "step": 576
1737
- },
1738
- {
1739
- "epoch": 0.07,
1740
- "learning_rate": 1.5988934993084373e-05,
1741
- "loss": 1.0494,
1742
- "step": 578
1743
- },
1744
- {
1745
- "epoch": 0.07,
1746
- "learning_rate": 1.604426002766252e-05,
1747
- "loss": 1.224,
1748
- "step": 580
1749
- },
1750
- {
1751
- "epoch": 0.07,
1752
- "learning_rate": 1.6099585062240665e-05,
1753
- "loss": 1.1624,
1754
- "step": 582
1755
- },
1756
- {
1757
- "epoch": 0.07,
1758
- "learning_rate": 1.615491009681881e-05,
1759
- "loss": 1.05,
1760
- "step": 584
1761
- },
1762
- {
1763
- "epoch": 0.07,
1764
- "learning_rate": 1.621023513139696e-05,
1765
- "loss": 1.1453,
1766
- "step": 586
1767
- },
1768
- {
1769
- "epoch": 0.07,
1770
- "learning_rate": 1.6265560165975106e-05,
1771
- "loss": 1.1319,
1772
- "step": 588
1773
- },
1774
- {
1775
- "epoch": 0.07,
1776
- "learning_rate": 1.6320885200553252e-05,
1777
- "loss": 1.1812,
1778
- "step": 590
1779
- },
1780
- {
1781
- "epoch": 0.07,
1782
- "learning_rate": 1.6376210235131398e-05,
1783
- "loss": 1.1411,
1784
- "step": 592
1785
- },
1786
- {
1787
- "epoch": 0.07,
1788
- "learning_rate": 1.6431535269709543e-05,
1789
- "loss": 1.1518,
1790
- "step": 594
1791
- },
1792
- {
1793
- "epoch": 0.07,
1794
- "learning_rate": 1.648686030428769e-05,
1795
- "loss": 0.9181,
1796
- "step": 596
1797
- },
1798
- {
1799
- "epoch": 0.07,
1800
- "learning_rate": 1.654218533886584e-05,
1801
- "loss": 1.1148,
1802
- "step": 598
1803
- },
1804
- {
1805
- "epoch": 0.07,
1806
- "learning_rate": 1.6597510373443984e-05,
1807
- "loss": 1.6634,
1808
- "step": 600
1809
  }
1810
  ],
1811
- "max_steps": 24090,
1812
  "num_train_epochs": 3,
1813
- "total_flos": 3.09662838849024e+16,
1814
  "trial_name": null,
1815
  "trial_params": null
1816
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 0.08125126955108673,
5
+ "global_step": 50,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
9
  "log_history": [
10
  {
11
  "epoch": 0.0,
12
+ "learning_rate": 2e-05,
13
+ "loss": 0.8133,
14
  "step": 2
15
  },
16
  {
17
+ "epoch": 0.01,
18
+ "learning_rate": 2e-05,
19
+ "loss": 0.9757,
20
  "step": 4
21
  },
22
  {
23
+ "epoch": 0.01,
24
+ "learning_rate": 2e-05,
25
+ "loss": 0.8953,
26
  "step": 6
27
  },
28
  {
29
+ "epoch": 0.01,
30
+ "learning_rate": 2e-05,
31
+ "loss": 0.9401,
32
  "step": 8
33
  },
34
  {
35
+ "epoch": 0.02,
36
+ "learning_rate": 2e-05,
37
+ "loss": 1.2914,
38
  "step": 10
39
  },
40
  {
41
+ "epoch": 0.02,
42
+ "learning_rate": 2e-05,
43
+ "loss": 1.2835,
44
  "step": 12
45
  },
46
  {
47
+ "epoch": 0.02,
48
+ "learning_rate": 2e-05,
49
+ "loss": 1.2533,
50
  "step": 14
51
  },
52
  {
53
+ "epoch": 0.03,
54
+ "learning_rate": 2e-05,
55
+ "loss": 1.4707,
56
  "step": 16
57
  },
58
  {
59
+ "epoch": 0.03,
60
+ "learning_rate": 2e-05,
61
+ "loss": 1.2831,
62
  "step": 18
63
  },
64
  {
65
+ "epoch": 0.03,
66
+ "learning_rate": 2e-05,
67
+ "loss": 1.2083,
68
  "step": 20
69
  },
70
  {
71
+ "epoch": 0.04,
72
+ "learning_rate": 2e-05,
73
+ "loss": 1.2233,
74
  "step": 22
75
  },
76
  {
77
+ "epoch": 0.04,
78
+ "learning_rate": 2e-05,
79
+ "loss": 1.2911,
80
  "step": 24
81
  },
82
  {
83
+ "epoch": 0.04,
84
+ "learning_rate": 2e-05,
85
+ "loss": 1.2881,
86
  "step": 26
87
  },
88
  {
89
+ "epoch": 0.05,
90
+ "learning_rate": 2e-05,
91
+ "loss": 1.4263,
92
  "step": 28
93
  },
94
  {
95
+ "epoch": 0.05,
96
+ "learning_rate": 2e-05,
97
+ "loss": 1.6368,
98
  "step": 30
99
  },
100
  {
101
+ "epoch": 0.05,
102
+ "learning_rate": 2e-05,
103
+ "loss": 1.4145,
104
  "step": 32
105
  },
106
  {
107
+ "epoch": 0.06,
108
+ "learning_rate": 2e-05,
109
+ "loss": 1.3531,
110
  "step": 34
111
  },
112
  {
113
+ "epoch": 0.06,
114
+ "learning_rate": 2e-05,
115
+ "loss": 1.4549,
116
  "step": 36
117
  },
118
  {
119
+ "epoch": 0.06,
120
+ "learning_rate": 2e-05,
121
+ "loss": 1.4699,
122
  "step": 38
123
  },
124
  {
125
+ "epoch": 0.07,
126
+ "learning_rate": 2e-05,
127
+ "loss": 1.5083,
128
  "step": 40
129
  },
130
  {
131
+ "epoch": 0.07,
132
+ "learning_rate": 2e-05,
133
+ "loss": 1.6162,
134
  "step": 42
135
  },
136
  {
137
+ "epoch": 0.07,
138
+ "learning_rate": 2e-05,
139
+ "loss": 1.8149,
140
  "step": 44
141
  },
142
  {
143
+ "epoch": 0.07,
144
+ "learning_rate": 2e-05,
145
+ "loss": 1.8081,
146
  "step": 46
147
  },
148
  {
149
+ "epoch": 0.08,
150
+ "learning_rate": 2e-05,
151
+ "loss": 1.8813,
152
  "step": 48
153
  },
154
  {
155
+ "epoch": 0.08,
156
+ "learning_rate": 2e-05,
157
+ "loss": 1.8284,
158
  "step": 50
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
159
  }
160
  ],
161
+ "max_steps": 1845,
162
  "num_train_epochs": 3,
163
+ "total_flos": 8177113966657536.0,
164
  "trial_name": null,
165
  "trial_params": null
166
  }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:669f116184b0888d969805711085a0151b986c44eaca8b4c15a79befd7e9dbc9
3
  size 3963
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a234f159887f9d3dc24bde19bae06b48d2e59bb51d4aed93cb3b05c0174d7aa
3
  size 3963