kevincstowe commited on
Commit
d8730db
1 Parent(s): 6fae8d2

updating to large prompt model

Browse files
Files changed (6) hide show
  1. optimizer.pt +3 -0
  2. pytorch_model.bin +1 -1
  3. rng_state.pth +3 -0
  4. scheduler.pt +3 -0
  5. trainer_state.json +784 -0
  6. training_args.bin +1 -1
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c7cdc5cc2eb5400801a7a5eea1611debd8c8352164897466b4ffb900d599677
3
+ size 1783207101
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d21560f2960d66411547cbde283bc6ed7d081b2ea0741c4fcdcda3a78e0efd95
3
  size 891614783
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c93bdc9f9bed379590d7714d08280520ec9ce0aa200d2ad509e89a2bfd05d175
3
  size 891614783
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63893504cff1cde0a5508a7d7da7fe24046ceb7a7e1829992528f073a51fcfde
3
+ size 14503
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a410640008efc44aada3aafd833d689ca9f11e896f4b86139798274a8b2852b7
3
+ size 623
trainer_state.json ADDED
@@ -0,0 +1,784 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.171730940548895,
5
+ "global_step": 64000,
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": 4.9977639200449365e-05,
13
+ "loss": 3.6579,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 4.995527840089873e-05,
19
+ "loss": 3.4615,
20
+ "step": 1000
21
+ },
22
+ {
23
+ "epoch": 0.0,
24
+ "learning_rate": 4.993291760134809e-05,
25
+ "loss": 3.3921,
26
+ "step": 1500
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 4.991055680179745e-05,
31
+ "loss": 3.3552,
32
+ "step": 2000
33
+ },
34
+ {
35
+ "epoch": 0.01,
36
+ "learning_rate": 4.9888196002246816e-05,
37
+ "loss": 3.3133,
38
+ "step": 2500
39
+ },
40
+ {
41
+ "epoch": 0.01,
42
+ "learning_rate": 4.986583520269618e-05,
43
+ "loss": 3.2847,
44
+ "step": 3000
45
+ },
46
+ {
47
+ "epoch": 0.01,
48
+ "learning_rate": 4.984347440314554e-05,
49
+ "loss": 3.261,
50
+ "step": 3500
51
+ },
52
+ {
53
+ "epoch": 0.01,
54
+ "learning_rate": 4.9821113603594904e-05,
55
+ "loss": 3.2291,
56
+ "step": 4000
57
+ },
58
+ {
59
+ "epoch": 0.01,
60
+ "learning_rate": 4.979875280404427e-05,
61
+ "loss": 3.2023,
62
+ "step": 4500
63
+ },
64
+ {
65
+ "epoch": 0.01,
66
+ "learning_rate": 4.977639200449363e-05,
67
+ "loss": 3.1846,
68
+ "step": 5000
69
+ },
70
+ {
71
+ "epoch": 0.01,
72
+ "learning_rate": 4.975403120494299e-05,
73
+ "loss": 3.1743,
74
+ "step": 5500
75
+ },
76
+ {
77
+ "epoch": 0.02,
78
+ "learning_rate": 4.9731670405392355e-05,
79
+ "loss": 3.1475,
80
+ "step": 6000
81
+ },
82
+ {
83
+ "epoch": 0.02,
84
+ "learning_rate": 4.970930960584171e-05,
85
+ "loss": 3.1315,
86
+ "step": 6500
87
+ },
88
+ {
89
+ "epoch": 0.02,
90
+ "learning_rate": 4.968694880629108e-05,
91
+ "loss": 3.1389,
92
+ "step": 7000
93
+ },
94
+ {
95
+ "epoch": 0.02,
96
+ "learning_rate": 4.9664588006740444e-05,
97
+ "loss": 3.0986,
98
+ "step": 7500
99
+ },
100
+ {
101
+ "epoch": 0.02,
102
+ "learning_rate": 4.9642227207189806e-05,
103
+ "loss": 3.0938,
104
+ "step": 8000
105
+ },
106
+ {
107
+ "epoch": 0.02,
108
+ "learning_rate": 4.961986640763916e-05,
109
+ "loss": 3.0844,
110
+ "step": 8500
111
+ },
112
+ {
113
+ "epoch": 0.02,
114
+ "learning_rate": 4.959750560808853e-05,
115
+ "loss": 3.0931,
116
+ "step": 9000
117
+ },
118
+ {
119
+ "epoch": 0.03,
120
+ "learning_rate": 4.9575144808537895e-05,
121
+ "loss": 3.0665,
122
+ "step": 9500
123
+ },
124
+ {
125
+ "epoch": 0.03,
126
+ "learning_rate": 4.955278400898726e-05,
127
+ "loss": 3.0465,
128
+ "step": 10000
129
+ },
130
+ {
131
+ "epoch": 0.03,
132
+ "learning_rate": 4.953042320943661e-05,
133
+ "loss": 3.0343,
134
+ "step": 10500
135
+ },
136
+ {
137
+ "epoch": 0.03,
138
+ "learning_rate": 4.950806240988598e-05,
139
+ "loss": 3.036,
140
+ "step": 11000
141
+ },
142
+ {
143
+ "epoch": 0.03,
144
+ "learning_rate": 4.9485701610335346e-05,
145
+ "loss": 3.0273,
146
+ "step": 11500
147
+ },
148
+ {
149
+ "epoch": 0.03,
150
+ "learning_rate": 4.94633408107847e-05,
151
+ "loss": 3.0231,
152
+ "step": 12000
153
+ },
154
+ {
155
+ "epoch": 0.03,
156
+ "learning_rate": 4.9440980011234064e-05,
157
+ "loss": 3.0177,
158
+ "step": 12500
159
+ },
160
+ {
161
+ "epoch": 0.03,
162
+ "learning_rate": 4.9418619211683434e-05,
163
+ "loss": 2.9969,
164
+ "step": 13000
165
+ },
166
+ {
167
+ "epoch": 0.04,
168
+ "learning_rate": 4.9396258412132796e-05,
169
+ "loss": 3.0019,
170
+ "step": 13500
171
+ },
172
+ {
173
+ "epoch": 0.04,
174
+ "learning_rate": 4.937389761258215e-05,
175
+ "loss": 2.9867,
176
+ "step": 14000
177
+ },
178
+ {
179
+ "epoch": 0.04,
180
+ "learning_rate": 4.9351536813031515e-05,
181
+ "loss": 2.9825,
182
+ "step": 14500
183
+ },
184
+ {
185
+ "epoch": 0.04,
186
+ "learning_rate": 4.9329176013480885e-05,
187
+ "loss": 2.9799,
188
+ "step": 15000
189
+ },
190
+ {
191
+ "epoch": 0.04,
192
+ "learning_rate": 4.930681521393025e-05,
193
+ "loss": 2.9578,
194
+ "step": 15500
195
+ },
196
+ {
197
+ "epoch": 0.04,
198
+ "learning_rate": 4.9284454414379603e-05,
199
+ "loss": 2.949,
200
+ "step": 16000
201
+ },
202
+ {
203
+ "epoch": 0.04,
204
+ "learning_rate": 4.9262093614828966e-05,
205
+ "loss": 2.9598,
206
+ "step": 16500
207
+ },
208
+ {
209
+ "epoch": 0.05,
210
+ "learning_rate": 4.9239732815278336e-05,
211
+ "loss": 2.9568,
212
+ "step": 17000
213
+ },
214
+ {
215
+ "epoch": 0.05,
216
+ "learning_rate": 4.921737201572769e-05,
217
+ "loss": 2.9395,
218
+ "step": 17500
219
+ },
220
+ {
221
+ "epoch": 0.05,
222
+ "learning_rate": 4.9195011216177054e-05,
223
+ "loss": 2.9499,
224
+ "step": 18000
225
+ },
226
+ {
227
+ "epoch": 0.05,
228
+ "learning_rate": 4.917265041662642e-05,
229
+ "loss": 2.9316,
230
+ "step": 18500
231
+ },
232
+ {
233
+ "epoch": 0.05,
234
+ "learning_rate": 4.915028961707579e-05,
235
+ "loss": 2.9356,
236
+ "step": 19000
237
+ },
238
+ {
239
+ "epoch": 0.05,
240
+ "learning_rate": 4.912792881752514e-05,
241
+ "loss": 2.9105,
242
+ "step": 19500
243
+ },
244
+ {
245
+ "epoch": 0.05,
246
+ "learning_rate": 4.9105568017974505e-05,
247
+ "loss": 2.9277,
248
+ "step": 20000
249
+ },
250
+ {
251
+ "epoch": 0.06,
252
+ "learning_rate": 4.908320721842387e-05,
253
+ "loss": 2.9224,
254
+ "step": 20500
255
+ },
256
+ {
257
+ "epoch": 0.06,
258
+ "learning_rate": 4.906084641887324e-05,
259
+ "loss": 2.9136,
260
+ "step": 21000
261
+ },
262
+ {
263
+ "epoch": 0.06,
264
+ "learning_rate": 4.9038485619322594e-05,
265
+ "loss": 2.9109,
266
+ "step": 21500
267
+ },
268
+ {
269
+ "epoch": 0.06,
270
+ "learning_rate": 4.9016124819771956e-05,
271
+ "loss": 2.896,
272
+ "step": 22000
273
+ },
274
+ {
275
+ "epoch": 0.06,
276
+ "learning_rate": 4.899376402022132e-05,
277
+ "loss": 2.8996,
278
+ "step": 22500
279
+ },
280
+ {
281
+ "epoch": 0.06,
282
+ "learning_rate": 4.897140322067068e-05,
283
+ "loss": 2.9059,
284
+ "step": 23000
285
+ },
286
+ {
287
+ "epoch": 0.06,
288
+ "learning_rate": 4.8949042421120045e-05,
289
+ "loss": 2.8849,
290
+ "step": 23500
291
+ },
292
+ {
293
+ "epoch": 0.06,
294
+ "learning_rate": 4.892668162156941e-05,
295
+ "loss": 2.8876,
296
+ "step": 24000
297
+ },
298
+ {
299
+ "epoch": 0.07,
300
+ "learning_rate": 4.890432082201877e-05,
301
+ "loss": 2.887,
302
+ "step": 24500
303
+ },
304
+ {
305
+ "epoch": 0.07,
306
+ "learning_rate": 4.888196002246813e-05,
307
+ "loss": 2.8687,
308
+ "step": 25000
309
+ },
310
+ {
311
+ "epoch": 0.07,
312
+ "learning_rate": 4.8859599222917496e-05,
313
+ "loss": 2.8734,
314
+ "step": 25500
315
+ },
316
+ {
317
+ "epoch": 0.07,
318
+ "learning_rate": 4.883723842336686e-05,
319
+ "loss": 2.8745,
320
+ "step": 26000
321
+ },
322
+ {
323
+ "epoch": 0.07,
324
+ "learning_rate": 4.881487762381622e-05,
325
+ "loss": 2.8904,
326
+ "step": 26500
327
+ },
328
+ {
329
+ "epoch": 0.07,
330
+ "learning_rate": 4.8792516824265584e-05,
331
+ "loss": 2.8652,
332
+ "step": 27000
333
+ },
334
+ {
335
+ "epoch": 0.07,
336
+ "learning_rate": 4.8770156024714947e-05,
337
+ "loss": 2.8715,
338
+ "step": 27500
339
+ },
340
+ {
341
+ "epoch": 0.08,
342
+ "learning_rate": 4.874779522516431e-05,
343
+ "loss": 2.8643,
344
+ "step": 28000
345
+ },
346
+ {
347
+ "epoch": 0.08,
348
+ "learning_rate": 4.872543442561367e-05,
349
+ "loss": 2.8483,
350
+ "step": 28500
351
+ },
352
+ {
353
+ "epoch": 0.08,
354
+ "learning_rate": 4.8703073626063035e-05,
355
+ "loss": 2.8537,
356
+ "step": 29000
357
+ },
358
+ {
359
+ "epoch": 0.08,
360
+ "learning_rate": 4.86807128265124e-05,
361
+ "loss": 2.8456,
362
+ "step": 29500
363
+ },
364
+ {
365
+ "epoch": 0.08,
366
+ "learning_rate": 4.865835202696176e-05,
367
+ "loss": 2.8494,
368
+ "step": 30000
369
+ },
370
+ {
371
+ "epoch": 0.08,
372
+ "learning_rate": 4.863599122741112e-05,
373
+ "loss": 2.8352,
374
+ "step": 30500
375
+ },
376
+ {
377
+ "epoch": 0.08,
378
+ "learning_rate": 4.8613630427860486e-05,
379
+ "loss": 2.8388,
380
+ "step": 31000
381
+ },
382
+ {
383
+ "epoch": 0.08,
384
+ "learning_rate": 4.859126962830985e-05,
385
+ "loss": 2.8381,
386
+ "step": 31500
387
+ },
388
+ {
389
+ "epoch": 0.09,
390
+ "learning_rate": 4.856890882875921e-05,
391
+ "loss": 2.82,
392
+ "step": 32000
393
+ },
394
+ {
395
+ "epoch": 0.09,
396
+ "learning_rate": 4.8546548029208574e-05,
397
+ "loss": 2.8428,
398
+ "step": 32500
399
+ },
400
+ {
401
+ "epoch": 0.09,
402
+ "learning_rate": 4.852418722965794e-05,
403
+ "loss": 2.8436,
404
+ "step": 33000
405
+ },
406
+ {
407
+ "epoch": 0.09,
408
+ "learning_rate": 4.85018264301073e-05,
409
+ "loss": 2.826,
410
+ "step": 33500
411
+ },
412
+ {
413
+ "epoch": 0.09,
414
+ "learning_rate": 4.8479465630556655e-05,
415
+ "loss": 2.8274,
416
+ "step": 34000
417
+ },
418
+ {
419
+ "epoch": 0.09,
420
+ "learning_rate": 4.8457104831006025e-05,
421
+ "loss": 2.8338,
422
+ "step": 34500
423
+ },
424
+ {
425
+ "epoch": 0.09,
426
+ "learning_rate": 4.843474403145539e-05,
427
+ "loss": 2.8228,
428
+ "step": 35000
429
+ },
430
+ {
431
+ "epoch": 0.1,
432
+ "learning_rate": 4.841238323190475e-05,
433
+ "loss": 2.8078,
434
+ "step": 35500
435
+ },
436
+ {
437
+ "epoch": 0.1,
438
+ "learning_rate": 4.8390022432354106e-05,
439
+ "loss": 2.8198,
440
+ "step": 36000
441
+ },
442
+ {
443
+ "epoch": 0.1,
444
+ "learning_rate": 4.8367661632803476e-05,
445
+ "loss": 2.8088,
446
+ "step": 36500
447
+ },
448
+ {
449
+ "epoch": 0.1,
450
+ "learning_rate": 4.834530083325284e-05,
451
+ "loss": 2.8287,
452
+ "step": 37000
453
+ },
454
+ {
455
+ "epoch": 0.1,
456
+ "learning_rate": 4.83229400337022e-05,
457
+ "loss": 2.7993,
458
+ "step": 37500
459
+ },
460
+ {
461
+ "epoch": 0.1,
462
+ "learning_rate": 4.830057923415156e-05,
463
+ "loss": 2.8223,
464
+ "step": 38000
465
+ },
466
+ {
467
+ "epoch": 0.1,
468
+ "learning_rate": 4.827821843460093e-05,
469
+ "loss": 2.8084,
470
+ "step": 38500
471
+ },
472
+ {
473
+ "epoch": 0.1,
474
+ "learning_rate": 4.825585763505029e-05,
475
+ "loss": 2.8119,
476
+ "step": 39000
477
+ },
478
+ {
479
+ "epoch": 0.11,
480
+ "learning_rate": 4.8233496835499646e-05,
481
+ "loss": 2.8131,
482
+ "step": 39500
483
+ },
484
+ {
485
+ "epoch": 0.11,
486
+ "learning_rate": 4.821113603594901e-05,
487
+ "loss": 2.7907,
488
+ "step": 40000
489
+ },
490
+ {
491
+ "epoch": 0.11,
492
+ "learning_rate": 4.818877523639838e-05,
493
+ "loss": 2.8071,
494
+ "step": 40500
495
+ },
496
+ {
497
+ "epoch": 0.11,
498
+ "learning_rate": 4.816641443684774e-05,
499
+ "loss": 2.7969,
500
+ "step": 41000
501
+ },
502
+ {
503
+ "epoch": 0.11,
504
+ "learning_rate": 4.81440536372971e-05,
505
+ "loss": 2.8045,
506
+ "step": 41500
507
+ },
508
+ {
509
+ "epoch": 0.11,
510
+ "learning_rate": 4.812169283774646e-05,
511
+ "loss": 2.7991,
512
+ "step": 42000
513
+ },
514
+ {
515
+ "epoch": 0.11,
516
+ "learning_rate": 4.809933203819583e-05,
517
+ "loss": 2.7845,
518
+ "step": 42500
519
+ },
520
+ {
521
+ "epoch": 0.12,
522
+ "learning_rate": 4.807697123864519e-05,
523
+ "loss": 2.8144,
524
+ "step": 43000
525
+ },
526
+ {
527
+ "epoch": 0.12,
528
+ "learning_rate": 4.805461043909455e-05,
529
+ "loss": 2.764,
530
+ "step": 43500
531
+ },
532
+ {
533
+ "epoch": 0.12,
534
+ "learning_rate": 4.803224963954391e-05,
535
+ "loss": 2.7744,
536
+ "step": 44000
537
+ },
538
+ {
539
+ "epoch": 0.12,
540
+ "learning_rate": 4.800988883999328e-05,
541
+ "loss": 2.7857,
542
+ "step": 44500
543
+ },
544
+ {
545
+ "epoch": 0.12,
546
+ "learning_rate": 4.7987528040442636e-05,
547
+ "loss": 2.8017,
548
+ "step": 45000
549
+ },
550
+ {
551
+ "epoch": 0.12,
552
+ "learning_rate": 4.7965167240892e-05,
553
+ "loss": 2.7958,
554
+ "step": 45500
555
+ },
556
+ {
557
+ "epoch": 0.12,
558
+ "learning_rate": 4.794280644134136e-05,
559
+ "loss": 2.7766,
560
+ "step": 46000
561
+ },
562
+ {
563
+ "epoch": 0.12,
564
+ "learning_rate": 4.792044564179073e-05,
565
+ "loss": 2.7614,
566
+ "step": 46500
567
+ },
568
+ {
569
+ "epoch": 0.13,
570
+ "learning_rate": 4.789808484224009e-05,
571
+ "loss": 2.7739,
572
+ "step": 47000
573
+ },
574
+ {
575
+ "epoch": 0.13,
576
+ "learning_rate": 4.787572404268945e-05,
577
+ "loss": 2.7736,
578
+ "step": 47500
579
+ },
580
+ {
581
+ "epoch": 0.13,
582
+ "learning_rate": 4.785336324313882e-05,
583
+ "loss": 2.7683,
584
+ "step": 48000
585
+ },
586
+ {
587
+ "epoch": 0.13,
588
+ "learning_rate": 4.783100244358818e-05,
589
+ "loss": 2.7765,
590
+ "step": 48500
591
+ },
592
+ {
593
+ "epoch": 0.13,
594
+ "learning_rate": 4.780864164403754e-05,
595
+ "loss": 2.752,
596
+ "step": 49000
597
+ },
598
+ {
599
+ "epoch": 0.13,
600
+ "learning_rate": 4.77862808444869e-05,
601
+ "loss": 2.7666,
602
+ "step": 49500
603
+ },
604
+ {
605
+ "epoch": 0.13,
606
+ "learning_rate": 4.776392004493627e-05,
607
+ "loss": 2.7481,
608
+ "step": 50000
609
+ },
610
+ {
611
+ "epoch": 0.14,
612
+ "learning_rate": 4.7741559245385626e-05,
613
+ "loss": 2.7724,
614
+ "step": 50500
615
+ },
616
+ {
617
+ "epoch": 0.14,
618
+ "learning_rate": 4.771919844583499e-05,
619
+ "loss": 2.7773,
620
+ "step": 51000
621
+ },
622
+ {
623
+ "epoch": 0.14,
624
+ "learning_rate": 4.769683764628435e-05,
625
+ "loss": 2.7659,
626
+ "step": 51500
627
+ },
628
+ {
629
+ "epoch": 0.14,
630
+ "learning_rate": 4.767447684673372e-05,
631
+ "loss": 2.7473,
632
+ "step": 52000
633
+ },
634
+ {
635
+ "epoch": 0.14,
636
+ "learning_rate": 4.765211604718308e-05,
637
+ "loss": 2.761,
638
+ "step": 52500
639
+ },
640
+ {
641
+ "epoch": 0.14,
642
+ "learning_rate": 4.762975524763244e-05,
643
+ "loss": 2.7289,
644
+ "step": 53000
645
+ },
646
+ {
647
+ "epoch": 0.14,
648
+ "learning_rate": 4.76073944480818e-05,
649
+ "loss": 2.7657,
650
+ "step": 53500
651
+ },
652
+ {
653
+ "epoch": 0.14,
654
+ "learning_rate": 4.758503364853117e-05,
655
+ "loss": 2.7719,
656
+ "step": 54000
657
+ },
658
+ {
659
+ "epoch": 0.15,
660
+ "learning_rate": 4.756267284898053e-05,
661
+ "loss": 2.7581,
662
+ "step": 54500
663
+ },
664
+ {
665
+ "epoch": 0.15,
666
+ "learning_rate": 4.754031204942989e-05,
667
+ "loss": 2.7548,
668
+ "step": 55000
669
+ },
670
+ {
671
+ "epoch": 0.15,
672
+ "learning_rate": 4.7517951249879253e-05,
673
+ "loss": 2.7509,
674
+ "step": 55500
675
+ },
676
+ {
677
+ "epoch": 0.15,
678
+ "learning_rate": 4.7495590450328616e-05,
679
+ "loss": 2.738,
680
+ "step": 56000
681
+ },
682
+ {
683
+ "epoch": 0.15,
684
+ "learning_rate": 4.747322965077798e-05,
685
+ "loss": 2.752,
686
+ "step": 56500
687
+ },
688
+ {
689
+ "epoch": 0.15,
690
+ "learning_rate": 4.745086885122734e-05,
691
+ "loss": 2.7373,
692
+ "step": 57000
693
+ },
694
+ {
695
+ "epoch": 0.15,
696
+ "learning_rate": 4.7428508051676704e-05,
697
+ "loss": 2.7541,
698
+ "step": 57500
699
+ },
700
+ {
701
+ "epoch": 0.16,
702
+ "learning_rate": 4.740614725212607e-05,
703
+ "loss": 2.7332,
704
+ "step": 58000
705
+ },
706
+ {
707
+ "epoch": 0.16,
708
+ "learning_rate": 4.738378645257543e-05,
709
+ "loss": 2.713,
710
+ "step": 58500
711
+ },
712
+ {
713
+ "epoch": 0.16,
714
+ "learning_rate": 4.736142565302479e-05,
715
+ "loss": 2.7534,
716
+ "step": 59000
717
+ },
718
+ {
719
+ "epoch": 0.16,
720
+ "learning_rate": 4.7339064853474155e-05,
721
+ "loss": 2.753,
722
+ "step": 59500
723
+ },
724
+ {
725
+ "epoch": 0.16,
726
+ "learning_rate": 4.731670405392352e-05,
727
+ "loss": 2.7357,
728
+ "step": 60000
729
+ },
730
+ {
731
+ "epoch": 0.16,
732
+ "learning_rate": 4.729434325437288e-05,
733
+ "loss": 2.7286,
734
+ "step": 60500
735
+ },
736
+ {
737
+ "epoch": 0.16,
738
+ "learning_rate": 4.7271982454822244e-05,
739
+ "loss": 2.7137,
740
+ "step": 61000
741
+ },
742
+ {
743
+ "epoch": 0.17,
744
+ "learning_rate": 4.7249621655271606e-05,
745
+ "loss": 2.73,
746
+ "step": 61500
747
+ },
748
+ {
749
+ "epoch": 0.17,
750
+ "learning_rate": 4.722726085572097e-05,
751
+ "loss": 2.7375,
752
+ "step": 62000
753
+ },
754
+ {
755
+ "epoch": 0.17,
756
+ "learning_rate": 4.720490005617033e-05,
757
+ "loss": 2.7257,
758
+ "step": 62500
759
+ },
760
+ {
761
+ "epoch": 0.17,
762
+ "learning_rate": 4.7182539256619695e-05,
763
+ "loss": 2.7162,
764
+ "step": 63000
765
+ },
766
+ {
767
+ "epoch": 0.17,
768
+ "learning_rate": 4.716017845706906e-05,
769
+ "loss": 2.7421,
770
+ "step": 63500
771
+ },
772
+ {
773
+ "epoch": 0.17,
774
+ "learning_rate": 4.713781765751842e-05,
775
+ "loss": 2.7255,
776
+ "step": 64000
777
+ }
778
+ ],
779
+ "max_steps": 1118028,
780
+ "num_train_epochs": 3,
781
+ "total_flos": 1.4482346926915584e+17,
782
+ "trial_name": null,
783
+ "trial_params": null
784
+ }
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a394cdaba86514556303cea6acab24470982f952cacc4b2621a58cd452347b3e
3
  size 3439
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d1a9344b2bce844f657a0b419d1ff06cc670c5cae8c6982797b4ba8cb64c4a78
3
  size 3439