aburnazy commited on
Commit
c5aaa98
1 Parent(s): ce2b932

Model save

Browse files
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 2,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 1,
6
+ "transformers_version": "4.31.0"
7
+ }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2612d4810df5545716580af0720fc2cb076f0a85c2fac5b4ac5d2960fe14fbf4
3
  size 1324902941
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:178c9f2adfcca351b15d0df89976b96627df87f40f21260ffc15c13424cf553d
3
  size 1324902941
runs/Jul24_13-49-34_e33fde8ab68f/events.out.tfevents.1690206682.e33fde8ab68f.3600.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:eb963feedf56d652cbaaabeb4f4214d98aa3d04dca9a39133797b8c4c536a476
3
- size 27750
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9175f805ccfa6dd32252519751eae2215d2591ae6cc7a2d9e2a5a0f9c4422f4a
3
+ size 29674
trainer_state.json ADDED
@@ -0,0 +1,985 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.9958002799813346,
5
+ "global_step": 1605,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.02,
12
+ "learning_rate": 4.081632653061225e-06,
13
+ "loss": 3.83,
14
+ "step": 10
15
+ },
16
+ {
17
+ "epoch": 0.04,
18
+ "learning_rate": 8.16326530612245e-06,
19
+ "loss": 3.7366,
20
+ "step": 20
21
+ },
22
+ {
23
+ "epoch": 0.06,
24
+ "learning_rate": 1.2244897959183674e-05,
25
+ "loss": 3.6074,
26
+ "step": 30
27
+ },
28
+ {
29
+ "epoch": 0.07,
30
+ "learning_rate": 1.63265306122449e-05,
31
+ "loss": 3.1722,
32
+ "step": 40
33
+ },
34
+ {
35
+ "epoch": 0.09,
36
+ "learning_rate": 1.9999979617830125e-05,
37
+ "loss": 3.1476,
38
+ "step": 50
39
+ },
40
+ {
41
+ "epoch": 0.11,
42
+ "learning_rate": 1.9997533857978314e-05,
43
+ "loss": 3.0026,
44
+ "step": 60
45
+ },
46
+ {
47
+ "epoch": 0.13,
48
+ "learning_rate": 1.999101280651531e-05,
49
+ "loss": 2.9246,
50
+ "step": 70
51
+ },
52
+ {
53
+ "epoch": 0.15,
54
+ "learning_rate": 1.99804191216153e-05,
55
+ "loss": 2.9575,
56
+ "step": 80
57
+ },
58
+ {
59
+ "epoch": 0.17,
60
+ "learning_rate": 1.996575712157875e-05,
61
+ "loss": 2.8573,
62
+ "step": 90
63
+ },
64
+ {
65
+ "epoch": 0.19,
66
+ "learning_rate": 1.9947032783072173e-05,
67
+ "loss": 2.8785,
68
+ "step": 100
69
+ },
70
+ {
71
+ "epoch": 0.21,
72
+ "learning_rate": 1.9924253738691836e-05,
73
+ "loss": 2.551,
74
+ "step": 110
75
+ },
76
+ {
77
+ "epoch": 0.22,
78
+ "learning_rate": 1.98974292738525e-05,
79
+ "loss": 2.6737,
80
+ "step": 120
81
+ },
82
+ {
83
+ "epoch": 0.24,
84
+ "learning_rate": 1.9866570323002413e-05,
85
+ "loss": 2.7509,
86
+ "step": 130
87
+ },
88
+ {
89
+ "epoch": 0.26,
90
+ "learning_rate": 1.983168946516609e-05,
91
+ "loss": 2.5926,
92
+ "step": 140
93
+ },
94
+ {
95
+ "epoch": 0.28,
96
+ "learning_rate": 1.9792800918816745e-05,
97
+ "loss": 2.5292,
98
+ "step": 150
99
+ },
100
+ {
101
+ "epoch": 0.3,
102
+ "learning_rate": 1.9749920536080403e-05,
103
+ "loss": 2.6191,
104
+ "step": 160
105
+ },
106
+ {
107
+ "epoch": 0.32,
108
+ "learning_rate": 1.9703065796274124e-05,
109
+ "loss": 2.6252,
110
+ "step": 170
111
+ },
112
+ {
113
+ "epoch": 0.34,
114
+ "learning_rate": 1.965225579878089e-05,
115
+ "loss": 2.4386,
116
+ "step": 180
117
+ },
118
+ {
119
+ "epoch": 0.35,
120
+ "learning_rate": 1.959751125526415e-05,
121
+ "loss": 2.5224,
122
+ "step": 190
123
+ },
124
+ {
125
+ "epoch": 0.37,
126
+ "learning_rate": 1.9538854481225114e-05,
127
+ "loss": 2.4643,
128
+ "step": 200
129
+ },
130
+ {
131
+ "epoch": 0.39,
132
+ "learning_rate": 1.9476309386906303e-05,
133
+ "loss": 2.3863,
134
+ "step": 210
135
+ },
136
+ {
137
+ "epoch": 0.41,
138
+ "learning_rate": 1.9409901467545038e-05,
139
+ "loss": 2.331,
140
+ "step": 220
141
+ },
142
+ {
143
+ "epoch": 0.43,
144
+ "learning_rate": 1.933965779298078e-05,
145
+ "loss": 2.2992,
146
+ "step": 230
147
+ },
148
+ {
149
+ "epoch": 0.45,
150
+ "learning_rate": 1.9265606996620708e-05,
151
+ "loss": 2.4181,
152
+ "step": 240
153
+ },
154
+ {
155
+ "epoch": 0.47,
156
+ "learning_rate": 1.9187779263767875e-05,
157
+ "loss": 2.3913,
158
+ "step": 250
159
+ },
160
+ {
161
+ "epoch": 0.49,
162
+ "learning_rate": 1.9106206319316757e-05,
163
+ "loss": 2.3832,
164
+ "step": 260
165
+ },
166
+ {
167
+ "epoch": 0.5,
168
+ "learning_rate": 1.9020921414821294e-05,
169
+ "loss": 2.4269,
170
+ "step": 270
171
+ },
172
+ {
173
+ "epoch": 0.52,
174
+ "learning_rate": 1.8931959314940523e-05,
175
+ "loss": 2.2566,
176
+ "step": 280
177
+ },
178
+ {
179
+ "epoch": 0.54,
180
+ "learning_rate": 1.88393562832675e-05,
181
+ "loss": 2.316,
182
+ "step": 290
183
+ },
184
+ {
185
+ "epoch": 0.56,
186
+ "learning_rate": 1.8743150067547168e-05,
187
+ "loss": 2.2473,
188
+ "step": 300
189
+ },
190
+ {
191
+ "epoch": 0.58,
192
+ "learning_rate": 1.8643379884289266e-05,
193
+ "loss": 2.3063,
194
+ "step": 310
195
+ },
196
+ {
197
+ "epoch": 0.6,
198
+ "learning_rate": 1.85400864027825e-05,
199
+ "loss": 2.187,
200
+ "step": 320
201
+ },
202
+ {
203
+ "epoch": 0.62,
204
+ "learning_rate": 1.8433311728516553e-05,
205
+ "loss": 2.141,
206
+ "step": 330
207
+ },
208
+ {
209
+ "epoch": 0.63,
210
+ "learning_rate": 1.832309938601862e-05,
211
+ "loss": 2.2089,
212
+ "step": 340
213
+ },
214
+ {
215
+ "epoch": 0.65,
216
+ "learning_rate": 1.8209494301111535e-05,
217
+ "loss": 2.2493,
218
+ "step": 350
219
+ },
220
+ {
221
+ "epoch": 0.67,
222
+ "learning_rate": 1.809254278260066e-05,
223
+ "loss": 2.2365,
224
+ "step": 360
225
+ },
226
+ {
227
+ "epoch": 0.69,
228
+ "learning_rate": 1.7972292503397036e-05,
229
+ "loss": 2.2406,
230
+ "step": 370
231
+ },
232
+ {
233
+ "epoch": 0.71,
234
+ "learning_rate": 1.7848792481084504e-05,
235
+ "loss": 2.1423,
236
+ "step": 380
237
+ },
238
+ {
239
+ "epoch": 0.73,
240
+ "learning_rate": 1.772209305793868e-05,
241
+ "loss": 2.1684,
242
+ "step": 390
243
+ },
244
+ {
245
+ "epoch": 0.75,
246
+ "learning_rate": 1.7592245880405934e-05,
247
+ "loss": 2.1613,
248
+ "step": 400
249
+ },
250
+ {
251
+ "epoch": 0.77,
252
+ "learning_rate": 1.7459303878050788e-05,
253
+ "loss": 2.1061,
254
+ "step": 410
255
+ },
256
+ {
257
+ "epoch": 0.78,
258
+ "learning_rate": 1.732332124198023e-05,
259
+ "loss": 2.1551,
260
+ "step": 420
261
+ },
262
+ {
263
+ "epoch": 0.8,
264
+ "learning_rate": 1.7184353402753832e-05,
265
+ "loss": 2.1107,
266
+ "step": 430
267
+ },
268
+ {
269
+ "epoch": 0.82,
270
+ "learning_rate": 1.7042457007788594e-05,
271
+ "loss": 2.2264,
272
+ "step": 440
273
+ },
274
+ {
275
+ "epoch": 0.84,
276
+ "learning_rate": 1.6897689898267774e-05,
277
+ "loss": 2.0884,
278
+ "step": 450
279
+ },
280
+ {
281
+ "epoch": 0.86,
282
+ "learning_rate": 1.6750111085563122e-05,
283
+ "loss": 2.1016,
284
+ "step": 460
285
+ },
286
+ {
287
+ "epoch": 0.88,
288
+ "learning_rate": 1.659978072718007e-05,
289
+ "loss": 2.138,
290
+ "step": 470
291
+ },
292
+ {
293
+ "epoch": 0.9,
294
+ "learning_rate": 1.644676010223576e-05,
295
+ "loss": 2.0713,
296
+ "step": 480
297
+ },
298
+ {
299
+ "epoch": 0.91,
300
+ "learning_rate": 1.6291111586479857e-05,
301
+ "loss": 2.1247,
302
+ "step": 490
303
+ },
304
+ {
305
+ "epoch": 0.93,
306
+ "learning_rate": 1.6132898626868336e-05,
307
+ "loss": 2.0748,
308
+ "step": 500
309
+ },
310
+ {
311
+ "epoch": 0.95,
312
+ "learning_rate": 1.5972185715700623e-05,
313
+ "loss": 2.0356,
314
+ "step": 510
315
+ },
316
+ {
317
+ "epoch": 0.97,
318
+ "learning_rate": 1.5809038364330617e-05,
319
+ "loss": 2.1173,
320
+ "step": 520
321
+ },
322
+ {
323
+ "epoch": 0.99,
324
+ "learning_rate": 1.5643523076462325e-05,
325
+ "loss": 2.1065,
326
+ "step": 530
327
+ },
328
+ {
329
+ "epoch": 1.01,
330
+ "learning_rate": 1.547570732104095e-05,
331
+ "loss": 2.1073,
332
+ "step": 540
333
+ },
334
+ {
335
+ "epoch": 1.03,
336
+ "learning_rate": 1.5305659504750595e-05,
337
+ "loss": 1.9632,
338
+ "step": 550
339
+ },
340
+ {
341
+ "epoch": 1.05,
342
+ "learning_rate": 1.5133448944129616e-05,
343
+ "loss": 2.0967,
344
+ "step": 560
345
+ },
346
+ {
347
+ "epoch": 1.06,
348
+ "learning_rate": 1.4959145837315182e-05,
349
+ "loss": 1.9325,
350
+ "step": 570
351
+ },
352
+ {
353
+ "epoch": 1.08,
354
+ "learning_rate": 1.4782821235428406e-05,
355
+ "loss": 2.0018,
356
+ "step": 580
357
+ },
358
+ {
359
+ "epoch": 1.1,
360
+ "learning_rate": 1.4604547013611813e-05,
361
+ "loss": 1.9439,
362
+ "step": 590
363
+ },
364
+ {
365
+ "epoch": 1.12,
366
+ "learning_rate": 1.4424395841730889e-05,
367
+ "loss": 1.9066,
368
+ "step": 600
369
+ },
370
+ {
371
+ "epoch": 1.14,
372
+ "learning_rate": 1.4242441154751691e-05,
373
+ "loss": 1.8828,
374
+ "step": 610
375
+ },
376
+ {
377
+ "epoch": 1.16,
378
+ "learning_rate": 1.4058757122806573e-05,
379
+ "loss": 1.9665,
380
+ "step": 620
381
+ },
382
+ {
383
+ "epoch": 1.18,
384
+ "learning_rate": 1.3873418620960237e-05,
385
+ "loss": 1.9528,
386
+ "step": 630
387
+ },
388
+ {
389
+ "epoch": 1.19,
390
+ "learning_rate": 1.3686501198688435e-05,
391
+ "loss": 1.9044,
392
+ "step": 640
393
+ },
394
+ {
395
+ "epoch": 1.21,
396
+ "learning_rate": 1.3498081049081773e-05,
397
+ "loss": 1.8582,
398
+ "step": 650
399
+ },
400
+ {
401
+ "epoch": 1.23,
402
+ "learning_rate": 1.330823497778711e-05,
403
+ "loss": 2.0053,
404
+ "step": 660
405
+ },
406
+ {
407
+ "epoch": 1.25,
408
+ "learning_rate": 1.3117040371699334e-05,
409
+ "loss": 1.9191,
410
+ "step": 670
411
+ },
412
+ {
413
+ "epoch": 1.27,
414
+ "learning_rate": 1.2924575167416147e-05,
415
+ "loss": 1.9297,
416
+ "step": 680
417
+ },
418
+ {
419
+ "epoch": 1.29,
420
+ "learning_rate": 1.2730917819468796e-05,
421
+ "loss": 1.8852,
422
+ "step": 690
423
+ },
424
+ {
425
+ "epoch": 1.31,
426
+ "learning_rate": 1.2536147268341676e-05,
427
+ "loss": 1.9083,
428
+ "step": 700
429
+ },
430
+ {
431
+ "epoch": 1.33,
432
+ "learning_rate": 1.2340342908293856e-05,
433
+ "loss": 1.8909,
434
+ "step": 710
435
+ },
436
+ {
437
+ "epoch": 1.34,
438
+ "learning_rate": 1.2143584554995604e-05,
439
+ "loss": 1.8984,
440
+ "step": 720
441
+ },
442
+ {
443
+ "epoch": 1.36,
444
+ "learning_rate": 1.1945952412993162e-05,
445
+ "loss": 1.9207,
446
+ "step": 730
447
+ },
448
+ {
449
+ "epoch": 1.38,
450
+ "learning_rate": 1.1747527043014997e-05,
451
+ "loss": 1.8663,
452
+ "step": 740
453
+ },
454
+ {
455
+ "epoch": 1.4,
456
+ "learning_rate": 1.1548389329132832e-05,
457
+ "loss": 1.9107,
458
+ "step": 750
459
+ },
460
+ {
461
+ "epoch": 1.42,
462
+ "learning_rate": 1.1348620445790926e-05,
463
+ "loss": 1.9066,
464
+ "step": 760
465
+ },
466
+ {
467
+ "epoch": 1.44,
468
+ "learning_rate": 1.114830182471695e-05,
469
+ "loss": 1.8737,
470
+ "step": 770
471
+ },
472
+ {
473
+ "epoch": 1.46,
474
+ "learning_rate": 1.0947515121727987e-05,
475
+ "loss": 1.935,
476
+ "step": 780
477
+ },
478
+ {
479
+ "epoch": 1.47,
480
+ "learning_rate": 1.0746342183445209e-05,
481
+ "loss": 1.8357,
482
+ "step": 790
483
+ },
484
+ {
485
+ "epoch": 1.49,
486
+ "learning_rate": 1.0544865013930763e-05,
487
+ "loss": 1.8526,
488
+ "step": 800
489
+ },
490
+ {
491
+ "epoch": 1.51,
492
+ "learning_rate": 1.0343165741260474e-05,
493
+ "loss": 1.9319,
494
+ "step": 810
495
+ },
496
+ {
497
+ "epoch": 1.53,
498
+ "learning_rate": 1.0141326584046012e-05,
499
+ "loss": 1.9127,
500
+ "step": 820
501
+ },
502
+ {
503
+ "epoch": 1.55,
504
+ "learning_rate": 9.93942981792013e-06,
505
+ "loss": 1.906,
506
+ "step": 830
507
+ },
508
+ {
509
+ "epoch": 1.57,
510
+ "learning_rate": 9.737557741998696e-06,
511
+ "loss": 1.7907,
512
+ "step": 840
513
+ },
514
+ {
515
+ "epoch": 1.59,
516
+ "learning_rate": 9.535792645333114e-06,
517
+ "loss": 1.8328,
518
+ "step": 850
519
+ },
520
+ {
521
+ "epoch": 1.61,
522
+ "learning_rate": 9.334216773366863e-06,
523
+ "loss": 1.8959,
524
+ "step": 860
525
+ },
526
+ {
527
+ "epoch": 1.62,
528
+ "learning_rate": 9.132912294409814e-06,
529
+ "loss": 1.8743,
530
+ "step": 870
531
+ },
532
+ {
533
+ "epoch": 1.64,
534
+ "learning_rate": 8.931961266144003e-06,
535
+ "loss": 1.7924,
536
+ "step": 880
537
+ },
538
+ {
539
+ "epoch": 1.66,
540
+ "learning_rate": 8.731445602174463e-06,
541
+ "loss": 1.8917,
542
+ "step": 890
543
+ },
544
+ {
545
+ "epoch": 1.68,
546
+ "learning_rate": 8.531447038638812e-06,
547
+ "loss": 1.8711,
548
+ "step": 900
549
+ },
550
+ {
551
+ "epoch": 1.7,
552
+ "learning_rate": 8.33204710088918e-06,
553
+ "loss": 1.8876,
554
+ "step": 910
555
+ },
556
+ {
557
+ "epoch": 1.72,
558
+ "learning_rate": 8.133327070260064e-06,
559
+ "loss": 1.8171,
560
+ "step": 920
561
+ },
562
+ {
563
+ "epoch": 1.74,
564
+ "learning_rate": 7.935367950935623e-06,
565
+ "loss": 1.84,
566
+ "step": 930
567
+ },
568
+ {
569
+ "epoch": 1.75,
570
+ "learning_rate": 7.738250436929977e-06,
571
+ "loss": 1.9742,
572
+ "step": 940
573
+ },
574
+ {
575
+ "epoch": 1.77,
576
+ "learning_rate": 7.5420548791939406e-06,
577
+ "loss": 1.8994,
578
+ "step": 950
579
+ },
580
+ {
581
+ "epoch": 1.79,
582
+ "learning_rate": 7.346861252861577e-06,
583
+ "loss": 1.886,
584
+ "step": 960
585
+ },
586
+ {
587
+ "epoch": 1.81,
588
+ "learning_rate": 7.152749124649961e-06,
589
+ "loss": 1.8094,
590
+ "step": 970
591
+ },
592
+ {
593
+ "epoch": 1.83,
594
+ "learning_rate": 6.959797620425435e-06,
595
+ "loss": 1.8344,
596
+ "step": 980
597
+ },
598
+ {
599
+ "epoch": 1.85,
600
+ "learning_rate": 6.768085392949527e-06,
601
+ "loss": 1.8566,
602
+ "step": 990
603
+ },
604
+ {
605
+ "epoch": 1.87,
606
+ "learning_rate": 6.5776905898177936e-06,
607
+ "loss": 1.7879,
608
+ "step": 1000
609
+ },
610
+ {
611
+ "epoch": 1.89,
612
+ "learning_rate": 6.388690821604505e-06,
613
+ "loss": 1.8543,
614
+ "step": 1010
615
+ },
616
+ {
617
+ "epoch": 1.9,
618
+ "learning_rate": 6.20116313022628e-06,
619
+ "loss": 1.8772,
620
+ "step": 1020
621
+ },
622
+ {
623
+ "epoch": 1.92,
624
+ "learning_rate": 6.015183957537523e-06,
625
+ "loss": 1.8362,
626
+ "step": 1030
627
+ },
628
+ {
629
+ "epoch": 1.94,
630
+ "learning_rate": 5.830829114170442e-06,
631
+ "loss": 1.8778,
632
+ "step": 1040
633
+ },
634
+ {
635
+ "epoch": 1.96,
636
+ "learning_rate": 5.6481737486323936e-06,
637
+ "loss": 1.8676,
638
+ "step": 1050
639
+ },
640
+ {
641
+ "epoch": 1.98,
642
+ "learning_rate": 5.467292316673114e-06,
643
+ "loss": 1.8362,
644
+ "step": 1060
645
+ },
646
+ {
647
+ "epoch": 2.0,
648
+ "learning_rate": 5.288258550934343e-06,
649
+ "loss": 1.8752,
650
+ "step": 1070
651
+ },
652
+ {
653
+ "epoch": 2.02,
654
+ "learning_rate": 5.111145430894229e-06,
655
+ "loss": 1.9466,
656
+ "step": 1080
657
+ },
658
+ {
659
+ "epoch": 2.03,
660
+ "learning_rate": 4.936025153118697e-06,
661
+ "loss": 1.7668,
662
+ "step": 1090
663
+ },
664
+ {
665
+ "epoch": 2.05,
666
+ "learning_rate": 4.762969101832011e-06,
667
+ "loss": 1.7815,
668
+ "step": 1100
669
+ },
670
+ {
671
+ "epoch": 2.07,
672
+ "learning_rate": 4.59204781981843e-06,
673
+ "loss": 1.72,
674
+ "step": 1110
675
+ },
676
+ {
677
+ "epoch": 2.09,
678
+ "learning_rate": 4.423330979666878e-06,
679
+ "loss": 1.7548,
680
+ "step": 1120
681
+ },
682
+ {
683
+ "epoch": 2.11,
684
+ "learning_rate": 4.256887355370333e-06,
685
+ "loss": 1.6771,
686
+ "step": 1130
687
+ },
688
+ {
689
+ "epoch": 2.13,
690
+ "learning_rate": 4.092784794291502e-06,
691
+ "loss": 1.8401,
692
+ "step": 1140
693
+ },
694
+ {
695
+ "epoch": 2.15,
696
+ "learning_rate": 3.931090189506222e-06,
697
+ "loss": 1.7071,
698
+ "step": 1150
699
+ },
700
+ {
701
+ "epoch": 2.17,
702
+ "learning_rate": 3.7718694525358534e-06,
703
+ "loss": 1.8095,
704
+ "step": 1160
705
+ },
706
+ {
707
+ "epoch": 2.18,
708
+ "learning_rate": 3.615187486479794e-06,
709
+ "loss": 1.7441,
710
+ "step": 1170
711
+ },
712
+ {
713
+ "epoch": 2.2,
714
+ "learning_rate": 3.4611081595590445e-06,
715
+ "loss": 1.7304,
716
+ "step": 1180
717
+ },
718
+ {
719
+ "epoch": 2.22,
720
+ "learning_rate": 3.30969427908162e-06,
721
+ "loss": 1.7527,
722
+ "step": 1190
723
+ },
724
+ {
725
+ "epoch": 2.24,
726
+ "learning_rate": 3.1610075658404436e-06,
727
+ "loss": 1.8059,
728
+ "step": 1200
729
+ },
730
+ {
731
+ "epoch": 2.26,
732
+ "learning_rate": 3.0151086289540997e-06,
733
+ "loss": 1.6814,
734
+ "step": 1210
735
+ },
736
+ {
737
+ "epoch": 2.28,
738
+ "learning_rate": 2.872056941160767e-06,
739
+ "loss": 1.7992,
740
+ "step": 1220
741
+ },
742
+ {
743
+ "epoch": 2.3,
744
+ "learning_rate": 2.7319108145753625e-06,
745
+ "loss": 1.6453,
746
+ "step": 1230
747
+ },
748
+ {
749
+ "epoch": 2.31,
750
+ "learning_rate": 2.594727376919798e-06,
751
+ "loss": 1.6889,
752
+ "step": 1240
753
+ },
754
+ {
755
+ "epoch": 2.33,
756
+ "learning_rate": 2.460562548236015e-06,
757
+ "loss": 1.7476,
758
+ "step": 1250
759
+ },
760
+ {
761
+ "epoch": 2.35,
762
+ "learning_rate": 2.329471018091323e-06,
763
+ "loss": 1.7296,
764
+ "step": 1260
765
+ },
766
+ {
767
+ "epoch": 2.37,
768
+ "learning_rate": 2.201506223285306e-06,
769
+ "loss": 1.8352,
770
+ "step": 1270
771
+ },
772
+ {
773
+ "epoch": 2.39,
774
+ "learning_rate": 2.0767203260673952e-06,
775
+ "loss": 1.718,
776
+ "step": 1280
777
+ },
778
+ {
779
+ "epoch": 2.41,
780
+ "learning_rate": 1.9551641928739886e-06,
781
+ "loss": 1.7623,
782
+ "step": 1290
783
+ },
784
+ {
785
+ "epoch": 2.43,
786
+ "learning_rate": 1.836887373593782e-06,
787
+ "loss": 1.7442,
788
+ "step": 1300
789
+ },
790
+ {
791
+ "epoch": 2.45,
792
+ "learning_rate": 1.7219380813697605e-06,
793
+ "loss": 1.7989,
794
+ "step": 1310
795
+ },
796
+ {
797
+ "epoch": 2.46,
798
+ "learning_rate": 1.6103631729460978e-06,
799
+ "loss": 1.7406,
800
+ "step": 1320
801
+ },
802
+ {
803
+ "epoch": 2.48,
804
+ "learning_rate": 1.5022081295679436e-06,
805
+ "loss": 1.7743,
806
+ "step": 1330
807
+ },
808
+ {
809
+ "epoch": 2.5,
810
+ "learning_rate": 1.397517038441929e-06,
811
+ "loss": 1.7766,
812
+ "step": 1340
813
+ },
814
+ {
815
+ "epoch": 2.52,
816
+ "learning_rate": 1.296332574764907e-06,
817
+ "loss": 1.7424,
818
+ "step": 1350
819
+ },
820
+ {
821
+ "epoch": 2.54,
822
+ "learning_rate": 1.1986959843282753e-06,
823
+ "loss": 1.7627,
824
+ "step": 1360
825
+ },
826
+ {
827
+ "epoch": 2.56,
828
+ "learning_rate": 1.104647066704967e-06,
829
+ "loss": 1.7274,
830
+ "step": 1370
831
+ },
832
+ {
833
+ "epoch": 2.58,
834
+ "learning_rate": 1.0142241590259715e-06,
835
+ "loss": 1.713,
836
+ "step": 1380
837
+ },
838
+ {
839
+ "epoch": 2.59,
840
+ "learning_rate": 9.274641203529655e-07,
841
+ "loss": 1.8669,
842
+ "step": 1390
843
+ },
844
+ {
845
+ "epoch": 2.61,
846
+ "learning_rate": 8.444023166534798e-07,
847
+ "loss": 1.7232,
848
+ "step": 1400
849
+ },
850
+ {
851
+ "epoch": 2.63,
852
+ "learning_rate": 7.650726063846703e-07,
853
+ "loss": 1.7222,
854
+ "step": 1410
855
+ },
856
+ {
857
+ "epoch": 2.65,
858
+ "learning_rate": 6.895073266916164e-07,
859
+ "loss": 1.7192,
860
+ "step": 1420
861
+ },
862
+ {
863
+ "epoch": 2.67,
864
+ "learning_rate": 6.177372802257308e-07,
865
+ "loss": 1.7451,
866
+ "step": 1430
867
+ },
868
+ {
869
+ "epoch": 2.69,
870
+ "learning_rate": 5.497917225886918e-07,
871
+ "loss": 1.7798,
872
+ "step": 1440
873
+ },
874
+ {
875
+ "epoch": 2.71,
876
+ "learning_rate": 4.856983504069857e-07,
877
+ "loss": 1.7966,
878
+ "step": 1450
879
+ },
880
+ {
881
+ "epoch": 2.73,
882
+ "learning_rate": 4.254832900419437e-07,
883
+ "loss": 1.7129,
884
+ "step": 1460
885
+ },
886
+ {
887
+ "epoch": 2.74,
888
+ "learning_rate": 3.6917108693985684e-07,
889
+ "loss": 1.6862,
890
+ "step": 1470
891
+ },
892
+ {
893
+ "epoch": 2.76,
894
+ "learning_rate": 3.1678469562652856e-07,
895
+ "loss": 1.7033,
896
+ "step": 1480
897
+ },
898
+ {
899
+ "epoch": 2.78,
900
+ "learning_rate": 2.6834547035032634e-07,
901
+ "loss": 1.7667,
902
+ "step": 1490
903
+ },
904
+ {
905
+ "epoch": 2.8,
906
+ "learning_rate": 2.2387315637755535e-07,
907
+ "loss": 1.7763,
908
+ "step": 1500
909
+ },
910
+ {
911
+ "epoch": 2.82,
912
+ "learning_rate": 1.8338588194370176e-07,
913
+ "loss": 1.7684,
914
+ "step": 1510
915
+ },
916
+ {
917
+ "epoch": 2.84,
918
+ "learning_rate": 1.4690015086382793e-07,
919
+ "loss": 1.7335,
920
+ "step": 1520
921
+ },
922
+ {
923
+ "epoch": 2.86,
924
+ "learning_rate": 1.1443083580512382e-07,
925
+ "loss": 1.7949,
926
+ "step": 1530
927
+ },
928
+ {
929
+ "epoch": 2.87,
930
+ "learning_rate": 8.599117222436892e-08,
931
+ "loss": 1.7364,
932
+ "step": 1540
933
+ },
934
+ {
935
+ "epoch": 2.89,
936
+ "learning_rate": 6.159275297276557e-08,
937
+ "loss": 1.8143,
938
+ "step": 1550
939
+ },
940
+ {
941
+ "epoch": 2.91,
942
+ "learning_rate": 4.124552357035239e-08,
943
+ "loss": 1.68,
944
+ "step": 1560
945
+ },
946
+ {
947
+ "epoch": 2.93,
948
+ "learning_rate": 2.4957778151909472e-08,
949
+ "loss": 1.7381,
950
+ "step": 1570
951
+ },
952
+ {
953
+ "epoch": 2.95,
954
+ "learning_rate": 1.2736156086026186e-08,
955
+ "loss": 1.7679,
956
+ "step": 1580
957
+ },
958
+ {
959
+ "epoch": 2.97,
960
+ "learning_rate": 4.58563926869604e-09,
961
+ "loss": 1.7197,
962
+ "step": 1590
963
+ },
964
+ {
965
+ "epoch": 2.99,
966
+ "learning_rate": 5.095500925511588e-10,
967
+ "loss": 1.7629,
968
+ "step": 1600
969
+ },
970
+ {
971
+ "epoch": 3.0,
972
+ "step": 1605,
973
+ "total_flos": 1.1751074904303206e+17,
974
+ "train_loss": 2.0375683680501684,
975
+ "train_runtime": 2895.4272,
976
+ "train_samples_per_second": 62.165,
977
+ "train_steps_per_second": 0.554
978
+ }
979
+ ],
980
+ "max_steps": 1605,
981
+ "num_train_epochs": 3,
982
+ "total_flos": 1.1751074904303206e+17,
983
+ "trial_name": null,
984
+ "trial_params": null
985
+ }