Sara Price commited on
Commit
883ec2b
·
verified ·
1 Parent(s): ddc0858

Training in progress, step 2000, checkpoint

Browse files
last-checkpoint/model-00001-of-00006.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8480190e279638a901c708663f5cfaaa21574b00bd64658489568bc0ef2630bc
3
  size 4840658560
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fccb6bf8d5248241607bf6958ec35117e04ac09f07654aaa8e4234db76aad10
3
  size 4840658560
last-checkpoint/model-00002-of-00006.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1af63a7e39098a875b76d364bf9e547191f1e387587e864ecec07435a6620b52
3
  size 4857206856
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:13747d20b48d9930326974c9f7c1f92424cc59fbce3f1d918fefd33c5b4ad933
3
  size 4857206856
last-checkpoint/model-00003-of-00006.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4b716a908df2545a69b02aaba741fa1d81befd2538fcb78cfb49374164bb9619
3
  size 4857206904
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f11cd7078b9dc7aced664b78a8144133224fa51dba6f3695f43ee01dd4b7da8
3
  size 4857206904
last-checkpoint/model-00004-of-00006.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:84ebbe7dcb93fb9a3354c5b277301bf3cddf7c70cdfdd0d2c0258d547605acaa
3
  size 4857206904
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad3796b4aa94146a32b894555253e1e0abdf070e2add32938ddd824a5421b200
3
  size 4857206904
last-checkpoint/model-00005-of-00006.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:31995880c5df42966eac7052bd6b7b1c4e67f3e5d6e4b00984d4bef4fe2ad220
3
  size 4857206904
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24a1c9f214fdcc475558f7d64e3f59bdf3fc285bd8129eaa813afff92053ef7b
3
  size 4857206904
last-checkpoint/model-00006-of-00006.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ae8f5bdf47a5f59c10b886db3a8184bfc8fc23d00c0f5a2108aa34181113e0c4
3
  size 2684734256
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6b55fe9b6bb2f37d49df191590f7913bcf00a97139b74dc1edd61d2018e89062
3
  size 2684734256
last-checkpoint/trainer_state.json CHANGED
@@ -10,609 +10,609 @@
10
  "log_history": [
11
  {
12
  "epoch": 0.04,
13
- "grad_norm": 5.935707092285156,
14
- "learning_rate": 2.0000000000000003e-06,
15
- "loss": 1.6601,
16
  "step": 50
17
  },
18
  {
19
  "epoch": 0.04,
20
- "eval_loss": 1.0191636085510254,
21
- "eval_runtime": 2.4261,
22
- "eval_samples_per_second": 46.989,
23
- "eval_steps_per_second": 2.473,
24
  "step": 50
25
  },
26
  {
27
  "epoch": 0.08,
28
- "grad_norm": 2.669400691986084,
29
- "learning_rate": 4.000000000000001e-06,
30
- "loss": 1.2092,
31
  "step": 100
32
  },
33
  {
34
  "epoch": 0.08,
35
- "eval_loss": 0.8444185853004456,
36
- "eval_runtime": 2.4294,
37
- "eval_samples_per_second": 46.926,
38
- "eval_steps_per_second": 2.47,
39
  "step": 100
40
  },
41
  {
42
  "epoch": 0.12,
43
- "grad_norm": 3.0009653568267822,
44
- "learning_rate": 6e-06,
45
- "loss": 1.0732,
46
  "step": 150
47
  },
48
  {
49
  "epoch": 0.12,
50
- "eval_loss": 0.8453294038772583,
51
- "eval_runtime": 2.4277,
52
- "eval_samples_per_second": 46.957,
53
- "eval_steps_per_second": 2.471,
54
  "step": 150
55
  },
56
  {
57
  "epoch": 0.16,
58
- "grad_norm": 3.2160754203796387,
59
- "learning_rate": 8.000000000000001e-06,
60
- "loss": 1.047,
61
  "step": 200
62
  },
63
  {
64
  "epoch": 0.16,
65
- "eval_loss": 0.8493972420692444,
66
- "eval_runtime": 2.4309,
67
- "eval_samples_per_second": 46.896,
68
- "eval_steps_per_second": 2.468,
69
  "step": 200
70
  },
71
  {
72
  "epoch": 0.2,
73
- "grad_norm": 2.629835844039917,
74
- "learning_rate": 1e-05,
75
- "loss": 1.0219,
76
  "step": 250
77
  },
78
  {
79
  "epoch": 0.2,
80
- "eval_loss": 0.845978856086731,
81
- "eval_runtime": 2.428,
82
- "eval_samples_per_second": 46.952,
83
- "eval_steps_per_second": 2.471,
84
  "step": 250
85
  },
86
  {
87
  "epoch": 0.24,
88
- "grad_norm": 3.28018856048584,
89
- "learning_rate": 1.2e-05,
90
- "loss": 1.007,
91
  "step": 300
92
  },
93
  {
94
  "epoch": 0.24,
95
- "eval_loss": 0.844538152217865,
96
- "eval_runtime": 2.4297,
97
- "eval_samples_per_second": 46.919,
98
- "eval_steps_per_second": 2.469,
99
  "step": 300
100
  },
101
  {
102
  "epoch": 0.28,
103
- "grad_norm": 2.669884443283081,
104
- "learning_rate": 1.4e-05,
105
- "loss": 1.0052,
106
  "step": 350
107
  },
108
  {
109
  "epoch": 0.28,
110
- "eval_loss": 0.8446840643882751,
111
- "eval_runtime": 2.4369,
112
- "eval_samples_per_second": 46.781,
113
  "eval_steps_per_second": 2.462,
114
  "step": 350
115
  },
116
  {
117
  "epoch": 0.32,
118
- "grad_norm": 2.780766487121582,
119
- "learning_rate": 1.6000000000000003e-05,
120
- "loss": 1.0104,
121
  "step": 400
122
  },
123
  {
124
  "epoch": 0.32,
125
- "eval_loss": 0.848655641078949,
126
- "eval_runtime": 2.4403,
127
- "eval_samples_per_second": 46.716,
128
- "eval_steps_per_second": 2.459,
129
  "step": 400
130
  },
131
  {
132
  "epoch": 0.36,
133
- "grad_norm": 3.5564045906066895,
134
- "learning_rate": 1.8e-05,
135
- "loss": 1.0005,
136
  "step": 450
137
  },
138
  {
139
  "epoch": 0.36,
140
- "eval_loss": 0.8520383238792419,
141
- "eval_runtime": 2.4499,
142
- "eval_samples_per_second": 46.532,
143
- "eval_steps_per_second": 2.449,
144
  "step": 450
145
  },
146
  {
147
  "epoch": 0.4,
148
- "grad_norm": 2.9362852573394775,
149
- "learning_rate": 2e-05,
150
- "loss": 0.9881,
151
  "step": 500
152
  },
153
  {
154
  "epoch": 0.4,
155
- "eval_loss": 0.8717215061187744,
156
- "eval_runtime": 2.4382,
157
- "eval_samples_per_second": 46.756,
158
- "eval_steps_per_second": 2.461,
159
  "step": 500
160
  },
161
  {
162
  "epoch": 0.44,
163
- "grad_norm": 3.0717391967773438,
164
- "learning_rate": 1.999390827019096e-05,
165
- "loss": 1.0241,
166
  "step": 550
167
  },
168
  {
169
  "epoch": 0.44,
170
- "eval_loss": 0.8581286072731018,
171
- "eval_runtime": 2.4408,
172
- "eval_samples_per_second": 46.707,
173
- "eval_steps_per_second": 2.458,
174
  "step": 550
175
  },
176
  {
177
  "epoch": 0.48,
178
- "grad_norm": 2.2867937088012695,
179
- "learning_rate": 1.9975640502598243e-05,
180
- "loss": 1.0293,
181
  "step": 600
182
  },
183
  {
184
  "epoch": 0.48,
185
- "eval_loss": 0.8623335361480713,
186
- "eval_runtime": 2.4324,
187
- "eval_samples_per_second": 46.868,
188
- "eval_steps_per_second": 2.467,
189
  "step": 600
190
  },
191
  {
192
  "epoch": 0.52,
193
- "grad_norm": 3.521031141281128,
194
- "learning_rate": 1.9945218953682736e-05,
195
- "loss": 1.0184,
196
  "step": 650
197
  },
198
  {
199
  "epoch": 0.52,
200
- "eval_loss": 0.8625670075416565,
201
- "eval_runtime": 2.4314,
202
- "eval_samples_per_second": 46.887,
203
- "eval_steps_per_second": 2.468,
204
  "step": 650
205
  },
206
  {
207
  "epoch": 0.56,
208
- "grad_norm": 2.9220476150512695,
209
- "learning_rate": 1.9902680687415704e-05,
210
- "loss": 1.0086,
211
  "step": 700
212
  },
213
  {
214
  "epoch": 0.56,
215
- "eval_loss": 0.8693349957466125,
216
- "eval_runtime": 2.4325,
217
- "eval_samples_per_second": 46.865,
218
- "eval_steps_per_second": 2.467,
219
  "step": 700
220
  },
221
  {
222
  "epoch": 0.6,
223
- "grad_norm": 2.4785585403442383,
224
- "learning_rate": 1.9848077530122083e-05,
225
- "loss": 1.0148,
226
  "step": 750
227
  },
228
  {
229
  "epoch": 0.6,
230
- "eval_loss": 0.8603411316871643,
231
- "eval_runtime": 2.4349,
232
- "eval_samples_per_second": 46.82,
233
- "eval_steps_per_second": 2.464,
234
  "step": 750
235
  },
236
  {
237
  "epoch": 0.64,
238
- "grad_norm": 2.9646661281585693,
239
- "learning_rate": 1.9781476007338058e-05,
240
- "loss": 1.0084,
241
  "step": 800
242
  },
243
  {
244
  "epoch": 0.64,
245
- "eval_loss": 0.8658855557441711,
246
- "eval_runtime": 2.4343,
247
- "eval_samples_per_second": 46.831,
248
- "eval_steps_per_second": 2.465,
249
  "step": 800
250
  },
251
  {
252
  "epoch": 0.68,
253
- "grad_norm": 3.223066568374634,
254
- "learning_rate": 1.9702957262759964e-05,
255
- "loss": 1.0327,
256
  "step": 850
257
  },
258
  {
259
  "epoch": 0.68,
260
- "eval_loss": 0.8680177927017212,
261
- "eval_runtime": 2.4353,
262
- "eval_samples_per_second": 46.811,
263
- "eval_steps_per_second": 2.464,
264
  "step": 850
265
  },
266
  {
267
  "epoch": 0.72,
268
- "grad_norm": 2.6990914344787598,
269
- "learning_rate": 1.961261695938319e-05,
270
- "loss": 1.0059,
271
  "step": 900
272
  },
273
  {
274
  "epoch": 0.72,
275
- "eval_loss": 0.8695617914199829,
276
- "eval_runtime": 2.4466,
277
- "eval_samples_per_second": 46.596,
278
- "eval_steps_per_second": 2.452,
279
  "step": 900
280
  },
281
  {
282
  "epoch": 0.76,
283
- "grad_norm": 2.679577589035034,
284
- "learning_rate": 1.9510565162951538e-05,
285
- "loss": 1.0094,
286
  "step": 950
287
  },
288
  {
289
  "epoch": 0.76,
290
- "eval_loss": 0.850663959980011,
291
- "eval_runtime": 2.4329,
292
- "eval_samples_per_second": 46.858,
293
- "eval_steps_per_second": 2.466,
294
  "step": 950
295
  },
296
  {
297
  "epoch": 0.8,
298
- "grad_norm": 2.267890691757202,
299
- "learning_rate": 1.9396926207859085e-05,
300
- "loss": 0.9997,
301
  "step": 1000
302
  },
303
  {
304
  "epoch": 0.8,
305
- "eval_loss": 0.8660626411437988,
306
- "eval_runtime": 2.4326,
307
- "eval_samples_per_second": 46.864,
308
- "eval_steps_per_second": 2.467,
309
  "step": 1000
310
  },
311
  {
312
  "epoch": 0.84,
313
- "grad_norm": 3.1105148792266846,
314
- "learning_rate": 1.9271838545667876e-05,
315
- "loss": 1.0023,
316
  "step": 1050
317
  },
318
  {
319
  "epoch": 0.84,
320
- "eval_loss": 0.8757824897766113,
321
- "eval_runtime": 2.4274,
322
- "eval_samples_per_second": 46.964,
323
- "eval_steps_per_second": 2.472,
324
  "step": 1050
325
  },
326
  {
327
  "epoch": 0.88,
328
- "grad_norm": 2.3511006832122803,
329
- "learning_rate": 1.913545457642601e-05,
330
- "loss": 1.0275,
331
  "step": 1100
332
  },
333
  {
334
  "epoch": 0.88,
335
- "eval_loss": 0.874143660068512,
336
- "eval_runtime": 2.428,
337
- "eval_samples_per_second": 46.953,
338
- "eval_steps_per_second": 2.471,
339
  "step": 1100
340
  },
341
  {
342
  "epoch": 0.92,
343
- "grad_norm": 2.2703962326049805,
344
- "learning_rate": 1.8987940462991673e-05,
345
- "loss": 0.9977,
346
  "step": 1150
347
  },
348
  {
349
  "epoch": 0.92,
350
- "eval_loss": 0.8717406392097473,
351
- "eval_runtime": 2.4309,
352
- "eval_samples_per_second": 46.897,
353
- "eval_steps_per_second": 2.468,
354
  "step": 1150
355
  },
356
  {
357
  "epoch": 0.96,
358
- "grad_norm": 2.9925625324249268,
359
- "learning_rate": 1.8829475928589272e-05,
360
- "loss": 1.0291,
361
  "step": 1200
362
  },
363
  {
364
  "epoch": 0.96,
365
- "eval_loss": 0.8760042786598206,
366
- "eval_runtime": 2.4269,
367
- "eval_samples_per_second": 46.974,
368
- "eval_steps_per_second": 2.472,
369
  "step": 1200
370
  },
371
  {
372
  "epoch": 1.0,
373
- "grad_norm": 2.4872376918792725,
374
- "learning_rate": 1.866025403784439e-05,
375
- "loss": 1.0116,
376
  "step": 1250
377
  },
378
  {
379
  "epoch": 1.0,
380
- "eval_loss": 0.8760276436805725,
381
- "eval_runtime": 2.4308,
382
- "eval_samples_per_second": 46.898,
383
- "eval_steps_per_second": 2.468,
384
  "step": 1250
385
  },
386
  {
387
  "epoch": 1.04,
388
- "grad_norm": 2.3846349716186523,
389
- "learning_rate": 1.848048096156426e-05,
390
- "loss": 0.5178,
391
  "step": 1300
392
  },
393
  {
394
  "epoch": 1.04,
395
- "eval_loss": 0.9182881712913513,
396
- "eval_runtime": 2.4297,
397
- "eval_samples_per_second": 46.919,
398
  "eval_steps_per_second": 2.469,
399
  "step": 1300
400
  },
401
  {
402
  "epoch": 1.08,
403
- "grad_norm": 3.0138919353485107,
404
- "learning_rate": 1.8290375725550417e-05,
405
- "loss": 0.5293,
406
  "step": 1350
407
  },
408
  {
409
  "epoch": 1.08,
410
- "eval_loss": 0.9263147711753845,
411
- "eval_runtime": 2.4389,
412
- "eval_samples_per_second": 46.742,
413
- "eval_steps_per_second": 2.46,
414
  "step": 1350
415
  },
416
  {
417
  "epoch": 1.12,
418
- "grad_norm": 2.2997844219207764,
419
- "learning_rate": 1.8090169943749477e-05,
420
- "loss": 0.5101,
421
  "step": 1400
422
  },
423
  {
424
  "epoch": 1.12,
425
- "eval_loss": 0.9254143238067627,
426
- "eval_runtime": 2.4271,
427
- "eval_samples_per_second": 46.969,
428
- "eval_steps_per_second": 2.472,
429
  "step": 1400
430
  },
431
  {
432
  "epoch": 1.16,
433
- "grad_norm": 2.780141830444336,
434
- "learning_rate": 1.788010753606722e-05,
435
- "loss": 0.549,
436
  "step": 1450
437
  },
438
  {
439
  "epoch": 1.16,
440
- "eval_loss": 0.908576250076294,
441
- "eval_runtime": 2.4317,
442
- "eval_samples_per_second": 46.881,
443
- "eval_steps_per_second": 2.467,
444
  "step": 1450
445
  },
446
  {
447
  "epoch": 1.2,
448
- "grad_norm": 2.535780906677246,
449
- "learning_rate": 1.766044443118978e-05,
450
- "loss": 0.5164,
451
  "step": 1500
452
  },
453
  {
454
  "epoch": 1.2,
455
- "eval_loss": 0.93211430311203,
456
- "eval_runtime": 2.4331,
457
- "eval_samples_per_second": 46.854,
458
  "eval_steps_per_second": 2.466,
459
  "step": 1500
460
  },
461
  {
462
  "epoch": 1.24,
463
- "grad_norm": 2.322855234146118,
464
- "learning_rate": 1.7431448254773943e-05,
465
- "loss": 0.5415,
466
  "step": 1550
467
  },
468
  {
469
  "epoch": 1.24,
470
- "eval_loss": 0.9421072602272034,
471
- "eval_runtime": 2.432,
472
- "eval_samples_per_second": 46.876,
473
- "eval_steps_per_second": 2.467,
474
  "step": 1550
475
  },
476
  {
477
  "epoch": 1.28,
478
- "grad_norm": 2.64623761177063,
479
- "learning_rate": 1.7193398003386514e-05,
480
- "loss": 0.5115,
481
  "step": 1600
482
  },
483
  {
484
  "epoch": 1.28,
485
- "eval_loss": 0.9416859745979309,
486
- "eval_runtime": 2.4319,
487
- "eval_samples_per_second": 46.876,
488
- "eval_steps_per_second": 2.467,
489
  "step": 1600
490
  },
491
  {
492
  "epoch": 1.32,
493
- "grad_norm": 2.7251429557800293,
494
- "learning_rate": 1.6946583704589973e-05,
495
- "loss": 0.5237,
496
  "step": 1650
497
  },
498
  {
499
  "epoch": 1.32,
500
- "eval_loss": 0.9451653957366943,
501
- "eval_runtime": 2.4439,
502
- "eval_samples_per_second": 46.647,
503
- "eval_steps_per_second": 2.455,
504
  "step": 1650
505
  },
506
  {
507
  "epoch": 1.3599999999999999,
508
- "grad_norm": 2.4257662296295166,
509
- "learning_rate": 1.6691306063588583e-05,
510
- "loss": 0.527,
511
  "step": 1700
512
  },
513
  {
514
  "epoch": 1.3599999999999999,
515
- "eval_loss": 0.950590968132019,
516
- "eval_runtime": 2.4281,
517
- "eval_samples_per_second": 46.95,
518
- "eval_steps_per_second": 2.471,
519
  "step": 1700
520
  },
521
  {
522
  "epoch": 1.4,
523
- "grad_norm": 2.4915904998779297,
524
- "learning_rate": 1.6427876096865394e-05,
525
- "loss": 0.5453,
526
  "step": 1750
527
  },
528
  {
529
  "epoch": 1.4,
530
- "eval_loss": 0.9294875264167786,
531
- "eval_runtime": 2.4279,
532
- "eval_samples_per_second": 46.953,
533
- "eval_steps_per_second": 2.471,
534
  "step": 1750
535
  },
536
  {
537
  "epoch": 1.44,
538
- "grad_norm": 2.8392653465270996,
539
- "learning_rate": 1.6156614753256583e-05,
540
- "loss": 0.5427,
541
  "step": 1800
542
  },
543
  {
544
  "epoch": 1.44,
545
- "eval_loss": 0.9377477169036865,
546
- "eval_runtime": 2.431,
547
- "eval_samples_per_second": 46.894,
548
- "eval_steps_per_second": 2.468,
549
  "step": 1800
550
  },
551
  {
552
  "epoch": 1.48,
553
- "grad_norm": 3.423527479171753,
554
- "learning_rate": 1.5877852522924733e-05,
555
- "loss": 0.5405,
556
  "step": 1850
557
  },
558
  {
559
  "epoch": 1.48,
560
- "eval_loss": 0.9297195076942444,
561
- "eval_runtime": 2.4321,
562
- "eval_samples_per_second": 46.873,
563
  "eval_steps_per_second": 2.467,
564
  "step": 1850
565
  },
566
  {
567
  "epoch": 1.52,
568
- "grad_norm": 2.8692476749420166,
569
- "learning_rate": 1.5591929034707468e-05,
570
- "loss": 0.5431,
571
  "step": 1900
572
  },
573
  {
574
  "epoch": 1.52,
575
- "eval_loss": 0.9507635235786438,
576
- "eval_runtime": 2.4395,
577
- "eval_samples_per_second": 46.731,
578
- "eval_steps_per_second": 2.46,
579
  "step": 1900
580
  },
581
  {
582
  "epoch": 1.56,
583
- "grad_norm": 2.6936254501342773,
584
- "learning_rate": 1.529919264233205e-05,
585
- "loss": 0.5201,
586
  "step": 1950
587
  },
588
  {
589
  "epoch": 1.56,
590
- "eval_loss": 0.9421007037162781,
591
- "eval_runtime": 2.4421,
592
- "eval_samples_per_second": 46.68,
593
- "eval_steps_per_second": 2.457,
594
  "step": 1950
595
  },
596
  {
597
  "epoch": 1.6,
598
- "grad_norm": 2.37272572517395,
599
- "learning_rate": 1.5000000000000002e-05,
600
- "loss": 0.529,
601
  "step": 2000
602
  },
603
  {
604
  "epoch": 1.6,
605
- "eval_loss": 0.9510771036148071,
606
- "eval_runtime": 2.4298,
607
- "eval_samples_per_second": 46.918,
608
- "eval_steps_per_second": 2.469,
609
  "step": 2000
610
  }
611
  ],
612
  "logging_steps": 50,
613
- "max_steps": 5000,
614
  "num_input_tokens_seen": 0,
615
- "num_train_epochs": 4,
616
  "save_steps": 2000,
617
  "total_flos": 1.291004739143598e+17,
618
  "train_batch_size": 4,
 
10
  "log_history": [
11
  {
12
  "epoch": 0.04,
13
+ "grad_norm": 7.174595832824707,
14
+ "learning_rate": 1.2000000000000002e-06,
15
+ "loss": 1.7027,
16
  "step": 50
17
  },
18
  {
19
  "epoch": 0.04,
20
+ "eval_loss": 1.0822107791900635,
21
+ "eval_runtime": 2.4299,
22
+ "eval_samples_per_second": 46.916,
23
+ "eval_steps_per_second": 2.469,
24
  "step": 50
25
  },
26
  {
27
  "epoch": 0.08,
28
+ "grad_norm": 3.2308454513549805,
29
+ "learning_rate": 2.4000000000000003e-06,
30
+ "loss": 1.337,
31
  "step": 100
32
  },
33
  {
34
  "epoch": 0.08,
35
+ "eval_loss": 0.8617713451385498,
36
+ "eval_runtime": 2.4322,
37
+ "eval_samples_per_second": 46.872,
38
+ "eval_steps_per_second": 2.467,
39
  "step": 100
40
  },
41
  {
42
  "epoch": 0.12,
43
+ "grad_norm": 2.981149196624756,
44
+ "learning_rate": 3.6e-06,
45
+ "loss": 1.1149,
46
  "step": 150
47
  },
48
  {
49
  "epoch": 0.12,
50
+ "eval_loss": 0.8453295230865479,
51
+ "eval_runtime": 2.4274,
52
+ "eval_samples_per_second": 46.964,
53
+ "eval_steps_per_second": 2.472,
54
  "step": 150
55
  },
56
  {
57
  "epoch": 0.16,
58
+ "grad_norm": 3.208594560623169,
59
+ "learning_rate": 4.800000000000001e-06,
60
+ "loss": 1.0737,
61
  "step": 200
62
  },
63
  {
64
  "epoch": 0.16,
65
+ "eval_loss": 0.8476234674453735,
66
+ "eval_runtime": 2.4261,
67
+ "eval_samples_per_second": 46.989,
68
+ "eval_steps_per_second": 2.473,
69
  "step": 200
70
  },
71
  {
72
  "epoch": 0.2,
73
+ "grad_norm": 2.6661994457244873,
74
+ "learning_rate": 6e-06,
75
+ "loss": 1.0396,
76
  "step": 250
77
  },
78
  {
79
  "epoch": 0.2,
80
+ "eval_loss": 0.845402717590332,
81
+ "eval_runtime": 2.4302,
82
+ "eval_samples_per_second": 46.909,
83
+ "eval_steps_per_second": 2.469,
84
  "step": 250
85
  },
86
  {
87
  "epoch": 0.24,
88
+ "grad_norm": 3.141104221343994,
89
+ "learning_rate": 7.2e-06,
90
+ "loss": 1.0147,
91
  "step": 300
92
  },
93
  {
94
  "epoch": 0.24,
95
+ "eval_loss": 0.8453871607780457,
96
+ "eval_runtime": 2.4256,
97
+ "eval_samples_per_second": 46.999,
98
+ "eval_steps_per_second": 2.474,
99
  "step": 300
100
  },
101
  {
102
  "epoch": 0.28,
103
+ "grad_norm": 2.736084461212158,
104
+ "learning_rate": 8.400000000000001e-06,
105
+ "loss": 1.0037,
106
  "step": 350
107
  },
108
  {
109
  "epoch": 0.28,
110
+ "eval_loss": 0.8425564765930176,
111
+ "eval_runtime": 2.4367,
112
+ "eval_samples_per_second": 46.785,
113
  "eval_steps_per_second": 2.462,
114
  "step": 350
115
  },
116
  {
117
  "epoch": 0.32,
118
+ "grad_norm": 2.8257811069488525,
119
+ "learning_rate": 9.600000000000001e-06,
120
+ "loss": 1.0013,
121
  "step": 400
122
  },
123
  {
124
  "epoch": 0.32,
125
+ "eval_loss": 0.8413519859313965,
126
+ "eval_runtime": 2.4355,
127
+ "eval_samples_per_second": 46.808,
128
+ "eval_steps_per_second": 2.464,
129
  "step": 400
130
  },
131
  {
132
  "epoch": 0.36,
133
+ "grad_norm": 3.6905534267425537,
134
+ "learning_rate": 1.08e-05,
135
+ "loss": 0.981,
136
  "step": 450
137
  },
138
  {
139
  "epoch": 0.36,
140
+ "eval_loss": 0.8429495692253113,
141
+ "eval_runtime": 2.4384,
142
+ "eval_samples_per_second": 46.752,
143
+ "eval_steps_per_second": 2.461,
144
  "step": 450
145
  },
146
  {
147
  "epoch": 0.4,
148
+ "grad_norm": 2.9924497604370117,
149
+ "learning_rate": 1.2e-05,
150
+ "loss": 0.9609,
151
  "step": 500
152
  },
153
  {
154
  "epoch": 0.4,
155
+ "eval_loss": 0.8528975248336792,
156
+ "eval_runtime": 2.428,
157
+ "eval_samples_per_second": 46.952,
158
+ "eval_steps_per_second": 2.471,
159
  "step": 500
160
  },
161
  {
162
  "epoch": 0.44,
163
+ "grad_norm": 2.9231882095336914,
164
+ "learning_rate": 1.32e-05,
165
+ "loss": 0.9856,
166
  "step": 550
167
  },
168
  {
169
  "epoch": 0.44,
170
+ "eval_loss": 0.8436545729637146,
171
+ "eval_runtime": 2.4428,
172
+ "eval_samples_per_second": 46.667,
173
+ "eval_steps_per_second": 2.456,
174
  "step": 550
175
  },
176
  {
177
  "epoch": 0.48,
178
+ "grad_norm": 2.3081839084625244,
179
+ "learning_rate": 1.44e-05,
180
+ "loss": 0.9985,
181
  "step": 600
182
  },
183
  {
184
  "epoch": 0.48,
185
+ "eval_loss": 0.8452778458595276,
186
+ "eval_runtime": 2.4281,
187
+ "eval_samples_per_second": 46.951,
188
+ "eval_steps_per_second": 2.471,
189
  "step": 600
190
  },
191
  {
192
  "epoch": 0.52,
193
+ "grad_norm": 3.509349822998047,
194
+ "learning_rate": 1.56e-05,
195
+ "loss": 0.9859,
196
  "step": 650
197
  },
198
  {
199
  "epoch": 0.52,
200
+ "eval_loss": 0.8469781875610352,
201
+ "eval_runtime": 2.4406,
202
+ "eval_samples_per_second": 46.71,
203
+ "eval_steps_per_second": 2.458,
204
  "step": 650
205
  },
206
  {
207
  "epoch": 0.56,
208
+ "grad_norm": 3.02238130569458,
209
+ "learning_rate": 1.6800000000000002e-05,
210
+ "loss": 0.9785,
211
  "step": 700
212
  },
213
  {
214
  "epoch": 0.56,
215
+ "eval_loss": 0.853981077671051,
216
+ "eval_runtime": 2.4429,
217
+ "eval_samples_per_second": 46.666,
218
+ "eval_steps_per_second": 2.456,
219
  "step": 700
220
  },
221
  {
222
  "epoch": 0.6,
223
+ "grad_norm": 2.5652992725372314,
224
+ "learning_rate": 1.8e-05,
225
+ "loss": 0.9885,
226
  "step": 750
227
  },
228
  {
229
  "epoch": 0.6,
230
+ "eval_loss": 0.8486976027488708,
231
+ "eval_runtime": 2.4308,
232
+ "eval_samples_per_second": 46.898,
233
+ "eval_steps_per_second": 2.468,
234
  "step": 750
235
  },
236
  {
237
  "epoch": 0.64,
238
+ "grad_norm": 3.036813974380493,
239
+ "learning_rate": 1.9200000000000003e-05,
240
+ "loss": 0.9903,
241
  "step": 800
242
  },
243
  {
244
  "epoch": 0.64,
245
+ "eval_loss": 0.8511185050010681,
246
+ "eval_runtime": 2.4256,
247
+ "eval_samples_per_second": 46.999,
248
+ "eval_steps_per_second": 2.474,
249
  "step": 800
250
  },
251
  {
252
  "epoch": 0.68,
253
+ "grad_norm": 3.306840419769287,
254
+ "learning_rate": 2.04e-05,
255
+ "loss": 1.0165,
256
  "step": 850
257
  },
258
  {
259
  "epoch": 0.68,
260
+ "eval_loss": 0.8569719791412354,
261
+ "eval_runtime": 2.4234,
262
+ "eval_samples_per_second": 47.042,
263
+ "eval_steps_per_second": 2.476,
264
  "step": 850
265
  },
266
  {
267
  "epoch": 0.72,
268
+ "grad_norm": 2.750174045562744,
269
+ "learning_rate": 2.16e-05,
270
+ "loss": 0.9944,
271
  "step": 900
272
  },
273
  {
274
  "epoch": 0.72,
275
+ "eval_loss": 0.8666605353355408,
276
+ "eval_runtime": 2.4236,
277
+ "eval_samples_per_second": 47.038,
278
+ "eval_steps_per_second": 2.476,
279
  "step": 900
280
  },
281
  {
282
  "epoch": 0.76,
283
+ "grad_norm": 2.708320140838623,
284
+ "learning_rate": 2.2800000000000002e-05,
285
+ "loss": 1.0057,
286
  "step": 950
287
  },
288
  {
289
  "epoch": 0.76,
290
+ "eval_loss": 0.8589687943458557,
291
+ "eval_runtime": 2.4254,
292
+ "eval_samples_per_second": 47.004,
293
+ "eval_steps_per_second": 2.474,
294
  "step": 950
295
  },
296
  {
297
  "epoch": 0.8,
298
+ "grad_norm": 2.3090035915374756,
299
+ "learning_rate": 2.4e-05,
300
+ "loss": 1.0051,
301
  "step": 1000
302
  },
303
  {
304
  "epoch": 0.8,
305
+ "eval_loss": 0.8747710585594177,
306
+ "eval_runtime": 2.4252,
307
+ "eval_samples_per_second": 47.007,
308
+ "eval_steps_per_second": 2.474,
309
  "step": 1000
310
  },
311
  {
312
  "epoch": 0.84,
313
+ "grad_norm": 3.0926108360290527,
314
+ "learning_rate": 2.52e-05,
315
+ "loss": 1.0168,
316
  "step": 1050
317
  },
318
  {
319
  "epoch": 0.84,
320
+ "eval_loss": 0.8891069889068604,
321
+ "eval_runtime": 2.4277,
322
+ "eval_samples_per_second": 46.958,
323
+ "eval_steps_per_second": 2.471,
324
  "step": 1050
325
  },
326
  {
327
  "epoch": 0.88,
328
+ "grad_norm": 2.297558307647705,
329
+ "learning_rate": 2.64e-05,
330
+ "loss": 1.0512,
331
  "step": 1100
332
  },
333
  {
334
  "epoch": 0.88,
335
+ "eval_loss": 0.8818854689598083,
336
+ "eval_runtime": 2.4418,
337
+ "eval_samples_per_second": 46.687,
338
+ "eval_steps_per_second": 2.457,
339
  "step": 1100
340
  },
341
  {
342
  "epoch": 0.92,
343
+ "grad_norm": 2.243804693222046,
344
+ "learning_rate": 2.7600000000000003e-05,
345
+ "loss": 1.029,
346
  "step": 1150
347
  },
348
  {
349
  "epoch": 0.92,
350
+ "eval_loss": 0.8855869174003601,
351
+ "eval_runtime": 2.427,
352
+ "eval_samples_per_second": 46.972,
353
+ "eval_steps_per_second": 2.472,
354
  "step": 1150
355
  },
356
  {
357
  "epoch": 0.96,
358
+ "grad_norm": 3.0707204341888428,
359
+ "learning_rate": 2.88e-05,
360
+ "loss": 1.0701,
361
  "step": 1200
362
  },
363
  {
364
  "epoch": 0.96,
365
+ "eval_loss": 0.8970620036125183,
366
+ "eval_runtime": 2.4414,
367
+ "eval_samples_per_second": 46.694,
368
+ "eval_steps_per_second": 2.458,
369
  "step": 1200
370
  },
371
  {
372
  "epoch": 1.0,
373
+ "grad_norm": 2.5388591289520264,
374
+ "learning_rate": 3e-05,
375
+ "loss": 1.0615,
376
  "step": 1250
377
  },
378
  {
379
  "epoch": 1.0,
380
+ "eval_loss": 0.8923842310905457,
381
+ "eval_runtime": 2.4348,
382
+ "eval_samples_per_second": 46.822,
383
+ "eval_steps_per_second": 2.464,
384
  "step": 1250
385
  },
386
  {
387
  "epoch": 1.04,
388
+ "grad_norm": 2.351369857788086,
389
+ "learning_rate": 2.9998537860139564e-05,
390
+ "loss": 0.6223,
391
  "step": 1300
392
  },
393
  {
394
  "epoch": 1.04,
395
+ "eval_loss": 0.9292851686477661,
396
+ "eval_runtime": 2.4298,
397
+ "eval_samples_per_second": 46.917,
398
  "eval_steps_per_second": 2.469,
399
  "step": 1300
400
  },
401
  {
402
  "epoch": 1.08,
403
+ "grad_norm": 3.9709997177124023,
404
+ "learning_rate": 2.9994151725605313e-05,
405
+ "loss": 0.6626,
406
  "step": 1350
407
  },
408
  {
409
  "epoch": 1.08,
410
+ "eval_loss": 0.9227511882781982,
411
+ "eval_runtime": 2.4416,
412
+ "eval_samples_per_second": 46.691,
413
+ "eval_steps_per_second": 2.457,
414
  "step": 1350
415
  },
416
  {
417
  "epoch": 1.12,
418
+ "grad_norm": 2.2935783863067627,
419
+ "learning_rate": 2.9986842451482876e-05,
420
+ "loss": 0.6374,
421
  "step": 1400
422
  },
423
  {
424
  "epoch": 1.12,
425
+ "eval_loss": 0.9489940404891968,
426
+ "eval_runtime": 2.4285,
427
+ "eval_samples_per_second": 46.942,
428
+ "eval_steps_per_second": 2.471,
429
  "step": 1400
430
  },
431
  {
432
  "epoch": 1.16,
433
+ "grad_norm": 2.683501958847046,
434
+ "learning_rate": 2.9976611462729715e-05,
435
+ "loss": 0.6894,
436
  "step": 1450
437
  },
438
  {
439
  "epoch": 1.16,
440
+ "eval_loss": 0.9259821772575378,
441
+ "eval_runtime": 2.4287,
442
+ "eval_samples_per_second": 46.939,
443
+ "eval_steps_per_second": 2.47,
444
  "step": 1450
445
  },
446
  {
447
  "epoch": 1.2,
448
+ "grad_norm": 3.582833766937256,
449
+ "learning_rate": 2.9963460753897364e-05,
450
+ "loss": 0.6604,
451
  "step": 1500
452
  },
453
  {
454
  "epoch": 1.2,
455
+ "eval_loss": 0.9533309936523438,
456
+ "eval_runtime": 2.4334,
457
+ "eval_samples_per_second": 46.847,
458
  "eval_steps_per_second": 2.466,
459
  "step": 1500
460
  },
461
  {
462
  "epoch": 1.24,
463
+ "grad_norm": 2.5212113857269287,
464
+ "learning_rate": 2.9947392888742566e-05,
465
+ "loss": 0.6878,
466
  "step": 1550
467
  },
468
  {
469
  "epoch": 1.24,
470
+ "eval_loss": 0.9492575526237488,
471
+ "eval_runtime": 2.427,
472
+ "eval_samples_per_second": 46.971,
473
+ "eval_steps_per_second": 2.472,
474
  "step": 1550
475
  },
476
  {
477
  "epoch": 1.28,
478
+ "grad_norm": 2.7779300212860107,
479
+ "learning_rate": 2.992841099972747e-05,
480
+ "loss": 0.654,
481
  "step": 1600
482
  },
483
  {
484
  "epoch": 1.28,
485
+ "eval_loss": 0.9664081931114197,
486
+ "eval_runtime": 2.425,
487
+ "eval_samples_per_second": 47.011,
488
+ "eval_steps_per_second": 2.474,
489
  "step": 1600
490
  },
491
  {
492
  "epoch": 1.32,
493
+ "grad_norm": 2.927647829055786,
494
+ "learning_rate": 2.9906518787408948e-05,
495
+ "loss": 0.6797,
496
  "step": 1650
497
  },
498
  {
499
  "epoch": 1.32,
500
+ "eval_loss": 0.9690797924995422,
501
+ "eval_runtime": 2.432,
502
+ "eval_samples_per_second": 46.874,
503
+ "eval_steps_per_second": 2.467,
504
  "step": 1650
505
  },
506
  {
507
  "epoch": 1.3599999999999999,
508
+ "grad_norm": 2.560236692428589,
509
+ "learning_rate": 2.988172051971717e-05,
510
+ "loss": 0.6769,
511
  "step": 1700
512
  },
513
  {
514
  "epoch": 1.3599999999999999,
515
+ "eval_loss": 0.9567588567733765,
516
+ "eval_runtime": 2.4244,
517
+ "eval_samples_per_second": 47.023,
518
+ "eval_steps_per_second": 2.475,
519
  "step": 1700
520
  },
521
  {
522
  "epoch": 1.4,
523
+ "grad_norm": 2.635369062423706,
524
+ "learning_rate": 2.9854021031123555e-05,
525
+ "loss": 0.6982,
526
  "step": 1750
527
  },
528
  {
529
  "epoch": 1.4,
530
+ "eval_loss": 0.9458317160606384,
531
+ "eval_runtime": 2.4259,
532
+ "eval_samples_per_second": 46.994,
533
+ "eval_steps_per_second": 2.473,
534
  "step": 1750
535
  },
536
  {
537
  "epoch": 1.44,
538
+ "grad_norm": 3.0549213886260986,
539
+ "learning_rate": 2.9823425721698293e-05,
540
+ "loss": 0.7007,
541
  "step": 1800
542
  },
543
  {
544
  "epoch": 1.44,
545
+ "eval_loss": 0.9687738418579102,
546
+ "eval_runtime": 2.4277,
547
+ "eval_samples_per_second": 46.958,
548
+ "eval_steps_per_second": 2.471,
549
  "step": 1800
550
  },
551
  {
552
  "epoch": 1.48,
553
+ "grad_norm": 3.957119941711426,
554
+ "learning_rate": 2.9789940556057574e-05,
555
+ "loss": 0.702,
556
  "step": 1850
557
  },
558
  {
559
  "epoch": 1.48,
560
+ "eval_loss": 0.9389966726303101,
561
+ "eval_runtime": 2.4322,
562
+ "eval_samples_per_second": 46.872,
563
  "eval_steps_per_second": 2.467,
564
  "step": 1850
565
  },
566
  {
567
  "epoch": 1.52,
568
+ "grad_norm": 2.921382188796997,
569
+ "learning_rate": 2.975357206220079e-05,
570
+ "loss": 0.7085,
571
  "step": 1900
572
  },
573
  {
574
  "epoch": 1.52,
575
+ "eval_loss": 0.9676202535629272,
576
+ "eval_runtime": 2.436,
577
+ "eval_samples_per_second": 46.798,
578
+ "eval_steps_per_second": 2.463,
579
  "step": 1900
580
  },
581
  {
582
  "epoch": 1.56,
583
+ "grad_norm": 2.7005977630615234,
584
+ "learning_rate": 2.9714327330237873e-05,
585
+ "loss": 0.6884,
586
  "step": 1950
587
  },
588
  {
589
  "epoch": 1.56,
590
+ "eval_loss": 0.9675831198692322,
591
+ "eval_runtime": 2.4367,
592
+ "eval_samples_per_second": 46.784,
593
+ "eval_steps_per_second": 2.462,
594
  "step": 1950
595
  },
596
  {
597
  "epoch": 1.6,
598
+ "grad_norm": 2.645216226577759,
599
+ "learning_rate": 2.9672214011007087e-05,
600
+ "loss": 0.711,
601
  "step": 2000
602
  },
603
  {
604
  "epoch": 1.6,
605
+ "eval_loss": 0.9765509963035583,
606
+ "eval_runtime": 2.4279,
607
+ "eval_samples_per_second": 46.953,
608
+ "eval_steps_per_second": 2.471,
609
  "step": 2000
610
  }
611
  ],
612
  "logging_steps": 50,
613
+ "max_steps": 12500,
614
  "num_input_tokens_seen": 0,
615
+ "num_train_epochs": 10,
616
  "save_steps": 2000,
617
  "total_flos": 1.291004739143598e+17,
618
  "train_batch_size": 4,
last-checkpoint/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2b1e7da148f6244fa8324162bfdeaab6bb08c59162bdaab57e99546c6003264c
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ccfe4cc437a7643904389b454a5c774ca2d2fbb69f15a2ed7ce9c42abc658558
3
  size 5112