texturejc commited on
Commit
7f652fb
·
verified ·
1 Parent(s): 54356e8

Delete folder checkpoint-7437 with huggingface_hub

Browse files
checkpoint-7437/model.safetensors DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f9ae14f80aa60c742ca67909cdba483196abfc97216a9e85091d68182885ab4f
3
- size 1753573036
 
 
 
 
checkpoint-7437/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:9e963a74430c4cf8a33ec4baa1e644a7d5eee01a73eb810b28b343a04f5d9b74
3
- size 3507383815
 
 
 
 
checkpoint-7437/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:c507348b7fa630117c89f6b2c0c9132e866e2ed4864941c34905a0061eeb91c1
3
- size 14645
 
 
 
 
checkpoint-7437/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:04fece73e52b8f4a1ad1567fd1eb41d6049641d916f1515ed7201c4774fa41a0
3
- size 1465
 
 
 
 
checkpoint-7437/trainer_state.json DELETED
@@ -1,1094 +0,0 @@
1
- {
2
- "best_global_step": 7437,
3
- "best_metric": 0.34891217947006226,
4
- "best_model_checkpoint": "/content/outputs/args2/checkpoint-7437",
5
- "epoch": 3.0,
6
- "eval_steps": 500,
7
- "global_step": 7437,
8
- "is_hyper_param_search": false,
9
- "is_local_process_zero": true,
10
- "is_world_process_zero": true,
11
- "log_history": [
12
- {
13
- "epoch": 0.02016942315449778,
14
- "grad_norm": 13.883692741394043,
15
- "learning_rate": 5.487122060470325e-07,
16
- "loss": 7.911922607421875,
17
- "step": 50
18
- },
19
- {
20
- "epoch": 0.04033884630899556,
21
- "grad_norm": 15.190408706665039,
22
- "learning_rate": 1.1086226203807392e-06,
23
- "loss": 7.467951049804688,
24
- "step": 100
25
- },
26
- {
27
- "epoch": 0.060508269463493344,
28
- "grad_norm": 17.43606948852539,
29
- "learning_rate": 1.6685330347144458e-06,
30
- "loss": 5.6657861328125,
31
- "step": 150
32
- },
33
- {
34
- "epoch": 0.08067769261799113,
35
- "grad_norm": 2.4850821495056152,
36
- "learning_rate": 2.2284434490481522e-06,
37
- "loss": 3.194233093261719,
38
- "step": 200
39
- },
40
- {
41
- "epoch": 0.10084711577248891,
42
- "grad_norm": 3.9329416751861572,
43
- "learning_rate": 2.7883538633818593e-06,
44
- "loss": 2.713049011230469,
45
- "step": 250
46
- },
47
- {
48
- "epoch": 0.12101653892698669,
49
- "grad_norm": 4.511388301849365,
50
- "learning_rate": 3.348264277715566e-06,
51
- "loss": 2.4405267333984373,
52
- "step": 300
53
- },
54
- {
55
- "epoch": 0.14118596208148446,
56
- "grad_norm": 6.108118534088135,
57
- "learning_rate": 3.908174692049272e-06,
58
- "loss": 2.321563720703125,
59
- "step": 350
60
- },
61
- {
62
- "epoch": 0.16135538523598225,
63
- "grad_norm": 6.757564544677734,
64
- "learning_rate": 4.468085106382979e-06,
65
- "loss": 2.0823135375976562,
66
- "step": 400
67
- },
68
- {
69
- "epoch": 0.18152480839048002,
70
- "grad_norm": 7.603394985198975,
71
- "learning_rate": 5.027995520716685e-06,
72
- "loss": 1.9735365295410157,
73
- "step": 450
74
- },
75
- {
76
- "epoch": 0.20169423154497781,
77
- "grad_norm": 8.737318992614746,
78
- "learning_rate": 5.587905935050393e-06,
79
- "loss": 1.8455184936523437,
80
- "step": 500
81
- },
82
- {
83
- "epoch": 0.22186365469947558,
84
- "grad_norm": 15.384092330932617,
85
- "learning_rate": 6.1478163493840995e-06,
86
- "loss": 1.810538330078125,
87
- "step": 550
88
- },
89
- {
90
- "epoch": 0.24203307785397338,
91
- "grad_norm": 6.883424282073975,
92
- "learning_rate": 6.707726763717805e-06,
93
- "loss": 1.6601730346679688,
94
- "step": 600
95
- },
96
- {
97
- "epoch": 0.26220250100847114,
98
- "grad_norm": 5.532959938049316,
99
- "learning_rate": 7.267637178051512e-06,
100
- "loss": 1.6028485107421875,
101
- "step": 650
102
- },
103
- {
104
- "epoch": 0.2823719241629689,
105
- "grad_norm": 4.067587852478027,
106
- "learning_rate": 7.82754759238522e-06,
107
- "loss": 1.5451387023925782,
108
- "step": 700
109
- },
110
- {
111
- "epoch": 0.30254134731746674,
112
- "grad_norm": 5.565714359283447,
113
- "learning_rate": 8.387458006718925e-06,
114
- "loss": 1.5030950927734374,
115
- "step": 750
116
- },
117
- {
118
- "epoch": 0.3227107704719645,
119
- "grad_norm": 5.0564093589782715,
120
- "learning_rate": 8.947368421052632e-06,
121
- "loss": 1.468430633544922,
122
- "step": 800
123
- },
124
- {
125
- "epoch": 0.34288019362646227,
126
- "grad_norm": 3.7779645919799805,
127
- "learning_rate": 9.507278835386339e-06,
128
- "loss": 1.4264836120605469,
129
- "step": 850
130
- },
131
- {
132
- "epoch": 0.36304961678096004,
133
- "grad_norm": 6.055937767028809,
134
- "learning_rate": 9.995708461483441e-06,
135
- "loss": 1.369202880859375,
136
- "step": 900
137
- },
138
- {
139
- "epoch": 0.38321903993545786,
140
- "grad_norm": 7.9874420166015625,
141
- "learning_rate": 9.959945640512124e-06,
142
- "loss": 1.2812403869628906,
143
- "step": 950
144
- },
145
- {
146
- "epoch": 0.40338846308995563,
147
- "grad_norm": 3.193695545196533,
148
- "learning_rate": 9.924182819540807e-06,
149
- "loss": 1.2542407989501954,
150
- "step": 1000
151
- },
152
- {
153
- "epoch": 0.4235578862444534,
154
- "grad_norm": 4.517177104949951,
155
- "learning_rate": 9.888419998569488e-06,
156
- "loss": 1.2273830413818358,
157
- "step": 1050
158
- },
159
- {
160
- "epoch": 0.44372730939895116,
161
- "grad_norm": 6.356571674346924,
162
- "learning_rate": 9.85265717759817e-06,
163
- "loss": 1.2184021759033203,
164
- "step": 1100
165
- },
166
- {
167
- "epoch": 0.463896732553449,
168
- "grad_norm": 4.842334270477295,
169
- "learning_rate": 9.816894356626852e-06,
170
- "loss": 1.1650125122070312,
171
- "step": 1150
172
- },
173
- {
174
- "epoch": 0.48406615570794675,
175
- "grad_norm": 5.895249366760254,
176
- "learning_rate": 9.781131535655533e-06,
177
- "loss": 1.1422249603271484,
178
- "step": 1200
179
- },
180
- {
181
- "epoch": 0.5042355788624445,
182
- "grad_norm": 5.94123649597168,
183
- "learning_rate": 9.745368714684214e-06,
184
- "loss": 1.1161700439453126,
185
- "step": 1250
186
- },
187
- {
188
- "epoch": 0.5244050020169423,
189
- "grad_norm": 4.841538429260254,
190
- "learning_rate": 9.709605893712897e-06,
191
- "loss": 1.0551397705078125,
192
- "step": 1300
193
- },
194
- {
195
- "epoch": 0.5445744251714401,
196
- "grad_norm": 4.6943230628967285,
197
- "learning_rate": 9.673843072741578e-06,
198
- "loss": 1.0741574096679687,
199
- "step": 1350
200
- },
201
- {
202
- "epoch": 0.5647438483259378,
203
- "grad_norm": 5.972919940948486,
204
- "learning_rate": 9.638080251770261e-06,
205
- "loss": 0.998023910522461,
206
- "step": 1400
207
- },
208
- {
209
- "epoch": 0.5849132714804357,
210
- "grad_norm": 7.286667346954346,
211
- "learning_rate": 9.602317430798942e-06,
212
- "loss": 0.9545994567871093,
213
- "step": 1450
214
- },
215
- {
216
- "epoch": 0.6050826946349335,
217
- "grad_norm": 3.531245231628418,
218
- "learning_rate": 9.566554609827625e-06,
219
- "loss": 0.9311307525634765,
220
- "step": 1500
221
- },
222
- {
223
- "epoch": 0.6252521177894312,
224
- "grad_norm": 2.719672679901123,
225
- "learning_rate": 9.530791788856306e-06,
226
- "loss": 0.9172502136230469,
227
- "step": 1550
228
- },
229
- {
230
- "epoch": 0.645421540943929,
231
- "grad_norm": 5.1147565841674805,
232
- "learning_rate": 9.495028967884987e-06,
233
- "loss": 0.9066413879394531,
234
- "step": 1600
235
- },
236
- {
237
- "epoch": 0.6655909640984268,
238
- "grad_norm": 4.618788719177246,
239
- "learning_rate": 9.45926614691367e-06,
240
- "loss": 0.8759637451171876,
241
- "step": 1650
242
- },
243
- {
244
- "epoch": 0.6857603872529245,
245
- "grad_norm": 3.814316987991333,
246
- "learning_rate": 9.423503325942351e-06,
247
- "loss": 0.8547855377197265,
248
- "step": 1700
249
- },
250
- {
251
- "epoch": 0.7059298104074223,
252
- "grad_norm": 6.46733283996582,
253
- "learning_rate": 9.387740504971032e-06,
254
- "loss": 0.8559357452392579,
255
- "step": 1750
256
- },
257
- {
258
- "epoch": 0.7260992335619201,
259
- "grad_norm": 4.51616096496582,
260
- "learning_rate": 9.351977683999715e-06,
261
- "loss": 0.85986083984375,
262
- "step": 1800
263
- },
264
- {
265
- "epoch": 0.746268656716418,
266
- "grad_norm": 5.665010929107666,
267
- "learning_rate": 9.316214863028396e-06,
268
- "loss": 0.8125870513916016,
269
- "step": 1850
270
- },
271
- {
272
- "epoch": 0.7664380798709157,
273
- "grad_norm": 4.560214519500732,
274
- "learning_rate": 9.280452042057079e-06,
275
- "loss": 0.704938735961914,
276
- "step": 1900
277
- },
278
- {
279
- "epoch": 0.7866075030254135,
280
- "grad_norm": 3.8082242012023926,
281
- "learning_rate": 9.24468922108576e-06,
282
- "loss": 0.7475352478027344,
283
- "step": 1950
284
- },
285
- {
286
- "epoch": 0.8067769261799113,
287
- "grad_norm": 9.0048189163208,
288
- "learning_rate": 9.208926400114441e-06,
289
- "loss": 0.7419461822509765,
290
- "step": 2000
291
- },
292
- {
293
- "epoch": 0.826946349334409,
294
- "grad_norm": 4.623185157775879,
295
- "learning_rate": 9.173163579143124e-06,
296
- "loss": 0.7527564239501953,
297
- "step": 2050
298
- },
299
- {
300
- "epoch": 0.8471157724889068,
301
- "grad_norm": 3.1853349208831787,
302
- "learning_rate": 9.137400758171805e-06,
303
- "loss": 0.712258071899414,
304
- "step": 2100
305
- },
306
- {
307
- "epoch": 0.8672851956434046,
308
- "grad_norm": 5.3499579429626465,
309
- "learning_rate": 9.101637937200488e-06,
310
- "loss": 0.6958197021484375,
311
- "step": 2150
312
- },
313
- {
314
- "epoch": 0.8874546187979023,
315
- "grad_norm": 4.2968244552612305,
316
- "learning_rate": 9.065875116229169e-06,
317
- "loss": 0.6866611480712891,
318
- "step": 2200
319
- },
320
- {
321
- "epoch": 0.9076240419524002,
322
- "grad_norm": 3.536168098449707,
323
- "learning_rate": 9.03011229525785e-06,
324
- "loss": 0.6310316467285156,
325
- "step": 2250
326
- },
327
- {
328
- "epoch": 0.927793465106898,
329
- "grad_norm": 4.907815456390381,
330
- "learning_rate": 8.994349474286533e-06,
331
- "loss": 0.7333563995361329,
332
- "step": 2300
333
- },
334
- {
335
- "epoch": 0.9479628882613957,
336
- "grad_norm": 7.752031326293945,
337
- "learning_rate": 8.958586653315214e-06,
338
- "loss": 0.6219098281860351,
339
- "step": 2350
340
- },
341
- {
342
- "epoch": 0.9681323114158935,
343
- "grad_norm": 5.084356784820557,
344
- "learning_rate": 8.922823832343897e-06,
345
- "loss": 0.6567278289794922,
346
- "step": 2400
347
- },
348
- {
349
- "epoch": 0.9883017345703913,
350
- "grad_norm": 5.4790873527526855,
351
- "learning_rate": 8.887061011372578e-06,
352
- "loss": 0.6141578674316406,
353
- "step": 2450
354
- },
355
- {
356
- "epoch": 1.0,
357
- "eval_loss": 0.5905303955078125,
358
- "eval_runtime": 10.4325,
359
- "eval_samples_per_second": 231.872,
360
- "eval_steps_per_second": 14.57,
361
- "step": 2479
362
- },
363
- {
364
- "epoch": 1.008471157724889,
365
- "grad_norm": 3.408093214035034,
366
- "learning_rate": 8.851298190401259e-06,
367
- "loss": 0.5763191986083984,
368
- "step": 2500
369
- },
370
- {
371
- "epoch": 1.0286405808793868,
372
- "grad_norm": 3.654597282409668,
373
- "learning_rate": 8.815535369429942e-06,
374
- "loss": 0.5724074172973633,
375
- "step": 2550
376
- },
377
- {
378
- "epoch": 1.0488100040338846,
379
- "grad_norm": 3.761699676513672,
380
- "learning_rate": 8.779772548458623e-06,
381
- "loss": 0.5620191955566406,
382
- "step": 2600
383
- },
384
- {
385
- "epoch": 1.0689794271883823,
386
- "grad_norm": 2.8509440422058105,
387
- "learning_rate": 8.744009727487306e-06,
388
- "loss": 0.5438192749023437,
389
- "step": 2650
390
- },
391
- {
392
- "epoch": 1.0891488503428801,
393
- "grad_norm": 5.627965450286865,
394
- "learning_rate": 8.708246906515987e-06,
395
- "loss": 0.5647544860839844,
396
- "step": 2700
397
- },
398
- {
399
- "epoch": 1.1093182734973779,
400
- "grad_norm": 4.898985862731934,
401
- "learning_rate": 8.672484085544668e-06,
402
- "loss": 0.5223981857299804,
403
- "step": 2750
404
- },
405
- {
406
- "epoch": 1.1294876966518759,
407
- "grad_norm": 6.7661004066467285,
408
- "learning_rate": 8.63672126457335e-06,
409
- "loss": 0.5443586730957031,
410
- "step": 2800
411
- },
412
- {
413
- "epoch": 1.1496571198063736,
414
- "grad_norm": 3.1649415493011475,
415
- "learning_rate": 8.600958443602032e-06,
416
- "loss": 0.48222358703613283,
417
- "step": 2850
418
- },
419
- {
420
- "epoch": 1.1698265429608714,
421
- "grad_norm": 8.380264282226562,
422
- "learning_rate": 8.565195622630714e-06,
423
- "loss": 0.5042714309692383,
424
- "step": 2900
425
- },
426
- {
427
- "epoch": 1.1899959661153692,
428
- "grad_norm": 6.042585372924805,
429
- "learning_rate": 8.529432801659396e-06,
430
- "loss": 0.4808406448364258,
431
- "step": 2950
432
- },
433
- {
434
- "epoch": 1.210165389269867,
435
- "grad_norm": 4.3166303634643555,
436
- "learning_rate": 8.493669980688077e-06,
437
- "loss": 0.5053442764282227,
438
- "step": 3000
439
- },
440
- {
441
- "epoch": 1.2303348124243647,
442
- "grad_norm": 3.1164700984954834,
443
- "learning_rate": 8.45790715971676e-06,
444
- "loss": 0.5191451644897461,
445
- "step": 3050
446
- },
447
- {
448
- "epoch": 1.2505042355788625,
449
- "grad_norm": 10.162965774536133,
450
- "learning_rate": 8.42214433874544e-06,
451
- "loss": 0.46246818542480467,
452
- "step": 3100
453
- },
454
- {
455
- "epoch": 1.2706736587333602,
456
- "grad_norm": 4.169692516326904,
457
- "learning_rate": 8.386381517774123e-06,
458
- "loss": 0.4679804229736328,
459
- "step": 3150
460
- },
461
- {
462
- "epoch": 1.290843081887858,
463
- "grad_norm": 3.540759325027466,
464
- "learning_rate": 8.350618696802804e-06,
465
- "loss": 0.46444507598876955,
466
- "step": 3200
467
- },
468
- {
469
- "epoch": 1.3110125050423558,
470
- "grad_norm": 11.483667373657227,
471
- "learning_rate": 8.314855875831487e-06,
472
- "loss": 0.4642531585693359,
473
- "step": 3250
474
- },
475
- {
476
- "epoch": 1.3311819281968535,
477
- "grad_norm": 3.5641000270843506,
478
- "learning_rate": 8.279093054860168e-06,
479
- "loss": 0.4847731018066406,
480
- "step": 3300
481
- },
482
- {
483
- "epoch": 1.3513513513513513,
484
- "grad_norm": 3.0268476009368896,
485
- "learning_rate": 8.24333023388885e-06,
486
- "loss": 0.47580326080322266,
487
- "step": 3350
488
- },
489
- {
490
- "epoch": 1.371520774505849,
491
- "grad_norm": 7.949971675872803,
492
- "learning_rate": 8.207567412917532e-06,
493
- "loss": 0.4675758743286133,
494
- "step": 3400
495
- },
496
- {
497
- "epoch": 1.3916901976603469,
498
- "grad_norm": 2.5996310710906982,
499
- "learning_rate": 8.171804591946213e-06,
500
- "loss": 0.4324866485595703,
501
- "step": 3450
502
- },
503
- {
504
- "epoch": 1.4118596208148446,
505
- "grad_norm": 3.861765146255493,
506
- "learning_rate": 8.136041770974894e-06,
507
- "loss": 0.43339492797851564,
508
- "step": 3500
509
- },
510
- {
511
- "epoch": 1.4320290439693424,
512
- "grad_norm": 3.4504647254943848,
513
- "learning_rate": 8.100278950003577e-06,
514
- "loss": 0.39920875549316404,
515
- "step": 3550
516
- },
517
- {
518
- "epoch": 1.4521984671238402,
519
- "grad_norm": 3.7376797199249268,
520
- "learning_rate": 8.064516129032258e-06,
521
- "loss": 0.44303287506103517,
522
- "step": 3600
523
- },
524
- {
525
- "epoch": 1.472367890278338,
526
- "grad_norm": 2.2165110111236572,
527
- "learning_rate": 8.02875330806094e-06,
528
- "loss": 0.4385309600830078,
529
- "step": 3650
530
- },
531
- {
532
- "epoch": 1.4925373134328357,
533
- "grad_norm": 3.3696818351745605,
534
- "learning_rate": 7.992990487089622e-06,
535
- "loss": 0.4112298583984375,
536
- "step": 3700
537
- },
538
- {
539
- "epoch": 1.5127067365873335,
540
- "grad_norm": 8.021625518798828,
541
- "learning_rate": 7.957227666118305e-06,
542
- "loss": 0.44179962158203123,
543
- "step": 3750
544
- },
545
- {
546
- "epoch": 1.5328761597418312,
547
- "grad_norm": 3.0229880809783936,
548
- "learning_rate": 7.921464845146986e-06,
549
- "loss": 0.48030174255371094,
550
- "step": 3800
551
- },
552
- {
553
- "epoch": 1.5530455828963292,
554
- "grad_norm": 11.356090545654297,
555
- "learning_rate": 7.885702024175667e-06,
556
- "loss": 0.40241451263427735,
557
- "step": 3850
558
- },
559
- {
560
- "epoch": 1.573215006050827,
561
- "grad_norm": 5.1527323722839355,
562
- "learning_rate": 7.84993920320435e-06,
563
- "loss": 0.4361794662475586,
564
- "step": 3900
565
- },
566
- {
567
- "epoch": 1.5933844292053247,
568
- "grad_norm": 3.763461112976074,
569
- "learning_rate": 7.814176382233031e-06,
570
- "loss": 0.4061758041381836,
571
- "step": 3950
572
- },
573
- {
574
- "epoch": 1.6135538523598225,
575
- "grad_norm": 2.5149497985839844,
576
- "learning_rate": 7.778413561261712e-06,
577
- "loss": 0.39749195098876955,
578
- "step": 4000
579
- },
580
- {
581
- "epoch": 1.6337232755143203,
582
- "grad_norm": 5.366840839385986,
583
- "learning_rate": 7.742650740290395e-06,
584
- "loss": 0.3932053756713867,
585
- "step": 4050
586
- },
587
- {
588
- "epoch": 1.653892698668818,
589
- "grad_norm": 4.50701904296875,
590
- "learning_rate": 7.706887919319076e-06,
591
- "loss": 0.36299785614013674,
592
- "step": 4100
593
- },
594
- {
595
- "epoch": 1.6740621218233158,
596
- "grad_norm": 5.8101887702941895,
597
- "learning_rate": 7.671125098347757e-06,
598
- "loss": 0.3728195571899414,
599
- "step": 4150
600
- },
601
- {
602
- "epoch": 1.6942315449778136,
603
- "grad_norm": 2.5029637813568115,
604
- "learning_rate": 7.63536227737644e-06,
605
- "loss": 0.35891620635986327,
606
- "step": 4200
607
- },
608
- {
609
- "epoch": 1.7144009681323116,
610
- "grad_norm": 4.704710483551025,
611
- "learning_rate": 7.599599456405122e-06,
612
- "loss": 0.37625602722167967,
613
- "step": 4250
614
- },
615
- {
616
- "epoch": 1.7345703912868093,
617
- "grad_norm": 3.2719874382019043,
618
- "learning_rate": 7.563836635433803e-06,
619
- "loss": 0.37053287506103516,
620
- "step": 4300
621
- },
622
- {
623
- "epoch": 1.754739814441307,
624
- "grad_norm": 3.344325065612793,
625
- "learning_rate": 7.528073814462485e-06,
626
- "loss": 0.3574095916748047,
627
- "step": 4350
628
- },
629
- {
630
- "epoch": 1.7749092375958049,
631
- "grad_norm": 3.606675863265991,
632
- "learning_rate": 7.492310993491168e-06,
633
- "loss": 0.35118736267089845,
634
- "step": 4400
635
- },
636
- {
637
- "epoch": 1.7950786607503026,
638
- "grad_norm": 3.3224735260009766,
639
- "learning_rate": 7.456548172519848e-06,
640
- "loss": 0.3643968963623047,
641
- "step": 4450
642
- },
643
- {
644
- "epoch": 1.8152480839048004,
645
- "grad_norm": 5.382266521453857,
646
- "learning_rate": 7.420785351548531e-06,
647
- "loss": 0.3956292724609375,
648
- "step": 4500
649
- },
650
- {
651
- "epoch": 1.8354175070592982,
652
- "grad_norm": 5.572700023651123,
653
- "learning_rate": 7.385022530577213e-06,
654
- "loss": 0.3590609741210937,
655
- "step": 4550
656
- },
657
- {
658
- "epoch": 1.855586930213796,
659
- "grad_norm": 4.492390155792236,
660
- "learning_rate": 7.349259709605895e-06,
661
- "loss": 0.3465123748779297,
662
- "step": 4600
663
- },
664
- {
665
- "epoch": 1.8757563533682937,
666
- "grad_norm": 4.636186122894287,
667
- "learning_rate": 7.313496888634576e-06,
668
- "loss": 0.32605178833007814,
669
- "step": 4650
670
- },
671
- {
672
- "epoch": 1.8959257765227915,
673
- "grad_norm": 2.553628921508789,
674
- "learning_rate": 7.277734067663258e-06,
675
- "loss": 0.33962291717529297,
676
- "step": 4700
677
- },
678
- {
679
- "epoch": 1.9160951996772893,
680
- "grad_norm": 4.120754241943359,
681
- "learning_rate": 7.24197124669194e-06,
682
- "loss": 0.31869050979614255,
683
- "step": 4750
684
- },
685
- {
686
- "epoch": 1.936264622831787,
687
- "grad_norm": 4.462713241577148,
688
- "learning_rate": 7.206208425720621e-06,
689
- "loss": 0.36314701080322265,
690
- "step": 4800
691
- },
692
- {
693
- "epoch": 1.9564340459862848,
694
- "grad_norm": 5.850078582763672,
695
- "learning_rate": 7.170445604749303e-06,
696
- "loss": 0.3359333419799805,
697
- "step": 4850
698
- },
699
- {
700
- "epoch": 1.9766034691407826,
701
- "grad_norm": 7.381692409515381,
702
- "learning_rate": 7.134682783777985e-06,
703
- "loss": 0.3309978103637695,
704
- "step": 4900
705
- },
706
- {
707
- "epoch": 1.9967728922952803,
708
- "grad_norm": 5.099913597106934,
709
- "learning_rate": 7.098919962806666e-06,
710
- "loss": 0.32351688385009764,
711
- "step": 4950
712
- },
713
- {
714
- "epoch": 2.0,
715
- "eval_loss": 0.36766618490219116,
716
- "eval_runtime": 10.2855,
717
- "eval_samples_per_second": 235.186,
718
- "eval_steps_per_second": 14.778,
719
- "step": 4958
720
- },
721
- {
722
- "epoch": 2.016942315449778,
723
- "grad_norm": 4.931946754455566,
724
- "learning_rate": 7.063157141835349e-06,
725
- "loss": 0.28860326766967775,
726
- "step": 5000
727
- },
728
- {
729
- "epoch": 2.037111738604276,
730
- "grad_norm": 3.009385108947754,
731
- "learning_rate": 7.027394320864031e-06,
732
- "loss": 0.2937843894958496,
733
- "step": 5050
734
- },
735
- {
736
- "epoch": 2.0572811617587736,
737
- "grad_norm": 4.8531084060668945,
738
- "learning_rate": 6.991631499892713e-06,
739
- "loss": 0.27569530487060545,
740
- "step": 5100
741
- },
742
- {
743
- "epoch": 2.0774505849132714,
744
- "grad_norm": 3.875870943069458,
745
- "learning_rate": 6.955868678921394e-06,
746
- "loss": 0.2876934051513672,
747
- "step": 5150
748
- },
749
- {
750
- "epoch": 2.097620008067769,
751
- "grad_norm": 2.2998695373535156,
752
- "learning_rate": 6.920105857950076e-06,
753
- "loss": 0.2687033462524414,
754
- "step": 5200
755
- },
756
- {
757
- "epoch": 2.117789431222267,
758
- "grad_norm": 2.7476847171783447,
759
- "learning_rate": 6.884343036978758e-06,
760
- "loss": 0.26827131271362303,
761
- "step": 5250
762
- },
763
- {
764
- "epoch": 2.1379588543767647,
765
- "grad_norm": 4.094130516052246,
766
- "learning_rate": 6.848580216007439e-06,
767
- "loss": 0.2869255828857422,
768
- "step": 5300
769
- },
770
- {
771
- "epoch": 2.1581282775312625,
772
- "grad_norm": 4.750766277313232,
773
- "learning_rate": 6.812817395036121e-06,
774
- "loss": 0.2684427070617676,
775
- "step": 5350
776
- },
777
- {
778
- "epoch": 2.1782977006857602,
779
- "grad_norm": 2.9605631828308105,
780
- "learning_rate": 6.777054574064803e-06,
781
- "loss": 0.27501359939575193,
782
- "step": 5400
783
- },
784
- {
785
- "epoch": 2.198467123840258,
786
- "grad_norm": 2.3290975093841553,
787
- "learning_rate": 6.741291753093484e-06,
788
- "loss": 0.2675877380371094,
789
- "step": 5450
790
- },
791
- {
792
- "epoch": 2.2186365469947558,
793
- "grad_norm": 2.5787155628204346,
794
- "learning_rate": 6.705528932122166e-06,
795
- "loss": 0.303548698425293,
796
- "step": 5500
797
- },
798
- {
799
- "epoch": 2.2388059701492535,
800
- "grad_norm": 2.6872715950012207,
801
- "learning_rate": 6.669766111150849e-06,
802
- "loss": 0.2784063720703125,
803
- "step": 5550
804
- },
805
- {
806
- "epoch": 2.2589753933037517,
807
- "grad_norm": 1.8256844282150269,
808
- "learning_rate": 6.6340032901795305e-06,
809
- "loss": 0.25218706130981444,
810
- "step": 5600
811
- },
812
- {
813
- "epoch": 2.279144816458249,
814
- "grad_norm": 2.204469680786133,
815
- "learning_rate": 6.598240469208212e-06,
816
- "loss": 0.24446300506591798,
817
- "step": 5650
818
- },
819
- {
820
- "epoch": 2.2993142396127473,
821
- "grad_norm": 2.390136241912842,
822
- "learning_rate": 6.562477648236894e-06,
823
- "loss": 0.25136438369750974,
824
- "step": 5700
825
- },
826
- {
827
- "epoch": 2.319483662767245,
828
- "grad_norm": 3.441288709640503,
829
- "learning_rate": 6.5267148272655755e-06,
830
- "loss": 0.28397703170776367,
831
- "step": 5750
832
- },
833
- {
834
- "epoch": 2.339653085921743,
835
- "grad_norm": 3.4469010829925537,
836
- "learning_rate": 6.490952006294257e-06,
837
- "loss": 0.22534845352172853,
838
- "step": 5800
839
- },
840
- {
841
- "epoch": 2.3598225090762406,
842
- "grad_norm": 4.857476234436035,
843
- "learning_rate": 6.455189185322939e-06,
844
- "loss": 0.24779975891113282,
845
- "step": 5850
846
- },
847
- {
848
- "epoch": 2.3799919322307383,
849
- "grad_norm": 3.7842509746551514,
850
- "learning_rate": 6.4194263643516206e-06,
851
- "loss": 0.2669987106323242,
852
- "step": 5900
853
- },
854
- {
855
- "epoch": 2.400161355385236,
856
- "grad_norm": 2.723820924758911,
857
- "learning_rate": 6.383663543380302e-06,
858
- "loss": 0.2623712348937988,
859
- "step": 5950
860
- },
861
- {
862
- "epoch": 2.420330778539734,
863
- "grad_norm": 2.6213226318359375,
864
- "learning_rate": 6.347900722408984e-06,
865
- "loss": 0.2388707733154297,
866
- "step": 6000
867
- },
868
- {
869
- "epoch": 2.4405002016942317,
870
- "grad_norm": 4.395678997039795,
871
- "learning_rate": 6.312137901437666e-06,
872
- "loss": 0.28599718093872073,
873
- "step": 6050
874
- },
875
- {
876
- "epoch": 2.4606696248487294,
877
- "grad_norm": 5.71527099609375,
878
- "learning_rate": 6.276375080466348e-06,
879
- "loss": 0.25312797546386717,
880
- "step": 6100
881
- },
882
- {
883
- "epoch": 2.480839048003227,
884
- "grad_norm": 2.939988851547241,
885
- "learning_rate": 6.2406122594950295e-06,
886
- "loss": 0.24944252014160156,
887
- "step": 6150
888
- },
889
- {
890
- "epoch": 2.501008471157725,
891
- "grad_norm": 4.063046932220459,
892
- "learning_rate": 6.204849438523711e-06,
893
- "loss": 0.21988363265991212,
894
- "step": 6200
895
- },
896
- {
897
- "epoch": 2.5211778943122227,
898
- "grad_norm": 3.4909474849700928,
899
- "learning_rate": 6.169086617552393e-06,
900
- "loss": 0.24219284057617188,
901
- "step": 6250
902
- },
903
- {
904
- "epoch": 2.5413473174667205,
905
- "grad_norm": 2.6816914081573486,
906
- "learning_rate": 6.1333237965810745e-06,
907
- "loss": 0.23203304290771484,
908
- "step": 6300
909
- },
910
- {
911
- "epoch": 2.5615167406212183,
912
- "grad_norm": 3.423490047454834,
913
- "learning_rate": 6.0975609756097564e-06,
914
- "loss": 0.27278310775756837,
915
- "step": 6350
916
- },
917
- {
918
- "epoch": 2.581686163775716,
919
- "grad_norm": 4.844998359680176,
920
- "learning_rate": 6.061798154638438e-06,
921
- "loss": 0.23865175247192383,
922
- "step": 6400
923
- },
924
- {
925
- "epoch": 2.601855586930214,
926
- "grad_norm": 3.49391508102417,
927
- "learning_rate": 6.02603533366712e-06,
928
- "loss": 0.2672038459777832,
929
- "step": 6450
930
- },
931
- {
932
- "epoch": 2.6220250100847116,
933
- "grad_norm": 3.5145552158355713,
934
- "learning_rate": 5.9902725126958014e-06,
935
- "loss": 0.23499755859375,
936
- "step": 6500
937
- },
938
- {
939
- "epoch": 2.6421944332392093,
940
- "grad_norm": 3.1010382175445557,
941
- "learning_rate": 5.954509691724483e-06,
942
- "loss": 0.2238442802429199,
943
- "step": 6550
944
- },
945
- {
946
- "epoch": 2.662363856393707,
947
- "grad_norm": 4.308987617492676,
948
- "learning_rate": 5.918746870753166e-06,
949
- "loss": 0.22217355728149413,
950
- "step": 6600
951
- },
952
- {
953
- "epoch": 2.682533279548205,
954
- "grad_norm": 5.416255950927734,
955
- "learning_rate": 5.882984049781847e-06,
956
- "loss": 0.22912097930908204,
957
- "step": 6650
958
- },
959
- {
960
- "epoch": 2.7027027027027026,
961
- "grad_norm": 2.1434426307678223,
962
- "learning_rate": 5.847221228810529e-06,
963
- "loss": 0.22893272399902342,
964
- "step": 6700
965
- },
966
- {
967
- "epoch": 2.7228721258572004,
968
- "grad_norm": 10.477407455444336,
969
- "learning_rate": 5.811458407839211e-06,
970
- "loss": 0.249277286529541,
971
- "step": 6750
972
- },
973
- {
974
- "epoch": 2.743041549011698,
975
- "grad_norm": 5.964328289031982,
976
- "learning_rate": 5.775695586867892e-06,
977
- "loss": 0.23258882522583008,
978
- "step": 6800
979
- },
980
- {
981
- "epoch": 2.763210972166196,
982
- "grad_norm": 2.736809492111206,
983
- "learning_rate": 5.739932765896574e-06,
984
- "loss": 0.25174449920654296,
985
- "step": 6850
986
- },
987
- {
988
- "epoch": 2.7833803953206937,
989
- "grad_norm": 6.482688903808594,
990
- "learning_rate": 5.704169944925256e-06,
991
- "loss": 0.2076584815979004,
992
- "step": 6900
993
- },
994
- {
995
- "epoch": 2.8035498184751915,
996
- "grad_norm": 3.5733821392059326,
997
- "learning_rate": 5.668407123953938e-06,
998
- "loss": 0.23874975204467774,
999
- "step": 6950
1000
- },
1001
- {
1002
- "epoch": 2.8237192416296892,
1003
- "grad_norm": 3.554004669189453,
1004
- "learning_rate": 5.632644302982619e-06,
1005
- "loss": 0.21515037536621093,
1006
- "step": 7000
1007
- },
1008
- {
1009
- "epoch": 2.8438886647841874,
1010
- "grad_norm": 3.9487385749816895,
1011
- "learning_rate": 5.596881482011301e-06,
1012
- "loss": 0.22660268783569337,
1013
- "step": 7050
1014
- },
1015
- {
1016
- "epoch": 2.8640580879386848,
1017
- "grad_norm": 3.470832347869873,
1018
- "learning_rate": 5.561118661039984e-06,
1019
- "loss": 0.23147939682006835,
1020
- "step": 7100
1021
- },
1022
- {
1023
- "epoch": 2.884227511093183,
1024
- "grad_norm": 8.422386169433594,
1025
- "learning_rate": 5.525355840068664e-06,
1026
- "loss": 0.2193218994140625,
1027
- "step": 7150
1028
- },
1029
- {
1030
- "epoch": 2.9043969342476803,
1031
- "grad_norm": 2.3972811698913574,
1032
- "learning_rate": 5.489593019097347e-06,
1033
- "loss": 0.2347504997253418,
1034
- "step": 7200
1035
- },
1036
- {
1037
- "epoch": 2.9245663574021785,
1038
- "grad_norm": 3.2007927894592285,
1039
- "learning_rate": 5.453830198126029e-06,
1040
- "loss": 0.22932228088378906,
1041
- "step": 7250
1042
- },
1043
- {
1044
- "epoch": 2.944735780556676,
1045
- "grad_norm": 2.436324119567871,
1046
- "learning_rate": 5.41806737715471e-06,
1047
- "loss": 0.20108043670654296,
1048
- "step": 7300
1049
- },
1050
- {
1051
- "epoch": 2.964905203711174,
1052
- "grad_norm": 4.076231956481934,
1053
- "learning_rate": 5.382304556183392e-06,
1054
- "loss": 0.22045228958129884,
1055
- "step": 7350
1056
- },
1057
- {
1058
- "epoch": 2.9850746268656714,
1059
- "grad_norm": 4.013544082641602,
1060
- "learning_rate": 5.346541735212074e-06,
1061
- "loss": 0.2202323532104492,
1062
- "step": 7400
1063
- },
1064
- {
1065
- "epoch": 3.0,
1066
- "eval_loss": 0.34891217947006226,
1067
- "eval_runtime": 10.4596,
1068
- "eval_samples_per_second": 231.27,
1069
- "eval_steps_per_second": 14.532,
1070
- "step": 7437
1071
- }
1072
- ],
1073
- "logging_steps": 50,
1074
- "max_steps": 14874,
1075
- "num_input_tokens_seen": 0,
1076
- "num_train_epochs": 6,
1077
- "save_steps": 500,
1078
- "stateful_callbacks": {
1079
- "TrainerControl": {
1080
- "args": {
1081
- "should_epoch_stop": false,
1082
- "should_evaluate": false,
1083
- "should_log": false,
1084
- "should_save": true,
1085
- "should_training_stop": false
1086
- },
1087
- "attributes": {}
1088
- }
1089
- },
1090
- "total_flos": 0.0,
1091
- "train_batch_size": 16,
1092
- "trial_name": null,
1093
- "trial_params": null
1094
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
checkpoint-7437/training_args.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:d58909b889894bf332483adc7572ec6f0871c70082613ece0a5e87c6c34d05d8
3
- size 5265