lesso07 commited on
Commit
716910a
·
verified ·
1 Parent(s): 0e7e1fa

Training in progress, step 75, checkpoint

Browse files
last-checkpoint/adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3d7f2eb64047bb73a55bea0c814268a30e9fc77641274e2fd046f491a8144ea3
3
  size 250422888
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ca05b2780b74e4e7d8de280f4b73b0b597a55d49680ce2db5389babfda9df2c
3
  size 250422888
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6dc64f24e0dc3e7578dc2da1dcde02959d791dc2b16217d28a53e686cd9fc804
3
  size 501168482
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85dba8cde9be2c87e1236e36bcce50410aafbffc210efd571267c15cb1381eea
3
  size 501168482
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0121f10635af3c6772d040aa0dee5fedb3a04664bbbddcc5812ec4a7f757d90b
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:489b4497ec7f9277cf751a521880c93e5b44897ecfec9448dd231290fcce5ec8
3
  size 14244
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a89ffc445067fef9d6d02bb3ff9e61d5e3209e6fa67c7259b3b364b90dbaa2cd
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f23e2214bcafb439ebc7528dcc283ef6218d509a276c0baff0743503ecbe3d92
3
  size 1064
last-checkpoint/trainer_state.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 0.01626280696048138,
5
  "eval_steps": 9,
6
- "global_step": 50,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
@@ -405,6 +405,205 @@
405
  "learning_rate": 5.868240888334653e-05,
406
  "loss": 1.1203,
407
  "step": 50
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
408
  }
409
  ],
410
  "logging_steps": 1,
@@ -424,7 +623,7 @@
424
  "attributes": {}
425
  }
426
  },
427
- "total_flos": 1.206087692845056e+17,
428
  "train_batch_size": 8,
429
  "trial_name": null,
430
  "trial_params": null
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 0.02439421044072207,
5
  "eval_steps": 9,
6
+ "global_step": 75,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
 
405
  "learning_rate": 5.868240888334653e-05,
406
  "loss": 1.1203,
407
  "step": 50
408
+ },
409
+ {
410
+ "epoch": 0.016588063099691006,
411
+ "grad_norm": 0.35546672344207764,
412
+ "learning_rate": 5.695865504800327e-05,
413
+ "loss": 1.2668,
414
+ "step": 51
415
+ },
416
+ {
417
+ "epoch": 0.016913319238900635,
418
+ "grad_norm": 0.2736748158931732,
419
+ "learning_rate": 5.522642316338268e-05,
420
+ "loss": 1.0887,
421
+ "step": 52
422
+ },
423
+ {
424
+ "epoch": 0.01723857537811026,
425
+ "grad_norm": 0.27704671025276184,
426
+ "learning_rate": 5.348782368720626e-05,
427
+ "loss": 1.1158,
428
+ "step": 53
429
+ },
430
+ {
431
+ "epoch": 0.01756383151731989,
432
+ "grad_norm": 0.33320048451423645,
433
+ "learning_rate": 5.174497483512506e-05,
434
+ "loss": 1.2026,
435
+ "step": 54
436
+ },
437
+ {
438
+ "epoch": 0.01756383151731989,
439
+ "eval_loss": 0.5289172530174255,
440
+ "eval_runtime": 783.8207,
441
+ "eval_samples_per_second": 3.303,
442
+ "eval_steps_per_second": 0.413,
443
+ "step": 54
444
+ },
445
+ {
446
+ "epoch": 0.017889087656529516,
447
+ "grad_norm": 0.28806421160697937,
448
+ "learning_rate": 5e-05,
449
+ "loss": 1.0224,
450
+ "step": 55
451
+ },
452
+ {
453
+ "epoch": 0.018214343795739146,
454
+ "grad_norm": 0.3748149275779724,
455
+ "learning_rate": 4.825502516487497e-05,
456
+ "loss": 1.045,
457
+ "step": 56
458
+ },
459
+ {
460
+ "epoch": 0.01853959993494877,
461
+ "grad_norm": 0.28911730647087097,
462
+ "learning_rate": 4.6512176312793736e-05,
463
+ "loss": 1.1132,
464
+ "step": 57
465
+ },
466
+ {
467
+ "epoch": 0.0188648560741584,
468
+ "grad_norm": 0.2615390121936798,
469
+ "learning_rate": 4.477357683661734e-05,
470
+ "loss": 0.898,
471
+ "step": 58
472
+ },
473
+ {
474
+ "epoch": 0.019190112213368027,
475
+ "grad_norm": 0.26514431834220886,
476
+ "learning_rate": 4.3041344951996746e-05,
477
+ "loss": 1.0557,
478
+ "step": 59
479
+ },
480
+ {
481
+ "epoch": 0.019515368352577656,
482
+ "grad_norm": 0.3049844801425934,
483
+ "learning_rate": 4.131759111665349e-05,
484
+ "loss": 1.0511,
485
+ "step": 60
486
+ },
487
+ {
488
+ "epoch": 0.019840624491787282,
489
+ "grad_norm": 0.3749238848686218,
490
+ "learning_rate": 3.960441545911204e-05,
491
+ "loss": 1.0863,
492
+ "step": 61
493
+ },
494
+ {
495
+ "epoch": 0.02016588063099691,
496
+ "grad_norm": 0.2859311103820801,
497
+ "learning_rate": 3.790390522001662e-05,
498
+ "loss": 0.9929,
499
+ "step": 62
500
+ },
501
+ {
502
+ "epoch": 0.020491136770206538,
503
+ "grad_norm": 0.2842947542667389,
504
+ "learning_rate": 3.6218132209150045e-05,
505
+ "loss": 0.9925,
506
+ "step": 63
507
+ },
508
+ {
509
+ "epoch": 0.020491136770206538,
510
+ "eval_loss": 0.523247480392456,
511
+ "eval_runtime": 783.5638,
512
+ "eval_samples_per_second": 3.304,
513
+ "eval_steps_per_second": 0.413,
514
+ "step": 63
515
+ },
516
+ {
517
+ "epoch": 0.020816392909416167,
518
+ "grad_norm": 0.2733912765979767,
519
+ "learning_rate": 3.4549150281252636e-05,
520
+ "loss": 1.0786,
521
+ "step": 64
522
+ },
523
+ {
524
+ "epoch": 0.021141649048625793,
525
+ "grad_norm": 0.29001614451408386,
526
+ "learning_rate": 3.289899283371657e-05,
527
+ "loss": 1.0672,
528
+ "step": 65
529
+ },
530
+ {
531
+ "epoch": 0.02146690518783542,
532
+ "grad_norm": 0.35740548372268677,
533
+ "learning_rate": 3.12696703292044e-05,
534
+ "loss": 1.0932,
535
+ "step": 66
536
+ },
537
+ {
538
+ "epoch": 0.021792161327045048,
539
+ "grad_norm": 0.2935461699962616,
540
+ "learning_rate": 2.9663167846209998e-05,
541
+ "loss": 0.8891,
542
+ "step": 67
543
+ },
544
+ {
545
+ "epoch": 0.022117417466254674,
546
+ "grad_norm": 0.2859397828578949,
547
+ "learning_rate": 2.8081442660546125e-05,
548
+ "loss": 1.0254,
549
+ "step": 68
550
+ },
551
+ {
552
+ "epoch": 0.022442673605464303,
553
+ "grad_norm": 0.29368773102760315,
554
+ "learning_rate": 2.6526421860705473e-05,
555
+ "loss": 0.9727,
556
+ "step": 69
557
+ },
558
+ {
559
+ "epoch": 0.02276792974467393,
560
+ "grad_norm": 0.34601661562919617,
561
+ "learning_rate": 2.500000000000001e-05,
562
+ "loss": 1.0418,
563
+ "step": 70
564
+ },
565
+ {
566
+ "epoch": 0.02309318588388356,
567
+ "grad_norm": 0.27021434903144836,
568
+ "learning_rate": 2.350403678833976e-05,
569
+ "loss": 0.9738,
570
+ "step": 71
571
+ },
572
+ {
573
+ "epoch": 0.023418442023093185,
574
+ "grad_norm": 0.3417051434516907,
575
+ "learning_rate": 2.2040354826462668e-05,
576
+ "loss": 1.0822,
577
+ "step": 72
578
+ },
579
+ {
580
+ "epoch": 0.023418442023093185,
581
+ "eval_loss": 0.5194622874259949,
582
+ "eval_runtime": 784.0065,
583
+ "eval_samples_per_second": 3.302,
584
+ "eval_steps_per_second": 0.413,
585
+ "step": 72
586
+ },
587
+ {
588
+ "epoch": 0.023743698162302814,
589
+ "grad_norm": 0.32264044880867004,
590
+ "learning_rate": 2.061073738537635e-05,
591
+ "loss": 0.941,
592
+ "step": 73
593
+ },
594
+ {
595
+ "epoch": 0.02406895430151244,
596
+ "grad_norm": 0.634362518787384,
597
+ "learning_rate": 1.9216926233717085e-05,
598
+ "loss": 1.2005,
599
+ "step": 74
600
+ },
601
+ {
602
+ "epoch": 0.02439421044072207,
603
+ "grad_norm": 0.3109305202960968,
604
+ "learning_rate": 1.7860619515673033e-05,
605
+ "loss": 1.012,
606
+ "step": 75
607
  }
608
  ],
609
  "logging_steps": 1,
 
623
  "attributes": {}
624
  }
625
  },
626
+ "total_flos": 1.8154793692299264e+17,
627
  "train_batch_size": 8,
628
  "trial_name": null,
629
  "trial_params": null