joelniklaus commited on
Commit
3ced673
1 Parent(s): 70055f0

Training in progress, step 100000

Browse files
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8da94f9b90f1ba4c0d3164e62b77b676f2b235652f6f692114e1b2bdb5273523
3
  size 2693742553
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3064095c68677a3c912b3b8c79971610fd83de1765a303d751d689310bd7b349
3
  size 2693742553
last-checkpoint/pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:de8a4f7df26972b1c1fe1546343f51ae325d6a1916a83baac53a80c81d4dd4ce
3
  size 1346893675
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0854eb9258505fe570a2fe172eeef34f6361865f6f12cc4b75ab455146992a0
3
  size 1346893675
last-checkpoint/rng_state_0.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5a9eaaa7cc088403f9a31ab70b2e1791d125e5bf39dbc05085cce4adba73595
3
  size 13611
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a766705dfa06584b04e13adde061e144056ae2a8ef43bca4589db57242d9e2e
3
  size 13611
last-checkpoint/rng_state_1.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5a9eaaa7cc088403f9a31ab70b2e1791d125e5bf39dbc05085cce4adba73595
3
  size 13611
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a766705dfa06584b04e13adde061e144056ae2a8ef43bca4589db57242d9e2e
3
  size 13611
last-checkpoint/rng_state_2.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5a9eaaa7cc088403f9a31ab70b2e1791d125e5bf39dbc05085cce4adba73595
3
  size 13611
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a766705dfa06584b04e13adde061e144056ae2a8ef43bca4589db57242d9e2e
3
  size 13611
last-checkpoint/rng_state_3.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5a9eaaa7cc088403f9a31ab70b2e1791d125e5bf39dbc05085cce4adba73595
3
  size 13611
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a766705dfa06584b04e13adde061e144056ae2a8ef43bca4589db57242d9e2e
3
  size 13611
last-checkpoint/rng_state_4.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5a9eaaa7cc088403f9a31ab70b2e1791d125e5bf39dbc05085cce4adba73595
3
  size 13611
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a766705dfa06584b04e13adde061e144056ae2a8ef43bca4589db57242d9e2e
3
  size 13611
last-checkpoint/rng_state_5.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5a9eaaa7cc088403f9a31ab70b2e1791d125e5bf39dbc05085cce4adba73595
3
  size 13611
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a766705dfa06584b04e13adde061e144056ae2a8ef43bca4589db57242d9e2e
3
  size 13611
last-checkpoint/rng_state_6.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5a9eaaa7cc088403f9a31ab70b2e1791d125e5bf39dbc05085cce4adba73595
3
  size 13611
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a766705dfa06584b04e13adde061e144056ae2a8ef43bca4589db57242d9e2e
3
  size 13611
last-checkpoint/rng_state_7.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c5a9eaaa7cc088403f9a31ab70b2e1791d125e5bf39dbc05085cce4adba73595
3
  size 13611
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a766705dfa06584b04e13adde061e144056ae2a8ef43bca4589db57242d9e2e
3
  size 13611
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:be1ccf49f4804619cd7d22b74b595a694a368e629a10492b4089d6536d07bdf2
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:647800033a1fe4aa55a6bc8c002ddb2326a52950bde89b878eccf8a697eacefa
3
  size 623
last-checkpoint/trainer_state.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 0.2,
5
- "global_step": 200000,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
@@ -318,931 +318,315 @@
318
  {
319
  "epoch": 0.05,
320
  "learning_rate": 9.999972660400536e-05,
321
- "loss": 0.9559,
322
  "step": 51000
323
  },
324
  {
325
  "epoch": 0.05,
326
  "learning_rate": 9.999890641901125e-05,
327
- "loss": 0.9156,
328
  "step": 52000
329
  },
330
  {
331
  "epoch": 0.05,
332
  "learning_rate": 9.999753945398704e-05,
333
- "loss": 0.9237,
334
  "step": 53000
335
  },
336
  {
337
  "epoch": 0.05,
338
  "learning_rate": 9.99956257238817e-05,
339
- "loss": 0.9444,
340
  "step": 54000
341
  },
342
  {
343
  "epoch": 0.06,
344
  "learning_rate": 9.999316524962345e-05,
345
- "loss": 0.9576,
346
  "step": 55000
347
  },
348
  {
349
  "epoch": 0.06,
350
  "learning_rate": 9.999015805811965e-05,
351
- "loss": 0.994,
352
  "step": 56000
353
  },
354
  {
355
  "epoch": 0.06,
356
  "learning_rate": 9.998660418225645e-05,
357
- "loss": 1.1439,
358
  "step": 57000
359
  },
360
  {
361
  "epoch": 0.06,
362
  "learning_rate": 9.998250366089848e-05,
363
- "loss": 1.0131,
364
  "step": 58000
365
  },
366
  {
367
  "epoch": 0.06,
368
  "learning_rate": 9.997785653888835e-05,
369
- "loss": 1.0879,
370
  "step": 59000
371
  },
372
  {
373
  "epoch": 0.06,
374
  "learning_rate": 9.997266286704631e-05,
375
- "loss": 0.925,
376
  "step": 60000
377
  },
378
  {
379
  "epoch": 0.06,
380
  "learning_rate": 9.996692270216947e-05,
381
- "loss": 0.9413,
382
  "step": 61000
383
  },
384
  {
385
  "epoch": 0.06,
386
  "learning_rate": 9.996063610703137e-05,
387
- "loss": 0.8956,
388
  "step": 62000
389
  },
390
  {
391
  "epoch": 0.06,
392
  "learning_rate": 9.995380315038119e-05,
393
- "loss": 0.8672,
394
  "step": 63000
395
  },
396
  {
397
  "epoch": 0.06,
398
  "learning_rate": 9.994642390694308e-05,
399
- "loss": 0.8875,
400
  "step": 64000
401
  },
402
  {
403
  "epoch": 0.07,
404
  "learning_rate": 9.993849845741524e-05,
405
- "loss": 0.9124,
406
  "step": 65000
407
  },
408
  {
409
  "epoch": 0.07,
410
  "learning_rate": 9.993002688846913e-05,
411
- "loss": 0.9428,
412
  "step": 66000
413
  },
414
  {
415
  "epoch": 0.07,
416
  "learning_rate": 9.992100929274846e-05,
417
- "loss": 0.9931,
418
  "step": 67000
419
  },
420
  {
421
  "epoch": 0.07,
422
  "learning_rate": 9.991144576886823e-05,
423
- "loss": 1.1243,
424
  "step": 68000
425
  },
426
  {
427
  "epoch": 0.07,
428
  "learning_rate": 9.990133642141359e-05,
429
- "loss": 0.9759,
430
  "step": 69000
431
  },
432
  {
433
  "epoch": 0.07,
434
  "learning_rate": 9.989068136093873e-05,
435
- "loss": 1.0443,
436
  "step": 70000
437
  },
438
  {
439
  "epoch": 0.07,
440
  "learning_rate": 9.987948070396571e-05,
441
- "loss": 0.9112,
442
  "step": 71000
443
  },
444
  {
445
  "epoch": 0.07,
446
  "learning_rate": 9.986773457298311e-05,
447
- "loss": 0.9212,
448
  "step": 72000
449
  },
450
  {
451
  "epoch": 0.07,
452
  "learning_rate": 9.985544309644475e-05,
453
- "loss": 0.8929,
454
  "step": 73000
455
  },
456
  {
457
  "epoch": 0.07,
458
  "learning_rate": 9.984260640876821e-05,
459
- "loss": 0.8609,
460
  "step": 74000
461
  },
462
  {
463
  "epoch": 0.07,
464
  "learning_rate": 9.98292246503335e-05,
465
- "loss": 0.8797,
466
  "step": 75000
467
  },
468
  {
469
  "epoch": 0.08,
470
  "learning_rate": 9.981529796748134e-05,
471
- "loss": 0.9014,
472
  "step": 76000
473
  },
474
  {
475
  "epoch": 0.08,
476
  "learning_rate": 9.980082651251175e-05,
477
- "loss": 0.9251,
478
  "step": 77000
479
  },
480
  {
481
  "epoch": 0.08,
482
  "learning_rate": 9.97858104436822e-05,
483
- "loss": 0.988,
484
  "step": 78000
485
  },
486
  {
487
  "epoch": 0.08,
488
  "learning_rate": 9.977024992520602e-05,
489
- "loss": 1.0805,
490
  "step": 79000
491
  },
492
  {
493
  "epoch": 0.08,
494
  "learning_rate": 9.975414512725057e-05,
495
- "loss": 0.9457,
496
  "step": 80000
497
  },
498
  {
499
  "epoch": 0.08,
500
  "learning_rate": 9.973749622593534e-05,
501
- "loss": 0.9887,
502
  "step": 81000
503
  },
504
  {
505
  "epoch": 0.08,
506
  "learning_rate": 9.972030340333001e-05,
507
- "loss": 0.8866,
508
  "step": 82000
509
  },
510
  {
511
  "epoch": 0.08,
512
  "learning_rate": 9.970256684745258e-05,
513
- "loss": 0.8843,
514
  "step": 83000
515
  },
516
  {
517
  "epoch": 0.08,
518
  "learning_rate": 9.968428675226714e-05,
519
- "loss": 0.8685,
520
  "step": 84000
521
  },
522
  {
523
  "epoch": 0.09,
524
  "learning_rate": 9.966546331768191e-05,
525
- "loss": 0.8369,
526
  "step": 85000
527
  },
528
  {
529
  "epoch": 0.09,
530
  "learning_rate": 9.964609674954696e-05,
531
- "loss": 0.8639,
532
  "step": 86000
533
  },
534
  {
535
  "epoch": 0.09,
536
  "learning_rate": 9.962618725965196e-05,
537
- "loss": 0.881,
538
  "step": 87000
539
  },
540
  {
541
  "epoch": 0.09,
542
  "learning_rate": 9.96057350657239e-05,
543
- "loss": 0.8954,
544
  "step": 88000
545
  },
546
  {
547
  "epoch": 0.09,
548
  "learning_rate": 9.95847403914247e-05,
549
- "loss": 0.9711,
550
  "step": 89000
551
  },
552
  {
553
  "epoch": 0.09,
554
  "learning_rate": 9.956320346634876e-05,
555
- "loss": 1.0358,
556
  "step": 90000
557
  },
558
  {
559
  "epoch": 0.09,
560
  "learning_rate": 9.954112452602045e-05,
561
- "loss": 0.927,
562
  "step": 91000
563
  },
564
  {
565
  "epoch": 0.09,
566
  "learning_rate": 9.95185038118915e-05,
567
- "loss": 0.9126,
568
  "step": 92000
569
  },
570
  {
571
  "epoch": 0.09,
572
  "learning_rate": 9.949534157133844e-05,
573
- "loss": 0.8692,
574
  "step": 93000
575
  },
576
  {
577
  "epoch": 0.09,
578
  "learning_rate": 9.94716380576598e-05,
579
- "loss": 0.8355,
580
  "step": 94000
581
  },
582
  {
583
  "epoch": 0.1,
584
  "learning_rate": 9.944739353007344e-05,
585
- "loss": 0.8358,
586
  "step": 95000
587
  },
588
  {
589
  "epoch": 0.1,
590
  "learning_rate": 9.942260825371358e-05,
591
- "loss": 0.8086,
592
  "step": 96000
593
  },
594
  {
595
  "epoch": 0.1,
596
  "learning_rate": 9.939728249962807e-05,
597
- "loss": 0.8348,
598
  "step": 97000
599
  },
600
  {
601
  "epoch": 0.1,
602
  "learning_rate": 9.937141654477528e-05,
603
- "loss": 0.8471,
604
  "step": 98000
605
  },
606
  {
607
  "epoch": 0.1,
608
  "learning_rate": 9.934501067202117e-05,
609
- "loss": 0.8733,
610
  "step": 99000
611
  },
612
  {
613
  "epoch": 0.1,
614
  "learning_rate": 9.931806517013612e-05,
615
- "loss": 0.9549,
616
  "step": 100000
617
  },
618
  {
619
  "epoch": 0.1,
620
- "eval_loss": 0.777112603187561,
621
- "eval_runtime": 32.3706,
622
- "eval_samples_per_second": 154.461,
623
- "eval_steps_per_second": 2.44,
624
  "step": 100000
625
- },
626
- {
627
- "epoch": 0.1,
628
- "learning_rate": 9.929058033379181e-05,
629
- "loss": 1.0084,
630
- "step": 101000
631
- },
632
- {
633
- "epoch": 0.1,
634
- "learning_rate": 9.926255646355804e-05,
635
- "loss": 0.9332,
636
- "step": 102000
637
- },
638
- {
639
- "epoch": 0.1,
640
- "learning_rate": 9.923399386589933e-05,
641
- "loss": 0.8486,
642
- "step": 103000
643
- },
644
- {
645
- "epoch": 0.1,
646
- "learning_rate": 9.92048928531717e-05,
647
- "loss": 0.8585,
648
- "step": 104000
649
- },
650
- {
651
- "epoch": 0.1,
652
- "learning_rate": 9.917525374361912e-05,
653
- "loss": 0.8168,
654
- "step": 105000
655
- },
656
- {
657
- "epoch": 0.11,
658
- "learning_rate": 9.914507686137019e-05,
659
- "loss": 0.8111,
660
- "step": 106000
661
- },
662
- {
663
- "epoch": 0.11,
664
- "learning_rate": 9.911436253643445e-05,
665
- "loss": 0.7897,
666
- "step": 107000
667
- },
668
- {
669
- "epoch": 0.11,
670
- "learning_rate": 9.90831111046988e-05,
671
- "loss": 0.8249,
672
- "step": 108000
673
- },
674
- {
675
- "epoch": 0.11,
676
- "learning_rate": 9.905132290792394e-05,
677
- "loss": 0.8336,
678
- "step": 109000
679
- },
680
- {
681
- "epoch": 0.11,
682
- "learning_rate": 9.901899829374047e-05,
683
- "loss": 0.8623,
684
- "step": 110000
685
- },
686
- {
687
- "epoch": 0.11,
688
- "learning_rate": 9.89861376156452e-05,
689
- "loss": 0.9608,
690
- "step": 111000
691
- },
692
- {
693
- "epoch": 0.11,
694
- "learning_rate": 9.895274123299723e-05,
695
- "loss": 0.9681,
696
- "step": 112000
697
- },
698
- {
699
- "epoch": 0.11,
700
- "learning_rate": 9.891880951101407e-05,
701
- "loss": 0.9459,
702
- "step": 113000
703
- },
704
- {
705
- "epoch": 0.11,
706
- "learning_rate": 9.888434282076758e-05,
707
- "loss": 0.7867,
708
- "step": 114000
709
- },
710
- {
711
- "epoch": 0.12,
712
- "learning_rate": 9.884934153917997e-05,
713
- "loss": 0.8544,
714
- "step": 115000
715
- },
716
- {
717
- "epoch": 0.12,
718
- "learning_rate": 9.881380604901964e-05,
719
- "loss": 0.8081,
720
- "step": 116000
721
- },
722
- {
723
- "epoch": 0.12,
724
- "learning_rate": 9.877773673889701e-05,
725
- "loss": 0.7888,
726
- "step": 117000
727
- },
728
- {
729
- "epoch": 0.12,
730
- "learning_rate": 9.87411340032603e-05,
731
- "loss": 0.7873,
732
- "step": 118000
733
- },
734
- {
735
- "epoch": 0.12,
736
- "learning_rate": 9.870399824239117e-05,
737
- "loss": 0.8214,
738
- "step": 119000
739
- },
740
- {
741
- "epoch": 0.12,
742
- "learning_rate": 9.86663298624003e-05,
743
- "loss": 0.8268,
744
- "step": 120000
745
- },
746
- {
747
- "epoch": 0.12,
748
- "learning_rate": 9.862812927522309e-05,
749
- "loss": 0.8562,
750
- "step": 121000
751
- },
752
- {
753
- "epoch": 0.12,
754
- "learning_rate": 9.858939689861506e-05,
755
- "loss": 0.977,
756
- "step": 122000
757
- },
758
- {
759
- "epoch": 0.12,
760
- "learning_rate": 9.855013315614725e-05,
761
- "loss": 0.9267,
762
- "step": 123000
763
- },
764
- {
765
- "epoch": 0.12,
766
- "learning_rate": 9.851033847720166e-05,
767
- "loss": 0.9484,
768
- "step": 124000
769
- },
770
- {
771
- "epoch": 0.12,
772
- "learning_rate": 9.847001329696653e-05,
773
- "loss": 0.7635,
774
- "step": 125000
775
- },
776
- {
777
- "epoch": 0.13,
778
- "learning_rate": 9.842915805643155e-05,
779
- "loss": 0.8435,
780
- "step": 126000
781
- },
782
- {
783
- "epoch": 0.13,
784
- "learning_rate": 9.838777320238312e-05,
785
- "loss": 0.7942,
786
- "step": 127000
787
- },
788
- {
789
- "epoch": 0.13,
790
- "learning_rate": 9.834585918739936e-05,
791
- "loss": 0.7748,
792
- "step": 128000
793
- },
794
- {
795
- "epoch": 0.13,
796
- "learning_rate": 9.830341646984521e-05,
797
- "loss": 0.7881,
798
- "step": 129000
799
- },
800
- {
801
- "epoch": 0.13,
802
- "learning_rate": 9.826044551386744e-05,
803
- "loss": 0.8121,
804
- "step": 130000
805
- },
806
- {
807
- "epoch": 0.13,
808
- "learning_rate": 9.821694678938953e-05,
809
- "loss": 0.8211,
810
- "step": 131000
811
- },
812
- {
813
- "epoch": 0.13,
814
- "learning_rate": 9.817292077210659e-05,
815
- "loss": 0.8477,
816
- "step": 132000
817
- },
818
- {
819
- "epoch": 0.13,
820
- "learning_rate": 9.812836794348004e-05,
821
- "loss": 0.9879,
822
- "step": 133000
823
- },
824
- {
825
- "epoch": 0.13,
826
- "learning_rate": 9.808328879073251e-05,
827
- "loss": 0.8801,
828
- "step": 134000
829
- },
830
- {
831
- "epoch": 0.14,
832
- "learning_rate": 9.803768380684242e-05,
833
- "loss": 0.9278,
834
- "step": 135000
835
- },
836
- {
837
- "epoch": 0.14,
838
- "learning_rate": 9.799155349053851e-05,
839
- "loss": 0.772,
840
- "step": 136000
841
- },
842
- {
843
- "epoch": 0.14,
844
- "learning_rate": 9.794489834629455e-05,
845
- "loss": 0.8189,
846
- "step": 137000
847
- },
848
- {
849
- "epoch": 0.14,
850
- "learning_rate": 9.789771888432375e-05,
851
- "loss": 0.7837,
852
- "step": 138000
853
- },
854
- {
855
- "epoch": 0.14,
856
- "learning_rate": 9.785001562057309e-05,
857
- "loss": 0.7684,
858
- "step": 139000
859
- },
860
- {
861
- "epoch": 0.14,
862
- "learning_rate": 9.780178907671789e-05,
863
- "loss": 0.7807,
864
- "step": 140000
865
- },
866
- {
867
- "epoch": 0.14,
868
- "learning_rate": 9.775303978015585e-05,
869
- "loss": 0.8037,
870
- "step": 141000
871
- },
872
- {
873
- "epoch": 0.14,
874
- "learning_rate": 9.77037682640015e-05,
875
- "loss": 0.8205,
876
- "step": 142000
877
- },
878
- {
879
- "epoch": 0.14,
880
- "learning_rate": 9.765397506708023e-05,
881
- "loss": 0.8609,
882
- "step": 143000
883
- },
884
- {
885
- "epoch": 0.14,
886
- "learning_rate": 9.760366073392246e-05,
887
- "loss": 0.9872,
888
- "step": 144000
889
- },
890
- {
891
- "epoch": 0.14,
892
- "learning_rate": 9.755282581475769e-05,
893
- "loss": 0.8561,
894
- "step": 145000
895
- },
896
- {
897
- "epoch": 0.15,
898
- "learning_rate": 9.750147086550844e-05,
899
- "loss": 0.9079,
900
- "step": 146000
901
- },
902
- {
903
- "epoch": 0.15,
904
- "learning_rate": 9.744959644778422e-05,
905
- "loss": 0.7751,
906
- "step": 147000
907
- },
908
- {
909
- "epoch": 0.15,
910
- "learning_rate": 9.739720312887535e-05,
911
- "loss": 0.7998,
912
- "step": 148000
913
- },
914
- {
915
- "epoch": 0.15,
916
- "learning_rate": 9.734429148174675e-05,
917
- "loss": 0.7823,
918
- "step": 149000
919
- },
920
- {
921
- "epoch": 0.15,
922
- "learning_rate": 9.729086208503174e-05,
923
- "loss": 0.7551,
924
- "step": 150000
925
- },
926
- {
927
- "epoch": 0.15,
928
- "eval_loss": 0.7697679400444031,
929
- "eval_runtime": 22.8281,
930
- "eval_samples_per_second": 219.028,
931
- "eval_steps_per_second": 3.461,
932
- "step": 150000
933
- },
934
- {
935
- "epoch": 0.15,
936
- "learning_rate": 9.723691552302562e-05,
937
- "loss": 0.7738,
938
- "step": 151000
939
- },
940
- {
941
- "epoch": 0.15,
942
- "learning_rate": 9.718245238567939e-05,
943
- "loss": 0.7963,
944
- "step": 152000
945
- },
946
- {
947
- "epoch": 0.15,
948
- "learning_rate": 9.712747326859315e-05,
949
- "loss": 0.8171,
950
- "step": 153000
951
- },
952
- {
953
- "epoch": 0.15,
954
- "learning_rate": 9.707197877300974e-05,
955
- "loss": 0.8773,
956
- "step": 154000
957
- },
958
- {
959
- "epoch": 0.15,
960
- "learning_rate": 9.701596950580806e-05,
961
- "loss": 0.9735,
962
- "step": 155000
963
- },
964
- {
965
- "epoch": 0.16,
966
- "learning_rate": 9.695944607949649e-05,
967
- "loss": 0.8433,
968
- "step": 156000
969
- },
970
- {
971
- "epoch": 0.16,
972
- "learning_rate": 9.690240911220618e-05,
973
- "loss": 0.8706,
974
- "step": 157000
975
- },
976
- {
977
- "epoch": 0.16,
978
- "learning_rate": 9.684485922768422e-05,
979
- "loss": 0.7855,
980
- "step": 158000
981
- },
982
- {
983
- "epoch": 0.16,
984
- "learning_rate": 9.6786797055287e-05,
985
- "loss": 0.785,
986
- "step": 159000
987
- },
988
- {
989
- "epoch": 0.16,
990
- "learning_rate": 9.672822322997305e-05,
991
- "loss": 0.765,
992
- "step": 160000
993
- },
994
- {
995
- "epoch": 0.16,
996
- "learning_rate": 9.66691383922964e-05,
997
- "loss": 0.7384,
998
- "step": 161000
999
- },
1000
- {
1001
- "epoch": 0.16,
1002
- "learning_rate": 9.660954318839933e-05,
1003
- "loss": 0.7583,
1004
- "step": 162000
1005
- },
1006
- {
1007
- "epoch": 0.16,
1008
- "learning_rate": 9.654943827000548e-05,
1009
- "loss": 0.7785,
1010
- "step": 163000
1011
- },
1012
- {
1013
- "epoch": 0.16,
1014
- "learning_rate": 9.648882429441257e-05,
1015
- "loss": 0.8002,
1016
- "step": 164000
1017
- },
1018
- {
1019
- "epoch": 0.17,
1020
- "learning_rate": 9.642770192448536e-05,
1021
- "loss": 0.8721,
1022
- "step": 165000
1023
- },
1024
- {
1025
- "epoch": 0.17,
1026
- "learning_rate": 9.636607182864827e-05,
1027
- "loss": 0.9439,
1028
- "step": 166000
1029
- },
1030
- {
1031
- "epoch": 0.17,
1032
- "learning_rate": 9.630393468087818e-05,
1033
- "loss": 0.8458,
1034
- "step": 167000
1035
- },
1036
- {
1037
- "epoch": 0.17,
1038
- "learning_rate": 9.624129116069694e-05,
1039
- "loss": 0.8241,
1040
- "step": 168000
1041
- },
1042
- {
1043
- "epoch": 0.17,
1044
- "learning_rate": 9.617814195316411e-05,
1045
- "loss": 0.7928,
1046
- "step": 169000
1047
- },
1048
- {
1049
- "epoch": 0.17,
1050
- "learning_rate": 9.611448774886924e-05,
1051
- "loss": 0.7656,
1052
- "step": 170000
1053
- },
1054
- {
1055
- "epoch": 0.17,
1056
- "learning_rate": 9.605032924392457e-05,
1057
- "loss": 0.7647,
1058
- "step": 171000
1059
- },
1060
- {
1061
- "epoch": 0.17,
1062
- "learning_rate": 9.598566713995718e-05,
1063
- "loss": 0.7523,
1064
- "step": 172000
1065
- },
1066
- {
1067
- "epoch": 0.17,
1068
- "learning_rate": 9.59205021441015e-05,
1069
- "loss": 0.7739,
1070
- "step": 173000
1071
- },
1072
- {
1073
- "epoch": 0.17,
1074
- "learning_rate": 9.58548349689915e-05,
1075
- "loss": 0.7857,
1076
- "step": 174000
1077
- },
1078
- {
1079
- "epoch": 0.17,
1080
- "learning_rate": 9.578866633275288e-05,
1081
- "loss": 0.8132,
1082
- "step": 175000
1083
- },
1084
- {
1085
- "epoch": 0.18,
1086
- "learning_rate": 9.572199695899522e-05,
1087
- "loss": 0.8928,
1088
- "step": 176000
1089
- },
1090
- {
1091
- "epoch": 0.18,
1092
- "learning_rate": 9.565482757680415e-05,
1093
- "loss": 0.9479,
1094
- "step": 177000
1095
- },
1096
- {
1097
- "epoch": 0.18,
1098
- "learning_rate": 9.558715892073323e-05,
1099
- "loss": 0.8774,
1100
- "step": 178000
1101
- },
1102
- {
1103
- "epoch": 0.18,
1104
- "learning_rate": 9.551899173079607e-05,
1105
- "loss": 0.7822,
1106
- "step": 179000
1107
- },
1108
- {
1109
- "epoch": 0.18,
1110
- "learning_rate": 9.545032675245813e-05,
1111
- "loss": 0.7981,
1112
- "step": 180000
1113
- },
1114
- {
1115
- "epoch": 0.18,
1116
- "learning_rate": 9.538116473662861e-05,
1117
- "loss": 0.7582,
1118
- "step": 181000
1119
- },
1120
- {
1121
- "epoch": 0.18,
1122
- "learning_rate": 9.531150643965223e-05,
1123
- "loss": 0.7543,
1124
- "step": 182000
1125
- },
1126
- {
1127
- "epoch": 0.18,
1128
- "learning_rate": 9.524135262330098e-05,
1129
- "loss": 0.741,
1130
- "step": 183000
1131
- },
1132
- {
1133
- "epoch": 0.18,
1134
- "learning_rate": 9.517070405476575e-05,
1135
- "loss": 0.7712,
1136
- "step": 184000
1137
- },
1138
- {
1139
- "epoch": 0.18,
1140
- "learning_rate": 9.509956150664796e-05,
1141
- "loss": 0.7824,
1142
- "step": 185000
1143
- },
1144
- {
1145
- "epoch": 0.19,
1146
- "learning_rate": 9.502792575695112e-05,
1147
- "loss": 0.8096,
1148
- "step": 186000
1149
- },
1150
- {
1151
- "epoch": 0.19,
1152
- "learning_rate": 9.49557975890723e-05,
1153
- "loss": 0.9033,
1154
- "step": 187000
1155
- },
1156
- {
1157
- "epoch": 0.19,
1158
- "learning_rate": 9.488317779179361e-05,
1159
- "loss": 0.9156,
1160
- "step": 188000
1161
- },
1162
- {
1163
- "epoch": 0.19,
1164
- "learning_rate": 9.481006715927351e-05,
1165
- "loss": 0.896,
1166
- "step": 189000
1167
- },
1168
- {
1169
- "epoch": 0.19,
1170
- "learning_rate": 9.473646649103818e-05,
1171
- "loss": 0.7307,
1172
- "step": 190000
1173
- },
1174
- {
1175
- "epoch": 0.19,
1176
- "learning_rate": 9.46623765919727e-05,
1177
- "loss": 0.8032,
1178
- "step": 191000
1179
- },
1180
- {
1181
- "epoch": 0.19,
1182
- "learning_rate": 9.458779827231237e-05,
1183
- "loss": 0.7586,
1184
- "step": 192000
1185
- },
1186
- {
1187
- "epoch": 0.19,
1188
- "learning_rate": 9.451273234763371e-05,
1189
- "loss": 0.7422,
1190
- "step": 193000
1191
- },
1192
- {
1193
- "epoch": 0.19,
1194
- "learning_rate": 9.443717963884569e-05,
1195
- "loss": 0.7412,
1196
- "step": 194000
1197
- },
1198
- {
1199
- "epoch": 0.2,
1200
- "learning_rate": 9.43611409721806e-05,
1201
- "loss": 0.7738,
1202
- "step": 195000
1203
- },
1204
- {
1205
- "epoch": 0.2,
1206
- "learning_rate": 9.428461717918511e-05,
1207
- "loss": 0.7806,
1208
- "step": 196000
1209
- },
1210
- {
1211
- "epoch": 0.2,
1212
- "learning_rate": 9.420760909671118e-05,
1213
- "loss": 0.807,
1214
- "step": 197000
1215
- },
1216
- {
1217
- "epoch": 0.2,
1218
- "learning_rate": 9.413011756690685e-05,
1219
- "loss": 0.9269,
1220
- "step": 198000
1221
- },
1222
- {
1223
- "epoch": 0.2,
1224
- "learning_rate": 9.405214343720707e-05,
1225
- "loss": 0.8776,
1226
- "step": 199000
1227
- },
1228
- {
1229
- "epoch": 0.2,
1230
- "learning_rate": 9.397368756032445e-05,
1231
- "loss": 0.8979,
1232
- "step": 200000
1233
- },
1234
- {
1235
- "epoch": 0.2,
1236
- "eval_loss": 0.7604823112487793,
1237
- "eval_runtime": 22.7684,
1238
- "eval_samples_per_second": 219.603,
1239
- "eval_steps_per_second": 3.47,
1240
- "step": 200000
1241
  }
1242
  ],
1243
  "max_steps": 1000000,
1244
  "num_train_epochs": 9223372036854775807,
1245
- "total_flos": 1.19299796631552e+19,
1246
  "trial_name": null,
1247
  "trial_params": null
1248
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 0.1,
5
+ "global_step": 100000,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
 
318
  {
319
  "epoch": 0.05,
320
  "learning_rate": 9.999972660400536e-05,
321
+ "loss": 1.0509,
322
  "step": 51000
323
  },
324
  {
325
  "epoch": 0.05,
326
  "learning_rate": 9.999890641901125e-05,
327
+ "loss": 1.2048,
328
  "step": 52000
329
  },
330
  {
331
  "epoch": 0.05,
332
  "learning_rate": 9.999753945398704e-05,
333
+ "loss": 1.2837,
334
  "step": 53000
335
  },
336
  {
337
  "epoch": 0.05,
338
  "learning_rate": 9.99956257238817e-05,
339
+ "loss": 1.1458,
340
  "step": 54000
341
  },
342
  {
343
  "epoch": 0.06,
344
  "learning_rate": 9.999316524962345e-05,
345
+ "loss": 1.0255,
346
  "step": 55000
347
  },
348
  {
349
  "epoch": 0.06,
350
  "learning_rate": 9.999015805811965e-05,
351
+ "loss": 1.2046,
352
  "step": 56000
353
  },
354
  {
355
  "epoch": 0.06,
356
  "learning_rate": 9.998660418225645e-05,
357
+ "loss": 0.9363,
358
  "step": 57000
359
  },
360
  {
361
  "epoch": 0.06,
362
  "learning_rate": 9.998250366089848e-05,
363
+ "loss": 0.6931,
364
  "step": 58000
365
  },
366
  {
367
  "epoch": 0.06,
368
  "learning_rate": 9.997785653888835e-05,
369
+ "loss": 0.7802,
370
  "step": 59000
371
  },
372
  {
373
  "epoch": 0.06,
374
  "learning_rate": 9.997266286704631e-05,
375
+ "loss": 1.0542,
376
  "step": 60000
377
  },
378
  {
379
  "epoch": 0.06,
380
  "learning_rate": 9.996692270216947e-05,
381
+ "loss": 1.1069,
382
  "step": 61000
383
  },
384
  {
385
  "epoch": 0.06,
386
  "learning_rate": 9.996063610703137e-05,
387
+ "loss": 1.0202,
388
  "step": 62000
389
  },
390
  {
391
  "epoch": 0.06,
392
  "learning_rate": 9.995380315038119e-05,
393
+ "loss": 1.0179,
394
  "step": 63000
395
  },
396
  {
397
  "epoch": 0.06,
398
  "learning_rate": 9.994642390694308e-05,
399
+ "loss": 1.0872,
400
  "step": 64000
401
  },
402
  {
403
  "epoch": 0.07,
404
  "learning_rate": 9.993849845741524e-05,
405
+ "loss": 1.0288,
406
  "step": 65000
407
  },
408
  {
409
  "epoch": 0.07,
410
  "learning_rate": 9.993002688846913e-05,
411
+ "loss": 1.0054,
412
  "step": 66000
413
  },
414
  {
415
  "epoch": 0.07,
416
  "learning_rate": 9.992100929274846e-05,
417
+ "loss": 1.0455,
418
  "step": 67000
419
  },
420
  {
421
  "epoch": 0.07,
422
  "learning_rate": 9.991144576886823e-05,
423
+ "loss": 0.9996,
424
  "step": 68000
425
  },
426
  {
427
  "epoch": 0.07,
428
  "learning_rate": 9.990133642141359e-05,
429
+ "loss": 1.0668,
430
  "step": 69000
431
  },
432
  {
433
  "epoch": 0.07,
434
  "learning_rate": 9.989068136093873e-05,
435
+ "loss": 0.993,
436
  "step": 70000
437
  },
438
  {
439
  "epoch": 0.07,
440
  "learning_rate": 9.987948070396571e-05,
441
+ "loss": 1.0315,
442
  "step": 71000
443
  },
444
  {
445
  "epoch": 0.07,
446
  "learning_rate": 9.986773457298311e-05,
447
+ "loss": 1.0934,
448
  "step": 72000
449
  },
450
  {
451
  "epoch": 0.07,
452
  "learning_rate": 9.985544309644475e-05,
453
+ "loss": 0.9016,
454
  "step": 73000
455
  },
456
  {
457
  "epoch": 0.07,
458
  "learning_rate": 9.984260640876821e-05,
459
+ "loss": 1.0367,
460
  "step": 74000
461
  },
462
  {
463
  "epoch": 0.07,
464
  "learning_rate": 9.98292246503335e-05,
465
+ "loss": 0.9656,
466
  "step": 75000
467
  },
468
  {
469
  "epoch": 0.08,
470
  "learning_rate": 9.981529796748134e-05,
471
+ "loss": 1.1794,
472
  "step": 76000
473
  },
474
  {
475
  "epoch": 0.08,
476
  "learning_rate": 9.980082651251175e-05,
477
+ "loss": 1.0076,
478
  "step": 77000
479
  },
480
  {
481
  "epoch": 0.08,
482
  "learning_rate": 9.97858104436822e-05,
483
+ "loss": 1.038,
484
  "step": 78000
485
  },
486
  {
487
  "epoch": 0.08,
488
  "learning_rate": 9.977024992520602e-05,
489
+ "loss": 1.1324,
490
  "step": 79000
491
  },
492
  {
493
  "epoch": 0.08,
494
  "learning_rate": 9.975414512725057e-05,
495
+ "loss": 0.9674,
496
  "step": 80000
497
  },
498
  {
499
  "epoch": 0.08,
500
  "learning_rate": 9.973749622593534e-05,
501
+ "loss": 0.9536,
502
  "step": 81000
503
  },
504
  {
505
  "epoch": 0.08,
506
  "learning_rate": 9.972030340333001e-05,
507
+ "loss": 0.9515,
508
  "step": 82000
509
  },
510
  {
511
  "epoch": 0.08,
512
  "learning_rate": 9.970256684745258e-05,
513
+ "loss": 1.0164,
514
  "step": 83000
515
  },
516
  {
517
  "epoch": 0.08,
518
  "learning_rate": 9.968428675226714e-05,
519
+ "loss": 0.9569,
520
  "step": 84000
521
  },
522
  {
523
  "epoch": 0.09,
524
  "learning_rate": 9.966546331768191e-05,
525
+ "loss": 0.9485,
526
  "step": 85000
527
  },
528
  {
529
  "epoch": 0.09,
530
  "learning_rate": 9.964609674954696e-05,
531
+ "loss": 0.9676,
532
  "step": 86000
533
  },
534
  {
535
  "epoch": 0.09,
536
  "learning_rate": 9.962618725965196e-05,
537
+ "loss": 0.9634,
538
  "step": 87000
539
  },
540
  {
541
  "epoch": 0.09,
542
  "learning_rate": 9.96057350657239e-05,
543
+ "loss": 0.9789,
544
  "step": 88000
545
  },
546
  {
547
  "epoch": 0.09,
548
  "learning_rate": 9.95847403914247e-05,
549
+ "loss": 1.011,
550
  "step": 89000
551
  },
552
  {
553
  "epoch": 0.09,
554
  "learning_rate": 9.956320346634876e-05,
555
+ "loss": 1.081,
556
  "step": 90000
557
  },
558
  {
559
  "epoch": 0.09,
560
  "learning_rate": 9.954112452602045e-05,
561
+ "loss": 1.1758,
562
  "step": 91000
563
  },
564
  {
565
  "epoch": 0.09,
566
  "learning_rate": 9.95185038118915e-05,
567
+ "loss": 1.1315,
568
  "step": 92000
569
  },
570
  {
571
  "epoch": 0.09,
572
  "learning_rate": 9.949534157133844e-05,
573
+ "loss": 1.0568,
574
  "step": 93000
575
  },
576
  {
577
  "epoch": 0.09,
578
  "learning_rate": 9.94716380576598e-05,
579
+ "loss": 1.0164,
580
  "step": 94000
581
  },
582
  {
583
  "epoch": 0.1,
584
  "learning_rate": 9.944739353007344e-05,
585
+ "loss": 0.9761,
586
  "step": 95000
587
  },
588
  {
589
  "epoch": 0.1,
590
  "learning_rate": 9.942260825371358e-05,
591
+ "loss": 1.0452,
592
  "step": 96000
593
  },
594
  {
595
  "epoch": 0.1,
596
  "learning_rate": 9.939728249962807e-05,
597
+ "loss": 1.0411,
598
  "step": 97000
599
  },
600
  {
601
  "epoch": 0.1,
602
  "learning_rate": 9.937141654477528e-05,
603
+ "loss": 1.003,
604
  "step": 98000
605
  },
606
  {
607
  "epoch": 0.1,
608
  "learning_rate": 9.934501067202117e-05,
609
+ "loss": 0.9932,
610
  "step": 99000
611
  },
612
  {
613
  "epoch": 0.1,
614
  "learning_rate": 9.931806517013612e-05,
615
+ "loss": 0.9302,
616
  "step": 100000
617
  },
618
  {
619
  "epoch": 0.1,
620
+ "eval_loss": 0.6663409471511841,
621
+ "eval_runtime": 34.9654,
622
+ "eval_samples_per_second": 142.999,
623
+ "eval_steps_per_second": 2.259,
624
  "step": 100000
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
625
  }
626
  ],
627
  "max_steps": 1000000,
628
  "num_train_epochs": 9223372036854775807,
629
+ "total_flos": 5.9649898315776e+18,
630
  "trial_name": null,
631
  "trial_params": null
632
  }
last-checkpoint/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1e8f2580c27a61098f2e908ec4e6aceadf949f5a10ea8072d23cfab52b1c267f
3
  size 3503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e856cc865a6d510c6ac37f67c764ae14fdb35ac890aa32010e6b6fa6f15544b4
3
  size 3503
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:de8a4f7df26972b1c1fe1546343f51ae325d6a1916a83baac53a80c81d4dd4ce
3
  size 1346893675
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0854eb9258505fe570a2fe172eeef34f6361865f6f12cc4b75ab455146992a0
3
  size 1346893675
runs/Feb16_16-14-57_t1v-n-eeadb94b-w-0/events.out.tfevents.1676564607.t1v-n-eeadb94b-w-0.3898595.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c86b06d1f8b18603f6ddeb5deb090948c3af173dea2f1dfd9bac06ba2cfd7d88
3
- size 28664
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b747c156fe81911dbb4b93b5fe395b4b8ee151e3049fedfb6a5ae19990e98f8f
3
+ size 29464
runs/Feb25_19-21-23_t1v-n-eeadb94b-w-0/1677352917.7689872/events.out.tfevents.1677352917.t1v-n-eeadb94b-w-0.615717.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95821b5b9ee743ab9ebf5df303dc91e663f24ebc92cdda1d082e0e3bb78e6b68
3
+ size 5499
runs/Feb25_19-21-23_t1v-n-eeadb94b-w-0/events.out.tfevents.1677352917.t1v-n-eeadb94b-w-0.615717.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a8e46c2b7595fef8d3e4e540a6841b9377b19692847bb255ce19bb8a0aaf911
3
+ size 12112
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1e8f2580c27a61098f2e908ec4e6aceadf949f5a10ea8072d23cfab52b1c267f
3
  size 3503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e856cc865a6d510c6ac37f67c764ae14fdb35ac890aa32010e6b6fa6f15544b4
3
  size 3503