longcld commited on
Commit
cc70666
1 Parent(s): b8335db
Files changed (5) hide show
  1. optimizer.pt +1 -1
  2. pytorch_model.bin +1 -1
  3. rng_state.pth +1 -1
  4. scheduler.pt +1 -1
  5. trainer_state.json +963 -3
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2bf0e3ee7ac6ef169f59f90f5da8b7e66a94dd788753c9662dffce7b6e63db3d
3
  size 1585984129
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7717cc77b9dfdc741573ac36af9500ebe261a63a279912b8f960b7c9e5242c99
3
  size 1585984129
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fedec05df74e3b73fee9612429bd48745ad27f3211de6fdab94b729619738f19
3
  size 1561347007
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1dbf7512968e9b49a9effc5e8b456b7594341266a4da30528a298080dd30eccb
3
  size 1561347007
rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e3efd8721122cb987ae899630b5d42b2f605030fcbc4cf665cf824ab7bc6d0dc
3
  size 14503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:076be2f9f01be38c4835c9081e22297aef567c05e412b1d30dc0e816e787e53a
3
  size 14503
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2ba03cf563c9e4596b104006b1c92df725ac84cdbd21e9c1fb2ccd8252896081
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c560416a495505caae4f068fd5c58b4aab552a43341f03d9f27bd27c64a28cc8
3
  size 623
trainer_state.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
- "epoch": 1.3285948363668572,
5
- "global_step": 7500,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
@@ -456,11 +456,971 @@
456
  "learning_rate": 4.33569530558016e-05,
457
  "loss": 2.3451,
458
  "step": 7500
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
459
  }
460
  ],
461
  "max_steps": 56450,
462
  "num_train_epochs": 10,
463
- "total_flos": 2.787290934878085e+17,
464
  "trial_name": null,
465
  "trial_params": null
466
  }
 
1
  {
2
  "best_metric": null,
3
  "best_model_checkpoint": null,
4
+ "epoch": 4.16296886763208,
5
+ "global_step": 23500,
6
  "is_hyper_param_search": false,
7
  "is_local_process_zero": true,
8
  "is_world_process_zero": true,
 
456
  "learning_rate": 4.33569530558016e-05,
457
  "loss": 2.3451,
458
  "step": 7500
459
+ },
460
+ {
461
+ "epoch": 1.35,
462
+ "learning_rate": 4.326837909654562e-05,
463
+ "loss": 2.3302,
464
+ "step": 7600
465
+ },
466
+ {
467
+ "epoch": 1.36,
468
+ "learning_rate": 4.317980513728964e-05,
469
+ "loss": 2.3673,
470
+ "step": 7700
471
+ },
472
+ {
473
+ "epoch": 1.38,
474
+ "learning_rate": 4.309123117803366e-05,
475
+ "loss": 2.29,
476
+ "step": 7800
477
+ },
478
+ {
479
+ "epoch": 1.4,
480
+ "learning_rate": 4.300265721877768e-05,
481
+ "loss": 2.3216,
482
+ "step": 7900
483
+ },
484
+ {
485
+ "epoch": 1.42,
486
+ "learning_rate": 4.29140832595217e-05,
487
+ "loss": 2.3616,
488
+ "step": 8000
489
+ },
490
+ {
491
+ "epoch": 1.43,
492
+ "learning_rate": 4.282550930026573e-05,
493
+ "loss": 2.3446,
494
+ "step": 8100
495
+ },
496
+ {
497
+ "epoch": 1.45,
498
+ "learning_rate": 4.273693534100974e-05,
499
+ "loss": 2.3276,
500
+ "step": 8200
501
+ },
502
+ {
503
+ "epoch": 1.47,
504
+ "learning_rate": 4.264836138175377e-05,
505
+ "loss": 2.3449,
506
+ "step": 8300
507
+ },
508
+ {
509
+ "epoch": 1.49,
510
+ "learning_rate": 4.255978742249779e-05,
511
+ "loss": 2.3174,
512
+ "step": 8400
513
+ },
514
+ {
515
+ "epoch": 1.51,
516
+ "learning_rate": 4.247121346324181e-05,
517
+ "loss": 2.3087,
518
+ "step": 8500
519
+ },
520
+ {
521
+ "epoch": 1.52,
522
+ "learning_rate": 4.238263950398583e-05,
523
+ "loss": 2.316,
524
+ "step": 8600
525
+ },
526
+ {
527
+ "epoch": 1.54,
528
+ "learning_rate": 4.229406554472985e-05,
529
+ "loss": 2.2905,
530
+ "step": 8700
531
+ },
532
+ {
533
+ "epoch": 1.56,
534
+ "learning_rate": 4.220549158547387e-05,
535
+ "loss": 2.3347,
536
+ "step": 8800
537
+ },
538
+ {
539
+ "epoch": 1.58,
540
+ "learning_rate": 4.211691762621789e-05,
541
+ "loss": 2.2881,
542
+ "step": 8900
543
+ },
544
+ {
545
+ "epoch": 1.59,
546
+ "learning_rate": 4.202834366696192e-05,
547
+ "loss": 2.299,
548
+ "step": 9000
549
+ },
550
+ {
551
+ "epoch": 1.61,
552
+ "learning_rate": 4.193976970770593e-05,
553
+ "loss": 2.2814,
554
+ "step": 9100
555
+ },
556
+ {
557
+ "epoch": 1.63,
558
+ "learning_rate": 4.185119574844996e-05,
559
+ "loss": 2.3295,
560
+ "step": 9200
561
+ },
562
+ {
563
+ "epoch": 1.65,
564
+ "learning_rate": 4.1762621789193976e-05,
565
+ "loss": 2.3077,
566
+ "step": 9300
567
+ },
568
+ {
569
+ "epoch": 1.67,
570
+ "learning_rate": 4.1674047829938e-05,
571
+ "loss": 2.2789,
572
+ "step": 9400
573
+ },
574
+ {
575
+ "epoch": 1.68,
576
+ "learning_rate": 4.158547387068202e-05,
577
+ "loss": 2.3245,
578
+ "step": 9500
579
+ },
580
+ {
581
+ "epoch": 1.7,
582
+ "learning_rate": 4.149689991142604e-05,
583
+ "loss": 2.3179,
584
+ "step": 9600
585
+ },
586
+ {
587
+ "epoch": 1.72,
588
+ "learning_rate": 4.140832595217006e-05,
589
+ "loss": 2.3154,
590
+ "step": 9700
591
+ },
592
+ {
593
+ "epoch": 1.74,
594
+ "learning_rate": 4.131975199291409e-05,
595
+ "loss": 2.307,
596
+ "step": 9800
597
+ },
598
+ {
599
+ "epoch": 1.75,
600
+ "learning_rate": 4.123117803365811e-05,
601
+ "loss": 2.2671,
602
+ "step": 9900
603
+ },
604
+ {
605
+ "epoch": 1.77,
606
+ "learning_rate": 4.114260407440213e-05,
607
+ "loss": 2.2912,
608
+ "step": 10000
609
+ },
610
+ {
611
+ "epoch": 1.79,
612
+ "learning_rate": 4.1054030115146146e-05,
613
+ "loss": 2.3039,
614
+ "step": 10100
615
+ },
616
+ {
617
+ "epoch": 1.81,
618
+ "learning_rate": 4.096545615589017e-05,
619
+ "loss": 2.3131,
620
+ "step": 10200
621
+ },
622
+ {
623
+ "epoch": 1.82,
624
+ "learning_rate": 4.087688219663419e-05,
625
+ "loss": 2.2901,
626
+ "step": 10300
627
+ },
628
+ {
629
+ "epoch": 1.84,
630
+ "learning_rate": 4.078830823737822e-05,
631
+ "loss": 2.3162,
632
+ "step": 10400
633
+ },
634
+ {
635
+ "epoch": 1.86,
636
+ "learning_rate": 4.069973427812223e-05,
637
+ "loss": 2.3314,
638
+ "step": 10500
639
+ },
640
+ {
641
+ "epoch": 1.88,
642
+ "learning_rate": 4.061116031886626e-05,
643
+ "loss": 2.3893,
644
+ "step": 10600
645
+ },
646
+ {
647
+ "epoch": 1.9,
648
+ "learning_rate": 4.0522586359610276e-05,
649
+ "loss": 2.4248,
650
+ "step": 10700
651
+ },
652
+ {
653
+ "epoch": 1.91,
654
+ "learning_rate": 4.0434012400354296e-05,
655
+ "loss": 2.4211,
656
+ "step": 10800
657
+ },
658
+ {
659
+ "epoch": 1.93,
660
+ "learning_rate": 4.034543844109832e-05,
661
+ "loss": 2.4191,
662
+ "step": 10900
663
+ },
664
+ {
665
+ "epoch": 1.95,
666
+ "learning_rate": 4.0256864481842335e-05,
667
+ "loss": 2.4139,
668
+ "step": 11000
669
+ },
670
+ {
671
+ "epoch": 1.97,
672
+ "learning_rate": 4.016829052258636e-05,
673
+ "loss": 2.3533,
674
+ "step": 11100
675
+ },
676
+ {
677
+ "epoch": 1.98,
678
+ "learning_rate": 4.007971656333038e-05,
679
+ "loss": 2.3789,
680
+ "step": 11200
681
+ },
682
+ {
683
+ "epoch": 2.0,
684
+ "learning_rate": 3.9991142604074407e-05,
685
+ "loss": 2.3682,
686
+ "step": 11300
687
+ },
688
+ {
689
+ "epoch": 2.02,
690
+ "learning_rate": 3.9902568644818426e-05,
691
+ "loss": 2.3127,
692
+ "step": 11400
693
+ },
694
+ {
695
+ "epoch": 2.04,
696
+ "learning_rate": 3.9813994685562446e-05,
697
+ "loss": 2.3951,
698
+ "step": 11500
699
+ },
700
+ {
701
+ "epoch": 2.05,
702
+ "learning_rate": 3.9725420726306465e-05,
703
+ "loss": 2.348,
704
+ "step": 11600
705
+ },
706
+ {
707
+ "epoch": 2.07,
708
+ "learning_rate": 3.963684676705049e-05,
709
+ "loss": 2.3298,
710
+ "step": 11700
711
+ },
712
+ {
713
+ "epoch": 2.09,
714
+ "learning_rate": 3.954827280779451e-05,
715
+ "loss": 2.348,
716
+ "step": 11800
717
+ },
718
+ {
719
+ "epoch": 2.11,
720
+ "learning_rate": 3.945969884853853e-05,
721
+ "loss": 2.2729,
722
+ "step": 11900
723
+ },
724
+ {
725
+ "epoch": 2.13,
726
+ "learning_rate": 3.937112488928255e-05,
727
+ "loss": 2.3321,
728
+ "step": 12000
729
+ },
730
+ {
731
+ "epoch": 2.14,
732
+ "learning_rate": 3.9282550930026576e-05,
733
+ "loss": 2.3571,
734
+ "step": 12100
735
+ },
736
+ {
737
+ "epoch": 2.16,
738
+ "learning_rate": 3.9193976970770595e-05,
739
+ "loss": 2.3534,
740
+ "step": 12200
741
+ },
742
+ {
743
+ "epoch": 2.18,
744
+ "learning_rate": 3.910540301151462e-05,
745
+ "loss": 2.3468,
746
+ "step": 12300
747
+ },
748
+ {
749
+ "epoch": 2.2,
750
+ "learning_rate": 3.9016829052258634e-05,
751
+ "loss": 2.3568,
752
+ "step": 12400
753
+ },
754
+ {
755
+ "epoch": 2.21,
756
+ "learning_rate": 3.892825509300266e-05,
757
+ "loss": 2.344,
758
+ "step": 12500
759
+ },
760
+ {
761
+ "epoch": 2.23,
762
+ "learning_rate": 3.883968113374668e-05,
763
+ "loss": 2.3459,
764
+ "step": 12600
765
+ },
766
+ {
767
+ "epoch": 2.25,
768
+ "learning_rate": 3.87511071744907e-05,
769
+ "loss": 2.34,
770
+ "step": 12700
771
+ },
772
+ {
773
+ "epoch": 2.27,
774
+ "learning_rate": 3.8662533215234726e-05,
775
+ "loss": 2.3478,
776
+ "step": 12800
777
+ },
778
+ {
779
+ "epoch": 2.29,
780
+ "learning_rate": 3.8573959255978745e-05,
781
+ "loss": 2.3262,
782
+ "step": 12900
783
+ },
784
+ {
785
+ "epoch": 2.3,
786
+ "learning_rate": 3.8485385296722765e-05,
787
+ "loss": 2.375,
788
+ "step": 13000
789
+ },
790
+ {
791
+ "epoch": 2.32,
792
+ "learning_rate": 3.8396811337466784e-05,
793
+ "loss": 2.3364,
794
+ "step": 13100
795
+ },
796
+ {
797
+ "epoch": 2.34,
798
+ "learning_rate": 3.830823737821081e-05,
799
+ "loss": 2.3101,
800
+ "step": 13200
801
+ },
802
+ {
803
+ "epoch": 2.36,
804
+ "learning_rate": 3.821966341895482e-05,
805
+ "loss": 2.3402,
806
+ "step": 13300
807
+ },
808
+ {
809
+ "epoch": 2.37,
810
+ "learning_rate": 3.813108945969885e-05,
811
+ "loss": 2.3722,
812
+ "step": 13400
813
+ },
814
+ {
815
+ "epoch": 2.39,
816
+ "learning_rate": 3.804251550044287e-05,
817
+ "loss": 2.3062,
818
+ "step": 13500
819
+ },
820
+ {
821
+ "epoch": 2.41,
822
+ "learning_rate": 3.7953941541186895e-05,
823
+ "loss": 2.295,
824
+ "step": 13600
825
+ },
826
+ {
827
+ "epoch": 2.43,
828
+ "learning_rate": 3.7865367581930915e-05,
829
+ "loss": 2.2954,
830
+ "step": 13700
831
+ },
832
+ {
833
+ "epoch": 2.44,
834
+ "learning_rate": 3.7776793622674934e-05,
835
+ "loss": 2.3376,
836
+ "step": 13800
837
+ },
838
+ {
839
+ "epoch": 2.46,
840
+ "learning_rate": 3.7688219663418954e-05,
841
+ "loss": 2.2626,
842
+ "step": 13900
843
+ },
844
+ {
845
+ "epoch": 2.48,
846
+ "learning_rate": 3.759964570416298e-05,
847
+ "loss": 2.3039,
848
+ "step": 14000
849
+ },
850
+ {
851
+ "epoch": 2.5,
852
+ "learning_rate": 3.7511071744907e-05,
853
+ "loss": 2.3199,
854
+ "step": 14100
855
+ },
856
+ {
857
+ "epoch": 2.52,
858
+ "learning_rate": 3.7422497785651026e-05,
859
+ "loss": 2.3196,
860
+ "step": 14200
861
+ },
862
+ {
863
+ "epoch": 2.53,
864
+ "learning_rate": 3.733392382639504e-05,
865
+ "loss": 2.2705,
866
+ "step": 14300
867
+ },
868
+ {
869
+ "epoch": 2.55,
870
+ "learning_rate": 3.7245349867139065e-05,
871
+ "loss": 2.316,
872
+ "step": 14400
873
+ },
874
+ {
875
+ "epoch": 2.57,
876
+ "learning_rate": 3.7156775907883084e-05,
877
+ "loss": 2.3111,
878
+ "step": 14500
879
+ },
880
+ {
881
+ "epoch": 2.59,
882
+ "learning_rate": 3.706820194862711e-05,
883
+ "loss": 2.3043,
884
+ "step": 14600
885
+ },
886
+ {
887
+ "epoch": 2.6,
888
+ "learning_rate": 3.697962798937112e-05,
889
+ "loss": 2.3409,
890
+ "step": 14700
891
+ },
892
+ {
893
+ "epoch": 2.62,
894
+ "learning_rate": 3.689105403011515e-05,
895
+ "loss": 2.3246,
896
+ "step": 14800
897
+ },
898
+ {
899
+ "epoch": 2.64,
900
+ "learning_rate": 3.680248007085917e-05,
901
+ "loss": 2.3275,
902
+ "step": 14900
903
+ },
904
+ {
905
+ "epoch": 2.66,
906
+ "learning_rate": 3.671390611160319e-05,
907
+ "loss": 2.2977,
908
+ "step": 15000
909
+ },
910
+ {
911
+ "epoch": 2.67,
912
+ "learning_rate": 3.6625332152347214e-05,
913
+ "loss": 2.3095,
914
+ "step": 15100
915
+ },
916
+ {
917
+ "epoch": 2.69,
918
+ "learning_rate": 3.6536758193091234e-05,
919
+ "loss": 2.3021,
920
+ "step": 15200
921
+ },
922
+ {
923
+ "epoch": 2.71,
924
+ "learning_rate": 3.6448184233835253e-05,
925
+ "loss": 2.2793,
926
+ "step": 15300
927
+ },
928
+ {
929
+ "epoch": 2.73,
930
+ "learning_rate": 3.635961027457927e-05,
931
+ "loss": 2.3148,
932
+ "step": 15400
933
+ },
934
+ {
935
+ "epoch": 2.75,
936
+ "learning_rate": 3.62710363153233e-05,
937
+ "loss": 2.3496,
938
+ "step": 15500
939
+ },
940
+ {
941
+ "epoch": 2.76,
942
+ "learning_rate": 3.618246235606732e-05,
943
+ "loss": 2.2763,
944
+ "step": 15600
945
+ },
946
+ {
947
+ "epoch": 2.78,
948
+ "learning_rate": 3.609388839681134e-05,
949
+ "loss": 2.3177,
950
+ "step": 15700
951
+ },
952
+ {
953
+ "epoch": 2.8,
954
+ "learning_rate": 3.600531443755536e-05,
955
+ "loss": 2.2661,
956
+ "step": 15800
957
+ },
958
+ {
959
+ "epoch": 2.82,
960
+ "learning_rate": 3.5916740478299384e-05,
961
+ "loss": 2.2856,
962
+ "step": 15900
963
+ },
964
+ {
965
+ "epoch": 2.83,
966
+ "learning_rate": 3.58281665190434e-05,
967
+ "loss": 2.294,
968
+ "step": 16000
969
+ },
970
+ {
971
+ "epoch": 2.85,
972
+ "learning_rate": 3.573959255978742e-05,
973
+ "loss": 2.2989,
974
+ "step": 16100
975
+ },
976
+ {
977
+ "epoch": 2.87,
978
+ "learning_rate": 3.565101860053144e-05,
979
+ "loss": 2.2981,
980
+ "step": 16200
981
+ },
982
+ {
983
+ "epoch": 2.89,
984
+ "learning_rate": 3.556244464127547e-05,
985
+ "loss": 2.2998,
986
+ "step": 16300
987
+ },
988
+ {
989
+ "epoch": 2.91,
990
+ "learning_rate": 3.547387068201949e-05,
991
+ "loss": 2.2688,
992
+ "step": 16400
993
+ },
994
+ {
995
+ "epoch": 2.92,
996
+ "learning_rate": 3.5385296722763514e-05,
997
+ "loss": 2.252,
998
+ "step": 16500
999
+ },
1000
+ {
1001
+ "epoch": 2.94,
1002
+ "learning_rate": 3.529672276350753e-05,
1003
+ "loss": 2.2984,
1004
+ "step": 16600
1005
+ },
1006
+ {
1007
+ "epoch": 2.96,
1008
+ "learning_rate": 3.520814880425155e-05,
1009
+ "loss": 2.2741,
1010
+ "step": 16700
1011
+ },
1012
+ {
1013
+ "epoch": 2.98,
1014
+ "learning_rate": 3.511957484499557e-05,
1015
+ "loss": 2.2973,
1016
+ "step": 16800
1017
+ },
1018
+ {
1019
+ "epoch": 2.99,
1020
+ "learning_rate": 3.503100088573959e-05,
1021
+ "loss": 2.3119,
1022
+ "step": 16900
1023
+ },
1024
+ {
1025
+ "epoch": 3.01,
1026
+ "learning_rate": 3.494242692648362e-05,
1027
+ "loss": 2.2745,
1028
+ "step": 17000
1029
+ },
1030
+ {
1031
+ "epoch": 3.03,
1032
+ "learning_rate": 3.485385296722764e-05,
1033
+ "loss": 2.2681,
1034
+ "step": 17100
1035
+ },
1036
+ {
1037
+ "epoch": 3.05,
1038
+ "learning_rate": 3.476527900797166e-05,
1039
+ "loss": 2.2529,
1040
+ "step": 17200
1041
+ },
1042
+ {
1043
+ "epoch": 3.06,
1044
+ "learning_rate": 3.467670504871568e-05,
1045
+ "loss": 2.219,
1046
+ "step": 17300
1047
+ },
1048
+ {
1049
+ "epoch": 3.08,
1050
+ "learning_rate": 3.45881310894597e-05,
1051
+ "loss": 2.2852,
1052
+ "step": 17400
1053
+ },
1054
+ {
1055
+ "epoch": 3.1,
1056
+ "learning_rate": 3.4499557130203716e-05,
1057
+ "loss": 2.2975,
1058
+ "step": 17500
1059
+ },
1060
+ {
1061
+ "epoch": 3.12,
1062
+ "learning_rate": 3.441098317094774e-05,
1063
+ "loss": 2.2538,
1064
+ "step": 17600
1065
+ },
1066
+ {
1067
+ "epoch": 3.14,
1068
+ "learning_rate": 3.432240921169176e-05,
1069
+ "loss": 2.2208,
1070
+ "step": 17700
1071
+ },
1072
+ {
1073
+ "epoch": 3.15,
1074
+ "learning_rate": 3.423383525243579e-05,
1075
+ "loss": 2.2279,
1076
+ "step": 17800
1077
+ },
1078
+ {
1079
+ "epoch": 3.17,
1080
+ "learning_rate": 3.414526129317981e-05,
1081
+ "loss": 2.2153,
1082
+ "step": 17900
1083
+ },
1084
+ {
1085
+ "epoch": 3.19,
1086
+ "learning_rate": 3.405668733392383e-05,
1087
+ "loss": 2.2567,
1088
+ "step": 18000
1089
+ },
1090
+ {
1091
+ "epoch": 3.21,
1092
+ "learning_rate": 3.3968113374667846e-05,
1093
+ "loss": 2.238,
1094
+ "step": 18100
1095
+ },
1096
+ {
1097
+ "epoch": 3.22,
1098
+ "learning_rate": 3.387953941541187e-05,
1099
+ "loss": 2.2724,
1100
+ "step": 18200
1101
+ },
1102
+ {
1103
+ "epoch": 3.24,
1104
+ "learning_rate": 3.379096545615589e-05,
1105
+ "loss": 2.2258,
1106
+ "step": 18300
1107
+ },
1108
+ {
1109
+ "epoch": 3.26,
1110
+ "learning_rate": 3.370239149689992e-05,
1111
+ "loss": 2.2305,
1112
+ "step": 18400
1113
+ },
1114
+ {
1115
+ "epoch": 3.28,
1116
+ "learning_rate": 3.361381753764393e-05,
1117
+ "loss": 2.2686,
1118
+ "step": 18500
1119
+ },
1120
+ {
1121
+ "epoch": 3.29,
1122
+ "learning_rate": 3.352524357838796e-05,
1123
+ "loss": 2.2579,
1124
+ "step": 18600
1125
+ },
1126
+ {
1127
+ "epoch": 3.31,
1128
+ "learning_rate": 3.343666961913198e-05,
1129
+ "loss": 2.2614,
1130
+ "step": 18700
1131
+ },
1132
+ {
1133
+ "epoch": 3.33,
1134
+ "learning_rate": 3.3348095659876e-05,
1135
+ "loss": 2.2328,
1136
+ "step": 18800
1137
+ },
1138
+ {
1139
+ "epoch": 3.35,
1140
+ "learning_rate": 3.3259521700620016e-05,
1141
+ "loss": 2.2333,
1142
+ "step": 18900
1143
+ },
1144
+ {
1145
+ "epoch": 3.37,
1146
+ "learning_rate": 3.317094774136404e-05,
1147
+ "loss": 2.2409,
1148
+ "step": 19000
1149
+ },
1150
+ {
1151
+ "epoch": 3.38,
1152
+ "learning_rate": 3.308237378210806e-05,
1153
+ "loss": 2.2329,
1154
+ "step": 19100
1155
+ },
1156
+ {
1157
+ "epoch": 3.4,
1158
+ "learning_rate": 3.299379982285208e-05,
1159
+ "loss": 2.2507,
1160
+ "step": 19200
1161
+ },
1162
+ {
1163
+ "epoch": 3.42,
1164
+ "learning_rate": 3.290522586359611e-05,
1165
+ "loss": 2.2749,
1166
+ "step": 19300
1167
+ },
1168
+ {
1169
+ "epoch": 3.44,
1170
+ "learning_rate": 3.2816651904340127e-05,
1171
+ "loss": 2.2754,
1172
+ "step": 19400
1173
+ },
1174
+ {
1175
+ "epoch": 3.45,
1176
+ "learning_rate": 3.2728077945084146e-05,
1177
+ "loss": 2.2784,
1178
+ "step": 19500
1179
+ },
1180
+ {
1181
+ "epoch": 3.47,
1182
+ "learning_rate": 3.2639503985828165e-05,
1183
+ "loss": 2.2722,
1184
+ "step": 19600
1185
+ },
1186
+ {
1187
+ "epoch": 3.49,
1188
+ "learning_rate": 3.255093002657219e-05,
1189
+ "loss": 2.2446,
1190
+ "step": 19700
1191
+ },
1192
+ {
1193
+ "epoch": 3.51,
1194
+ "learning_rate": 3.246235606731621e-05,
1195
+ "loss": 2.2728,
1196
+ "step": 19800
1197
+ },
1198
+ {
1199
+ "epoch": 3.53,
1200
+ "learning_rate": 3.237378210806023e-05,
1201
+ "loss": 2.2227,
1202
+ "step": 19900
1203
+ },
1204
+ {
1205
+ "epoch": 3.54,
1206
+ "learning_rate": 3.228520814880425e-05,
1207
+ "loss": 2.2475,
1208
+ "step": 20000
1209
+ },
1210
+ {
1211
+ "epoch": 3.56,
1212
+ "learning_rate": 3.2196634189548276e-05,
1213
+ "loss": 2.2331,
1214
+ "step": 20100
1215
+ },
1216
+ {
1217
+ "epoch": 3.58,
1218
+ "learning_rate": 3.2108060230292296e-05,
1219
+ "loss": 2.2311,
1220
+ "step": 20200
1221
+ },
1222
+ {
1223
+ "epoch": 3.6,
1224
+ "learning_rate": 3.2019486271036315e-05,
1225
+ "loss": 2.2233,
1226
+ "step": 20300
1227
+ },
1228
+ {
1229
+ "epoch": 3.61,
1230
+ "learning_rate": 3.1930912311780335e-05,
1231
+ "loss": 2.2881,
1232
+ "step": 20400
1233
+ },
1234
+ {
1235
+ "epoch": 3.63,
1236
+ "learning_rate": 3.184233835252436e-05,
1237
+ "loss": 2.2643,
1238
+ "step": 20500
1239
+ },
1240
+ {
1241
+ "epoch": 3.65,
1242
+ "learning_rate": 3.175376439326838e-05,
1243
+ "loss": 2.2226,
1244
+ "step": 20600
1245
+ },
1246
+ {
1247
+ "epoch": 3.67,
1248
+ "learning_rate": 3.166519043401241e-05,
1249
+ "loss": 2.241,
1250
+ "step": 20700
1251
+ },
1252
+ {
1253
+ "epoch": 3.68,
1254
+ "learning_rate": 3.157661647475642e-05,
1255
+ "loss": 2.2243,
1256
+ "step": 20800
1257
+ },
1258
+ {
1259
+ "epoch": 3.7,
1260
+ "learning_rate": 3.1488042515500446e-05,
1261
+ "loss": 2.2507,
1262
+ "step": 20900
1263
+ },
1264
+ {
1265
+ "epoch": 3.72,
1266
+ "learning_rate": 3.1399468556244465e-05,
1267
+ "loss": 2.2178,
1268
+ "step": 21000
1269
+ },
1270
+ {
1271
+ "epoch": 3.74,
1272
+ "learning_rate": 3.131089459698849e-05,
1273
+ "loss": 2.2204,
1274
+ "step": 21100
1275
+ },
1276
+ {
1277
+ "epoch": 3.76,
1278
+ "learning_rate": 3.122232063773251e-05,
1279
+ "loss": 2.2293,
1280
+ "step": 21200
1281
+ },
1282
+ {
1283
+ "epoch": 3.77,
1284
+ "learning_rate": 3.113374667847653e-05,
1285
+ "loss": 2.2273,
1286
+ "step": 21300
1287
+ },
1288
+ {
1289
+ "epoch": 3.79,
1290
+ "learning_rate": 3.104517271922055e-05,
1291
+ "loss": 2.2294,
1292
+ "step": 21400
1293
+ },
1294
+ {
1295
+ "epoch": 3.81,
1296
+ "learning_rate": 3.095659875996457e-05,
1297
+ "loss": 2.1951,
1298
+ "step": 21500
1299
+ },
1300
+ {
1301
+ "epoch": 3.83,
1302
+ "learning_rate": 3.0868024800708596e-05,
1303
+ "loss": 2.1944,
1304
+ "step": 21600
1305
+ },
1306
+ {
1307
+ "epoch": 3.84,
1308
+ "learning_rate": 3.077945084145261e-05,
1309
+ "loss": 2.2176,
1310
+ "step": 21700
1311
+ },
1312
+ {
1313
+ "epoch": 3.86,
1314
+ "learning_rate": 3.0690876882196635e-05,
1315
+ "loss": 2.1841,
1316
+ "step": 21800
1317
+ },
1318
+ {
1319
+ "epoch": 3.88,
1320
+ "learning_rate": 3.0602302922940654e-05,
1321
+ "loss": 2.259,
1322
+ "step": 21900
1323
+ },
1324
+ {
1325
+ "epoch": 3.9,
1326
+ "learning_rate": 3.051372896368468e-05,
1327
+ "loss": 2.2386,
1328
+ "step": 22000
1329
+ },
1330
+ {
1331
+ "epoch": 3.91,
1332
+ "learning_rate": 3.04251550044287e-05,
1333
+ "loss": 2.2392,
1334
+ "step": 22100
1335
+ },
1336
+ {
1337
+ "epoch": 3.93,
1338
+ "learning_rate": 3.033658104517272e-05,
1339
+ "loss": 2.2674,
1340
+ "step": 22200
1341
+ },
1342
+ {
1343
+ "epoch": 3.95,
1344
+ "learning_rate": 3.0248007085916742e-05,
1345
+ "loss": 2.2529,
1346
+ "step": 22300
1347
+ },
1348
+ {
1349
+ "epoch": 3.97,
1350
+ "learning_rate": 3.015943312666076e-05,
1351
+ "loss": 2.2233,
1352
+ "step": 22400
1353
+ },
1354
+ {
1355
+ "epoch": 3.99,
1356
+ "learning_rate": 3.0070859167404785e-05,
1357
+ "loss": 2.1672,
1358
+ "step": 22500
1359
+ },
1360
+ {
1361
+ "epoch": 4.0,
1362
+ "learning_rate": 2.9982285208148807e-05,
1363
+ "loss": 2.1987,
1364
+ "step": 22600
1365
+ },
1366
+ {
1367
+ "epoch": 4.02,
1368
+ "learning_rate": 2.9893711248892823e-05,
1369
+ "loss": 2.2067,
1370
+ "step": 22700
1371
+ },
1372
+ {
1373
+ "epoch": 4.04,
1374
+ "learning_rate": 2.9805137289636846e-05,
1375
+ "loss": 2.185,
1376
+ "step": 22800
1377
+ },
1378
+ {
1379
+ "epoch": 4.06,
1380
+ "learning_rate": 2.971656333038087e-05,
1381
+ "loss": 2.1959,
1382
+ "step": 22900
1383
+ },
1384
+ {
1385
+ "epoch": 4.07,
1386
+ "learning_rate": 2.9627989371124892e-05,
1387
+ "loss": 2.1773,
1388
+ "step": 23000
1389
+ },
1390
+ {
1391
+ "epoch": 4.09,
1392
+ "learning_rate": 2.9539415411868908e-05,
1393
+ "loss": 2.1577,
1394
+ "step": 23100
1395
+ },
1396
+ {
1397
+ "epoch": 4.11,
1398
+ "learning_rate": 2.945084145261293e-05,
1399
+ "loss": 2.1838,
1400
+ "step": 23200
1401
+ },
1402
+ {
1403
+ "epoch": 4.13,
1404
+ "learning_rate": 2.9362267493356954e-05,
1405
+ "loss": 2.1854,
1406
+ "step": 23300
1407
+ },
1408
+ {
1409
+ "epoch": 4.15,
1410
+ "learning_rate": 2.9273693534100977e-05,
1411
+ "loss": 2.1923,
1412
+ "step": 23400
1413
+ },
1414
+ {
1415
+ "epoch": 4.16,
1416
+ "learning_rate": 2.9185119574845e-05,
1417
+ "loss": 2.191,
1418
+ "step": 23500
1419
  }
1420
  ],
1421
  "max_steps": 56450,
1422
  "num_train_epochs": 10,
1423
+ "total_flos": 8.72923785124373e+17,
1424
  "trial_name": null,
1425
  "trial_params": null
1426
  }