qubvel-hf HF staff commited on
Commit
fe3eb32
1 Parent(s): 6d7435b

End of training

Browse files
README.md CHANGED
@@ -3,6 +3,8 @@ library_name: transformers
3
  license: apache-2.0
4
  base_model: timm/resnet18.a1_in1k
5
  tags:
 
 
6
  - generated_from_trainer
7
  metrics:
8
  - accuracy
@@ -16,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  # vit-base-beans
18
 
19
- This model is a fine-tuned version of [timm/resnet18.a1_in1k](https://huggingface.co/timm/resnet18.a1_in1k) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.8550
22
  - Accuracy: 0.7895
 
3
  license: apache-2.0
4
  base_model: timm/resnet18.a1_in1k
5
  tags:
6
+ - image-classification
7
+ - vision
8
  - generated_from_trainer
9
  metrics:
10
  - accuracy
 
18
 
19
  # vit-base-beans
20
 
21
+ This model is a fine-tuned version of [timm/resnet18.a1_in1k](https://huggingface.co/timm/resnet18.a1_in1k) on the beans dataset.
22
  It achieves the following results on the evaluation set:
23
  - Loss: 0.8550
24
  - Accuracy: 0.7895
all_results.json CHANGED
@@ -1,13 +1,13 @@
1
  {
2
- "epoch": 5.0,
3
- "eval_accuracy": 0.6766917293233082,
4
- "eval_loss": 1.0332472324371338,
5
- "eval_runtime": 0.7765,
6
- "eval_samples_per_second": 171.282,
7
- "eval_steps_per_second": 21.893,
8
- "total_flos": 5.219455168198656e+16,
9
- "train_loss": 1.0677934587918796,
10
- "train_runtime": 48.1154,
11
- "train_samples_per_second": 107.45,
12
- "train_steps_per_second": 13.509
13
  }
 
1
  {
2
+ "epoch": 15.0,
3
+ "eval_accuracy": 0.7894736842105263,
4
+ "eval_loss": 0.8550169467926025,
5
+ "eval_runtime": 0.7658,
6
+ "eval_samples_per_second": 173.669,
7
+ "eval_steps_per_second": 22.198,
8
+ "total_flos": 1.5658365504595968e+17,
9
+ "train_loss": 0.6299933981284117,
10
+ "train_runtime": 92.772,
11
+ "train_samples_per_second": 167.184,
12
+ "train_steps_per_second": 21.019
13
  }
eval_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 5.0,
3
- "eval_accuracy": 0.6766917293233082,
4
- "eval_loss": 1.0332472324371338,
5
- "eval_runtime": 0.7765,
6
- "eval_samples_per_second": 171.282,
7
- "eval_steps_per_second": 21.893
8
  }
 
1
  {
2
+ "epoch": 15.0,
3
+ "eval_accuracy": 0.7894736842105263,
4
+ "eval_loss": 0.8550169467926025,
5
+ "eval_runtime": 0.7658,
6
+ "eval_samples_per_second": 173.669,
7
+ "eval_steps_per_second": 22.198
8
  }
runs/Nov01_15-31-14_ip-10-90-1-182/events.out.tfevents.1730475174.ip-10-90-1-182.137679.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5dfb2648b394157822070d8c34e8294be4a258f14b7fabf6baad529b3d9466c
3
+ size 411
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 5.0,
3
- "total_flos": 5.219455168198656e+16,
4
- "train_loss": 1.0677934587918796,
5
- "train_runtime": 48.1154,
6
- "train_samples_per_second": 107.45,
7
- "train_steps_per_second": 13.509
8
  }
 
1
  {
2
+ "epoch": 15.0,
3
+ "total_flos": 1.5658365504595968e+17,
4
+ "train_loss": 0.6299933981284117,
5
+ "train_runtime": 92.772,
6
+ "train_samples_per_second": 167.184,
7
+ "train_steps_per_second": 21.019
8
  }
trainer_state.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
- "best_metric": 1.0332472324371338,
3
- "best_model_checkpoint": "./beans_outputs/checkpoint-650",
4
- "epoch": 5.0,
5
  "eval_steps": 500,
6
- "global_step": 650,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
@@ -509,19 +509,1019 @@
509
  "step": 650
510
  },
511
  {
512
- "epoch": 5.0,
513
- "step": 650,
514
- "total_flos": 5.219455168198656e+16,
515
- "train_loss": 1.0677934587918796,
516
- "train_runtime": 48.1154,
517
- "train_samples_per_second": 107.45,
518
- "train_steps_per_second": 13.509
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
519
  }
520
  ],
521
  "logging_steps": 10,
522
- "max_steps": 650,
523
  "num_input_tokens_seen": 0,
524
- "num_train_epochs": 5,
525
  "save_steps": 500,
526
  "stateful_callbacks": {
527
  "TrainerControl": {
@@ -535,7 +1535,7 @@
535
  "attributes": {}
536
  }
537
  },
538
- "total_flos": 5.219455168198656e+16,
539
  "train_batch_size": 8,
540
  "trial_name": null,
541
  "trial_params": null
 
1
  {
2
+ "best_metric": 0.8550169467926025,
3
+ "best_model_checkpoint": "./beans_outputs/checkpoint-1950",
4
+ "epoch": 15.0,
5
  "eval_steps": 500,
6
+ "global_step": 1950,
7
  "is_hyper_param_search": false,
8
  "is_local_process_zero": true,
9
  "is_world_process_zero": true,
 
509
  "step": 650
510
  },
511
  {
512
+ "epoch": 5.076923076923077,
513
+ "grad_norm": 1.9310355186462402,
514
+ "learning_rate": 1.3230769230769231e-05,
515
+ "loss": 1.0281,
516
+ "step": 660
517
+ },
518
+ {
519
+ "epoch": 5.153846153846154,
520
+ "grad_norm": 2.36603045463562,
521
+ "learning_rate": 1.312820512820513e-05,
522
+ "loss": 1.0587,
523
+ "step": 670
524
+ },
525
+ {
526
+ "epoch": 5.230769230769231,
527
+ "grad_norm": 2.0758917331695557,
528
+ "learning_rate": 1.3025641025641027e-05,
529
+ "loss": 1.0534,
530
+ "step": 680
531
+ },
532
+ {
533
+ "epoch": 5.3076923076923075,
534
+ "grad_norm": 2.549725294113159,
535
+ "learning_rate": 1.2923076923076925e-05,
536
+ "loss": 1.0421,
537
+ "step": 690
538
+ },
539
+ {
540
+ "epoch": 5.384615384615385,
541
+ "grad_norm": 2.7183680534362793,
542
+ "learning_rate": 1.2820512820512823e-05,
543
+ "loss": 1.035,
544
+ "step": 700
545
+ },
546
+ {
547
+ "epoch": 5.461538461538462,
548
+ "grad_norm": 1.7176955938339233,
549
+ "learning_rate": 1.2717948717948718e-05,
550
+ "loss": 1.0268,
551
+ "step": 710
552
+ },
553
+ {
554
+ "epoch": 5.538461538461538,
555
+ "grad_norm": 1.8257861137390137,
556
+ "learning_rate": 1.2615384615384616e-05,
557
+ "loss": 1.0302,
558
+ "step": 720
559
+ },
560
+ {
561
+ "epoch": 5.615384615384615,
562
+ "grad_norm": 2.50368595123291,
563
+ "learning_rate": 1.2512820512820514e-05,
564
+ "loss": 1.0331,
565
+ "step": 730
566
+ },
567
+ {
568
+ "epoch": 5.6923076923076925,
569
+ "grad_norm": 2.4315121173858643,
570
+ "learning_rate": 1.2410256410256412e-05,
571
+ "loss": 1.0208,
572
+ "step": 740
573
+ },
574
+ {
575
+ "epoch": 5.769230769230769,
576
+ "grad_norm": 2.043854236602783,
577
+ "learning_rate": 1.230769230769231e-05,
578
+ "loss": 1.0241,
579
+ "step": 750
580
+ },
581
+ {
582
+ "epoch": 5.846153846153846,
583
+ "grad_norm": 2.0800740718841553,
584
+ "learning_rate": 1.2205128205128208e-05,
585
+ "loss": 1.0251,
586
+ "step": 760
587
+ },
588
+ {
589
+ "epoch": 5.923076923076923,
590
+ "grad_norm": 3.4356396198272705,
591
+ "learning_rate": 1.2102564102564102e-05,
592
+ "loss": 1.0104,
593
+ "step": 770
594
+ },
595
+ {
596
+ "epoch": 6.0,
597
+ "grad_norm": 4.49416971206665,
598
+ "learning_rate": 1.2e-05,
599
+ "loss": 1.0357,
600
+ "step": 780
601
+ },
602
+ {
603
+ "epoch": 6.0,
604
+ "eval_accuracy": 0.6541353383458647,
605
+ "eval_loss": 1.0109117031097412,
606
+ "eval_runtime": 0.7801,
607
+ "eval_samples_per_second": 170.483,
608
+ "eval_steps_per_second": 21.791,
609
+ "step": 780
610
+ },
611
+ {
612
+ "epoch": 6.076923076923077,
613
+ "grad_norm": 2.650513172149658,
614
+ "learning_rate": 1.1897435897435898e-05,
615
+ "loss": 1.0148,
616
+ "step": 790
617
+ },
618
+ {
619
+ "epoch": 6.153846153846154,
620
+ "grad_norm": 2.3682632446289062,
621
+ "learning_rate": 1.1794871794871796e-05,
622
+ "loss": 1.0094,
623
+ "step": 800
624
+ },
625
+ {
626
+ "epoch": 6.230769230769231,
627
+ "grad_norm": 1.6716077327728271,
628
+ "learning_rate": 1.1692307692307694e-05,
629
+ "loss": 1.0097,
630
+ "step": 810
631
+ },
632
+ {
633
+ "epoch": 6.3076923076923075,
634
+ "grad_norm": 2.4839890003204346,
635
+ "learning_rate": 1.1589743589743592e-05,
636
+ "loss": 1.01,
637
+ "step": 820
638
+ },
639
+ {
640
+ "epoch": 6.384615384615385,
641
+ "grad_norm": 2.4004769325256348,
642
+ "learning_rate": 1.1487179487179487e-05,
643
+ "loss": 1.0104,
644
+ "step": 830
645
+ },
646
+ {
647
+ "epoch": 6.461538461538462,
648
+ "grad_norm": 2.9597084522247314,
649
+ "learning_rate": 1.1384615384615385e-05,
650
+ "loss": 1.0137,
651
+ "step": 840
652
+ },
653
+ {
654
+ "epoch": 6.538461538461538,
655
+ "grad_norm": 2.680335760116577,
656
+ "learning_rate": 1.1282051282051283e-05,
657
+ "loss": 1.0147,
658
+ "step": 850
659
+ },
660
+ {
661
+ "epoch": 6.615384615384615,
662
+ "grad_norm": 1.7677160501480103,
663
+ "learning_rate": 1.117948717948718e-05,
664
+ "loss": 0.9947,
665
+ "step": 860
666
+ },
667
+ {
668
+ "epoch": 6.6923076923076925,
669
+ "grad_norm": 2.0484132766723633,
670
+ "learning_rate": 1.1076923076923079e-05,
671
+ "loss": 1.0169,
672
+ "step": 870
673
+ },
674
+ {
675
+ "epoch": 6.769230769230769,
676
+ "grad_norm": 2.1910479068756104,
677
+ "learning_rate": 1.0974358974358977e-05,
678
+ "loss": 1.0024,
679
+ "step": 880
680
+ },
681
+ {
682
+ "epoch": 6.846153846153846,
683
+ "grad_norm": 2.181236743927002,
684
+ "learning_rate": 1.0871794871794871e-05,
685
+ "loss": 0.9962,
686
+ "step": 890
687
+ },
688
+ {
689
+ "epoch": 6.923076923076923,
690
+ "grad_norm": 2.898885488510132,
691
+ "learning_rate": 1.076923076923077e-05,
692
+ "loss": 1.012,
693
+ "step": 900
694
+ },
695
+ {
696
+ "epoch": 7.0,
697
+ "grad_norm": 4.950052738189697,
698
+ "learning_rate": 1.0666666666666667e-05,
699
+ "loss": 1.0012,
700
+ "step": 910
701
+ },
702
+ {
703
+ "epoch": 7.0,
704
+ "eval_accuracy": 0.7368421052631579,
705
+ "eval_loss": 0.981479287147522,
706
+ "eval_runtime": 0.7693,
707
+ "eval_samples_per_second": 172.886,
708
+ "eval_steps_per_second": 22.098,
709
+ "step": 910
710
+ },
711
+ {
712
+ "epoch": 7.076923076923077,
713
+ "grad_norm": 2.692753553390503,
714
+ "learning_rate": 1.0564102564102565e-05,
715
+ "loss": 0.9889,
716
+ "step": 920
717
+ },
718
+ {
719
+ "epoch": 7.153846153846154,
720
+ "grad_norm": 2.9175124168395996,
721
+ "learning_rate": 1.0461538461538463e-05,
722
+ "loss": 0.9911,
723
+ "step": 930
724
+ },
725
+ {
726
+ "epoch": 7.230769230769231,
727
+ "grad_norm": 3.221527099609375,
728
+ "learning_rate": 1.0358974358974361e-05,
729
+ "loss": 0.9827,
730
+ "step": 940
731
+ },
732
+ {
733
+ "epoch": 7.3076923076923075,
734
+ "grad_norm": 2.507923126220703,
735
+ "learning_rate": 1.0256410256410256e-05,
736
+ "loss": 0.9919,
737
+ "step": 950
738
+ },
739
+ {
740
+ "epoch": 7.384615384615385,
741
+ "grad_norm": 2.4533870220184326,
742
+ "learning_rate": 1.0153846153846154e-05,
743
+ "loss": 0.9962,
744
+ "step": 960
745
+ },
746
+ {
747
+ "epoch": 7.461538461538462,
748
+ "grad_norm": 2.1032631397247314,
749
+ "learning_rate": 1.0051282051282052e-05,
750
+ "loss": 0.9751,
751
+ "step": 970
752
+ },
753
+ {
754
+ "epoch": 7.538461538461538,
755
+ "grad_norm": 2.5848186016082764,
756
+ "learning_rate": 9.94871794871795e-06,
757
+ "loss": 0.9939,
758
+ "step": 980
759
+ },
760
+ {
761
+ "epoch": 7.615384615384615,
762
+ "grad_norm": 2.17742919921875,
763
+ "learning_rate": 9.846153846153848e-06,
764
+ "loss": 0.9745,
765
+ "step": 990
766
+ },
767
+ {
768
+ "epoch": 7.6923076923076925,
769
+ "grad_norm": 1.9953967332839966,
770
+ "learning_rate": 9.743589743589744e-06,
771
+ "loss": 0.9665,
772
+ "step": 1000
773
+ },
774
+ {
775
+ "epoch": 7.769230769230769,
776
+ "grad_norm": 3.0263218879699707,
777
+ "learning_rate": 9.641025641025642e-06,
778
+ "loss": 0.9646,
779
+ "step": 1010
780
+ },
781
+ {
782
+ "epoch": 7.846153846153846,
783
+ "grad_norm": 2.3735406398773193,
784
+ "learning_rate": 9.53846153846154e-06,
785
+ "loss": 0.9836,
786
+ "step": 1020
787
+ },
788
+ {
789
+ "epoch": 7.923076923076923,
790
+ "grad_norm": 2.548480272293091,
791
+ "learning_rate": 9.435897435897436e-06,
792
+ "loss": 0.9546,
793
+ "step": 1030
794
+ },
795
+ {
796
+ "epoch": 8.0,
797
+ "grad_norm": 3.8450028896331787,
798
+ "learning_rate": 9.333333333333334e-06,
799
+ "loss": 0.9932,
800
+ "step": 1040
801
+ },
802
+ {
803
+ "epoch": 8.0,
804
+ "eval_accuracy": 0.7669172932330827,
805
+ "eval_loss": 0.9549766778945923,
806
+ "eval_runtime": 0.7536,
807
+ "eval_samples_per_second": 176.486,
808
+ "eval_steps_per_second": 22.558,
809
+ "step": 1040
810
+ },
811
+ {
812
+ "epoch": 8.076923076923077,
813
+ "grad_norm": 1.6870847940444946,
814
+ "learning_rate": 9.230769230769232e-06,
815
+ "loss": 0.9752,
816
+ "step": 1050
817
+ },
818
+ {
819
+ "epoch": 8.153846153846153,
820
+ "grad_norm": 2.1122217178344727,
821
+ "learning_rate": 9.128205128205129e-06,
822
+ "loss": 0.9573,
823
+ "step": 1060
824
+ },
825
+ {
826
+ "epoch": 8.23076923076923,
827
+ "grad_norm": 4.1552886962890625,
828
+ "learning_rate": 9.025641025641027e-06,
829
+ "loss": 0.9764,
830
+ "step": 1070
831
+ },
832
+ {
833
+ "epoch": 8.307692307692308,
834
+ "grad_norm": 1.7864203453063965,
835
+ "learning_rate": 8.923076923076925e-06,
836
+ "loss": 0.9434,
837
+ "step": 1080
838
+ },
839
+ {
840
+ "epoch": 8.384615384615385,
841
+ "grad_norm": 2.2091946601867676,
842
+ "learning_rate": 8.820512820512821e-06,
843
+ "loss": 0.974,
844
+ "step": 1090
845
+ },
846
+ {
847
+ "epoch": 8.461538461538462,
848
+ "grad_norm": 2.4063644409179688,
849
+ "learning_rate": 8.717948717948719e-06,
850
+ "loss": 0.9576,
851
+ "step": 1100
852
+ },
853
+ {
854
+ "epoch": 8.538461538461538,
855
+ "grad_norm": 1.6061931848526,
856
+ "learning_rate": 8.615384615384617e-06,
857
+ "loss": 0.9588,
858
+ "step": 1110
859
+ },
860
+ {
861
+ "epoch": 8.615384615384615,
862
+ "grad_norm": 2.8999595642089844,
863
+ "learning_rate": 8.512820512820513e-06,
864
+ "loss": 0.9791,
865
+ "step": 1120
866
+ },
867
+ {
868
+ "epoch": 8.692307692307692,
869
+ "grad_norm": 3.6554131507873535,
870
+ "learning_rate": 8.410256410256411e-06,
871
+ "loss": 0.9629,
872
+ "step": 1130
873
+ },
874
+ {
875
+ "epoch": 8.76923076923077,
876
+ "grad_norm": 1.7246966361999512,
877
+ "learning_rate": 8.307692307692309e-06,
878
+ "loss": 0.9707,
879
+ "step": 1140
880
+ },
881
+ {
882
+ "epoch": 8.846153846153847,
883
+ "grad_norm": 2.3160033226013184,
884
+ "learning_rate": 8.205128205128205e-06,
885
+ "loss": 0.9578,
886
+ "step": 1150
887
+ },
888
+ {
889
+ "epoch": 8.923076923076923,
890
+ "grad_norm": 2.685718059539795,
891
+ "learning_rate": 8.102564102564103e-06,
892
+ "loss": 0.972,
893
+ "step": 1160
894
+ },
895
+ {
896
+ "epoch": 9.0,
897
+ "grad_norm": 3.6465442180633545,
898
+ "learning_rate": 8.000000000000001e-06,
899
+ "loss": 0.9748,
900
+ "step": 1170
901
+ },
902
+ {
903
+ "epoch": 9.0,
904
+ "eval_accuracy": 0.7669172932330827,
905
+ "eval_loss": 0.9408761858940125,
906
+ "eval_runtime": 0.7615,
907
+ "eval_samples_per_second": 174.644,
908
+ "eval_steps_per_second": 22.323,
909
+ "step": 1170
910
+ },
911
+ {
912
+ "epoch": 9.076923076923077,
913
+ "grad_norm": 1.8668149709701538,
914
+ "learning_rate": 7.897435897435898e-06,
915
+ "loss": 0.9712,
916
+ "step": 1180
917
+ },
918
+ {
919
+ "epoch": 9.153846153846153,
920
+ "grad_norm": 2.042644739151001,
921
+ "learning_rate": 7.794871794871796e-06,
922
+ "loss": 0.9407,
923
+ "step": 1190
924
+ },
925
+ {
926
+ "epoch": 9.23076923076923,
927
+ "grad_norm": 1.967020869255066,
928
+ "learning_rate": 7.692307692307694e-06,
929
+ "loss": 0.9457,
930
+ "step": 1200
931
+ },
932
+ {
933
+ "epoch": 9.307692307692308,
934
+ "grad_norm": 2.147862672805786,
935
+ "learning_rate": 7.58974358974359e-06,
936
+ "loss": 0.9442,
937
+ "step": 1210
938
+ },
939
+ {
940
+ "epoch": 9.384615384615385,
941
+ "grad_norm": 1.8528053760528564,
942
+ "learning_rate": 7.487179487179488e-06,
943
+ "loss": 0.9526,
944
+ "step": 1220
945
+ },
946
+ {
947
+ "epoch": 9.461538461538462,
948
+ "grad_norm": 3.2000551223754883,
949
+ "learning_rate": 7.384615384615386e-06,
950
+ "loss": 0.9465,
951
+ "step": 1230
952
+ },
953
+ {
954
+ "epoch": 9.538461538461538,
955
+ "grad_norm": 2.259323835372925,
956
+ "learning_rate": 7.282051282051282e-06,
957
+ "loss": 0.9503,
958
+ "step": 1240
959
+ },
960
+ {
961
+ "epoch": 9.615384615384615,
962
+ "grad_norm": 2.4054858684539795,
963
+ "learning_rate": 7.17948717948718e-06,
964
+ "loss": 0.9274,
965
+ "step": 1250
966
+ },
967
+ {
968
+ "epoch": 9.692307692307692,
969
+ "grad_norm": 3.4811408519744873,
970
+ "learning_rate": 7.076923076923078e-06,
971
+ "loss": 0.943,
972
+ "step": 1260
973
+ },
974
+ {
975
+ "epoch": 9.76923076923077,
976
+ "grad_norm": 1.7080141305923462,
977
+ "learning_rate": 6.974358974358974e-06,
978
+ "loss": 0.9247,
979
+ "step": 1270
980
+ },
981
+ {
982
+ "epoch": 9.846153846153847,
983
+ "grad_norm": 2.0476508140563965,
984
+ "learning_rate": 6.871794871794872e-06,
985
+ "loss": 0.9194,
986
+ "step": 1280
987
+ },
988
+ {
989
+ "epoch": 9.923076923076923,
990
+ "grad_norm": 2.149641990661621,
991
+ "learning_rate": 6.76923076923077e-06,
992
+ "loss": 0.9269,
993
+ "step": 1290
994
+ },
995
+ {
996
+ "epoch": 10.0,
997
+ "grad_norm": 5.121323108673096,
998
+ "learning_rate": 6.666666666666667e-06,
999
+ "loss": 0.9113,
1000
+ "step": 1300
1001
+ },
1002
+ {
1003
+ "epoch": 10.0,
1004
+ "eval_accuracy": 0.7819548872180451,
1005
+ "eval_loss": 0.9149269461631775,
1006
+ "eval_runtime": 0.7775,
1007
+ "eval_samples_per_second": 171.06,
1008
+ "eval_steps_per_second": 21.865,
1009
+ "step": 1300
1010
+ },
1011
+ {
1012
+ "epoch": 10.076923076923077,
1013
+ "grad_norm": 2.06109881401062,
1014
+ "learning_rate": 6.564102564102565e-06,
1015
+ "loss": 0.925,
1016
+ "step": 1310
1017
+ },
1018
+ {
1019
+ "epoch": 10.153846153846153,
1020
+ "grad_norm": 1.9137018918991089,
1021
+ "learning_rate": 6.461538461538463e-06,
1022
+ "loss": 0.9657,
1023
+ "step": 1320
1024
+ },
1025
+ {
1026
+ "epoch": 10.23076923076923,
1027
+ "grad_norm": 2.0686280727386475,
1028
+ "learning_rate": 6.358974358974359e-06,
1029
+ "loss": 0.9565,
1030
+ "step": 1330
1031
+ },
1032
+ {
1033
+ "epoch": 10.307692307692308,
1034
+ "grad_norm": 2.046623945236206,
1035
+ "learning_rate": 6.256410256410257e-06,
1036
+ "loss": 0.918,
1037
+ "step": 1340
1038
+ },
1039
+ {
1040
+ "epoch": 10.384615384615385,
1041
+ "grad_norm": 2.281343936920166,
1042
+ "learning_rate": 6.153846153846155e-06,
1043
+ "loss": 0.9118,
1044
+ "step": 1350
1045
+ },
1046
+ {
1047
+ "epoch": 10.461538461538462,
1048
+ "grad_norm": 2.694427728652954,
1049
+ "learning_rate": 6.051282051282051e-06,
1050
+ "loss": 0.9377,
1051
+ "step": 1360
1052
+ },
1053
+ {
1054
+ "epoch": 10.538461538461538,
1055
+ "grad_norm": 2.3148765563964844,
1056
+ "learning_rate": 5.948717948717949e-06,
1057
+ "loss": 0.911,
1058
+ "step": 1370
1059
+ },
1060
+ {
1061
+ "epoch": 10.615384615384615,
1062
+ "grad_norm": 2.595669746398926,
1063
+ "learning_rate": 5.846153846153847e-06,
1064
+ "loss": 0.9146,
1065
+ "step": 1380
1066
+ },
1067
+ {
1068
+ "epoch": 10.692307692307692,
1069
+ "grad_norm": 2.136301279067993,
1070
+ "learning_rate": 5.743589743589743e-06,
1071
+ "loss": 0.9061,
1072
+ "step": 1390
1073
+ },
1074
+ {
1075
+ "epoch": 10.76923076923077,
1076
+ "grad_norm": 3.0159363746643066,
1077
+ "learning_rate": 5.641025641025641e-06,
1078
+ "loss": 0.9365,
1079
+ "step": 1400
1080
+ },
1081
+ {
1082
+ "epoch": 10.846153846153847,
1083
+ "grad_norm": 2.1136507987976074,
1084
+ "learning_rate": 5.538461538461539e-06,
1085
+ "loss": 0.9232,
1086
+ "step": 1410
1087
+ },
1088
+ {
1089
+ "epoch": 10.923076923076923,
1090
+ "grad_norm": 1.713663101196289,
1091
+ "learning_rate": 5.435897435897436e-06,
1092
+ "loss": 0.9344,
1093
+ "step": 1420
1094
+ },
1095
+ {
1096
+ "epoch": 11.0,
1097
+ "grad_norm": 4.04538631439209,
1098
+ "learning_rate": 5.333333333333334e-06,
1099
+ "loss": 0.9255,
1100
+ "step": 1430
1101
+ },
1102
+ {
1103
+ "epoch": 11.0,
1104
+ "eval_accuracy": 0.7894736842105263,
1105
+ "eval_loss": 0.8905543088912964,
1106
+ "eval_runtime": 0.7659,
1107
+ "eval_samples_per_second": 173.659,
1108
+ "eval_steps_per_second": 22.197,
1109
+ "step": 1430
1110
+ },
1111
+ {
1112
+ "epoch": 11.076923076923077,
1113
+ "grad_norm": 2.04194974899292,
1114
+ "learning_rate": 5.230769230769232e-06,
1115
+ "loss": 0.9333,
1116
+ "step": 1440
1117
+ },
1118
+ {
1119
+ "epoch": 11.153846153846153,
1120
+ "grad_norm": 3.108344554901123,
1121
+ "learning_rate": 5.128205128205128e-06,
1122
+ "loss": 0.9174,
1123
+ "step": 1450
1124
+ },
1125
+ {
1126
+ "epoch": 11.23076923076923,
1127
+ "grad_norm": 2.406233072280884,
1128
+ "learning_rate": 5.025641025641026e-06,
1129
+ "loss": 0.8948,
1130
+ "step": 1460
1131
+ },
1132
+ {
1133
+ "epoch": 11.307692307692308,
1134
+ "grad_norm": 2.4100501537323,
1135
+ "learning_rate": 4.923076923076924e-06,
1136
+ "loss": 0.9155,
1137
+ "step": 1470
1138
+ },
1139
+ {
1140
+ "epoch": 11.384615384615385,
1141
+ "grad_norm": 2.7117860317230225,
1142
+ "learning_rate": 4.820512820512821e-06,
1143
+ "loss": 0.9075,
1144
+ "step": 1480
1145
+ },
1146
+ {
1147
+ "epoch": 11.461538461538462,
1148
+ "grad_norm": 2.0159695148468018,
1149
+ "learning_rate": 4.717948717948718e-06,
1150
+ "loss": 0.9338,
1151
+ "step": 1490
1152
+ },
1153
+ {
1154
+ "epoch": 11.538461538461538,
1155
+ "grad_norm": 3.280245304107666,
1156
+ "learning_rate": 4.615384615384616e-06,
1157
+ "loss": 0.9243,
1158
+ "step": 1500
1159
+ },
1160
+ {
1161
+ "epoch": 11.615384615384615,
1162
+ "grad_norm": 3.1355690956115723,
1163
+ "learning_rate": 4.512820512820513e-06,
1164
+ "loss": 0.9185,
1165
+ "step": 1510
1166
+ },
1167
+ {
1168
+ "epoch": 11.692307692307692,
1169
+ "grad_norm": 3.0900094509124756,
1170
+ "learning_rate": 4.4102564102564104e-06,
1171
+ "loss": 0.937,
1172
+ "step": 1520
1173
+ },
1174
+ {
1175
+ "epoch": 11.76923076923077,
1176
+ "grad_norm": 1.8758033514022827,
1177
+ "learning_rate": 4.307692307692308e-06,
1178
+ "loss": 0.9052,
1179
+ "step": 1530
1180
+ },
1181
+ {
1182
+ "epoch": 11.846153846153847,
1183
+ "grad_norm": 2.0586955547332764,
1184
+ "learning_rate": 4.2051282051282055e-06,
1185
+ "loss": 0.8874,
1186
+ "step": 1540
1187
+ },
1188
+ {
1189
+ "epoch": 11.923076923076923,
1190
+ "grad_norm": 2.0720062255859375,
1191
+ "learning_rate": 4.102564102564103e-06,
1192
+ "loss": 0.9141,
1193
+ "step": 1550
1194
+ },
1195
+ {
1196
+ "epoch": 12.0,
1197
+ "grad_norm": 3.183523416519165,
1198
+ "learning_rate": 4.000000000000001e-06,
1199
+ "loss": 0.8877,
1200
+ "step": 1560
1201
+ },
1202
+ {
1203
+ "epoch": 12.0,
1204
+ "eval_accuracy": 0.7894736842105263,
1205
+ "eval_loss": 0.8749483823776245,
1206
+ "eval_runtime": 0.7374,
1207
+ "eval_samples_per_second": 180.372,
1208
+ "eval_steps_per_second": 23.055,
1209
+ "step": 1560
1210
+ },
1211
+ {
1212
+ "epoch": 12.076923076923077,
1213
+ "grad_norm": 2.0058720111846924,
1214
+ "learning_rate": 3.897435897435898e-06,
1215
+ "loss": 0.8829,
1216
+ "step": 1570
1217
+ },
1218
+ {
1219
+ "epoch": 12.153846153846153,
1220
+ "grad_norm": 2.2991676330566406,
1221
+ "learning_rate": 3.794871794871795e-06,
1222
+ "loss": 0.9152,
1223
+ "step": 1580
1224
+ },
1225
+ {
1226
+ "epoch": 12.23076923076923,
1227
+ "grad_norm": 1.5903538465499878,
1228
+ "learning_rate": 3.692307692307693e-06,
1229
+ "loss": 0.9149,
1230
+ "step": 1590
1231
+ },
1232
+ {
1233
+ "epoch": 12.307692307692308,
1234
+ "grad_norm": 1.7883615493774414,
1235
+ "learning_rate": 3.58974358974359e-06,
1236
+ "loss": 0.9163,
1237
+ "step": 1600
1238
+ },
1239
+ {
1240
+ "epoch": 12.384615384615385,
1241
+ "grad_norm": 2.2841601371765137,
1242
+ "learning_rate": 3.487179487179487e-06,
1243
+ "loss": 0.8958,
1244
+ "step": 1610
1245
+ },
1246
+ {
1247
+ "epoch": 12.461538461538462,
1248
+ "grad_norm": 2.3814501762390137,
1249
+ "learning_rate": 3.384615384615385e-06,
1250
+ "loss": 0.8918,
1251
+ "step": 1620
1252
+ },
1253
+ {
1254
+ "epoch": 12.538461538461538,
1255
+ "grad_norm": 1.9848734140396118,
1256
+ "learning_rate": 3.2820512820512823e-06,
1257
+ "loss": 0.889,
1258
+ "step": 1630
1259
+ },
1260
+ {
1261
+ "epoch": 12.615384615384615,
1262
+ "grad_norm": 1.7236778736114502,
1263
+ "learning_rate": 3.1794871794871795e-06,
1264
+ "loss": 0.8979,
1265
+ "step": 1640
1266
+ },
1267
+ {
1268
+ "epoch": 12.692307692307692,
1269
+ "grad_norm": 3.340665102005005,
1270
+ "learning_rate": 3.0769230769230774e-06,
1271
+ "loss": 0.8695,
1272
+ "step": 1650
1273
+ },
1274
+ {
1275
+ "epoch": 12.76923076923077,
1276
+ "grad_norm": 2.127927780151367,
1277
+ "learning_rate": 2.9743589743589746e-06,
1278
+ "loss": 0.9323,
1279
+ "step": 1660
1280
+ },
1281
+ {
1282
+ "epoch": 12.846153846153847,
1283
+ "grad_norm": 1.8213707208633423,
1284
+ "learning_rate": 2.8717948717948717e-06,
1285
+ "loss": 0.9178,
1286
+ "step": 1670
1287
+ },
1288
+ {
1289
+ "epoch": 12.923076923076923,
1290
+ "grad_norm": 2.0011963844299316,
1291
+ "learning_rate": 2.7692307692307697e-06,
1292
+ "loss": 0.8872,
1293
+ "step": 1680
1294
+ },
1295
+ {
1296
+ "epoch": 13.0,
1297
+ "grad_norm": 3.812871217727661,
1298
+ "learning_rate": 2.666666666666667e-06,
1299
+ "loss": 0.9032,
1300
+ "step": 1690
1301
+ },
1302
+ {
1303
+ "epoch": 13.0,
1304
+ "eval_accuracy": 0.7969924812030075,
1305
+ "eval_loss": 0.8698711395263672,
1306
+ "eval_runtime": 0.7423,
1307
+ "eval_samples_per_second": 179.165,
1308
+ "eval_steps_per_second": 22.901,
1309
+ "step": 1690
1310
+ },
1311
+ {
1312
+ "epoch": 13.076923076923077,
1313
+ "grad_norm": 2.8741540908813477,
1314
+ "learning_rate": 2.564102564102564e-06,
1315
+ "loss": 0.8842,
1316
+ "step": 1700
1317
+ },
1318
+ {
1319
+ "epoch": 13.153846153846153,
1320
+ "grad_norm": 2.3278818130493164,
1321
+ "learning_rate": 2.461538461538462e-06,
1322
+ "loss": 0.9131,
1323
+ "step": 1710
1324
+ },
1325
+ {
1326
+ "epoch": 13.23076923076923,
1327
+ "grad_norm": 2.8419501781463623,
1328
+ "learning_rate": 2.358974358974359e-06,
1329
+ "loss": 0.8965,
1330
+ "step": 1720
1331
+ },
1332
+ {
1333
+ "epoch": 13.307692307692308,
1334
+ "grad_norm": 1.8506221771240234,
1335
+ "learning_rate": 2.2564102564102566e-06,
1336
+ "loss": 0.8967,
1337
+ "step": 1730
1338
+ },
1339
+ {
1340
+ "epoch": 13.384615384615385,
1341
+ "grad_norm": 2.6166839599609375,
1342
+ "learning_rate": 2.153846153846154e-06,
1343
+ "loss": 0.8785,
1344
+ "step": 1740
1345
+ },
1346
+ {
1347
+ "epoch": 13.461538461538462,
1348
+ "grad_norm": 4.287515640258789,
1349
+ "learning_rate": 2.0512820512820513e-06,
1350
+ "loss": 0.914,
1351
+ "step": 1750
1352
+ },
1353
+ {
1354
+ "epoch": 13.538461538461538,
1355
+ "grad_norm": 2.516889810562134,
1356
+ "learning_rate": 1.948717948717949e-06,
1357
+ "loss": 0.9286,
1358
+ "step": 1760
1359
+ },
1360
+ {
1361
+ "epoch": 13.615384615384615,
1362
+ "grad_norm": 1.8332946300506592,
1363
+ "learning_rate": 1.8461538461538465e-06,
1364
+ "loss": 0.8995,
1365
+ "step": 1770
1366
+ },
1367
+ {
1368
+ "epoch": 13.692307692307692,
1369
+ "grad_norm": 2.2418551445007324,
1370
+ "learning_rate": 1.7435897435897436e-06,
1371
+ "loss": 0.8818,
1372
+ "step": 1780
1373
+ },
1374
+ {
1375
+ "epoch": 13.76923076923077,
1376
+ "grad_norm": 1.794832706451416,
1377
+ "learning_rate": 1.6410256410256412e-06,
1378
+ "loss": 0.9044,
1379
+ "step": 1790
1380
+ },
1381
+ {
1382
+ "epoch": 13.846153846153847,
1383
+ "grad_norm": 3.0142152309417725,
1384
+ "learning_rate": 1.5384615384615387e-06,
1385
+ "loss": 0.8826,
1386
+ "step": 1800
1387
+ },
1388
+ {
1389
+ "epoch": 13.923076923076923,
1390
+ "grad_norm": 2.5891315937042236,
1391
+ "learning_rate": 1.4358974358974359e-06,
1392
+ "loss": 0.8387,
1393
+ "step": 1810
1394
+ },
1395
+ {
1396
+ "epoch": 14.0,
1397
+ "grad_norm": 5.37412691116333,
1398
+ "learning_rate": 1.3333333333333334e-06,
1399
+ "loss": 0.9001,
1400
+ "step": 1820
1401
+ },
1402
+ {
1403
+ "epoch": 14.0,
1404
+ "eval_accuracy": 0.7819548872180451,
1405
+ "eval_loss": 0.8673797845840454,
1406
+ "eval_runtime": 0.7642,
1407
+ "eval_samples_per_second": 174.027,
1408
+ "eval_steps_per_second": 22.244,
1409
+ "step": 1820
1410
+ },
1411
+ {
1412
+ "epoch": 14.076923076923077,
1413
+ "grad_norm": 1.8213236331939697,
1414
+ "learning_rate": 1.230769230769231e-06,
1415
+ "loss": 0.9047,
1416
+ "step": 1830
1417
+ },
1418
+ {
1419
+ "epoch": 14.153846153846153,
1420
+ "grad_norm": 1.8006333112716675,
1421
+ "learning_rate": 1.1282051282051283e-06,
1422
+ "loss": 0.8768,
1423
+ "step": 1840
1424
+ },
1425
+ {
1426
+ "epoch": 14.23076923076923,
1427
+ "grad_norm": 2.691574811935425,
1428
+ "learning_rate": 1.0256410256410257e-06,
1429
+ "loss": 0.8757,
1430
+ "step": 1850
1431
+ },
1432
+ {
1433
+ "epoch": 14.307692307692308,
1434
+ "grad_norm": 5.015848636627197,
1435
+ "learning_rate": 9.230769230769232e-07,
1436
+ "loss": 0.8734,
1437
+ "step": 1860
1438
+ },
1439
+ {
1440
+ "epoch": 14.384615384615385,
1441
+ "grad_norm": 2.5233821868896484,
1442
+ "learning_rate": 8.205128205128206e-07,
1443
+ "loss": 0.8787,
1444
+ "step": 1870
1445
+ },
1446
+ {
1447
+ "epoch": 14.461538461538462,
1448
+ "grad_norm": 2.1718924045562744,
1449
+ "learning_rate": 7.179487179487179e-07,
1450
+ "loss": 0.8767,
1451
+ "step": 1880
1452
+ },
1453
+ {
1454
+ "epoch": 14.538461538461538,
1455
+ "grad_norm": 3.0364015102386475,
1456
+ "learning_rate": 6.153846153846155e-07,
1457
+ "loss": 0.873,
1458
+ "step": 1890
1459
+ },
1460
+ {
1461
+ "epoch": 14.615384615384615,
1462
+ "grad_norm": 2.5152034759521484,
1463
+ "learning_rate": 5.128205128205128e-07,
1464
+ "loss": 0.9096,
1465
+ "step": 1900
1466
+ },
1467
+ {
1468
+ "epoch": 14.692307692307692,
1469
+ "grad_norm": 1.819096565246582,
1470
+ "learning_rate": 4.102564102564103e-07,
1471
+ "loss": 0.892,
1472
+ "step": 1910
1473
+ },
1474
+ {
1475
+ "epoch": 14.76923076923077,
1476
+ "grad_norm": 3.512732982635498,
1477
+ "learning_rate": 3.0769230769230774e-07,
1478
+ "loss": 0.8937,
1479
+ "step": 1920
1480
+ },
1481
+ {
1482
+ "epoch": 14.846153846153847,
1483
+ "grad_norm": 2.917677879333496,
1484
+ "learning_rate": 2.0512820512820514e-07,
1485
+ "loss": 0.9231,
1486
+ "step": 1930
1487
+ },
1488
+ {
1489
+ "epoch": 14.923076923076923,
1490
+ "grad_norm": 2.0683395862579346,
1491
+ "learning_rate": 1.0256410256410257e-07,
1492
+ "loss": 0.8613,
1493
+ "step": 1940
1494
+ },
1495
+ {
1496
+ "epoch": 15.0,
1497
+ "grad_norm": 4.704519271850586,
1498
+ "learning_rate": 0.0,
1499
+ "loss": 0.8842,
1500
+ "step": 1950
1501
+ },
1502
+ {
1503
+ "epoch": 15.0,
1504
+ "eval_accuracy": 0.7894736842105263,
1505
+ "eval_loss": 0.8550169467926025,
1506
+ "eval_runtime": 0.7718,
1507
+ "eval_samples_per_second": 172.329,
1508
+ "eval_steps_per_second": 22.027,
1509
+ "step": 1950
1510
+ },
1511
+ {
1512
+ "epoch": 15.0,
1513
+ "step": 1950,
1514
+ "total_flos": 1.5658365504595968e+17,
1515
+ "train_loss": 0.6299933981284117,
1516
+ "train_runtime": 92.772,
1517
+ "train_samples_per_second": 167.184,
1518
+ "train_steps_per_second": 21.019
1519
  }
1520
  ],
1521
  "logging_steps": 10,
1522
+ "max_steps": 1950,
1523
  "num_input_tokens_seen": 0,
1524
+ "num_train_epochs": 15,
1525
  "save_steps": 500,
1526
  "stateful_callbacks": {
1527
  "TrainerControl": {
 
1535
  "attributes": {}
1536
  }
1537
  },
1538
+ "total_flos": 1.5658365504595968e+17,
1539
  "train_batch_size": 8,
1540
  "trial_name": null,
1541
  "trial_params": null