Training in progress, epoch 1, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 203456160
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4054fc3341df3ec62ad23eb40df504bf559ae3fe109c9358a6fe2fbe13c2b5db
|
| 3 |
size 203456160
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 407127126
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7de5b56bb44d80a59d57ae0d74443cf71e7c3a04fcea2547755d8ae3f6ab98b
|
| 3 |
size 407127126
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d4d2e3c9c2c299605f839dd9681034a9a825180b34e15be15b76272f4fb00418
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6dfefdd350a9c50a5ce8f17f77222cc43ade97bd1bc4dfe20825b1649c6776a6
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -637,6 +637,643 @@
|
|
| 637 |
"learning_rate": 6.861468292009727e-05,
|
| 638 |
"loss": 0.7384,
|
| 639 |
"step": 630
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 640 |
}
|
| 641 |
],
|
| 642 |
"logging_steps": 7,
|
|
@@ -656,7 +1293,7 @@
|
|
| 656 |
"attributes": {}
|
| 657 |
}
|
| 658 |
},
|
| 659 |
-
"total_flos":
|
| 660 |
"train_batch_size": 2,
|
| 661 |
"trial_name": null,
|
| 662 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9988177339901478,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1268,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 637 |
"learning_rate": 6.861468292009727e-05,
|
| 638 |
"loss": 0.7384,
|
| 639 |
"step": 630
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.0041379310344827,
|
| 643 |
+
"grad_norm": 0.28151699900627136,
|
| 644 |
+
"learning_rate": 6.788347664153447e-05,
|
| 645 |
+
"loss": 0.7612,
|
| 646 |
+
"step": 637
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.0151724137931035,
|
| 650 |
+
"grad_norm": 0.2590833604335785,
|
| 651 |
+
"learning_rate": 6.714785788270658e-05,
|
| 652 |
+
"loss": 0.6778,
|
| 653 |
+
"step": 644
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.0262068965517241,
|
| 657 |
+
"grad_norm": 0.268964558839798,
|
| 658 |
+
"learning_rate": 6.640800814653503e-05,
|
| 659 |
+
"loss": 0.6337,
|
| 660 |
+
"step": 651
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.0372413793103448,
|
| 664 |
+
"grad_norm": 0.26795223355293274,
|
| 665 |
+
"learning_rate": 6.566410997987163e-05,
|
| 666 |
+
"loss": 0.6548,
|
| 667 |
+
"step": 658
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.0482758620689656,
|
| 671 |
+
"grad_norm": 0.25789910554885864,
|
| 672 |
+
"learning_rate": 6.49163469284578e-05,
|
| 673 |
+
"loss": 0.6494,
|
| 674 |
+
"step": 665
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.0593103448275862,
|
| 678 |
+
"grad_norm": 0.27773144841194153,
|
| 679 |
+
"learning_rate": 6.416490349163748e-05,
|
| 680 |
+
"loss": 0.6541,
|
| 681 |
+
"step": 672
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.0703448275862069,
|
| 685 |
+
"grad_norm": 0.28650370240211487,
|
| 686 |
+
"learning_rate": 6.340996507683458e-05,
|
| 687 |
+
"loss": 0.6677,
|
| 688 |
+
"step": 679
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.0813793103448275,
|
| 692 |
+
"grad_norm": 0.2758565843105316,
|
| 693 |
+
"learning_rate": 6.265171795380659e-05,
|
| 694 |
+
"loss": 0.6299,
|
| 695 |
+
"step": 686
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.0924137931034483,
|
| 699 |
+
"grad_norm": 0.2588457465171814,
|
| 700 |
+
"learning_rate": 6.189034920868522e-05,
|
| 701 |
+
"loss": 0.6391,
|
| 702 |
+
"step": 693
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.103448275862069,
|
| 706 |
+
"grad_norm": 0.2821789085865021,
|
| 707 |
+
"learning_rate": 6.112604669781572e-05,
|
| 708 |
+
"loss": 0.6239,
|
| 709 |
+
"step": 700
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.1144827586206896,
|
| 713 |
+
"grad_norm": 0.2735387086868286,
|
| 714 |
+
"learning_rate": 6.0358999001406156e-05,
|
| 715 |
+
"loss": 0.6581,
|
| 716 |
+
"step": 707
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.1255172413793104,
|
| 720 |
+
"grad_norm": 0.2622285485267639,
|
| 721 |
+
"learning_rate": 5.9589395376998e-05,
|
| 722 |
+
"loss": 0.6396,
|
| 723 |
+
"step": 714
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.136551724137931,
|
| 727 |
+
"grad_norm": 0.27206316590309143,
|
| 728 |
+
"learning_rate": 5.8817425712769794e-05,
|
| 729 |
+
"loss": 0.6305,
|
| 730 |
+
"step": 721
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.1475862068965517,
|
| 734 |
+
"grad_norm": 0.2703116536140442,
|
| 735 |
+
"learning_rate": 5.804328048068492e-05,
|
| 736 |
+
"loss": 0.6498,
|
| 737 |
+
"step": 728
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.1586206896551725,
|
| 741 |
+
"grad_norm": 0.2688596844673157,
|
| 742 |
+
"learning_rate": 5.7267150689495644e-05,
|
| 743 |
+
"loss": 0.6453,
|
| 744 |
+
"step": 735
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.1696551724137931,
|
| 748 |
+
"grad_norm": 0.27926501631736755,
|
| 749 |
+
"learning_rate": 5.648922783761443e-05,
|
| 750 |
+
"loss": 0.6715,
|
| 751 |
+
"step": 742
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.1806896551724138,
|
| 755 |
+
"grad_norm": 0.2714243531227112,
|
| 756 |
+
"learning_rate": 5.570970386586469e-05,
|
| 757 |
+
"loss": 0.6502,
|
| 758 |
+
"step": 749
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.1917241379310344,
|
| 762 |
+
"grad_norm": 0.27098771929740906,
|
| 763 |
+
"learning_rate": 5.492877111012218e-05,
|
| 764 |
+
"loss": 0.6382,
|
| 765 |
+
"step": 756
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.2027586206896552,
|
| 769 |
+
"grad_norm": 0.28337958455085754,
|
| 770 |
+
"learning_rate": 5.414662225385903e-05,
|
| 771 |
+
"loss": 0.6383,
|
| 772 |
+
"step": 763
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.2137931034482758,
|
| 776 |
+
"grad_norm": 0.28135356307029724,
|
| 777 |
+
"learning_rate": 5.336345028060199e-05,
|
| 778 |
+
"loss": 0.6433,
|
| 779 |
+
"step": 770
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.2248275862068965,
|
| 783 |
+
"grad_norm": 0.268600195646286,
|
| 784 |
+
"learning_rate": 5.257944842631658e-05,
|
| 785 |
+
"loss": 0.6503,
|
| 786 |
+
"step": 777
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.2358620689655173,
|
| 790 |
+
"grad_norm": 0.268028199672699,
|
| 791 |
+
"learning_rate": 5.179481013172912e-05,
|
| 792 |
+
"loss": 0.64,
|
| 793 |
+
"step": 784
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.246896551724138,
|
| 797 |
+
"grad_norm": 0.26292359828948975,
|
| 798 |
+
"learning_rate": 5.100972899459796e-05,
|
| 799 |
+
"loss": 0.6211,
|
| 800 |
+
"step": 791
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.2579310344827586,
|
| 804 |
+
"grad_norm": 0.2784179151058197,
|
| 805 |
+
"learning_rate": 5.022439872194629e-05,
|
| 806 |
+
"loss": 0.6466,
|
| 807 |
+
"step": 798
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.2689655172413792,
|
| 811 |
+
"grad_norm": 0.28252890706062317,
|
| 812 |
+
"learning_rate": 4.943901308226771e-05,
|
| 813 |
+
"loss": 0.6368,
|
| 814 |
+
"step": 805
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.28,
|
| 818 |
+
"grad_norm": 0.2834304869174957,
|
| 819 |
+
"learning_rate": 4.865376585771687e-05,
|
| 820 |
+
"loss": 0.6336,
|
| 821 |
+
"step": 812
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.2910344827586206,
|
| 825 |
+
"grad_norm": 0.2688741981983185,
|
| 826 |
+
"learning_rate": 4.7868850796296495e-05,
|
| 827 |
+
"loss": 0.653,
|
| 828 |
+
"step": 819
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.3020689655172415,
|
| 832 |
+
"grad_norm": 0.2756083905696869,
|
| 833 |
+
"learning_rate": 4.708446156405307e-05,
|
| 834 |
+
"loss": 0.6581,
|
| 835 |
+
"step": 826
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.3131034482758621,
|
| 839 |
+
"grad_norm": 0.2754676640033722,
|
| 840 |
+
"learning_rate": 4.630079169729257e-05,
|
| 841 |
+
"loss": 0.6393,
|
| 842 |
+
"step": 833
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.3241379310344827,
|
| 846 |
+
"grad_norm": 0.26543307304382324,
|
| 847 |
+
"learning_rate": 4.551803455482833e-05,
|
| 848 |
+
"loss": 0.6264,
|
| 849 |
+
"step": 840
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.3351724137931034,
|
| 853 |
+
"grad_norm": 0.2671962380409241,
|
| 854 |
+
"learning_rate": 4.473638327027259e-05,
|
| 855 |
+
"loss": 0.6266,
|
| 856 |
+
"step": 847
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.3462068965517242,
|
| 860 |
+
"grad_norm": 0.2819383442401886,
|
| 861 |
+
"learning_rate": 4.395603070438373e-05,
|
| 862 |
+
"loss": 0.6426,
|
| 863 |
+
"step": 854
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.3572413793103448,
|
| 867 |
+
"grad_norm": 0.2790848910808563,
|
| 868 |
+
"learning_rate": 4.31771693974807e-05,
|
| 869 |
+
"loss": 0.633,
|
| 870 |
+
"step": 861
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.3682758620689655,
|
| 874 |
+
"grad_norm": 0.2653134763240814,
|
| 875 |
+
"learning_rate": 4.239999152193664e-05,
|
| 876 |
+
"loss": 0.637,
|
| 877 |
+
"step": 868
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.3793103448275863,
|
| 881 |
+
"grad_norm": 0.27828529477119446,
|
| 882 |
+
"learning_rate": 4.162468883476319e-05,
|
| 883 |
+
"loss": 0.6273,
|
| 884 |
+
"step": 875
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.390344827586207,
|
| 888 |
+
"grad_norm": 0.27933961153030396,
|
| 889 |
+
"learning_rate": 4.085145263029726e-05,
|
| 890 |
+
"loss": 0.6477,
|
| 891 |
+
"step": 882
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.4013793103448275,
|
| 895 |
+
"grad_norm": 0.2893773317337036,
|
| 896 |
+
"learning_rate": 4.008047369300218e-05,
|
| 897 |
+
"loss": 0.6374,
|
| 898 |
+
"step": 889
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.4124137931034482,
|
| 902 |
+
"grad_norm": 0.29266366362571716,
|
| 903 |
+
"learning_rate": 3.9311942250394276e-05,
|
| 904 |
+
"loss": 0.6051,
|
| 905 |
+
"step": 896
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.423448275862069,
|
| 909 |
+
"grad_norm": 0.28573766350746155,
|
| 910 |
+
"learning_rate": 3.8546047926107256e-05,
|
| 911 |
+
"loss": 0.6371,
|
| 912 |
+
"step": 903
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.4344827586206896,
|
| 916 |
+
"grad_norm": 0.2910100817680359,
|
| 917 |
+
"learning_rate": 3.778297969310529e-05,
|
| 918 |
+
"loss": 0.6491,
|
| 919 |
+
"step": 910
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.4455172413793105,
|
| 923 |
+
"grad_norm": 0.2896603047847748,
|
| 924 |
+
"learning_rate": 3.7022925827056884e-05,
|
| 925 |
+
"loss": 0.6332,
|
| 926 |
+
"step": 917
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.456551724137931,
|
| 930 |
+
"grad_norm": 0.26403993368148804,
|
| 931 |
+
"learning_rate": 3.62660738598805e-05,
|
| 932 |
+
"loss": 0.6387,
|
| 933 |
+
"step": 924
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.4675862068965517,
|
| 937 |
+
"grad_norm": 0.27510949969291687,
|
| 938 |
+
"learning_rate": 3.551261053347404e-05,
|
| 939 |
+
"loss": 0.6217,
|
| 940 |
+
"step": 931
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.4786206896551723,
|
| 944 |
+
"grad_norm": 0.28636762499809265,
|
| 945 |
+
"learning_rate": 3.4762721753638995e-05,
|
| 946 |
+
"loss": 0.6186,
|
| 947 |
+
"step": 938
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.489655172413793,
|
| 951 |
+
"grad_norm": 0.2784745991230011,
|
| 952 |
+
"learning_rate": 3.401659254421094e-05,
|
| 953 |
+
"loss": 0.6392,
|
| 954 |
+
"step": 945
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.5006896551724138,
|
| 958 |
+
"grad_norm": 0.2818538546562195,
|
| 959 |
+
"learning_rate": 3.3274407001407735e-05,
|
| 960 |
+
"loss": 0.6311,
|
| 961 |
+
"step": 952
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.5117241379310344,
|
| 965 |
+
"grad_norm": 0.2834908068180084,
|
| 966 |
+
"learning_rate": 3.2536348248406534e-05,
|
| 967 |
+
"loss": 0.6205,
|
| 968 |
+
"step": 959
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.5227586206896553,
|
| 972 |
+
"grad_norm": 0.27675577998161316,
|
| 973 |
+
"learning_rate": 3.1802598390160784e-05,
|
| 974 |
+
"loss": 0.6191,
|
| 975 |
+
"step": 966
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.533793103448276,
|
| 979 |
+
"grad_norm": 0.27893996238708496,
|
| 980 |
+
"learning_rate": 3.107333846846872e-05,
|
| 981 |
+
"loss": 0.6263,
|
| 982 |
+
"step": 973
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.5448275862068965,
|
| 986 |
+
"grad_norm": 0.2641468048095703,
|
| 987 |
+
"learning_rate": 3.0348748417303823e-05,
|
| 988 |
+
"loss": 0.6163,
|
| 989 |
+
"step": 980
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.5558620689655172,
|
| 993 |
+
"grad_norm": 0.2818867862224579,
|
| 994 |
+
"learning_rate": 2.9629007018418985e-05,
|
| 995 |
+
"loss": 0.6201,
|
| 996 |
+
"step": 987
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.5668965517241378,
|
| 1000 |
+
"grad_norm": 0.28501781821250916,
|
| 1001 |
+
"learning_rate": 2.8914291857234636e-05,
|
| 1002 |
+
"loss": 0.6308,
|
| 1003 |
+
"step": 994
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.5779310344827586,
|
| 1007 |
+
"grad_norm": 0.2769615352153778,
|
| 1008 |
+
"learning_rate": 2.8204779279022276e-05,
|
| 1009 |
+
"loss": 0.6063,
|
| 1010 |
+
"step": 1001
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.5889655172413795,
|
| 1014 |
+
"grad_norm": 0.2790084481239319,
|
| 1015 |
+
"learning_rate": 2.7500644345393943e-05,
|
| 1016 |
+
"loss": 0.6325,
|
| 1017 |
+
"step": 1008
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.6,
|
| 1021 |
+
"grad_norm": 0.2709619402885437,
|
| 1022 |
+
"learning_rate": 2.68020607911083e-05,
|
| 1023 |
+
"loss": 0.5982,
|
| 1024 |
+
"step": 1015
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.6110344827586207,
|
| 1028 |
+
"grad_norm": 0.2806256115436554,
|
| 1029 |
+
"learning_rate": 2.610920098120424e-05,
|
| 1030 |
+
"loss": 0.6453,
|
| 1031 |
+
"step": 1022
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.6220689655172413,
|
| 1035 |
+
"grad_norm": 0.2840186059474945,
|
| 1036 |
+
"learning_rate": 2.5422235868472345e-05,
|
| 1037 |
+
"loss": 0.6243,
|
| 1038 |
+
"step": 1029
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.633103448275862,
|
| 1042 |
+
"grad_norm": 0.2787083089351654,
|
| 1043 |
+
"learning_rate": 2.4741334951274947e-05,
|
| 1044 |
+
"loss": 0.6097,
|
| 1045 |
+
"step": 1036
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.6441379310344828,
|
| 1049 |
+
"grad_norm": 0.27798426151275635,
|
| 1050 |
+
"learning_rate": 2.40666662317248e-05,
|
| 1051 |
+
"loss": 0.616,
|
| 1052 |
+
"step": 1043
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.6551724137931034,
|
| 1056 |
+
"grad_norm": 0.27529022097587585,
|
| 1057 |
+
"learning_rate": 2.3398396174233178e-05,
|
| 1058 |
+
"loss": 0.6207,
|
| 1059 |
+
"step": 1050
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.6662068965517243,
|
| 1063 |
+
"grad_norm": 0.2800416946411133,
|
| 1064 |
+
"learning_rate": 2.2736689664437217e-05,
|
| 1065 |
+
"loss": 0.6273,
|
| 1066 |
+
"step": 1057
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.677241379310345,
|
| 1070 |
+
"grad_norm": 0.2721683979034424,
|
| 1071 |
+
"learning_rate": 2.2081709968516866e-05,
|
| 1072 |
+
"loss": 0.6,
|
| 1073 |
+
"step": 1064
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.6882758620689655,
|
| 1077 |
+
"grad_norm": 0.29215288162231445,
|
| 1078 |
+
"learning_rate": 2.1433618692911467e-05,
|
| 1079 |
+
"loss": 0.6105,
|
| 1080 |
+
"step": 1071
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.6993103448275861,
|
| 1084 |
+
"grad_norm": 0.2790670692920685,
|
| 1085 |
+
"learning_rate": 2.0792575744445653e-05,
|
| 1086 |
+
"loss": 0.6028,
|
| 1087 |
+
"step": 1078
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.7103448275862068,
|
| 1091 |
+
"grad_norm": 0.28281012177467346,
|
| 1092 |
+
"learning_rate": 2.015873929087482e-05,
|
| 1093 |
+
"loss": 0.6168,
|
| 1094 |
+
"step": 1085
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.7213793103448276,
|
| 1098 |
+
"grad_norm": 0.2883938252925873,
|
| 1099 |
+
"learning_rate": 1.95322657218596e-05,
|
| 1100 |
+
"loss": 0.6045,
|
| 1101 |
+
"step": 1092
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.7324137931034482,
|
| 1105 |
+
"grad_norm": 0.27805426716804504,
|
| 1106 |
+
"learning_rate": 1.8913309610379015e-05,
|
| 1107 |
+
"loss": 0.6154,
|
| 1108 |
+
"step": 1099
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.743448275862069,
|
| 1112 |
+
"grad_norm": 0.27731558680534363,
|
| 1113 |
+
"learning_rate": 1.8302023674591935e-05,
|
| 1114 |
+
"loss": 0.6098,
|
| 1115 |
+
"step": 1106
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.7544827586206897,
|
| 1119 |
+
"grad_norm": 0.2910196781158447,
|
| 1120 |
+
"learning_rate": 1.7698558740156135e-05,
|
| 1121 |
+
"loss": 0.6106,
|
| 1122 |
+
"step": 1113
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.7655172413793103,
|
| 1126 |
+
"grad_norm": 0.27517473697662354,
|
| 1127 |
+
"learning_rate": 1.7103063703014372e-05,
|
| 1128 |
+
"loss": 0.6239,
|
| 1129 |
+
"step": 1120
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.776551724137931,
|
| 1133 |
+
"grad_norm": 0.26991525292396545,
|
| 1134 |
+
"learning_rate": 1.6515685492656467e-05,
|
| 1135 |
+
"loss": 0.602,
|
| 1136 |
+
"step": 1127
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.7875862068965516,
|
| 1140 |
+
"grad_norm": 0.28398916125297546,
|
| 1141 |
+
"learning_rate": 1.59365690358667e-05,
|
| 1142 |
+
"loss": 0.6138,
|
| 1143 |
+
"step": 1134
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.7986206896551724,
|
| 1147 |
+
"grad_norm": 0.2737048268318176,
|
| 1148 |
+
"learning_rate": 1.5365857220965275e-05,
|
| 1149 |
+
"loss": 0.6034,
|
| 1150 |
+
"step": 1141
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.8096551724137933,
|
| 1154 |
+
"grad_norm": 0.27919578552246094,
|
| 1155 |
+
"learning_rate": 1.4803690862552755e-05,
|
| 1156 |
+
"loss": 0.6126,
|
| 1157 |
+
"step": 1148
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.8206896551724139,
|
| 1161 |
+
"grad_norm": 0.2809952199459076,
|
| 1162 |
+
"learning_rate": 1.4250208666766235e-05,
|
| 1163 |
+
"loss": 0.607,
|
| 1164 |
+
"step": 1155
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.8317241379310345,
|
| 1168 |
+
"grad_norm": 0.28192582726478577,
|
| 1169 |
+
"learning_rate": 1.3705547197055584e-05,
|
| 1170 |
+
"loss": 0.6064,
|
| 1171 |
+
"step": 1162
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.8427586206896551,
|
| 1175 |
+
"grad_norm": 0.2873040735721588,
|
| 1176 |
+
"learning_rate": 1.3169840840488501e-05,
|
| 1177 |
+
"loss": 0.6173,
|
| 1178 |
+
"step": 1169
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.8537931034482757,
|
| 1182 |
+
"grad_norm": 0.2937031388282776,
|
| 1183 |
+
"learning_rate": 1.2643221774592518e-05,
|
| 1184 |
+
"loss": 0.6148,
|
| 1185 |
+
"step": 1176
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.8648275862068966,
|
| 1189 |
+
"grad_norm": 0.28401997685432434,
|
| 1190 |
+
"learning_rate": 1.2125819934742188e-05,
|
| 1191 |
+
"loss": 0.6027,
|
| 1192 |
+
"step": 1183
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.8758620689655172,
|
| 1196 |
+
"grad_norm": 0.2690780758857727,
|
| 1197 |
+
"learning_rate": 1.1617762982099446e-05,
|
| 1198 |
+
"loss": 0.6089,
|
| 1199 |
+
"step": 1190
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.886896551724138,
|
| 1203 |
+
"grad_norm": 0.28273099660873413,
|
| 1204 |
+
"learning_rate": 1.1119176272115128e-05,
|
| 1205 |
+
"loss": 0.6017,
|
| 1206 |
+
"step": 1197
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.8979310344827587,
|
| 1210 |
+
"grad_norm": 0.2761295437812805,
|
| 1211 |
+
"learning_rate": 1.0630182823599399e-05,
|
| 1212 |
+
"loss": 0.6156,
|
| 1213 |
+
"step": 1204
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.9089655172413793,
|
| 1217 |
+
"grad_norm": 0.2870751917362213,
|
| 1218 |
+
"learning_rate": 1.0150903288368741e-05,
|
| 1219 |
+
"loss": 0.6038,
|
| 1220 |
+
"step": 1211
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.92,
|
| 1224 |
+
"grad_norm": 0.2764483690261841,
|
| 1225 |
+
"learning_rate": 9.681455921476839e-06,
|
| 1226 |
+
"loss": 0.603,
|
| 1227 |
+
"step": 1218
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.9310344827586206,
|
| 1231 |
+
"grad_norm": 0.2793201208114624,
|
| 1232 |
+
"learning_rate": 9.221956552036992e-06,
|
| 1233 |
+
"loss": 0.6194,
|
| 1234 |
+
"step": 1225
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.9420689655172414,
|
| 1238 |
+
"grad_norm": 0.27977249026298523,
|
| 1239 |
+
"learning_rate": 8.772518554642973e-06,
|
| 1240 |
+
"loss": 0.622,
|
| 1241 |
+
"step": 1232
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.953103448275862,
|
| 1245 |
+
"grad_norm": 0.27671122550964355,
|
| 1246 |
+
"learning_rate": 8.333252821395526e-06,
|
| 1247 |
+
"loss": 0.6083,
|
| 1248 |
+
"step": 1239
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.9641379310344829,
|
| 1252 |
+
"grad_norm": 0.2808462679386139,
|
| 1253 |
+
"learning_rate": 7.904267734541498e-06,
|
| 1254 |
+
"loss": 0.6155,
|
| 1255 |
+
"step": 1246
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.9751724137931035,
|
| 1259 |
+
"grad_norm": 0.2745875418186188,
|
| 1260 |
+
"learning_rate": 7.485669139732004e-06,
|
| 1261 |
+
"loss": 0.5934,
|
| 1262 |
+
"step": 1253
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.986206896551724,
|
| 1266 |
+
"grad_norm": 0.2853251099586487,
|
| 1267 |
+
"learning_rate": 7.077560319906695e-06,
|
| 1268 |
+
"loss": 0.5992,
|
| 1269 |
+
"step": 1260
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.9972413793103447,
|
| 1273 |
+
"grad_norm": 0.2884569764137268,
|
| 1274 |
+
"learning_rate": 6.680041969810203e-06,
|
| 1275 |
+
"loss": 0.5927,
|
| 1276 |
+
"step": 1267
|
| 1277 |
}
|
| 1278 |
],
|
| 1279 |
"logging_steps": 7,
|
|
|
|
| 1293 |
"attributes": {}
|
| 1294 |
}
|
| 1295 |
},
|
| 1296 |
+
"total_flos": 8.41863001276416e+17,
|
| 1297 |
"train_batch_size": 2,
|
| 1298 |
"trial_name": null,
|
| 1299 |
"trial_params": null
|