Upload folder using huggingface_hub
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +0 -0
- scheduler.pt +0 -0
- trainer_state.json +633 -3
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1426462208
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3e16bc3c19db986ac4ca912dd200a709fa26a7b0cce2c7ad804756f07f9764ae
|
| 3 |
size 1426462208
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2853107898
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:39de36962e52b3f73461bf26e39adafd9fc4118dc96dfa826b517795df14d8da
|
| 3 |
size 2853107898
|
rng_state.pth
CHANGED
|
Binary files a/rng_state.pth and b/rng_state.pth differ
|
|
|
scheduler.pt
CHANGED
|
Binary files a/scheduler.pt and b/scheduler.pt differ
|
|
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 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,
|
|
@@ -707,6 +707,636 @@
|
|
| 707 |
"learning_rate": 2.6694910512455145e-05,
|
| 708 |
"loss": 2.4431,
|
| 709 |
"step": 1000000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 710 |
}
|
| 711 |
],
|
| 712 |
"logging_steps": 10000,
|
|
@@ -726,7 +1356,7 @@
|
|
| 726 |
"attributes": {}
|
| 727 |
}
|
| 728 |
},
|
| 729 |
-
"total_flos":
|
| 730 |
"train_batch_size": 4,
|
| 731 |
"trial_name": null,
|
| 732 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.8859229638692638,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1900000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 707 |
"learning_rate": 2.6694910512455145e-05,
|
| 708 |
"loss": 2.4431,
|
| 709 |
"step": 1000000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.470937996583135,
|
| 713 |
+
"grad_norm": 0.8998690843582153,
|
| 714 |
+
"learning_rate": 2.6461866145433116e-05,
|
| 715 |
+
"loss": 2.4419,
|
| 716 |
+
"step": 1010000
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.4756007490245522,
|
| 720 |
+
"grad_norm": 0.9611771702766418,
|
| 721 |
+
"learning_rate": 2.6228821778411084e-05,
|
| 722 |
+
"loss": 2.4425,
|
| 723 |
+
"step": 1020000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.48026350146596936,
|
| 727 |
+
"grad_norm": 1.0314167737960815,
|
| 728 |
+
"learning_rate": 2.5995754097626846e-05,
|
| 729 |
+
"loss": 2.4393,
|
| 730 |
+
"step": 1030000
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.48492625390738653,
|
| 734 |
+
"grad_norm": 1.0637314319610596,
|
| 735 |
+
"learning_rate": 2.5762709730604817e-05,
|
| 736 |
+
"loss": 2.4373,
|
| 737 |
+
"step": 1040000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.4895890063488037,
|
| 741 |
+
"grad_norm": 1.1488375663757324,
|
| 742 |
+
"learning_rate": 2.552964204982058e-05,
|
| 743 |
+
"loss": 2.4348,
|
| 744 |
+
"step": 1050000
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.49425175879022093,
|
| 748 |
+
"grad_norm": 1.1408883333206177,
|
| 749 |
+
"learning_rate": 2.529657436903634e-05,
|
| 750 |
+
"loss": 2.4352,
|
| 751 |
+
"step": 1060000
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.4989145112316381,
|
| 755 |
+
"grad_norm": 0.9368888735771179,
|
| 756 |
+
"learning_rate": 2.5063530002014312e-05,
|
| 757 |
+
"loss": 2.4321,
|
| 758 |
+
"step": 1070000
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.5035772636730552,
|
| 762 |
+
"grad_norm": 0.9766519069671631,
|
| 763 |
+
"learning_rate": 2.483046232123007e-05,
|
| 764 |
+
"loss": 2.4319,
|
| 765 |
+
"step": 1080000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.5082400161144724,
|
| 769 |
+
"grad_norm": 0.9681417942047119,
|
| 770 |
+
"learning_rate": 2.4597417954208042e-05,
|
| 771 |
+
"loss": 2.4329,
|
| 772 |
+
"step": 1090000
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.5129027685558896,
|
| 776 |
+
"grad_norm": 1.0913608074188232,
|
| 777 |
+
"learning_rate": 2.4364373587186014e-05,
|
| 778 |
+
"loss": 2.428,
|
| 779 |
+
"step": 1100000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.5175655209973068,
|
| 783 |
+
"grad_norm": 0.9350466132164001,
|
| 784 |
+
"learning_rate": 2.413132922016398e-05,
|
| 785 |
+
"loss": 2.4241,
|
| 786 |
+
"step": 1110000
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.522228273438724,
|
| 790 |
+
"grad_norm": 0.8676067590713501,
|
| 791 |
+
"learning_rate": 2.3898261539379743e-05,
|
| 792 |
+
"loss": 2.4227,
|
| 793 |
+
"step": 1120000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.5268910258801411,
|
| 797 |
+
"grad_norm": 0.8786779046058655,
|
| 798 |
+
"learning_rate": 2.366521717235771e-05,
|
| 799 |
+
"loss": 2.4208,
|
| 800 |
+
"step": 1130000
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.5315537783215584,
|
| 804 |
+
"grad_norm": 0.9234575629234314,
|
| 805 |
+
"learning_rate": 2.3432126177811267e-05,
|
| 806 |
+
"loss": 2.4249,
|
| 807 |
+
"step": 1140000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.5362165307629755,
|
| 811 |
+
"grad_norm": 1.3854731321334839,
|
| 812 |
+
"learning_rate": 2.3199105124551445e-05,
|
| 813 |
+
"loss": 2.4245,
|
| 814 |
+
"step": 1150000
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.5408792832043927,
|
| 818 |
+
"grad_norm": 1.0635942220687866,
|
| 819 |
+
"learning_rate": 2.2966060757529416e-05,
|
| 820 |
+
"loss": 2.4202,
|
| 821 |
+
"step": 1160000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.5455420356458098,
|
| 825 |
+
"grad_norm": 0.8858787417411804,
|
| 826 |
+
"learning_rate": 2.2732993076745178e-05,
|
| 827 |
+
"loss": 2.418,
|
| 828 |
+
"step": 1170000
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.5502047880872271,
|
| 832 |
+
"grad_norm": 1.309348702430725,
|
| 833 |
+
"learning_rate": 2.2499948709723146e-05,
|
| 834 |
+
"loss": 2.4148,
|
| 835 |
+
"step": 1180000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.5548675405286442,
|
| 839 |
+
"grad_norm": 0.9109322428703308,
|
| 840 |
+
"learning_rate": 2.2266904342701114e-05,
|
| 841 |
+
"loss": 2.4151,
|
| 842 |
+
"step": 1190000
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.5595302929700614,
|
| 846 |
+
"grad_norm": 1.0102558135986328,
|
| 847 |
+
"learning_rate": 2.2033836661916876e-05,
|
| 848 |
+
"loss": 2.4134,
|
| 849 |
+
"step": 1200000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.5641930454114785,
|
| 853 |
+
"grad_norm": 1.0807286500930786,
|
| 854 |
+
"learning_rate": 2.1800815608657053e-05,
|
| 855 |
+
"loss": 2.4135,
|
| 856 |
+
"step": 1210000
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.5688557978528958,
|
| 860 |
+
"grad_norm": 0.9256259799003601,
|
| 861 |
+
"learning_rate": 2.156774792787282e-05,
|
| 862 |
+
"loss": 2.4106,
|
| 863 |
+
"step": 1220000
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.573518550294313,
|
| 867 |
+
"grad_norm": 0.9988642930984497,
|
| 868 |
+
"learning_rate": 2.1334703560850786e-05,
|
| 869 |
+
"loss": 2.4127,
|
| 870 |
+
"step": 1230000
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.5781813027357301,
|
| 874 |
+
"grad_norm": 1.1274610757827759,
|
| 875 |
+
"learning_rate": 2.1101659193828754e-05,
|
| 876 |
+
"loss": 2.4096,
|
| 877 |
+
"step": 1240000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.5828440551771473,
|
| 881 |
+
"grad_norm": 1.0102494955062866,
|
| 882 |
+
"learning_rate": 2.0868591513044516e-05,
|
| 883 |
+
"loss": 2.4066,
|
| 884 |
+
"step": 1250000
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.5875068076185644,
|
| 888 |
+
"grad_norm": 1.011261224746704,
|
| 889 |
+
"learning_rate": 2.0635547146022484e-05,
|
| 890 |
+
"loss": 2.4046,
|
| 891 |
+
"step": 1260000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.5921695600599817,
|
| 895 |
+
"grad_norm": 1.1067317724227905,
|
| 896 |
+
"learning_rate": 2.0402502779000456e-05,
|
| 897 |
+
"loss": 2.4054,
|
| 898 |
+
"step": 1270000
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.5968323125013988,
|
| 902 |
+
"grad_norm": 1.117375135421753,
|
| 903 |
+
"learning_rate": 2.0169435098216217e-05,
|
| 904 |
+
"loss": 2.4057,
|
| 905 |
+
"step": 1280000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.601495064942816,
|
| 909 |
+
"grad_norm": 1.067470669746399,
|
| 910 |
+
"learning_rate": 1.993636741743198e-05,
|
| 911 |
+
"loss": 2.4066,
|
| 912 |
+
"step": 1290000
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.6061578173842331,
|
| 916 |
+
"grad_norm": 1.0158133506774902,
|
| 917 |
+
"learning_rate": 1.9703346364172157e-05,
|
| 918 |
+
"loss": 2.4029,
|
| 919 |
+
"step": 1300000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.6108205698256504,
|
| 923 |
+
"grad_norm": 1.1179207563400269,
|
| 924 |
+
"learning_rate": 1.947027868338792e-05,
|
| 925 |
+
"loss": 2.4006,
|
| 926 |
+
"step": 1310000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.6154833222670676,
|
| 930 |
+
"grad_norm": 0.8885159492492676,
|
| 931 |
+
"learning_rate": 1.923721100260368e-05,
|
| 932 |
+
"loss": 2.4008,
|
| 933 |
+
"step": 1320000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.6201460747084847,
|
| 937 |
+
"grad_norm": 0.9562169313430786,
|
| 938 |
+
"learning_rate": 1.9004143321819446e-05,
|
| 939 |
+
"loss": 2.4014,
|
| 940 |
+
"step": 1330000
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.6248088271499019,
|
| 944 |
+
"grad_norm": 1.0893275737762451,
|
| 945 |
+
"learning_rate": 1.8771098954797414e-05,
|
| 946 |
+
"loss": 2.3992,
|
| 947 |
+
"step": 1340000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.629471579591319,
|
| 951 |
+
"grad_norm": 1.1396783590316772,
|
| 952 |
+
"learning_rate": 1.853807790153759e-05,
|
| 953 |
+
"loss": 2.3961,
|
| 954 |
+
"step": 1350000
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.6341343320327363,
|
| 958 |
+
"grad_norm": 0.894639790058136,
|
| 959 |
+
"learning_rate": 1.830503353451556e-05,
|
| 960 |
+
"loss": 2.3949,
|
| 961 |
+
"step": 1360000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.6387970844741534,
|
| 965 |
+
"grad_norm": 1.0523122549057007,
|
| 966 |
+
"learning_rate": 1.807196585373132e-05,
|
| 967 |
+
"loss": 2.3924,
|
| 968 |
+
"step": 1370000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.6434598369155706,
|
| 972 |
+
"grad_norm": 1.4329748153686523,
|
| 973 |
+
"learning_rate": 1.7838898172947086e-05,
|
| 974 |
+
"loss": 2.3965,
|
| 975 |
+
"step": 1380000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.6481225893569877,
|
| 979 |
+
"grad_norm": 0.9407207369804382,
|
| 980 |
+
"learning_rate": 1.7605853805925054e-05,
|
| 981 |
+
"loss": 2.3944,
|
| 982 |
+
"step": 1390000
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.652785341798405,
|
| 986 |
+
"grad_norm": 1.1153851747512817,
|
| 987 |
+
"learning_rate": 1.7372809438903022e-05,
|
| 988 |
+
"loss": 2.3951,
|
| 989 |
+
"step": 1400000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.6574480942398221,
|
| 993 |
+
"grad_norm": 1.4270461797714233,
|
| 994 |
+
"learning_rate": 1.7139741758118784e-05,
|
| 995 |
+
"loss": 2.3903,
|
| 996 |
+
"step": 1410000
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.6621108466812393,
|
| 1000 |
+
"grad_norm": 0.9156707525253296,
|
| 1001 |
+
"learning_rate": 1.6906697391096756e-05,
|
| 1002 |
+
"loss": 2.387,
|
| 1003 |
+
"step": 1420000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.6667735991226565,
|
| 1007 |
+
"grad_norm": 1.0517213344573975,
|
| 1008 |
+
"learning_rate": 1.6673653024074724e-05,
|
| 1009 |
+
"loss": 2.3908,
|
| 1010 |
+
"step": 1430000
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.6714363515640737,
|
| 1014 |
+
"grad_norm": 1.1789027452468872,
|
| 1015 |
+
"learning_rate": 1.64406319708149e-05,
|
| 1016 |
+
"loss": 2.3857,
|
| 1017 |
+
"step": 1440000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.6760991040054909,
|
| 1021 |
+
"grad_norm": 0.9410611391067505,
|
| 1022 |
+
"learning_rate": 1.6207564290030663e-05,
|
| 1023 |
+
"loss": 2.3851,
|
| 1024 |
+
"step": 1450000
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.680761856446908,
|
| 1028 |
+
"grad_norm": 1.2597123384475708,
|
| 1029 |
+
"learning_rate": 1.597451992300863e-05,
|
| 1030 |
+
"loss": 2.3853,
|
| 1031 |
+
"step": 1460000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.6854246088883252,
|
| 1035 |
+
"grad_norm": 1.111659288406372,
|
| 1036 |
+
"learning_rate": 1.5741452242224393e-05,
|
| 1037 |
+
"loss": 2.3849,
|
| 1038 |
+
"step": 1470000
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.6900873613297424,
|
| 1042 |
+
"grad_norm": 1.114686131477356,
|
| 1043 |
+
"learning_rate": 1.5508384561440158e-05,
|
| 1044 |
+
"loss": 2.3844,
|
| 1045 |
+
"step": 1480000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.6947501137711596,
|
| 1049 |
+
"grad_norm": 1.3087519407272339,
|
| 1050 |
+
"learning_rate": 1.527531688065592e-05,
|
| 1051 |
+
"loss": 2.3811,
|
| 1052 |
+
"step": 1490000
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.6994128662125767,
|
| 1056 |
+
"grad_norm": 1.2704778909683228,
|
| 1057 |
+
"learning_rate": 1.5042319141158304e-05,
|
| 1058 |
+
"loss": 2.3793,
|
| 1059 |
+
"step": 1500000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.7040756186539939,
|
| 1063 |
+
"grad_norm": 1.0817821025848389,
|
| 1064 |
+
"learning_rate": 1.4809251460374065e-05,
|
| 1065 |
+
"loss": 2.3793,
|
| 1066 |
+
"step": 1510000
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.7087383710954112,
|
| 1070 |
+
"grad_norm": 1.1640921831130981,
|
| 1071 |
+
"learning_rate": 1.4576230407114241e-05,
|
| 1072 |
+
"loss": 2.3826,
|
| 1073 |
+
"step": 1520000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.7134011235368283,
|
| 1077 |
+
"grad_norm": 1.5091464519500732,
|
| 1078 |
+
"learning_rate": 1.4343162726330003e-05,
|
| 1079 |
+
"loss": 2.379,
|
| 1080 |
+
"step": 1530000
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.7180638759782455,
|
| 1084 |
+
"grad_norm": 1.3562886714935303,
|
| 1085 |
+
"learning_rate": 1.4110118359307974e-05,
|
| 1086 |
+
"loss": 2.3748,
|
| 1087 |
+
"step": 1540000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.7227266284196626,
|
| 1091 |
+
"grad_norm": 0.9998787641525269,
|
| 1092 |
+
"learning_rate": 1.3877073992285944e-05,
|
| 1093 |
+
"loss": 2.375,
|
| 1094 |
+
"step": 1550000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.7273893808610798,
|
| 1098 |
+
"grad_norm": 1.163294792175293,
|
| 1099 |
+
"learning_rate": 1.3644006311501706e-05,
|
| 1100 |
+
"loss": 2.3776,
|
| 1101 |
+
"step": 1560000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.732052133302497,
|
| 1105 |
+
"grad_norm": 1.0799118280410767,
|
| 1106 |
+
"learning_rate": 1.3410985258241882e-05,
|
| 1107 |
+
"loss": 2.3732,
|
| 1108 |
+
"step": 1570000
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.7367148857439142,
|
| 1112 |
+
"grad_norm": 0.9467183351516724,
|
| 1113 |
+
"learning_rate": 1.317794089121985e-05,
|
| 1114 |
+
"loss": 2.3705,
|
| 1115 |
+
"step": 1580000
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.7413776381853313,
|
| 1119 |
+
"grad_norm": 1.2810046672821045,
|
| 1120 |
+
"learning_rate": 1.2944873210435612e-05,
|
| 1121 |
+
"loss": 2.3721,
|
| 1122 |
+
"step": 1590000
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.7460403906267485,
|
| 1126 |
+
"grad_norm": 1.2798866033554077,
|
| 1127 |
+
"learning_rate": 1.2711828843413585e-05,
|
| 1128 |
+
"loss": 2.3738,
|
| 1129 |
+
"step": 1600000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.7507031430681657,
|
| 1133 |
+
"grad_norm": 1.221845030784607,
|
| 1134 |
+
"learning_rate": 1.2478784476391553e-05,
|
| 1135 |
+
"loss": 2.3683,
|
| 1136 |
+
"step": 1610000
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.7553658955095829,
|
| 1140 |
+
"grad_norm": 1.2743821144104004,
|
| 1141 |
+
"learning_rate": 1.2245740109369522e-05,
|
| 1142 |
+
"loss": 2.3724,
|
| 1143 |
+
"step": 1620000
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.7600286479510001,
|
| 1147 |
+
"grad_norm": 1.1069179773330688,
|
| 1148 |
+
"learning_rate": 1.201269574234749e-05,
|
| 1149 |
+
"loss": 2.3662,
|
| 1150 |
+
"step": 1630000
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.7646914003924172,
|
| 1154 |
+
"grad_norm": 1.4689267873764038,
|
| 1155 |
+
"learning_rate": 1.177965137532546e-05,
|
| 1156 |
+
"loss": 2.3713,
|
| 1157 |
+
"step": 1640000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.7693541528338345,
|
| 1161 |
+
"grad_norm": 1.0129334926605225,
|
| 1162 |
+
"learning_rate": 1.154660700830343e-05,
|
| 1163 |
+
"loss": 2.3689,
|
| 1164 |
+
"step": 1650000
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.7740169052752516,
|
| 1168 |
+
"grad_norm": 0.9776953458786011,
|
| 1169 |
+
"learning_rate": 1.1313539327519193e-05,
|
| 1170 |
+
"loss": 2.363,
|
| 1171 |
+
"step": 1660000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.7786796577166688,
|
| 1175 |
+
"grad_norm": 1.1849191188812256,
|
| 1176 |
+
"learning_rate": 1.1080494960497161e-05,
|
| 1177 |
+
"loss": 2.3671,
|
| 1178 |
+
"step": 1670000
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.7833424101580859,
|
| 1182 |
+
"grad_norm": 1.0659184455871582,
|
| 1183 |
+
"learning_rate": 1.0847427279712923e-05,
|
| 1184 |
+
"loss": 2.363,
|
| 1185 |
+
"step": 1680000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.7880051625995032,
|
| 1189 |
+
"grad_norm": 1.0228557586669922,
|
| 1190 |
+
"learning_rate": 1.0614382912690895e-05,
|
| 1191 |
+
"loss": 2.362,
|
| 1192 |
+
"step": 1690000
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.7926679150409203,
|
| 1196 |
+
"grad_norm": 0.9540805816650391,
|
| 1197 |
+
"learning_rate": 1.0381315231906656e-05,
|
| 1198 |
+
"loss": 2.366,
|
| 1199 |
+
"step": 1700000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.7973306674823375,
|
| 1203 |
+
"grad_norm": 1.1381940841674805,
|
| 1204 |
+
"learning_rate": 1.0148270864884624e-05,
|
| 1205 |
+
"loss": 2.3592,
|
| 1206 |
+
"step": 1710000
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.8019934199237547,
|
| 1210 |
+
"grad_norm": 1.1460505723953247,
|
| 1211 |
+
"learning_rate": 9.915226497862596e-06,
|
| 1212 |
+
"loss": 2.3591,
|
| 1213 |
+
"step": 1720000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.8066561723651718,
|
| 1217 |
+
"grad_norm": 1.0586894750595093,
|
| 1218 |
+
"learning_rate": 9.682182130840564e-06,
|
| 1219 |
+
"loss": 2.3592,
|
| 1220 |
+
"step": 1730000
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.8113189248065891,
|
| 1224 |
+
"grad_norm": 1.3877402544021606,
|
| 1225 |
+
"learning_rate": 9.449114450056326e-06,
|
| 1226 |
+
"loss": 2.3635,
|
| 1227 |
+
"step": 1740000
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.8159816772480062,
|
| 1231 |
+
"grad_norm": 1.2622848749160767,
|
| 1232 |
+
"learning_rate": 9.216046769272089e-06,
|
| 1233 |
+
"loss": 2.3577,
|
| 1234 |
+
"step": 1750000
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.8206444296894234,
|
| 1238 |
+
"grad_norm": 1.1290611028671265,
|
| 1239 |
+
"learning_rate": 8.983002402250059e-06,
|
| 1240 |
+
"loss": 2.3587,
|
| 1241 |
+
"step": 1760000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.8253071821308405,
|
| 1245 |
+
"grad_norm": 1.0407214164733887,
|
| 1246 |
+
"learning_rate": 8.74993472146582e-06,
|
| 1247 |
+
"loss": 2.3562,
|
| 1248 |
+
"step": 1770000
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.8299699345722578,
|
| 1252 |
+
"grad_norm": 1.1062073707580566,
|
| 1253 |
+
"learning_rate": 8.51689035444379e-06,
|
| 1254 |
+
"loss": 2.358,
|
| 1255 |
+
"step": 1780000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.8346326870136749,
|
| 1259 |
+
"grad_norm": 1.04072904586792,
|
| 1260 |
+
"learning_rate": 8.28384598742176e-06,
|
| 1261 |
+
"loss": 2.3518,
|
| 1262 |
+
"step": 1790000
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.8392954394550921,
|
| 1266 |
+
"grad_norm": 1.0454237461090088,
|
| 1267 |
+
"learning_rate": 8.050801620399728e-06,
|
| 1268 |
+
"loss": 2.3587,
|
| 1269 |
+
"step": 1800000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.8439581918965092,
|
| 1273 |
+
"grad_norm": 1.2492414712905884,
|
| 1274 |
+
"learning_rate": 7.817780567139906e-06,
|
| 1275 |
+
"loss": 2.3552,
|
| 1276 |
+
"step": 1810000
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.8486209443379265,
|
| 1280 |
+
"grad_norm": 1.2101612091064453,
|
| 1281 |
+
"learning_rate": 7.584712886355667e-06,
|
| 1282 |
+
"loss": 2.358,
|
| 1283 |
+
"step": 1820000
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.8532836967793437,
|
| 1287 |
+
"grad_norm": 1.0315169095993042,
|
| 1288 |
+
"learning_rate": 7.351668519333638e-06,
|
| 1289 |
+
"loss": 2.3515,
|
| 1290 |
+
"step": 1830000
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.8579464492207608,
|
| 1294 |
+
"grad_norm": 1.130194902420044,
|
| 1295 |
+
"learning_rate": 7.118624152311607e-06,
|
| 1296 |
+
"loss": 2.3535,
|
| 1297 |
+
"step": 1840000
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.862609201662178,
|
| 1301 |
+
"grad_norm": 1.1591068506240845,
|
| 1302 |
+
"learning_rate": 6.885579785289576e-06,
|
| 1303 |
+
"loss": 2.3484,
|
| 1304 |
+
"step": 1850000
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.8672719541035951,
|
| 1308 |
+
"grad_norm": 1.1694544553756714,
|
| 1309 |
+
"learning_rate": 6.652535418267545e-06,
|
| 1310 |
+
"loss": 2.3473,
|
| 1311 |
+
"step": 1860000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.8719347065450124,
|
| 1315 |
+
"grad_norm": 1.2773854732513428,
|
| 1316 |
+
"learning_rate": 6.419491051245514e-06,
|
| 1317 |
+
"loss": 2.3464,
|
| 1318 |
+
"step": 1870000
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.8765974589864295,
|
| 1322 |
+
"grad_norm": 1.0938977003097534,
|
| 1323 |
+
"learning_rate": 6.186423370461277e-06,
|
| 1324 |
+
"loss": 2.3468,
|
| 1325 |
+
"step": 1880000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.8812602114278467,
|
| 1329 |
+
"grad_norm": 1.178916573524475,
|
| 1330 |
+
"learning_rate": 5.9534023172014535e-06,
|
| 1331 |
+
"loss": 2.3455,
|
| 1332 |
+
"step": 1890000
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.8859229638692638,
|
| 1336 |
+
"grad_norm": 1.2058972120285034,
|
| 1337 |
+
"learning_rate": 5.720334636417215e-06,
|
| 1338 |
+
"loss": 2.3433,
|
| 1339 |
+
"step": 1900000
|
| 1340 |
}
|
| 1341 |
],
|
| 1342 |
"logging_steps": 10000,
|
|
|
|
| 1356 |
"attributes": {}
|
| 1357 |
}
|
| 1358 |
},
|
| 1359 |
+
"total_flos": 4.684660451731086e+19,
|
| 1360 |
"train_batch_size": 4,
|
| 1361 |
"trial_name": null,
|
| 1362 |
"trial_params": null
|