Training in progress, step 184, 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 30322120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0839408bc64ce86e532ac0bee107a132e47a4332c875723159b5adf437750782
|
| 3 |
size 30322120
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15768084
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:04fe6d3b5df6ad8af2c0860e258597d5366d0ac01ea9a4752bdfcb1af6eb4ab6
|
| 3 |
size 15768084
|
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:e5859d87ccb3cb657e795c86ddefdc1a9bc81fa2143585b1404f7bd03a8a8a8c
|
| 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 sha256:3c92fc546447cfc1151596b19e46f7088e39100f9de60871cd8a0f9b1a8d1df9
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 2.2546868324279785,
|
| 3 |
"best_model_checkpoint": "miner_id_24/checkpoint-100",
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 100,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -723,6 +723,594 @@
|
|
| 723 |
"eval_samples_per_second": 52.416,
|
| 724 |
"eval_steps_per_second": 13.188,
|
| 725 |
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 726 |
}
|
| 727 |
],
|
| 728 |
"logging_steps": 1,
|
|
@@ -746,12 +1334,12 @@
|
|
| 746 |
"should_evaluate": false,
|
| 747 |
"should_log": false,
|
| 748 |
"should_save": true,
|
| 749 |
-
"should_training_stop":
|
| 750 |
},
|
| 751 |
"attributes": {}
|
| 752 |
}
|
| 753 |
},
|
| 754 |
-
"total_flos":
|
| 755 |
"train_batch_size": 4,
|
| 756 |
"trial_name": null,
|
| 757 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": 2.2546868324279785,
|
| 3 |
"best_model_checkpoint": "miner_id_24/checkpoint-100",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 184,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 723 |
"eval_samples_per_second": 52.416,
|
| 724 |
"eval_steps_per_second": 13.188,
|
| 725 |
"step": 100
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.0978260869565217,
|
| 729 |
+
"grad_norm": 0.39603304862976074,
|
| 730 |
+
"learning_rate": 9.27842244199384e-05,
|
| 731 |
+
"loss": 2.2077,
|
| 732 |
+
"step": 101
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.108695652173913,
|
| 736 |
+
"grad_norm": 0.34627410769462585,
|
| 737 |
+
"learning_rate": 9.098469188879349e-05,
|
| 738 |
+
"loss": 2.3818,
|
| 739 |
+
"step": 102
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.1195652173913044,
|
| 743 |
+
"grad_norm": 0.39001986384391785,
|
| 744 |
+
"learning_rate": 8.918809815760585e-05,
|
| 745 |
+
"loss": 2.2167,
|
| 746 |
+
"step": 103
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.1304347826086956,
|
| 750 |
+
"grad_norm": 0.38835734128952026,
|
| 751 |
+
"learning_rate": 8.739502887797107e-05,
|
| 752 |
+
"loss": 2.2521,
|
| 753 |
+
"step": 104
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.141304347826087,
|
| 757 |
+
"grad_norm": 0.417492151260376,
|
| 758 |
+
"learning_rate": 8.560606855258808e-05,
|
| 759 |
+
"loss": 2.2448,
|
| 760 |
+
"step": 105
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.1521739130434783,
|
| 764 |
+
"grad_norm": 0.3795416057109833,
|
| 765 |
+
"learning_rate": 8.382180034472353e-05,
|
| 766 |
+
"loss": 1.9877,
|
| 767 |
+
"step": 106
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.1630434782608696,
|
| 771 |
+
"grad_norm": 0.33009013533592224,
|
| 772 |
+
"learning_rate": 8.204280588811283e-05,
|
| 773 |
+
"loss": 1.96,
|
| 774 |
+
"step": 107
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.1739130434782608,
|
| 778 |
+
"grad_norm": 0.3021019995212555,
|
| 779 |
+
"learning_rate": 8.026966509736001e-05,
|
| 780 |
+
"loss": 2.0639,
|
| 781 |
+
"step": 108
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.184782608695652,
|
| 785 |
+
"grad_norm": 0.34734079241752625,
|
| 786 |
+
"learning_rate": 7.85029559788976e-05,
|
| 787 |
+
"loss": 2.1691,
|
| 788 |
+
"step": 109
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.1956521739130435,
|
| 792 |
+
"grad_norm": 0.38481855392456055,
|
| 793 |
+
"learning_rate": 7.674325444256899e-05,
|
| 794 |
+
"loss": 1.9582,
|
| 795 |
+
"step": 110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.2065217391304348,
|
| 799 |
+
"grad_norm": 0.37570053339004517,
|
| 800 |
+
"learning_rate": 7.499113411389371e-05,
|
| 801 |
+
"loss": 2.1185,
|
| 802 |
+
"step": 111
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.2173913043478262,
|
| 806 |
+
"grad_norm": 0.35231196880340576,
|
| 807 |
+
"learning_rate": 7.324716614707793e-05,
|
| 808 |
+
"loss": 1.9297,
|
| 809 |
+
"step": 112
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.2282608695652173,
|
| 813 |
+
"grad_norm": 0.5351376533508301,
|
| 814 |
+
"learning_rate": 7.151191903883001e-05,
|
| 815 |
+
"loss": 2.4002,
|
| 816 |
+
"step": 113
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.2391304347826086,
|
| 820 |
+
"grad_norm": 0.3427269160747528,
|
| 821 |
+
"learning_rate": 6.978595844304271e-05,
|
| 822 |
+
"loss": 2.0233,
|
| 823 |
+
"step": 114
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.25,
|
| 827 |
+
"grad_norm": 0.33183610439300537,
|
| 828 |
+
"learning_rate": 6.806984698640202e-05,
|
| 829 |
+
"loss": 2.1097,
|
| 830 |
+
"step": 115
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.2608695652173914,
|
| 834 |
+
"grad_norm": 0.41684049367904663,
|
| 835 |
+
"learning_rate": 6.636414408498249e-05,
|
| 836 |
+
"loss": 2.0766,
|
| 837 |
+
"step": 116
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.2717391304347827,
|
| 841 |
+
"grad_norm": 0.48687830567359924,
|
| 842 |
+
"learning_rate": 6.466940576188977e-05,
|
| 843 |
+
"loss": 2.2554,
|
| 844 |
+
"step": 117
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.2826086956521738,
|
| 848 |
+
"grad_norm": 0.32962566614151,
|
| 849 |
+
"learning_rate": 6.298618446600856e-05,
|
| 850 |
+
"loss": 2.119,
|
| 851 |
+
"step": 118
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.2934782608695652,
|
| 855 |
+
"grad_norm": 0.4224526584148407,
|
| 856 |
+
"learning_rate": 6.13150288919161e-05,
|
| 857 |
+
"loss": 2.1045,
|
| 858 |
+
"step": 119
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.3043478260869565,
|
| 862 |
+
"grad_norm": 0.3788648545742035,
|
| 863 |
+
"learning_rate": 5.965648380101916e-05,
|
| 864 |
+
"loss": 2.091,
|
| 865 |
+
"step": 120
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.315217391304348,
|
| 869 |
+
"grad_norm": 0.4119998514652252,
|
| 870 |
+
"learning_rate": 5.801108984397354e-05,
|
| 871 |
+
"loss": 2.2618,
|
| 872 |
+
"step": 121
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.3260869565217392,
|
| 876 |
+
"grad_norm": 0.43901023268699646,
|
| 877 |
+
"learning_rate": 5.6379383384443255e-05,
|
| 878 |
+
"loss": 2.1975,
|
| 879 |
+
"step": 122
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.3369565217391304,
|
| 883 |
+
"grad_norm": 0.38879257440567017,
|
| 884 |
+
"learning_rate": 5.476189632425732e-05,
|
| 885 |
+
"loss": 2.1746,
|
| 886 |
+
"step": 123
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.3478260869565217,
|
| 890 |
+
"grad_norm": 0.46310365200042725,
|
| 891 |
+
"learning_rate": 5.3159155930021e-05,
|
| 892 |
+
"loss": 2.2617,
|
| 893 |
+
"step": 124
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.358695652173913,
|
| 897 |
+
"grad_norm": 0.32008013129234314,
|
| 898 |
+
"learning_rate": 5.1571684661238075e-05,
|
| 899 |
+
"loss": 1.5438,
|
| 900 |
+
"step": 125
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.3695652173913042,
|
| 904 |
+
"grad_norm": 0.47657638788223267,
|
| 905 |
+
"learning_rate": 5.000000000000002e-05,
|
| 906 |
+
"loss": 2.2612,
|
| 907 |
+
"step": 126
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.3804347826086958,
|
| 911 |
+
"grad_norm": 0.45614132285118103,
|
| 912 |
+
"learning_rate": 4.844461428229782e-05,
|
| 913 |
+
"loss": 2.3454,
|
| 914 |
+
"step": 127
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.391304347826087,
|
| 918 |
+
"grad_norm": 0.3681153655052185,
|
| 919 |
+
"learning_rate": 4.6906034531011346e-05,
|
| 920 |
+
"loss": 2.1628,
|
| 921 |
+
"step": 128
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.4021739130434783,
|
| 925 |
+
"grad_norm": 0.44740939140319824,
|
| 926 |
+
"learning_rate": 4.53847622906303e-05,
|
| 927 |
+
"loss": 2.1763,
|
| 928 |
+
"step": 129
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.4130434782608696,
|
| 932 |
+
"grad_norm": 0.441448837518692,
|
| 933 |
+
"learning_rate": 4.388129346376178e-05,
|
| 934 |
+
"loss": 2.423,
|
| 935 |
+
"step": 130
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.4239130434782608,
|
| 939 |
+
"grad_norm": 0.37970641255378723,
|
| 940 |
+
"learning_rate": 4.239611814947605e-05,
|
| 941 |
+
"loss": 2.078,
|
| 942 |
+
"step": 131
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.434782608695652,
|
| 946 |
+
"grad_norm": 0.48924893140792847,
|
| 947 |
+
"learning_rate": 4.092972048354491e-05,
|
| 948 |
+
"loss": 2.073,
|
| 949 |
+
"step": 132
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.4456521739130435,
|
| 953 |
+
"grad_norm": 0.39348116517066956,
|
| 954 |
+
"learning_rate": 3.948257848062351e-05,
|
| 955 |
+
"loss": 1.7686,
|
| 956 |
+
"step": 133
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.4565217391304348,
|
| 960 |
+
"grad_norm": 0.6471808552742004,
|
| 961 |
+
"learning_rate": 3.80551638784277e-05,
|
| 962 |
+
"loss": 1.7646,
|
| 963 |
+
"step": 134
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.4673913043478262,
|
| 967 |
+
"grad_norm": 0.396930456161499,
|
| 968 |
+
"learning_rate": 3.664794198395764e-05,
|
| 969 |
+
"loss": 2.3518,
|
| 970 |
+
"step": 135
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.4782608695652173,
|
| 974 |
+
"grad_norm": 0.45122313499450684,
|
| 975 |
+
"learning_rate": 3.5261371521817244e-05,
|
| 976 |
+
"loss": 2.1385,
|
| 977 |
+
"step": 136
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.4891304347826086,
|
| 981 |
+
"grad_norm": 0.3892301023006439,
|
| 982 |
+
"learning_rate": 3.3895904484679984e-05,
|
| 983 |
+
"loss": 2.0135,
|
| 984 |
+
"step": 137
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.5,
|
| 988 |
+
"grad_norm": 0.4577631950378418,
|
| 989 |
+
"learning_rate": 3.2551985985948616e-05,
|
| 990 |
+
"loss": 2.1586,
|
| 991 |
+
"step": 138
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.5108695652173914,
|
| 995 |
+
"grad_norm": 0.39805516600608826,
|
| 996 |
+
"learning_rate": 3.123005411465766e-05,
|
| 997 |
+
"loss": 2.3181,
|
| 998 |
+
"step": 139
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.5217391304347827,
|
| 1002 |
+
"grad_norm": 0.36367368698120117,
|
| 1003 |
+
"learning_rate": 2.993053979266577e-05,
|
| 1004 |
+
"loss": 1.9977,
|
| 1005 |
+
"step": 140
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.5326086956521738,
|
| 1009 |
+
"grad_norm": 0.3784072697162628,
|
| 1010 |
+
"learning_rate": 2.8653866634184e-05,
|
| 1011 |
+
"loss": 1.8181,
|
| 1012 |
+
"step": 141
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.5434782608695652,
|
| 1016 |
+
"grad_norm": 0.3953021466732025,
|
| 1017 |
+
"learning_rate": 2.7400450807686938e-05,
|
| 1018 |
+
"loss": 2.2361,
|
| 1019 |
+
"step": 142
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.5543478260869565,
|
| 1023 |
+
"grad_norm": 0.42295852303504944,
|
| 1024 |
+
"learning_rate": 2.6170700900250144e-05,
|
| 1025 |
+
"loss": 1.9388,
|
| 1026 |
+
"step": 143
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.5652173913043477,
|
| 1030 |
+
"grad_norm": 0.43486106395721436,
|
| 1031 |
+
"learning_rate": 2.496501778435977e-05,
|
| 1032 |
+
"loss": 2.1987,
|
| 1033 |
+
"step": 144
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.5760869565217392,
|
| 1037 |
+
"grad_norm": 0.4382745027542114,
|
| 1038 |
+
"learning_rate": 2.3783794487236365e-05,
|
| 1039 |
+
"loss": 2.1836,
|
| 1040 |
+
"step": 145
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.5869565217391304,
|
| 1044 |
+
"grad_norm": 0.4367092251777649,
|
| 1045 |
+
"learning_rate": 2.2627416062716366e-05,
|
| 1046 |
+
"loss": 2.3411,
|
| 1047 |
+
"step": 146
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.5978260869565217,
|
| 1051 |
+
"grad_norm": 0.38458558917045593,
|
| 1052 |
+
"learning_rate": 2.1496259465732783e-05,
|
| 1053 |
+
"loss": 2.109,
|
| 1054 |
+
"step": 147
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.608695652173913,
|
| 1058 |
+
"grad_norm": 0.36800533533096313,
|
| 1059 |
+
"learning_rate": 2.0390693429435627e-05,
|
| 1060 |
+
"loss": 2.1868,
|
| 1061 |
+
"step": 148
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.6195652173913042,
|
| 1065 |
+
"grad_norm": 0.3944242596626282,
|
| 1066 |
+
"learning_rate": 1.931107834499296e-05,
|
| 1067 |
+
"loss": 2.3074,
|
| 1068 |
+
"step": 149
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.6304347826086958,
|
| 1072 |
+
"grad_norm": 0.35989850759506226,
|
| 1073 |
+
"learning_rate": 1.825776614411082e-05,
|
| 1074 |
+
"loss": 1.9306,
|
| 1075 |
+
"step": 150
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.641304347826087,
|
| 1079 |
+
"grad_norm": 0.35803163051605225,
|
| 1080 |
+
"learning_rate": 1.7231100184310956e-05,
|
| 1081 |
+
"loss": 1.761,
|
| 1082 |
+
"step": 151
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.6521739130434783,
|
| 1086 |
+
"grad_norm": 0.46954256296157837,
|
| 1087 |
+
"learning_rate": 1.6231415137003537e-05,
|
| 1088 |
+
"loss": 1.8994,
|
| 1089 |
+
"step": 152
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.6630434782608696,
|
| 1093 |
+
"grad_norm": 0.3517175018787384,
|
| 1094 |
+
"learning_rate": 1.5259036878391343e-05,
|
| 1095 |
+
"loss": 1.9668,
|
| 1096 |
+
"step": 153
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.6739130434782608,
|
| 1100 |
+
"grad_norm": 0.3753047585487366,
|
| 1101 |
+
"learning_rate": 1.4314282383241096e-05,
|
| 1102 |
+
"loss": 2.008,
|
| 1103 |
+
"step": 154
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.6847826086956523,
|
| 1107 |
+
"grad_norm": 0.4350530803203583,
|
| 1108 |
+
"learning_rate": 1.339745962155613e-05,
|
| 1109 |
+
"loss": 2.0114,
|
| 1110 |
+
"step": 155
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.6956521739130435,
|
| 1114 |
+
"grad_norm": 0.4570012092590332,
|
| 1115 |
+
"learning_rate": 1.2508867458185037e-05,
|
| 1116 |
+
"loss": 2.1608,
|
| 1117 |
+
"step": 156
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.7065217391304348,
|
| 1121 |
+
"grad_norm": 0.41090089082717896,
|
| 1122 |
+
"learning_rate": 1.1648795555397719e-05,
|
| 1123 |
+
"loss": 1.9537,
|
| 1124 |
+
"step": 157
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.7173913043478262,
|
| 1128 |
+
"grad_norm": 0.4272090196609497,
|
| 1129 |
+
"learning_rate": 1.0817524278461776e-05,
|
| 1130 |
+
"loss": 1.9259,
|
| 1131 |
+
"step": 158
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.7282608695652173,
|
| 1135 |
+
"grad_norm": 0.42349883913993835,
|
| 1136 |
+
"learning_rate": 1.0015324604249343e-05,
|
| 1137 |
+
"loss": 2.2443,
|
| 1138 |
+
"step": 159
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.7391304347826086,
|
| 1142 |
+
"grad_norm": 0.44883671402931213,
|
| 1143 |
+
"learning_rate": 9.242458032904311e-06,
|
| 1144 |
+
"loss": 2.2694,
|
| 1145 |
+
"step": 160
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.75,
|
| 1149 |
+
"grad_norm": 0.377241849899292,
|
| 1150 |
+
"learning_rate": 8.499176502598782e-06,
|
| 1151 |
+
"loss": 2.0696,
|
| 1152 |
+
"step": 161
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.7608695652173914,
|
| 1156 |
+
"grad_norm": 0.4020577073097229,
|
| 1157 |
+
"learning_rate": 7.785722307406684e-06,
|
| 1158 |
+
"loss": 2.0756,
|
| 1159 |
+
"step": 162
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.7717391304347827,
|
| 1163 |
+
"grad_norm": 0.38903653621673584,
|
| 1164 |
+
"learning_rate": 7.102328018320858e-06,
|
| 1165 |
+
"loss": 2.0419,
|
| 1166 |
+
"step": 163
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.7826086956521738,
|
| 1170 |
+
"grad_norm": 0.4010683000087738,
|
| 1171 |
+
"learning_rate": 6.4492164074399065e-06,
|
| 1172 |
+
"loss": 2.0915,
|
| 1173 |
+
"step": 164
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.7934782608695652,
|
| 1177 |
+
"grad_norm": 0.49136319756507874,
|
| 1178 |
+
"learning_rate": 5.826600375349201e-06,
|
| 1179 |
+
"loss": 2.0298,
|
| 1180 |
+
"step": 165
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.8043478260869565,
|
| 1184 |
+
"grad_norm": 0.4226118326187134,
|
| 1185 |
+
"learning_rate": 5.2346828817197655e-06,
|
| 1186 |
+
"loss": 2.2148,
|
| 1187 |
+
"step": 166
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.8152173913043477,
|
| 1191 |
+
"grad_norm": 0.36731913685798645,
|
| 1192 |
+
"learning_rate": 4.673656879147737e-06,
|
| 1193 |
+
"loss": 1.969,
|
| 1194 |
+
"step": 167
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.8260869565217392,
|
| 1198 |
+
"grad_norm": 0.37431615591049194,
|
| 1199 |
+
"learning_rate": 4.143705250255869e-06,
|
| 1200 |
+
"loss": 2.142,
|
| 1201 |
+
"step": 168
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.8369565217391304,
|
| 1205 |
+
"grad_norm": 0.41436100006103516,
|
| 1206 |
+
"learning_rate": 3.6450007480777093e-06,
|
| 1207 |
+
"loss": 2.0905,
|
| 1208 |
+
"step": 169
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.8478260869565217,
|
| 1212 |
+
"grad_norm": 0.4659545123577118,
|
| 1213 |
+
"learning_rate": 3.1777059397436692e-06,
|
| 1214 |
+
"loss": 2.3771,
|
| 1215 |
+
"step": 170
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.858695652173913,
|
| 1219 |
+
"grad_norm": 0.3477473258972168,
|
| 1220 |
+
"learning_rate": 2.741973153487687e-06,
|
| 1221 |
+
"loss": 1.8993,
|
| 1222 |
+
"step": 171
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 1.8695652173913042,
|
| 1226 |
+
"grad_norm": 0.43065935373306274,
|
| 1227 |
+
"learning_rate": 2.3379444289913342e-06,
|
| 1228 |
+
"loss": 1.9872,
|
| 1229 |
+
"step": 172
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.8804347826086958,
|
| 1233 |
+
"grad_norm": 0.4017411768436432,
|
| 1234 |
+
"learning_rate": 1.965751471081978e-06,
|
| 1235 |
+
"loss": 2.1586,
|
| 1236 |
+
"step": 173
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.891304347826087,
|
| 1240 |
+
"grad_norm": 0.38099217414855957,
|
| 1241 |
+
"learning_rate": 1.6255156067997323e-06,
|
| 1242 |
+
"loss": 2.111,
|
| 1243 |
+
"step": 174
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 1.9021739130434783,
|
| 1247 |
+
"grad_norm": 0.49076399207115173,
|
| 1248 |
+
"learning_rate": 1.317347745847386e-06,
|
| 1249 |
+
"loss": 2.1897,
|
| 1250 |
+
"step": 175
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.9130434782608696,
|
| 1254 |
+
"grad_norm": 0.3371427357196808,
|
| 1255 |
+
"learning_rate": 1.0413483444362771e-06,
|
| 1256 |
+
"loss": 1.9407,
|
| 1257 |
+
"step": 176
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.9239130434782608,
|
| 1261 |
+
"grad_norm": 0.3599741756916046,
|
| 1262 |
+
"learning_rate": 7.976073725395661e-07,
|
| 1263 |
+
"loss": 2.1587,
|
| 1264 |
+
"step": 177
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1.9347826086956523,
|
| 1268 |
+
"grad_norm": 0.4170662462711334,
|
| 1269 |
+
"learning_rate": 5.862042845640403e-07,
|
| 1270 |
+
"loss": 2.1134,
|
| 1271 |
+
"step": 178
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.9456521739130435,
|
| 1275 |
+
"grad_norm": 0.399311363697052,
|
| 1276 |
+
"learning_rate": 4.072079934495765e-07,
|
| 1277 |
+
"loss": 2.1783,
|
| 1278 |
+
"step": 179
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.9565217391304348,
|
| 1282 |
+
"grad_norm": 0.3565449118614197,
|
| 1283 |
+
"learning_rate": 2.606768482050215e-07,
|
| 1284 |
+
"loss": 2.323,
|
| 1285 |
+
"step": 180
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.9673913043478262,
|
| 1289 |
+
"grad_norm": 0.41395241022109985,
|
| 1290 |
+
"learning_rate": 1.4665861488761813e-07,
|
| 1291 |
+
"loss": 1.9735,
|
| 1292 |
+
"step": 181
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.9782608695652173,
|
| 1296 |
+
"grad_norm": 0.373086541891098,
|
| 1297 |
+
"learning_rate": 6.519046103230508e-08,
|
| 1298 |
+
"loss": 2.1872,
|
| 1299 |
+
"step": 182
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.9891304347826086,
|
| 1303 |
+
"grad_norm": 0.4048343002796173,
|
| 1304 |
+
"learning_rate": 1.6298943535875134e-08,
|
| 1305 |
+
"loss": 2.1919,
|
| 1306 |
+
"step": 183
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 2.0,
|
| 1310 |
+
"grad_norm": 0.46199992299079895,
|
| 1311 |
+
"learning_rate": 0.0,
|
| 1312 |
+
"loss": 2.1255,
|
| 1313 |
+
"step": 184
|
| 1314 |
}
|
| 1315 |
],
|
| 1316 |
"logging_steps": 1,
|
|
|
|
| 1334 |
"should_evaluate": false,
|
| 1335 |
"should_log": false,
|
| 1336 |
"should_save": true,
|
| 1337 |
+
"should_training_stop": true
|
| 1338 |
},
|
| 1339 |
"attributes": {}
|
| 1340 |
}
|
| 1341 |
},
|
| 1342 |
+
"total_flos": 1.3670974633279488e+16,
|
| 1343 |
"train_batch_size": 4,
|
| 1344 |
"trial_name": null,
|
| 1345 |
"trial_params": null
|