Training in progress, step 155, 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 50624
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26657c3db366efc7fa6d92372aa3219664b8571ca133e00feda4d1f205b15d90
|
| 3 |
size 50624
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 111142
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88fd526672df865f81c4176a06dc3da1929c1ce8317f76c85d58e8c3304b2eb9
|
| 3 |
size 111142
|
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 sha256:14a3afbab8271bee510684da18634a513067dc3f106f7cf4fa953ce19ce526db
|
| 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:fad1344d0d68cec31f4379da8eb730afca29401bda64dafa2bff69b67bba283b
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 10.353182792663574,
|
| 3 |
"best_model_checkpoint": "miner_id_24/checkpoint-100",
|
| 4 |
-
"epoch": 0.
|
| 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 +723,391 @@
|
|
| 723 |
"eval_samples_per_second": 328.204,
|
| 724 |
"eval_steps_per_second": 82.159,
|
| 725 |
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 726 |
}
|
| 727 |
],
|
| 728 |
"logging_steps": 1,
|
|
@@ -746,12 +1131,12 @@
|
|
| 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": 10.353182792663574,
|
| 3 |
"best_model_checkpoint": "miner_id_24/checkpoint-100",
|
| 4 |
+
"epoch": 0.34315760343157603,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 155,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 723 |
"eval_samples_per_second": 328.204,
|
| 724 |
"eval_steps_per_second": 82.159,
|
| 725 |
"step": 100
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.2236059222360592,
|
| 729 |
+
"grad_norm": 0.03327897563576698,
|
| 730 |
+
"learning_rate": 6.0982293673944544e-05,
|
| 731 |
+
"loss": 10.3556,
|
| 732 |
+
"step": 101
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.22581984225819843,
|
| 736 |
+
"grad_norm": 0.03691767901182175,
|
| 737 |
+
"learning_rate": 5.899671794785839e-05,
|
| 738 |
+
"loss": 10.3553,
|
| 739 |
+
"step": 102
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.22803376228033762,
|
| 743 |
+
"grad_norm": 0.030172044411301613,
|
| 744 |
+
"learning_rate": 5.703038932486484e-05,
|
| 745 |
+
"loss": 10.3542,
|
| 746 |
+
"step": 103
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.23024768230247683,
|
| 750 |
+
"grad_norm": 0.021300997585058212,
|
| 751 |
+
"learning_rate": 5.5084230807412126e-05,
|
| 752 |
+
"loss": 10.3543,
|
| 753 |
+
"step": 104
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.23246160232461602,
|
| 757 |
+
"grad_norm": 0.03279775753617287,
|
| 758 |
+
"learning_rate": 5.3159155930021e-05,
|
| 759 |
+
"loss": 10.3539,
|
| 760 |
+
"step": 105
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.23467552234675523,
|
| 764 |
+
"grad_norm": 0.02924364060163498,
|
| 765 |
+
"learning_rate": 5.12560683304681e-05,
|
| 766 |
+
"loss": 10.3543,
|
| 767 |
+
"step": 106
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.23688944236889442,
|
| 771 |
+
"grad_norm": 0.03262259438633919,
|
| 772 |
+
"learning_rate": 4.9375861325614606e-05,
|
| 773 |
+
"loss": 10.355,
|
| 774 |
+
"step": 107
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.23910336239103364,
|
| 778 |
+
"grad_norm": 0.034073278307914734,
|
| 779 |
+
"learning_rate": 4.751941749207995e-05,
|
| 780 |
+
"loss": 10.357,
|
| 781 |
+
"step": 108
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.24131728241317282,
|
| 785 |
+
"grad_norm": 0.03489963710308075,
|
| 786 |
+
"learning_rate": 4.5687608251956714e-05,
|
| 787 |
+
"loss": 10.355,
|
| 788 |
+
"step": 109
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.243531202435312,
|
| 792 |
+
"grad_norm": 0.02581014297902584,
|
| 793 |
+
"learning_rate": 4.388129346376178e-05,
|
| 794 |
+
"loss": 10.3537,
|
| 795 |
+
"step": 110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.24574512245745123,
|
| 799 |
+
"grad_norm": 0.023335812613368034,
|
| 800 |
+
"learning_rate": 4.210132101881516e-05,
|
| 801 |
+
"loss": 10.3553,
|
| 802 |
+
"step": 111
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.2479590424795904,
|
| 806 |
+
"grad_norm": 0.03612133115530014,
|
| 807 |
+
"learning_rate": 4.034852644323661e-05,
|
| 808 |
+
"loss": 10.3534,
|
| 809 |
+
"step": 112
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.2501729625017296,
|
| 813 |
+
"grad_norm": 0.02821163646876812,
|
| 814 |
+
"learning_rate": 3.862373250574626e-05,
|
| 815 |
+
"loss": 10.3556,
|
| 816 |
+
"step": 113
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.2523868825238688,
|
| 820 |
+
"grad_norm": 0.0300295390188694,
|
| 821 |
+
"learning_rate": 3.6927748831453836e-05,
|
| 822 |
+
"loss": 10.3551,
|
| 823 |
+
"step": 114
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.25460080254600803,
|
| 827 |
+
"grad_norm": 0.025033898651599884,
|
| 828 |
+
"learning_rate": 3.5261371521817244e-05,
|
| 829 |
+
"loss": 10.3525,
|
| 830 |
+
"step": 115
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.25681472256814725,
|
| 834 |
+
"grad_norm": 0.02417595498263836,
|
| 835 |
+
"learning_rate": 3.3625382780949574e-05,
|
| 836 |
+
"loss": 10.3542,
|
| 837 |
+
"step": 116
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.2590286425902864,
|
| 841 |
+
"grad_norm": 0.02323267050087452,
|
| 842 |
+
"learning_rate": 3.202055054844921e-05,
|
| 843 |
+
"loss": 10.3521,
|
| 844 |
+
"step": 117
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.2612425626124256,
|
| 848 |
+
"grad_norm": 0.026705941185355186,
|
| 849 |
+
"learning_rate": 3.0447628138926156e-05,
|
| 850 |
+
"loss": 10.3536,
|
| 851 |
+
"step": 118
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.26345648263456484,
|
| 855 |
+
"grad_norm": 0.033880215138196945,
|
| 856 |
+
"learning_rate": 2.890735388839295e-05,
|
| 857 |
+
"loss": 10.3554,
|
| 858 |
+
"step": 119
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.26567040265670405,
|
| 862 |
+
"grad_norm": 0.029678767547011375,
|
| 863 |
+
"learning_rate": 2.7400450807686938e-05,
|
| 864 |
+
"loss": 10.353,
|
| 865 |
+
"step": 120
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.2678843226788432,
|
| 869 |
+
"grad_norm": 0.02781762182712555,
|
| 870 |
+
"learning_rate": 2.59276262430861e-05,
|
| 871 |
+
"loss": 10.3553,
|
| 872 |
+
"step": 121
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.2700982427009824,
|
| 876 |
+
"grad_norm": 0.027942579239606857,
|
| 877 |
+
"learning_rate": 2.4489571544277945e-05,
|
| 878 |
+
"loss": 10.3556,
|
| 879 |
+
"step": 122
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.27231216272312164,
|
| 883 |
+
"grad_norm": 0.02813226915895939,
|
| 884 |
+
"learning_rate": 2.308696173983711e-05,
|
| 885 |
+
"loss": 10.3528,
|
| 886 |
+
"step": 123
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.2745260827452608,
|
| 890 |
+
"grad_norm": 0.030037103220820427,
|
| 891 |
+
"learning_rate": 2.1720455220364444e-05,
|
| 892 |
+
"loss": 10.353,
|
| 893 |
+
"step": 124
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.2767400027674,
|
| 897 |
+
"grad_norm": 0.030153660103678703,
|
| 898 |
+
"learning_rate": 2.0390693429435627e-05,
|
| 899 |
+
"loss": 10.3552,
|
| 900 |
+
"step": 125
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.27895392278953923,
|
| 904 |
+
"grad_norm": 0.037454936653375626,
|
| 905 |
+
"learning_rate": 1.9098300562505266e-05,
|
| 906 |
+
"loss": 10.3531,
|
| 907 |
+
"step": 126
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 0.28116784281167845,
|
| 911 |
+
"grad_norm": 0.029572051018476486,
|
| 912 |
+
"learning_rate": 1.784388327390687e-05,
|
| 913 |
+
"loss": 10.3504,
|
| 914 |
+
"step": 127
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.2833817628338176,
|
| 918 |
+
"grad_norm": 0.027182403951883316,
|
| 919 |
+
"learning_rate": 1.6628030392087e-05,
|
| 920 |
+
"loss": 10.3504,
|
| 921 |
+
"step": 128
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.2855956828559568,
|
| 925 |
+
"grad_norm": 0.03518354520201683,
|
| 926 |
+
"learning_rate": 1.5451312643206827e-05,
|
| 927 |
+
"loss": 10.3536,
|
| 928 |
+
"step": 129
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.28780960287809604,
|
| 932 |
+
"grad_norm": 0.024024929851293564,
|
| 933 |
+
"learning_rate": 1.4314282383241096e-05,
|
| 934 |
+
"loss": 10.3533,
|
| 935 |
+
"step": 130
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.29002352290023525,
|
| 939 |
+
"grad_norm": 0.020221339538693428,
|
| 940 |
+
"learning_rate": 1.3217473338699859e-05,
|
| 941 |
+
"loss": 10.3521,
|
| 942 |
+
"step": 131
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 0.2922374429223744,
|
| 946 |
+
"grad_norm": 0.02575266920030117,
|
| 947 |
+
"learning_rate": 1.2161400356095375e-05,
|
| 948 |
+
"loss": 10.3528,
|
| 949 |
+
"step": 132
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 0.29445136294451363,
|
| 953 |
+
"grad_norm": 0.019320376217365265,
|
| 954 |
+
"learning_rate": 1.1146559160270875e-05,
|
| 955 |
+
"loss": 10.3529,
|
| 956 |
+
"step": 133
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.29666528296665284,
|
| 960 |
+
"grad_norm": 0.022721335291862488,
|
| 961 |
+
"learning_rate": 1.0173426121705576e-05,
|
| 962 |
+
"loss": 10.3534,
|
| 963 |
+
"step": 134
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.298879202988792,
|
| 967 |
+
"grad_norm": 0.030306054279208183,
|
| 968 |
+
"learning_rate": 9.242458032904311e-06,
|
| 969 |
+
"loss": 10.3529,
|
| 970 |
+
"step": 135
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.3010931230109312,
|
| 974 |
+
"grad_norm": 0.03088550828397274,
|
| 975 |
+
"learning_rate": 8.354091893977401e-06,
|
| 976 |
+
"loss": 10.3542,
|
| 977 |
+
"step": 136
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.30330704303307043,
|
| 981 |
+
"grad_norm": 0.02875097282230854,
|
| 982 |
+
"learning_rate": 7.508744707511117e-06,
|
| 983 |
+
"loss": 10.3556,
|
| 984 |
+
"step": 137
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.30552096305520965,
|
| 988 |
+
"grad_norm": 0.022224275395274162,
|
| 989 |
+
"learning_rate": 6.70681328282492e-06,
|
| 990 |
+
"loss": 10.3536,
|
| 991 |
+
"step": 138
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.3077348830773488,
|
| 995 |
+
"grad_norm": 0.028994860127568245,
|
| 996 |
+
"learning_rate": 5.948674049707603e-06,
|
| 997 |
+
"loss": 10.3531,
|
| 998 |
+
"step": 139
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.309948803099488,
|
| 1002 |
+
"grad_norm": 0.03301481530070305,
|
| 1003 |
+
"learning_rate": 5.2346828817197655e-06,
|
| 1004 |
+
"loss": 10.3539,
|
| 1005 |
+
"step": 140
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.31216272312162724,
|
| 1009 |
+
"grad_norm": 0.031127754598855972,
|
| 1010 |
+
"learning_rate": 4.565174929145188e-06,
|
| 1011 |
+
"loss": 10.3541,
|
| 1012 |
+
"step": 141
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.31437664314376645,
|
| 1016 |
+
"grad_norm": 0.03324393928050995,
|
| 1017 |
+
"learning_rate": 3.940464461670135e-06,
|
| 1018 |
+
"loss": 10.3558,
|
| 1019 |
+
"step": 142
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.3165905631659056,
|
| 1023 |
+
"grad_norm": 0.025115065276622772,
|
| 1024 |
+
"learning_rate": 3.360844720863765e-06,
|
| 1025 |
+
"loss": 10.3528,
|
| 1026 |
+
"step": 143
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.31880448318804483,
|
| 1030 |
+
"grad_norm": 0.02497878670692444,
|
| 1031 |
+
"learning_rate": 2.826587782529444e-06,
|
| 1032 |
+
"loss": 10.353,
|
| 1033 |
+
"step": 144
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.32101840321018404,
|
| 1037 |
+
"grad_norm": 0.026006096974015236,
|
| 1038 |
+
"learning_rate": 2.3379444289913342e-06,
|
| 1039 |
+
"loss": 10.3548,
|
| 1040 |
+
"step": 145
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.32323232323232326,
|
| 1044 |
+
"grad_norm": 0.028647800907492638,
|
| 1045 |
+
"learning_rate": 1.8951440313760837e-06,
|
| 1046 |
+
"loss": 10.355,
|
| 1047 |
+
"step": 146
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.3254462432544624,
|
| 1051 |
+
"grad_norm": 0.03259943798184395,
|
| 1052 |
+
"learning_rate": 1.4983944419451613e-06,
|
| 1053 |
+
"loss": 10.3541,
|
| 1054 |
+
"step": 147
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.32766016327660163,
|
| 1058 |
+
"grad_norm": 0.033153582364320755,
|
| 1059 |
+
"learning_rate": 1.1478818965281911e-06,
|
| 1060 |
+
"loss": 10.3529,
|
| 1061 |
+
"step": 148
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 0.32987408329874085,
|
| 1065 |
+
"grad_norm": 0.026517199352383614,
|
| 1066 |
+
"learning_rate": 8.437709271030603e-07,
|
| 1067 |
+
"loss": 10.3537,
|
| 1068 |
+
"step": 149
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.33208800332088,
|
| 1072 |
+
"grad_norm": 0.02658323012292385,
|
| 1073 |
+
"learning_rate": 5.862042845640403e-07,
|
| 1074 |
+
"loss": 10.3527,
|
| 1075 |
+
"step": 150
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.3343019233430192,
|
| 1079 |
+
"grad_norm": 0.024668825790286064,
|
| 1080 |
+
"learning_rate": 3.7530287171387843e-07,
|
| 1081 |
+
"loss": 10.3539,
|
| 1082 |
+
"step": 151
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.33651584336515844,
|
| 1086 |
+
"grad_norm": 0.030534988269209862,
|
| 1087 |
+
"learning_rate": 2.1116568651156076e-07,
|
| 1088 |
+
"loss": 10.3545,
|
| 1089 |
+
"step": 152
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.33872976338729766,
|
| 1093 |
+
"grad_norm": 0.024195190519094467,
|
| 1094 |
+
"learning_rate": 9.386977560232879e-08,
|
| 1095 |
+
"loss": 10.3539,
|
| 1096 |
+
"step": 153
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.3409436834094368,
|
| 1100 |
+
"grad_norm": 0.027664266526699066,
|
| 1101 |
+
"learning_rate": 2.347019815158724e-08,
|
| 1102 |
+
"loss": 10.3528,
|
| 1103 |
+
"step": 154
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.34315760343157603,
|
| 1107 |
+
"grad_norm": 0.03868336230516434,
|
| 1108 |
+
"learning_rate": 0.0,
|
| 1109 |
+
"loss": 10.3539,
|
| 1110 |
+
"step": 155
|
| 1111 |
}
|
| 1112 |
],
|
| 1113 |
"logging_steps": 1,
|
|
|
|
| 1131 |
"should_evaluate": false,
|
| 1132 |
"should_log": false,
|
| 1133 |
"should_save": true,
|
| 1134 |
+
"should_training_stop": true
|
| 1135 |
},
|
| 1136 |
"attributes": {}
|
| 1137 |
}
|
| 1138 |
},
|
| 1139 |
+
"total_flos": 64855033774080.0,
|
| 1140 |
"train_batch_size": 4,
|
| 1141 |
"trial_name": null,
|
| 1142 |
"trial_params": null
|