Upload Step 150/4099
Browse files- adapter_model.safetensors +1 -1
- trainer_state.json +393 -3
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 193637776
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d26843ab8a1a0b401b682a499f810279290ca09ca1df8f16170e4cf3676cd1a5
|
| 3 |
size 193637776
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 10,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -796,6 +796,396 @@
|
|
| 796 |
"eval_samples_per_second": 1.622,
|
| 797 |
"eval_steps_per_second": 0.811,
|
| 798 |
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 799 |
}
|
| 800 |
],
|
| 801 |
"logging_steps": 1,
|
|
@@ -815,7 +1205,7 @@
|
|
| 815 |
"attributes": {}
|
| 816 |
}
|
| 817 |
},
|
| 818 |
-
"total_flos":
|
| 819 |
"train_batch_size": 2,
|
| 820 |
"trial_name": null,
|
| 821 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.036594291290558675,
|
| 6 |
"eval_steps": 10,
|
| 7 |
+
"global_step": 150,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 796 |
"eval_samples_per_second": 1.622,
|
| 797 |
"eval_steps_per_second": 0.811,
|
| 798 |
"step": 100
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.02464015613564284,
|
| 802 |
+
"grad_norm": 0.5069964528083801,
|
| 803 |
+
"learning_rate": 2.494015244130635e-07,
|
| 804 |
+
"loss": 1.8013,
|
| 805 |
+
"step": 101
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.024884118077579898,
|
| 809 |
+
"grad_norm": 0.7139447927474976,
|
| 810 |
+
"learning_rate": 2.493954631663007e-07,
|
| 811 |
+
"loss": 1.8216,
|
| 812 |
+
"step": 102
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.025128080019516955,
|
| 816 |
+
"grad_norm": 0.48631080985069275,
|
| 817 |
+
"learning_rate": 2.493893991962659e-07,
|
| 818 |
+
"loss": 1.9325,
|
| 819 |
+
"step": 103
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.02537204196145401,
|
| 823 |
+
"grad_norm": 0.5576779842376709,
|
| 824 |
+
"learning_rate": 2.493833325011235e-07,
|
| 825 |
+
"loss": 2.0052,
|
| 826 |
+
"step": 104
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.025616003903391072,
|
| 830 |
+
"grad_norm": 0.6407865285873413,
|
| 831 |
+
"learning_rate": 2.4937726307903606e-07,
|
| 832 |
+
"loss": 1.9411,
|
| 833 |
+
"step": 105
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.02585996584532813,
|
| 837 |
+
"grad_norm": 0.7654765248298645,
|
| 838 |
+
"learning_rate": 2.493711909281646e-07,
|
| 839 |
+
"loss": 1.9438,
|
| 840 |
+
"step": 106
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 0.026103927787265186,
|
| 844 |
+
"grad_norm": 1.2607905864715576,
|
| 845 |
+
"learning_rate": 2.493651160466685e-07,
|
| 846 |
+
"loss": 2.0134,
|
| 847 |
+
"step": 107
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 0.026347889729202243,
|
| 851 |
+
"grad_norm": 0.8633036017417908,
|
| 852 |
+
"learning_rate": 2.493590384327053e-07,
|
| 853 |
+
"loss": 1.9775,
|
| 854 |
+
"step": 108
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.026591851671139304,
|
| 858 |
+
"grad_norm": 0.7568155527114868,
|
| 859 |
+
"learning_rate": 2.49352958084431e-07,
|
| 860 |
+
"loss": 1.9074,
|
| 861 |
+
"step": 109
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.02683581361307636,
|
| 865 |
+
"grad_norm": 0.5505961179733276,
|
| 866 |
+
"learning_rate": 2.49346875e-07,
|
| 867 |
+
"loss": 1.8467,
|
| 868 |
+
"step": 110
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 0.02683581361307636,
|
| 872 |
+
"eval_loss": 1.8828259706497192,
|
| 873 |
+
"eval_runtime": 158.4116,
|
| 874 |
+
"eval_samples_per_second": 1.616,
|
| 875 |
+
"eval_steps_per_second": 0.808,
|
| 876 |
+
"step": 110
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 0.027079775555013418,
|
| 880 |
+
"grad_norm": 0.5095446109771729,
|
| 881 |
+
"learning_rate": 2.49340789177565e-07,
|
| 882 |
+
"loss": 1.9961,
|
| 883 |
+
"step": 111
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 0.027323737496950475,
|
| 887 |
+
"grad_norm": 1.7097959518432617,
|
| 888 |
+
"learning_rate": 2.4933470061527687e-07,
|
| 889 |
+
"loss": 1.9335,
|
| 890 |
+
"step": 112
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.027567699438887535,
|
| 894 |
+
"grad_norm": 1.0115768909454346,
|
| 895 |
+
"learning_rate": 2.493286093112851e-07,
|
| 896 |
+
"loss": 1.8118,
|
| 897 |
+
"step": 113
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.027811661380824592,
|
| 901 |
+
"grad_norm": 0.6412175297737122,
|
| 902 |
+
"learning_rate": 2.493225152637374e-07,
|
| 903 |
+
"loss": 1.9623,
|
| 904 |
+
"step": 114
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 0.02805562332276165,
|
| 908 |
+
"grad_norm": 0.5357053875923157,
|
| 909 |
+
"learning_rate": 2.4931641847077963e-07,
|
| 910 |
+
"loss": 1.8131,
|
| 911 |
+
"step": 115
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.028299585264698706,
|
| 915 |
+
"grad_norm": 0.6828150153160095,
|
| 916 |
+
"learning_rate": 2.493103189305562e-07,
|
| 917 |
+
"loss": 1.767,
|
| 918 |
+
"step": 116
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.028543547206635766,
|
| 922 |
+
"grad_norm": 0.5804136395454407,
|
| 923 |
+
"learning_rate": 2.493042166412099e-07,
|
| 924 |
+
"loss": 1.9831,
|
| 925 |
+
"step": 117
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.028787509148572824,
|
| 929 |
+
"grad_norm": 0.6375969052314758,
|
| 930 |
+
"learning_rate": 2.492981116008816e-07,
|
| 931 |
+
"loss": 1.9651,
|
| 932 |
+
"step": 118
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.02903147109050988,
|
| 936 |
+
"grad_norm": 0.6621755957603455,
|
| 937 |
+
"learning_rate": 2.492920038077106e-07,
|
| 938 |
+
"loss": 2.1064,
|
| 939 |
+
"step": 119
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.029275433032446938,
|
| 943 |
+
"grad_norm": 0.7436494827270508,
|
| 944 |
+
"learning_rate": 2.492858932598346e-07,
|
| 945 |
+
"loss": 1.8961,
|
| 946 |
+
"step": 120
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 0.029275433032446938,
|
| 950 |
+
"eval_loss": 1.8782259225845337,
|
| 951 |
+
"eval_runtime": 158.1634,
|
| 952 |
+
"eval_samples_per_second": 1.619,
|
| 953 |
+
"eval_steps_per_second": 0.809,
|
| 954 |
+
"step": 120
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.029519394974383995,
|
| 958 |
+
"grad_norm": 0.5152058005332947,
|
| 959 |
+
"learning_rate": 2.4927977995538954e-07,
|
| 960 |
+
"loss": 1.875,
|
| 961 |
+
"step": 121
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.029763356916321055,
|
| 965 |
+
"grad_norm": 0.4640464782714844,
|
| 966 |
+
"learning_rate": 2.4927366389250973e-07,
|
| 967 |
+
"loss": 1.8429,
|
| 968 |
+
"step": 122
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.030007318858258112,
|
| 972 |
+
"grad_norm": 0.6126062273979187,
|
| 973 |
+
"learning_rate": 2.4926754506932774e-07,
|
| 974 |
+
"loss": 1.9581,
|
| 975 |
+
"step": 123
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.03025128080019517,
|
| 979 |
+
"grad_norm": 0.5338674187660217,
|
| 980 |
+
"learning_rate": 2.4926142348397453e-07,
|
| 981 |
+
"loss": 1.9682,
|
| 982 |
+
"step": 124
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.030495242742132226,
|
| 986 |
+
"grad_norm": 0.48220378160476685,
|
| 987 |
+
"learning_rate": 2.492552991345792e-07,
|
| 988 |
+
"loss": 1.9316,
|
| 989 |
+
"step": 125
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.030739204684069286,
|
| 993 |
+
"grad_norm": 1.0571016073226929,
|
| 994 |
+
"learning_rate": 2.4924917201926936e-07,
|
| 995 |
+
"loss": 1.9837,
|
| 996 |
+
"step": 126
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.030983166626006343,
|
| 1000 |
+
"grad_norm": 0.5729621052742004,
|
| 1001 |
+
"learning_rate": 2.492430421361708e-07,
|
| 1002 |
+
"loss": 1.7242,
|
| 1003 |
+
"step": 127
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.0312271285679434,
|
| 1007 |
+
"grad_norm": 0.9092426896095276,
|
| 1008 |
+
"learning_rate": 2.4923690948340783e-07,
|
| 1009 |
+
"loss": 1.8327,
|
| 1010 |
+
"step": 128
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.03147109050988046,
|
| 1014 |
+
"grad_norm": 0.44636791944503784,
|
| 1015 |
+
"learning_rate": 2.4923077405910264e-07,
|
| 1016 |
+
"loss": 2.0464,
|
| 1017 |
+
"step": 129
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.031715052451817514,
|
| 1021 |
+
"grad_norm": 0.6733670830726624,
|
| 1022 |
+
"learning_rate": 2.4922463586137616e-07,
|
| 1023 |
+
"loss": 1.8564,
|
| 1024 |
+
"step": 130
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.031715052451817514,
|
| 1028 |
+
"eval_loss": 1.873685359954834,
|
| 1029 |
+
"eval_runtime": 158.2193,
|
| 1030 |
+
"eval_samples_per_second": 1.618,
|
| 1031 |
+
"eval_steps_per_second": 0.809,
|
| 1032 |
+
"step": 130
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.03195901439375457,
|
| 1036 |
+
"grad_norm": 0.6245723366737366,
|
| 1037 |
+
"learning_rate": 2.4921849488834745e-07,
|
| 1038 |
+
"loss": 2.0072,
|
| 1039 |
+
"step": 131
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.032202976335691635,
|
| 1043 |
+
"grad_norm": 0.47369739413261414,
|
| 1044 |
+
"learning_rate": 2.4921235113813376e-07,
|
| 1045 |
+
"loss": 2.0033,
|
| 1046 |
+
"step": 132
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.03244693827762869,
|
| 1050 |
+
"grad_norm": 0.6961667537689209,
|
| 1051 |
+
"learning_rate": 2.492062046088508e-07,
|
| 1052 |
+
"loss": 1.8175,
|
| 1053 |
+
"step": 133
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.03269090021956575,
|
| 1057 |
+
"grad_norm": 0.7953224182128906,
|
| 1058 |
+
"learning_rate": 2.4920005529861254e-07,
|
| 1059 |
+
"loss": 1.8035,
|
| 1060 |
+
"step": 134
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.032934862161502806,
|
| 1064 |
+
"grad_norm": 0.516058087348938,
|
| 1065 |
+
"learning_rate": 2.491939032055311e-07,
|
| 1066 |
+
"loss": 1.8855,
|
| 1067 |
+
"step": 135
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.03317882410343986,
|
| 1071 |
+
"grad_norm": 0.6488027572631836,
|
| 1072 |
+
"learning_rate": 2.491877483277171e-07,
|
| 1073 |
+
"loss": 1.9622,
|
| 1074 |
+
"step": 136
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.03342278604537692,
|
| 1078 |
+
"grad_norm": 0.6827359199523926,
|
| 1079 |
+
"learning_rate": 2.4918159066327943e-07,
|
| 1080 |
+
"loss": 1.847,
|
| 1081 |
+
"step": 137
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.03366674798731398,
|
| 1085 |
+
"grad_norm": 0.4918162226676941,
|
| 1086 |
+
"learning_rate": 2.49175430210325e-07,
|
| 1087 |
+
"loss": 1.9214,
|
| 1088 |
+
"step": 138
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.033910709929251034,
|
| 1092 |
+
"grad_norm": 0.7824620008468628,
|
| 1093 |
+
"learning_rate": 2.491692669669594e-07,
|
| 1094 |
+
"loss": 1.8472,
|
| 1095 |
+
"step": 139
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.03415467187118809,
|
| 1099 |
+
"grad_norm": 0.7084971070289612,
|
| 1100 |
+
"learning_rate": 2.4916310093128616e-07,
|
| 1101 |
+
"loss": 1.8638,
|
| 1102 |
+
"step": 140
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.03415467187118809,
|
| 1106 |
+
"eval_loss": 1.869973063468933,
|
| 1107 |
+
"eval_runtime": 157.6522,
|
| 1108 |
+
"eval_samples_per_second": 1.624,
|
| 1109 |
+
"eval_steps_per_second": 0.812,
|
| 1110 |
+
"step": 140
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 0.034398633813125155,
|
| 1114 |
+
"grad_norm": 0.4873005747795105,
|
| 1115 |
+
"learning_rate": 2.491569321014073e-07,
|
| 1116 |
+
"loss": 1.9326,
|
| 1117 |
+
"step": 141
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 0.03464259575506221,
|
| 1121 |
+
"grad_norm": 0.6483212113380432,
|
| 1122 |
+
"learning_rate": 2.49150760475423e-07,
|
| 1123 |
+
"loss": 1.9035,
|
| 1124 |
+
"step": 142
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.03488655769699927,
|
| 1128 |
+
"grad_norm": 0.46081703901290894,
|
| 1129 |
+
"learning_rate": 2.4914458605143187e-07,
|
| 1130 |
+
"loss": 1.9746,
|
| 1131 |
+
"step": 143
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.035130519638936326,
|
| 1135 |
+
"grad_norm": 0.683131754398346,
|
| 1136 |
+
"learning_rate": 2.491384088275306e-07,
|
| 1137 |
+
"loss": 1.8517,
|
| 1138 |
+
"step": 144
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.03537448158087338,
|
| 1142 |
+
"grad_norm": 0.4871167242527008,
|
| 1143 |
+
"learning_rate": 2.491322288018143e-07,
|
| 1144 |
+
"loss": 1.7198,
|
| 1145 |
+
"step": 145
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.03561844352281044,
|
| 1149 |
+
"grad_norm": 0.6227270364761353,
|
| 1150 |
+
"learning_rate": 2.4912604597237626e-07,
|
| 1151 |
+
"loss": 1.8555,
|
| 1152 |
+
"step": 146
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 0.0358624054647475,
|
| 1156 |
+
"grad_norm": 0.5372536182403564,
|
| 1157 |
+
"learning_rate": 2.4911986033730807e-07,
|
| 1158 |
+
"loss": 1.8245,
|
| 1159 |
+
"step": 147
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 0.036106367406684554,
|
| 1163 |
+
"grad_norm": 0.7428392171859741,
|
| 1164 |
+
"learning_rate": 2.491136718946997e-07,
|
| 1165 |
+
"loss": 2.0657,
|
| 1166 |
+
"step": 148
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.03635032934862162,
|
| 1170 |
+
"grad_norm": 0.9103279709815979,
|
| 1171 |
+
"learning_rate": 2.4910748064263914e-07,
|
| 1172 |
+
"loss": 1.9042,
|
| 1173 |
+
"step": 149
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.036594291290558675,
|
| 1177 |
+
"grad_norm": 1.1896861791610718,
|
| 1178 |
+
"learning_rate": 2.491012865792129e-07,
|
| 1179 |
+
"loss": 1.8883,
|
| 1180 |
+
"step": 150
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 0.036594291290558675,
|
| 1184 |
+
"eval_loss": 1.86661696434021,
|
| 1185 |
+
"eval_runtime": 158.3624,
|
| 1186 |
+
"eval_samples_per_second": 1.617,
|
| 1187 |
+
"eval_steps_per_second": 0.808,
|
| 1188 |
+
"step": 150
|
| 1189 |
}
|
| 1190 |
],
|
| 1191 |
"logging_steps": 1,
|
|
|
|
| 1205 |
"attributes": {}
|
| 1206 |
}
|
| 1207 |
},
|
| 1208 |
+
"total_flos": 3.859442244255744e+17,
|
| 1209 |
"train_batch_size": 2,
|
| 1210 |
"trial_name": null,
|
| 1211 |
"trial_params": null
|