loss 1.1
Browse files- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +543 -3
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1585984129
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b103cf3481b1442713482cd3f0abee34eb33e5c137d46fe5b7cc1e76c8cdc2dc
|
| 3 |
size 1585984129
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1561347007
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28bcb2b4c3cf5d47f95b0f7492cf068fc8a87a6d6e036c7f54376f6bb96b4fba
|
| 3 |
size 1561347007
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fc73e9a0017399222f8b738572d466ab73752e0ba8499f5e427cacf49cf26963
|
| 3 |
size 14503
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:913b5468f28d3c2ff51c474c3dfe5ccf74915831df26f09fc32ab01cf4f5196b
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -846,11 +846,551 @@
|
|
| 846 |
"learning_rate": 7.519929140832596e-06,
|
| 847 |
"loss": 1.1611,
|
| 848 |
"step": 14000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 849 |
}
|
| 850 |
],
|
| 851 |
"max_steps": 56450,
|
| 852 |
"num_train_epochs": 10,
|
| 853 |
-
"total_flos":
|
| 854 |
"trial_name": null,
|
| 855 |
"trial_params": null
|
| 856 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.074398830875515,
|
| 5 |
+
"global_step": 23000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 846 |
"learning_rate": 7.519929140832596e-06,
|
| 847 |
"loss": 1.1611,
|
| 848 |
"step": 14000
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 2.5,
|
| 852 |
+
"learning_rate": 7.5022143489814e-06,
|
| 853 |
+
"loss": 1.1703,
|
| 854 |
+
"step": 14100
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 2.52,
|
| 858 |
+
"learning_rate": 7.484499557130205e-06,
|
| 859 |
+
"loss": 1.1642,
|
| 860 |
+
"step": 14200
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.53,
|
| 864 |
+
"learning_rate": 7.466784765279008e-06,
|
| 865 |
+
"loss": 1.1285,
|
| 866 |
+
"step": 14300
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.55,
|
| 870 |
+
"learning_rate": 7.449069973427812e-06,
|
| 871 |
+
"loss": 1.1627,
|
| 872 |
+
"step": 14400
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 2.57,
|
| 876 |
+
"learning_rate": 7.431355181576617e-06,
|
| 877 |
+
"loss": 1.1667,
|
| 878 |
+
"step": 14500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.59,
|
| 882 |
+
"learning_rate": 7.413640389725421e-06,
|
| 883 |
+
"loss": 1.162,
|
| 884 |
+
"step": 14600
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.6,
|
| 888 |
+
"learning_rate": 7.3959255978742254e-06,
|
| 889 |
+
"loss": 1.1835,
|
| 890 |
+
"step": 14700
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 2.62,
|
| 894 |
+
"learning_rate": 7.3782108060230295e-06,
|
| 895 |
+
"loss": 1.1697,
|
| 896 |
+
"step": 14800
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 2.64,
|
| 900 |
+
"learning_rate": 7.3604960141718344e-06,
|
| 901 |
+
"loss": 1.18,
|
| 902 |
+
"step": 14900
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.66,
|
| 906 |
+
"learning_rate": 7.3427812223206385e-06,
|
| 907 |
+
"loss": 1.1594,
|
| 908 |
+
"step": 15000
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.67,
|
| 912 |
+
"learning_rate": 7.325066430469443e-06,
|
| 913 |
+
"loss": 1.165,
|
| 914 |
+
"step": 15100
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 2.69,
|
| 918 |
+
"learning_rate": 7.307351638618247e-06,
|
| 919 |
+
"loss": 1.1598,
|
| 920 |
+
"step": 15200
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.71,
|
| 924 |
+
"learning_rate": 7.289636846767051e-06,
|
| 925 |
+
"loss": 1.1412,
|
| 926 |
+
"step": 15300
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.73,
|
| 930 |
+
"learning_rate": 7.271922054915856e-06,
|
| 931 |
+
"loss": 1.1745,
|
| 932 |
+
"step": 15400
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 2.75,
|
| 936 |
+
"learning_rate": 7.25420726306466e-06,
|
| 937 |
+
"loss": 1.1961,
|
| 938 |
+
"step": 15500
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 2.76,
|
| 942 |
+
"learning_rate": 7.2364924712134646e-06,
|
| 943 |
+
"loss": 1.1557,
|
| 944 |
+
"step": 15600
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.78,
|
| 948 |
+
"learning_rate": 7.218777679362268e-06,
|
| 949 |
+
"loss": 1.171,
|
| 950 |
+
"step": 15700
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 2.8,
|
| 954 |
+
"learning_rate": 7.201062887511072e-06,
|
| 955 |
+
"loss": 1.146,
|
| 956 |
+
"step": 15800
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 2.82,
|
| 960 |
+
"learning_rate": 7.183348095659877e-06,
|
| 961 |
+
"loss": 1.157,
|
| 962 |
+
"step": 15900
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.83,
|
| 966 |
+
"learning_rate": 7.165633303808681e-06,
|
| 967 |
+
"loss": 1.1622,
|
| 968 |
+
"step": 16000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.85,
|
| 972 |
+
"learning_rate": 7.147918511957485e-06,
|
| 973 |
+
"loss": 1.1661,
|
| 974 |
+
"step": 16100
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 2.87,
|
| 978 |
+
"learning_rate": 7.130203720106289e-06,
|
| 979 |
+
"loss": 1.1694,
|
| 980 |
+
"step": 16200
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 2.89,
|
| 984 |
+
"learning_rate": 7.112488928255094e-06,
|
| 985 |
+
"loss": 1.1689,
|
| 986 |
+
"step": 16300
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.91,
|
| 990 |
+
"learning_rate": 7.094774136403898e-06,
|
| 991 |
+
"loss": 1.1394,
|
| 992 |
+
"step": 16400
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 2.92,
|
| 996 |
+
"learning_rate": 7.077059344552703e-06,
|
| 997 |
+
"loss": 1.137,
|
| 998 |
+
"step": 16500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.94,
|
| 1002 |
+
"learning_rate": 7.059344552701506e-06,
|
| 1003 |
+
"loss": 1.1639,
|
| 1004 |
+
"step": 16600
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.96,
|
| 1008 |
+
"learning_rate": 7.04162976085031e-06,
|
| 1009 |
+
"loss": 1.1545,
|
| 1010 |
+
"step": 16700
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.98,
|
| 1014 |
+
"learning_rate": 7.023914968999115e-06,
|
| 1015 |
+
"loss": 1.1742,
|
| 1016 |
+
"step": 16800
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 2.99,
|
| 1020 |
+
"learning_rate": 7.006200177147919e-06,
|
| 1021 |
+
"loss": 1.161,
|
| 1022 |
+
"step": 16900
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 3.01,
|
| 1026 |
+
"learning_rate": 6.988485385296724e-06,
|
| 1027 |
+
"loss": 1.1506,
|
| 1028 |
+
"step": 17000
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 3.03,
|
| 1032 |
+
"learning_rate": 6.970770593445527e-06,
|
| 1033 |
+
"loss": 1.154,
|
| 1034 |
+
"step": 17100
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 3.05,
|
| 1038 |
+
"learning_rate": 6.953055801594331e-06,
|
| 1039 |
+
"loss": 1.1414,
|
| 1040 |
+
"step": 17200
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 3.06,
|
| 1044 |
+
"learning_rate": 6.935341009743136e-06,
|
| 1045 |
+
"loss": 1.1289,
|
| 1046 |
+
"step": 17300
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 3.08,
|
| 1050 |
+
"learning_rate": 6.91762621789194e-06,
|
| 1051 |
+
"loss": 1.1664,
|
| 1052 |
+
"step": 17400
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 3.1,
|
| 1056 |
+
"learning_rate": 6.899911426040744e-06,
|
| 1057 |
+
"loss": 1.1751,
|
| 1058 |
+
"step": 17500
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 3.12,
|
| 1062 |
+
"learning_rate": 6.882196634189548e-06,
|
| 1063 |
+
"loss": 1.1514,
|
| 1064 |
+
"step": 17600
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 3.14,
|
| 1068 |
+
"learning_rate": 6.864481842338353e-06,
|
| 1069 |
+
"loss": 1.1219,
|
| 1070 |
+
"step": 17700
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 3.15,
|
| 1074 |
+
"learning_rate": 6.846767050487157e-06,
|
| 1075 |
+
"loss": 1.1343,
|
| 1076 |
+
"step": 17800
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 3.17,
|
| 1080 |
+
"learning_rate": 6.829052258635962e-06,
|
| 1081 |
+
"loss": 1.1209,
|
| 1082 |
+
"step": 17900
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 3.19,
|
| 1086 |
+
"learning_rate": 6.8113374667847655e-06,
|
| 1087 |
+
"loss": 1.1612,
|
| 1088 |
+
"step": 18000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.21,
|
| 1092 |
+
"learning_rate": 6.79362267493357e-06,
|
| 1093 |
+
"loss": 1.1427,
|
| 1094 |
+
"step": 18100
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 3.22,
|
| 1098 |
+
"learning_rate": 6.7759078830823745e-06,
|
| 1099 |
+
"loss": 1.1621,
|
| 1100 |
+
"step": 18200
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 3.24,
|
| 1104 |
+
"learning_rate": 6.7581930912311786e-06,
|
| 1105 |
+
"loss": 1.1434,
|
| 1106 |
+
"step": 18300
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 3.26,
|
| 1110 |
+
"learning_rate": 6.7404782993799835e-06,
|
| 1111 |
+
"loss": 1.1324,
|
| 1112 |
+
"step": 18400
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 3.28,
|
| 1116 |
+
"learning_rate": 6.722763507528787e-06,
|
| 1117 |
+
"loss": 1.1566,
|
| 1118 |
+
"step": 18500
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 3.29,
|
| 1122 |
+
"learning_rate": 6.705048715677592e-06,
|
| 1123 |
+
"loss": 1.1393,
|
| 1124 |
+
"step": 18600
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 3.31,
|
| 1128 |
+
"learning_rate": 6.687333923826396e-06,
|
| 1129 |
+
"loss": 1.1642,
|
| 1130 |
+
"step": 18700
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 3.33,
|
| 1134 |
+
"learning_rate": 6.6696191319752006e-06,
|
| 1135 |
+
"loss": 1.1419,
|
| 1136 |
+
"step": 18800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 3.35,
|
| 1140 |
+
"learning_rate": 6.651904340124004e-06,
|
| 1141 |
+
"loss": 1.1293,
|
| 1142 |
+
"step": 18900
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 3.37,
|
| 1146 |
+
"learning_rate": 6.634189548272808e-06,
|
| 1147 |
+
"loss": 1.1467,
|
| 1148 |
+
"step": 19000
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 3.38,
|
| 1152 |
+
"learning_rate": 6.616474756421613e-06,
|
| 1153 |
+
"loss": 1.1619,
|
| 1154 |
+
"step": 19100
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.4,
|
| 1158 |
+
"learning_rate": 6.598759964570417e-06,
|
| 1159 |
+
"loss": 1.1566,
|
| 1160 |
+
"step": 19200
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 3.42,
|
| 1164 |
+
"learning_rate": 6.581045172719222e-06,
|
| 1165 |
+
"loss": 1.1741,
|
| 1166 |
+
"step": 19300
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 3.44,
|
| 1170 |
+
"learning_rate": 6.563330380868025e-06,
|
| 1171 |
+
"loss": 1.1763,
|
| 1172 |
+
"step": 19400
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 3.45,
|
| 1176 |
+
"learning_rate": 6.545615589016829e-06,
|
| 1177 |
+
"loss": 1.1628,
|
| 1178 |
+
"step": 19500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 3.47,
|
| 1182 |
+
"learning_rate": 6.527900797165634e-06,
|
| 1183 |
+
"loss": 1.1585,
|
| 1184 |
+
"step": 19600
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 3.49,
|
| 1188 |
+
"learning_rate": 6.510186005314438e-06,
|
| 1189 |
+
"loss": 1.1546,
|
| 1190 |
+
"step": 19700
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 3.51,
|
| 1194 |
+
"learning_rate": 6.492471213463243e-06,
|
| 1195 |
+
"loss": 1.1684,
|
| 1196 |
+
"step": 19800
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 3.53,
|
| 1200 |
+
"learning_rate": 6.474756421612046e-06,
|
| 1201 |
+
"loss": 1.1274,
|
| 1202 |
+
"step": 19900
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 3.54,
|
| 1206 |
+
"learning_rate": 6.457041629760851e-06,
|
| 1207 |
+
"loss": 1.1432,
|
| 1208 |
+
"step": 20000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 3.56,
|
| 1212 |
+
"learning_rate": 6.439326837909655e-06,
|
| 1213 |
+
"loss": 1.1391,
|
| 1214 |
+
"step": 20100
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 3.58,
|
| 1218 |
+
"learning_rate": 6.42161204605846e-06,
|
| 1219 |
+
"loss": 1.154,
|
| 1220 |
+
"step": 20200
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 3.6,
|
| 1224 |
+
"learning_rate": 6.403897254207263e-06,
|
| 1225 |
+
"loss": 1.128,
|
| 1226 |
+
"step": 20300
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 3.61,
|
| 1230 |
+
"learning_rate": 6.386182462356067e-06,
|
| 1231 |
+
"loss": 1.1623,
|
| 1232 |
+
"step": 20400
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 3.63,
|
| 1236 |
+
"learning_rate": 6.368467670504872e-06,
|
| 1237 |
+
"loss": 1.1661,
|
| 1238 |
+
"step": 20500
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 3.65,
|
| 1242 |
+
"learning_rate": 6.350752878653676e-06,
|
| 1243 |
+
"loss": 1.1328,
|
| 1244 |
+
"step": 20600
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 3.67,
|
| 1248 |
+
"learning_rate": 6.333038086802481e-06,
|
| 1249 |
+
"loss": 1.1457,
|
| 1250 |
+
"step": 20700
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 3.68,
|
| 1254 |
+
"learning_rate": 6.315323294951284e-06,
|
| 1255 |
+
"loss": 1.1347,
|
| 1256 |
+
"step": 20800
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 3.7,
|
| 1260 |
+
"learning_rate": 6.2976085031000885e-06,
|
| 1261 |
+
"loss": 1.1434,
|
| 1262 |
+
"step": 20900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 3.72,
|
| 1266 |
+
"learning_rate": 6.279893711248893e-06,
|
| 1267 |
+
"loss": 1.1342,
|
| 1268 |
+
"step": 21000
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 3.74,
|
| 1272 |
+
"learning_rate": 6.2621789193976975e-06,
|
| 1273 |
+
"loss": 1.1377,
|
| 1274 |
+
"step": 21100
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 3.76,
|
| 1278 |
+
"learning_rate": 6.244464127546502e-06,
|
| 1279 |
+
"loss": 1.1321,
|
| 1280 |
+
"step": 21200
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 3.77,
|
| 1284 |
+
"learning_rate": 6.226749335695306e-06,
|
| 1285 |
+
"loss": 1.138,
|
| 1286 |
+
"step": 21300
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 3.79,
|
| 1290 |
+
"learning_rate": 6.2090345438441105e-06,
|
| 1291 |
+
"loss": 1.1458,
|
| 1292 |
+
"step": 21400
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 3.81,
|
| 1296 |
+
"learning_rate": 6.1913197519929146e-06,
|
| 1297 |
+
"loss": 1.1134,
|
| 1298 |
+
"step": 21500
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 3.83,
|
| 1302 |
+
"learning_rate": 6.1736049601417195e-06,
|
| 1303 |
+
"loss": 1.1179,
|
| 1304 |
+
"step": 21600
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 3.84,
|
| 1308 |
+
"learning_rate": 6.155890168290523e-06,
|
| 1309 |
+
"loss": 1.1396,
|
| 1310 |
+
"step": 21700
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 3.86,
|
| 1314 |
+
"learning_rate": 6.138175376439327e-06,
|
| 1315 |
+
"loss": 1.122,
|
| 1316 |
+
"step": 21800
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 3.88,
|
| 1320 |
+
"learning_rate": 6.120460584588132e-06,
|
| 1321 |
+
"loss": 1.1576,
|
| 1322 |
+
"step": 21900
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 3.9,
|
| 1326 |
+
"learning_rate": 6.102745792736936e-06,
|
| 1327 |
+
"loss": 1.1513,
|
| 1328 |
+
"step": 22000
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 3.91,
|
| 1332 |
+
"learning_rate": 6.085031000885741e-06,
|
| 1333 |
+
"loss": 1.14,
|
| 1334 |
+
"step": 22100
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 3.93,
|
| 1338 |
+
"learning_rate": 6.067316209034544e-06,
|
| 1339 |
+
"loss": 1.1657,
|
| 1340 |
+
"step": 22200
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 3.95,
|
| 1344 |
+
"learning_rate": 6.049601417183349e-06,
|
| 1345 |
+
"loss": 1.1452,
|
| 1346 |
+
"step": 22300
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 3.97,
|
| 1350 |
+
"learning_rate": 6.031886625332153e-06,
|
| 1351 |
+
"loss": 1.1379,
|
| 1352 |
+
"step": 22400
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.99,
|
| 1356 |
+
"learning_rate": 6.014171833480958e-06,
|
| 1357 |
+
"loss": 1.1167,
|
| 1358 |
+
"step": 22500
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 4.0,
|
| 1362 |
+
"learning_rate": 5.996457041629762e-06,
|
| 1363 |
+
"loss": 1.1221,
|
| 1364 |
+
"step": 22600
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 4.02,
|
| 1368 |
+
"learning_rate": 5.978742249778565e-06,
|
| 1369 |
+
"loss": 1.1274,
|
| 1370 |
+
"step": 22700
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 4.04,
|
| 1374 |
+
"learning_rate": 5.96102745792737e-06,
|
| 1375 |
+
"loss": 1.1267,
|
| 1376 |
+
"step": 22800
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 4.06,
|
| 1380 |
+
"learning_rate": 5.943312666076174e-06,
|
| 1381 |
+
"loss": 1.1295,
|
| 1382 |
+
"step": 22900
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 4.07,
|
| 1386 |
+
"learning_rate": 5.925597874224979e-06,
|
| 1387 |
+
"loss": 1.1073,
|
| 1388 |
+
"step": 23000
|
| 1389 |
}
|
| 1390 |
],
|
| 1391 |
"max_steps": 56450,
|
| 1392 |
"num_train_epochs": 10,
|
| 1393 |
+
"total_flos": 4.338858046644173e+17,
|
| 1394 |
"trial_name": null,
|
| 1395 |
"trial_params": null
|
| 1396 |
}
|