Training in progress, step 50000
Browse files- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/pytorch_model.bin +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +203 -3
- pytorch_model.bin +1 -1
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 50044241
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:44519022516715e3b4aa9fbebd26a8f16d69cc1037c43939223444845c039b2a
|
3 |
size 50044241
|
last-checkpoint/pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 25761253
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c31ddb9e289da6dc236cf93f3881df915ba1b345fb3bc4108ca78271c87c7807
|
3 |
size 25761253
|
last-checkpoint/rng_state_0.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:e95236c288314cf7e95c084d5233e1be1e022395a42503d56c52a71f4cff32b4
|
3 |
size 14503
|
last-checkpoint/rng_state_1.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:e95236c288314cf7e95c084d5233e1be1e022395a42503d56c52a71f4cff32b4
|
3 |
size 14503
|
last-checkpoint/rng_state_2.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:e95236c288314cf7e95c084d5233e1be1e022395a42503d56c52a71f4cff32b4
|
3 |
size 14503
|
last-checkpoint/rng_state_3.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:e95236c288314cf7e95c084d5233e1be1e022395a42503d56c52a71f4cff32b4
|
3 |
size 14503
|
last-checkpoint/rng_state_4.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:e95236c288314cf7e95c084d5233e1be1e022395a42503d56c52a71f4cff32b4
|
3 |
size 14503
|
last-checkpoint/rng_state_5.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:e95236c288314cf7e95c084d5233e1be1e022395a42503d56c52a71f4cff32b4
|
3 |
size 14503
|
last-checkpoint/rng_state_6.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:e95236c288314cf7e95c084d5233e1be1e022395a42503d56c52a71f4cff32b4
|
3 |
size 14503
|
last-checkpoint/rng_state_7.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:e95236c288314cf7e95c084d5233e1be1e022395a42503d56c52a71f4cff32b4
|
3 |
size 14503
|
last-checkpoint/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:f9511edec0a698219a4fec94630f2bf1ec60055a31bda0393e6aadb5a36db13d
|
3 |
size 623
|
last-checkpoint/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,
|
@@ -806,11 +806,211 @@
|
|
806 |
"eval_samples_per_second": 1057.979,
|
807 |
"eval_steps_per_second": 16.581,
|
808 |
"step": 40000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
809 |
}
|
810 |
],
|
811 |
"max_steps": 250000,
|
812 |
"num_train_epochs": 12,
|
813 |
-
"total_flos":
|
814 |
"trial_name": null,
|
815 |
"trial_params": null
|
816 |
}
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 2.230549607423269,
|
5 |
+
"global_step": 50000,
|
6 |
"is_hyper_param_search": false,
|
7 |
"is_local_process_zero": true,
|
8 |
"is_world_process_zero": true,
|
|
|
806 |
"eval_samples_per_second": 1057.979,
|
807 |
"eval_steps_per_second": 16.581,
|
808 |
"step": 40000
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 1.81,
|
812 |
+
"learning_rate": 0.0005799963048115559,
|
813 |
+
"loss": 0.6757,
|
814 |
+
"step": 40500
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 1.83,
|
818 |
+
"learning_rate": 0.0005792840633370341,
|
819 |
+
"loss": 0.6575,
|
820 |
+
"step": 41000
|
821 |
+
},
|
822 |
+
{
|
823 |
+
"epoch": 1.83,
|
824 |
+
"eval_loss": 0.6465174555778503,
|
825 |
+
"eval_runtime": 2.2243,
|
826 |
+
"eval_samples_per_second": 1032.694,
|
827 |
+
"eval_steps_per_second": 16.185,
|
828 |
+
"step": 41000
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 1.85,
|
832 |
+
"learning_rate": 0.0005785598237890247,
|
833 |
+
"loss": 0.6461,
|
834 |
+
"step": 41500
|
835 |
+
},
|
836 |
+
{
|
837 |
+
"epoch": 1.87,
|
838 |
+
"learning_rate": 0.0005778236178481119,
|
839 |
+
"loss": 0.6373,
|
840 |
+
"step": 42000
|
841 |
+
},
|
842 |
+
{
|
843 |
+
"epoch": 1.87,
|
844 |
+
"eval_loss": 0.631793737411499,
|
845 |
+
"eval_runtime": 2.22,
|
846 |
+
"eval_samples_per_second": 1034.694,
|
847 |
+
"eval_steps_per_second": 16.216,
|
848 |
+
"step": 42000
|
849 |
+
},
|
850 |
+
{
|
851 |
+
"epoch": 1.9,
|
852 |
+
"learning_rate": 0.0005770754777183285,
|
853 |
+
"loss": 0.6309,
|
854 |
+
"step": 42500
|
855 |
+
},
|
856 |
+
{
|
857 |
+
"epoch": 1.92,
|
858 |
+
"learning_rate": 0.0005763154361257473,
|
859 |
+
"loss": 0.6257,
|
860 |
+
"step": 43000
|
861 |
+
},
|
862 |
+
{
|
863 |
+
"epoch": 1.92,
|
864 |
+
"eval_loss": 0.6184197664260864,
|
865 |
+
"eval_runtime": 2.2848,
|
866 |
+
"eval_samples_per_second": 1005.32,
|
867 |
+
"eval_steps_per_second": 15.756,
|
868 |
+
"step": 43000
|
869 |
+
},
|
870 |
+
{
|
871 |
+
"epoch": 1.94,
|
872 |
+
"learning_rate": 0.0005755435263170498,
|
873 |
+
"loss": 0.623,
|
874 |
+
"step": 43500
|
875 |
+
},
|
876 |
+
{
|
877 |
+
"epoch": 1.96,
|
878 |
+
"learning_rate": 0.0005747597820580717,
|
879 |
+
"loss": 0.621,
|
880 |
+
"step": 44000
|
881 |
+
},
|
882 |
+
{
|
883 |
+
"epoch": 1.96,
|
884 |
+
"eval_loss": 0.6136025786399841,
|
885 |
+
"eval_runtime": 2.2706,
|
886 |
+
"eval_samples_per_second": 1011.621,
|
887 |
+
"eval_steps_per_second": 15.855,
|
888 |
+
"step": 44000
|
889 |
+
},
|
890 |
+
{
|
891 |
+
"epoch": 1.99,
|
892 |
+
"learning_rate": 0.000573964237632326,
|
893 |
+
"loss": 0.6195,
|
894 |
+
"step": 44500
|
895 |
+
},
|
896 |
+
{
|
897 |
+
"epoch": 2.01,
|
898 |
+
"learning_rate": 0.0005731569278395029,
|
899 |
+
"loss": 0.6183,
|
900 |
+
"step": 45000
|
901 |
+
},
|
902 |
+
{
|
903 |
+
"epoch": 2.01,
|
904 |
+
"eval_loss": 0.6127315163612366,
|
905 |
+
"eval_runtime": 2.1317,
|
906 |
+
"eval_samples_per_second": 1077.566,
|
907 |
+
"eval_steps_per_second": 16.888,
|
908 |
+
"step": 45000
|
909 |
+
},
|
910 |
+
{
|
911 |
+
"epoch": 2.03,
|
912 |
+
"learning_rate": 0.0005723378879939481,
|
913 |
+
"loss": 0.6172,
|
914 |
+
"step": 45500
|
915 |
+
},
|
916 |
+
{
|
917 |
+
"epoch": 2.05,
|
918 |
+
"learning_rate": 0.0005715071539231178,
|
919 |
+
"loss": 0.6165,
|
920 |
+
"step": 46000
|
921 |
+
},
|
922 |
+
{
|
923 |
+
"epoch": 2.05,
|
924 |
+
"eval_loss": 0.6103396415710449,
|
925 |
+
"eval_runtime": 2.1687,
|
926 |
+
"eval_samples_per_second": 1059.145,
|
927 |
+
"eval_steps_per_second": 16.6,
|
928 |
+
"step": 46000
|
929 |
+
},
|
930 |
+
{
|
931 |
+
"epoch": 2.07,
|
932 |
+
"learning_rate": 0.0005706647619660116,
|
933 |
+
"loss": 0.6155,
|
934 |
+
"step": 46500
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 2.1,
|
938 |
+
"learning_rate": 0.0005698107489715823,
|
939 |
+
"loss": 0.612,
|
940 |
+
"step": 47000
|
941 |
+
},
|
942 |
+
{
|
943 |
+
"epoch": 2.1,
|
944 |
+
"eval_loss": 0.6013069748878479,
|
945 |
+
"eval_runtime": 2.2612,
|
946 |
+
"eval_samples_per_second": 1015.821,
|
947 |
+
"eval_steps_per_second": 15.921,
|
948 |
+
"step": 47000
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 2.12,
|
952 |
+
"learning_rate": 0.0005689451522971252,
|
953 |
+
"loss": 0.6068,
|
954 |
+
"step": 47500
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 2.14,
|
958 |
+
"learning_rate": 0.0005680680098066429,
|
959 |
+
"loss": 0.6037,
|
960 |
+
"step": 48000
|
961 |
+
},
|
962 |
+
{
|
963 |
+
"epoch": 2.14,
|
964 |
+
"eval_loss": 0.5942632555961609,
|
965 |
+
"eval_runtime": 2.2238,
|
966 |
+
"eval_samples_per_second": 1032.912,
|
967 |
+
"eval_steps_per_second": 16.188,
|
968 |
+
"step": 48000
|
969 |
+
},
|
970 |
+
{
|
971 |
+
"epoch": 2.16,
|
972 |
+
"learning_rate": 0.0005671793598691895,
|
973 |
+
"loss": 0.6015,
|
974 |
+
"step": 48500
|
975 |
+
},
|
976 |
+
{
|
977 |
+
"epoch": 2.19,
|
978 |
+
"learning_rate": 0.0005662792413571921,
|
979 |
+
"loss": 0.6,
|
980 |
+
"step": 49000
|
981 |
+
},
|
982 |
+
{
|
983 |
+
"epoch": 2.19,
|
984 |
+
"eval_loss": 0.5915176868438721,
|
985 |
+
"eval_runtime": 2.2389,
|
986 |
+
"eval_samples_per_second": 1025.951,
|
987 |
+
"eval_steps_per_second": 16.079,
|
988 |
+
"step": 49000
|
989 |
+
},
|
990 |
+
{
|
991 |
+
"epoch": 2.21,
|
992 |
+
"learning_rate": 0.0005653676936447504,
|
993 |
+
"loss": 0.5986,
|
994 |
+
"step": 49500
|
995 |
+
},
|
996 |
+
{
|
997 |
+
"epoch": 2.23,
|
998 |
+
"learning_rate": 0.0005644447566059142,
|
999 |
+
"loss": 0.5973,
|
1000 |
+
"step": 50000
|
1001 |
+
},
|
1002 |
+
{
|
1003 |
+
"epoch": 2.23,
|
1004 |
+
"eval_loss": 0.5880685448646545,
|
1005 |
+
"eval_runtime": 2.2056,
|
1006 |
+
"eval_samples_per_second": 1041.459,
|
1007 |
+
"eval_steps_per_second": 16.322,
|
1008 |
+
"step": 50000
|
1009 |
}
|
1010 |
],
|
1011 |
"max_steps": 250000,
|
1012 |
"num_train_epochs": 12,
|
1013 |
+
"total_flos": 8.008257503646733e+20,
|
1014 |
"trial_name": null,
|
1015 |
"trial_params": null
|
1016 |
}
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 25761253
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c31ddb9e289da6dc236cf93f3881df915ba1b345fb3bc4108ca78271c87c7807
|
3 |
size 25761253
|