Johannes Garstenauer commited on
Commit ·
e40687c
1
Parent(s): 73c7fcb
Training in progress, step 80624
Browse files- last-checkpoint/optimizer.pt +2 -2
- last-checkpoint/pytorch_model.bin +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +257 -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
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1a32601442637b42f7a6eb10ebea11d5d313b374e4bc71dab957e52f81ba30d
|
| 3 |
+
size 133906309
|
last-checkpoint/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 266387761
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a81cbf5d8cd63bef63639bc99cfc0f2477944d78d679cea7a6a8bf8465875fd
|
| 3 |
size 266387761
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14511
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99abdf87e091757a4c26601a35007bfa738d7dc783b6dec2f7e9dcc551f7dba0
|
| 3 |
size 14511
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62e8dbbdc09c6ae689f3a38febec079d787f5f9cfadb6ee31fcc9975cbbabc68
|
| 3 |
size 627
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -758,11 +758,265 @@
|
|
| 758 |
"eval_samples_per_second": 388.927,
|
| 759 |
"eval_steps_per_second": 6.077,
|
| 760 |
"step": 60000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 761 |
}
|
| 762 |
],
|
| 763 |
"max_steps": 806202,
|
| 764 |
"num_train_epochs": 3,
|
| 765 |
-
"total_flos": 1.
|
| 766 |
"trial_name": null,
|
| 767 |
"trial_params": null
|
| 768 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.3000135821786931,
|
| 5 |
+
"global_step": 80624,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 758 |
"eval_samples_per_second": 388.927,
|
| 759 |
"eval_steps_per_second": 6.077,
|
| 760 |
"step": 60000
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.23,
|
| 764 |
+
"learning_rate": 2.3594167338706678e-05,
|
| 765 |
+
"loss": 2.9852,
|
| 766 |
+
"step": 60500
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.23,
|
| 770 |
+
"learning_rate": 2.3578347234067888e-05,
|
| 771 |
+
"loss": 3.1677,
|
| 772 |
+
"step": 61000
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.23,
|
| 776 |
+
"learning_rate": 2.35625271294291e-05,
|
| 777 |
+
"loss": 3.0343,
|
| 778 |
+
"step": 61500
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 0.23,
|
| 782 |
+
"learning_rate": 2.3546707024790315e-05,
|
| 783 |
+
"loss": 2.9848,
|
| 784 |
+
"step": 62000
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 0.23,
|
| 788 |
+
"learning_rate": 2.3530886920151525e-05,
|
| 789 |
+
"loss": 3.0706,
|
| 790 |
+
"step": 62500
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 0.23,
|
| 794 |
+
"learning_rate": 2.351506681551274e-05,
|
| 795 |
+
"loss": 3.036,
|
| 796 |
+
"step": 63000
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.24,
|
| 800 |
+
"learning_rate": 2.349924671087395e-05,
|
| 801 |
+
"loss": 2.9671,
|
| 802 |
+
"step": 63500
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.24,
|
| 806 |
+
"learning_rate": 2.3483426606235166e-05,
|
| 807 |
+
"loss": 3.0112,
|
| 808 |
+
"step": 64000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.24,
|
| 812 |
+
"learning_rate": 2.3467606501596376e-05,
|
| 813 |
+
"loss": 2.9627,
|
| 814 |
+
"step": 64500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.24,
|
| 818 |
+
"learning_rate": 2.3451786396957586e-05,
|
| 819 |
+
"loss": 3.0045,
|
| 820 |
+
"step": 65000
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 0.24,
|
| 824 |
+
"learning_rate": 2.34359662923188e-05,
|
| 825 |
+
"loss": 3.0389,
|
| 826 |
+
"step": 65500
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.25,
|
| 830 |
+
"learning_rate": 2.3420146187680013e-05,
|
| 831 |
+
"loss": 2.9675,
|
| 832 |
+
"step": 66000
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 0.25,
|
| 836 |
+
"learning_rate": 2.3404326083041223e-05,
|
| 837 |
+
"loss": 2.9137,
|
| 838 |
+
"step": 66500
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.25,
|
| 842 |
+
"learning_rate": 2.3388505978402436e-05,
|
| 843 |
+
"loss": 2.9304,
|
| 844 |
+
"step": 67000
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.25,
|
| 848 |
+
"learning_rate": 2.3372685873763647e-05,
|
| 849 |
+
"loss": 2.931,
|
| 850 |
+
"step": 67500
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.25,
|
| 854 |
+
"learning_rate": 2.3356865769124863e-05,
|
| 855 |
+
"loss": 2.8961,
|
| 856 |
+
"step": 68000
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.25,
|
| 860 |
+
"learning_rate": 2.3341045664486074e-05,
|
| 861 |
+
"loss": 2.9894,
|
| 862 |
+
"step": 68500
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 0.26,
|
| 866 |
+
"learning_rate": 2.3325225559847284e-05,
|
| 867 |
+
"loss": 2.8711,
|
| 868 |
+
"step": 69000
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 0.26,
|
| 872 |
+
"learning_rate": 2.3309405455208497e-05,
|
| 873 |
+
"loss": 2.8287,
|
| 874 |
+
"step": 69500
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.26,
|
| 878 |
+
"learning_rate": 2.329358535056971e-05,
|
| 879 |
+
"loss": 2.899,
|
| 880 |
+
"step": 70000
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.26,
|
| 884 |
+
"learning_rate": 2.327776524593092e-05,
|
| 885 |
+
"loss": 2.8684,
|
| 886 |
+
"step": 70500
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.26,
|
| 890 |
+
"learning_rate": 2.3261945141292134e-05,
|
| 891 |
+
"loss": 2.8482,
|
| 892 |
+
"step": 71000
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.27,
|
| 896 |
+
"learning_rate": 2.3246125036653344e-05,
|
| 897 |
+
"loss": 2.8913,
|
| 898 |
+
"step": 71500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.27,
|
| 902 |
+
"learning_rate": 2.3230304932014558e-05,
|
| 903 |
+
"loss": 2.8641,
|
| 904 |
+
"step": 72000
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 0.27,
|
| 908 |
+
"learning_rate": 2.321448482737577e-05,
|
| 909 |
+
"loss": 2.861,
|
| 910 |
+
"step": 72500
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.27,
|
| 914 |
+
"learning_rate": 2.319866472273698e-05,
|
| 915 |
+
"loss": 2.8647,
|
| 916 |
+
"step": 73000
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 0.27,
|
| 920 |
+
"learning_rate": 2.3182844618098195e-05,
|
| 921 |
+
"loss": 2.8619,
|
| 922 |
+
"step": 73500
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.28,
|
| 926 |
+
"learning_rate": 2.316702451345941e-05,
|
| 927 |
+
"loss": 2.8267,
|
| 928 |
+
"step": 74000
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.28,
|
| 932 |
+
"learning_rate": 2.3151204408820622e-05,
|
| 933 |
+
"loss": 2.8155,
|
| 934 |
+
"step": 74500
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 0.28,
|
| 938 |
+
"learning_rate": 2.3135384304181832e-05,
|
| 939 |
+
"loss": 2.876,
|
| 940 |
+
"step": 75000
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.28,
|
| 944 |
+
"eval_loss": 2.817479133605957,
|
| 945 |
+
"eval_runtime": 15565.5566,
|
| 946 |
+
"eval_samples_per_second": 389.979,
|
| 947 |
+
"eval_steps_per_second": 6.093,
|
| 948 |
+
"step": 75000
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.28,
|
| 952 |
+
"learning_rate": 2.3119564199543042e-05,
|
| 953 |
+
"loss": 2.7945,
|
| 954 |
+
"step": 75500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.28,
|
| 958 |
+
"learning_rate": 2.3103744094904256e-05,
|
| 959 |
+
"loss": 2.8052,
|
| 960 |
+
"step": 76000
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.28,
|
| 964 |
+
"learning_rate": 2.308792399026547e-05,
|
| 965 |
+
"loss": 2.7997,
|
| 966 |
+
"step": 76500
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.29,
|
| 970 |
+
"learning_rate": 2.307210388562668e-05,
|
| 971 |
+
"loss": 2.7941,
|
| 972 |
+
"step": 77000
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 0.29,
|
| 976 |
+
"learning_rate": 2.3056283780987893e-05,
|
| 977 |
+
"loss": 2.8057,
|
| 978 |
+
"step": 77500
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.29,
|
| 982 |
+
"learning_rate": 2.3040463676349103e-05,
|
| 983 |
+
"loss": 2.8142,
|
| 984 |
+
"step": 78000
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.29,
|
| 988 |
+
"learning_rate": 2.302464357171032e-05,
|
| 989 |
+
"loss": 2.8154,
|
| 990 |
+
"step": 78500
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.29,
|
| 994 |
+
"learning_rate": 2.300882346707153e-05,
|
| 995 |
+
"loss": 2.7736,
|
| 996 |
+
"step": 79000
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.3,
|
| 1000 |
+
"learning_rate": 2.299300336243274e-05,
|
| 1001 |
+
"loss": 2.8183,
|
| 1002 |
+
"step": 79500
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 0.3,
|
| 1006 |
+
"learning_rate": 2.2977183257793953e-05,
|
| 1007 |
+
"loss": 2.7761,
|
| 1008 |
+
"step": 80000
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.3,
|
| 1012 |
+
"learning_rate": 2.2961363153155167e-05,
|
| 1013 |
+
"loss": 2.747,
|
| 1014 |
+
"step": 80500
|
| 1015 |
}
|
| 1016 |
],
|
| 1017 |
"max_steps": 806202,
|
| 1018 |
"num_train_epochs": 3,
|
| 1019 |
+
"total_flos": 1.3679977728143524e+18,
|
| 1020 |
"trial_name": null,
|
| 1021 |
"trial_params": null
|
| 1022 |
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 266387761
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a81cbf5d8cd63bef63639bc99cfc0f2477944d78d679cea7a6a8bf8465875fd
|
| 3 |
size 266387761
|