Training in progress, step 4000, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 966995080
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0abaa86fc4833d57f7fe503c09526ace445919e1b76f14b35ac06c12e186254f
|
| 3 |
size 966995080
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1925064044
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:77a61b8fba58f2ecec15e248785b8d20ba3be4672b027d63e4ae27a50a6e05e6
|
| 3 |
size 1925064044
|
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:d4a887c0679a244fb3578da62fc4230274c5d38de547b25494a50298ffcd112e
|
| 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 sha256:4a76e193687c482f6cc875caf45cbf094edc541bfbe3eb9f8259fd2d597d2f4e
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "./whisper-small-taiwanese-hanzi/checkpoint-
|
| 4 |
-
"epoch": 1.
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -874,6 +874,295 @@
|
|
| 874 |
"eval_samples_per_second": 2.265,
|
| 875 |
"eval_steps_per_second": 0.283,
|
| 876 |
"step": 3000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 877 |
}
|
| 878 |
],
|
| 879 |
"logging_steps": 25,
|
|
@@ -881,7 +1170,7 @@
|
|
| 881 |
"num_input_tokens_seen": 0,
|
| 882 |
"num_train_epochs": 2,
|
| 883 |
"save_steps": 1000,
|
| 884 |
-
"total_flos": 1.
|
| 885 |
"train_batch_size": 8,
|
| 886 |
"trial_name": null,
|
| 887 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 18.183709992924626,
|
| 3 |
+
"best_model_checkpoint": "./whisper-small-taiwanese-hanzi/checkpoint-4000",
|
| 4 |
+
"epoch": 1.6,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 4000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 874 |
"eval_samples_per_second": 2.265,
|
| 875 |
"eval_steps_per_second": 0.283,
|
| 876 |
"step": 3000
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 1.21,
|
| 880 |
+
"grad_norm": 7.052708148956299,
|
| 881 |
+
"learning_rate": 4.388888888888889e-06,
|
| 882 |
+
"loss": 0.2062,
|
| 883 |
+
"step": 3025
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 1.22,
|
| 887 |
+
"grad_norm": 9.39709186553955,
|
| 888 |
+
"learning_rate": 4.333333333333334e-06,
|
| 889 |
+
"loss": 0.2342,
|
| 890 |
+
"step": 3050
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.23,
|
| 894 |
+
"grad_norm": 5.614500522613525,
|
| 895 |
+
"learning_rate": 4.277777777777778e-06,
|
| 896 |
+
"loss": 0.218,
|
| 897 |
+
"step": 3075
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 1.24,
|
| 901 |
+
"grad_norm": 7.228747367858887,
|
| 902 |
+
"learning_rate": 4.222222222222223e-06,
|
| 903 |
+
"loss": 0.2154,
|
| 904 |
+
"step": 3100
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 1.25,
|
| 908 |
+
"grad_norm": 10.118727684020996,
|
| 909 |
+
"learning_rate": 4.166666666666667e-06,
|
| 910 |
+
"loss": 0.2519,
|
| 911 |
+
"step": 3125
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 1.26,
|
| 915 |
+
"grad_norm": 9.65976333618164,
|
| 916 |
+
"learning_rate": 4.111111111111111e-06,
|
| 917 |
+
"loss": 0.2336,
|
| 918 |
+
"step": 3150
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 1.27,
|
| 922 |
+
"grad_norm": 9.654402732849121,
|
| 923 |
+
"learning_rate": 4.055555555555556e-06,
|
| 924 |
+
"loss": 0.2474,
|
| 925 |
+
"step": 3175
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 1.28,
|
| 929 |
+
"grad_norm": 6.564414024353027,
|
| 930 |
+
"learning_rate": 4.000000000000001e-06,
|
| 931 |
+
"loss": 0.2329,
|
| 932 |
+
"step": 3200
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 1.29,
|
| 936 |
+
"grad_norm": 7.532024383544922,
|
| 937 |
+
"learning_rate": 3.944444444444445e-06,
|
| 938 |
+
"loss": 0.2135,
|
| 939 |
+
"step": 3225
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 1.3,
|
| 943 |
+
"grad_norm": 5.012012004852295,
|
| 944 |
+
"learning_rate": 3.88888888888889e-06,
|
| 945 |
+
"loss": 0.2278,
|
| 946 |
+
"step": 3250
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 1.31,
|
| 950 |
+
"grad_norm": 7.632811069488525,
|
| 951 |
+
"learning_rate": 3.833333333333334e-06,
|
| 952 |
+
"loss": 0.2238,
|
| 953 |
+
"step": 3275
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 1.32,
|
| 957 |
+
"grad_norm": 6.163909912109375,
|
| 958 |
+
"learning_rate": 3.777777777777778e-06,
|
| 959 |
+
"loss": 0.2338,
|
| 960 |
+
"step": 3300
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 1.33,
|
| 964 |
+
"grad_norm": 7.474846839904785,
|
| 965 |
+
"learning_rate": 3.7222222222222225e-06,
|
| 966 |
+
"loss": 0.2217,
|
| 967 |
+
"step": 3325
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 1.34,
|
| 971 |
+
"grad_norm": 9.621731758117676,
|
| 972 |
+
"learning_rate": 3.6666666666666666e-06,
|
| 973 |
+
"loss": 0.226,
|
| 974 |
+
"step": 3350
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 1.35,
|
| 978 |
+
"grad_norm": 7.167582035064697,
|
| 979 |
+
"learning_rate": 3.6111111111111115e-06,
|
| 980 |
+
"loss": 0.2219,
|
| 981 |
+
"step": 3375
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 1.3599999999999999,
|
| 985 |
+
"grad_norm": 8.028946876525879,
|
| 986 |
+
"learning_rate": 3.555555555555556e-06,
|
| 987 |
+
"loss": 0.2243,
|
| 988 |
+
"step": 3400
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 1.37,
|
| 992 |
+
"grad_norm": 9.479345321655273,
|
| 993 |
+
"learning_rate": 3.5e-06,
|
| 994 |
+
"loss": 0.2486,
|
| 995 |
+
"step": 3425
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 1.38,
|
| 999 |
+
"grad_norm": 7.631903648376465,
|
| 1000 |
+
"learning_rate": 3.444444444444445e-06,
|
| 1001 |
+
"loss": 0.219,
|
| 1002 |
+
"step": 3450
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 1.3900000000000001,
|
| 1006 |
+
"grad_norm": 8.521742820739746,
|
| 1007 |
+
"learning_rate": 3.3888888888888893e-06,
|
| 1008 |
+
"loss": 0.2106,
|
| 1009 |
+
"step": 3475
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 1.4,
|
| 1013 |
+
"grad_norm": 9.257588386535645,
|
| 1014 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 1015 |
+
"loss": 0.225,
|
| 1016 |
+
"step": 3500
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.41,
|
| 1020 |
+
"grad_norm": 10.871211051940918,
|
| 1021 |
+
"learning_rate": 3.277777777777778e-06,
|
| 1022 |
+
"loss": 0.2248,
|
| 1023 |
+
"step": 3525
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 1.42,
|
| 1027 |
+
"grad_norm": 5.058568954467773,
|
| 1028 |
+
"learning_rate": 3.2222222222222227e-06,
|
| 1029 |
+
"loss": 0.2315,
|
| 1030 |
+
"step": 3550
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.43,
|
| 1034 |
+
"grad_norm": 5.446198463439941,
|
| 1035 |
+
"learning_rate": 3.1666666666666667e-06,
|
| 1036 |
+
"loss": 0.2213,
|
| 1037 |
+
"step": 3575
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 1.44,
|
| 1041 |
+
"grad_norm": 7.046563148498535,
|
| 1042 |
+
"learning_rate": 3.1111111111111116e-06,
|
| 1043 |
+
"loss": 0.2029,
|
| 1044 |
+
"step": 3600
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 1.45,
|
| 1048 |
+
"grad_norm": 6.039666175842285,
|
| 1049 |
+
"learning_rate": 3.055555555555556e-06,
|
| 1050 |
+
"loss": 0.2046,
|
| 1051 |
+
"step": 3625
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 1.46,
|
| 1055 |
+
"grad_norm": 6.638487815856934,
|
| 1056 |
+
"learning_rate": 3e-06,
|
| 1057 |
+
"loss": 0.2218,
|
| 1058 |
+
"step": 3650
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 1.47,
|
| 1062 |
+
"grad_norm": 9.455551147460938,
|
| 1063 |
+
"learning_rate": 2.944444444444445e-06,
|
| 1064 |
+
"loss": 0.2252,
|
| 1065 |
+
"step": 3675
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 1.48,
|
| 1069 |
+
"grad_norm": 9.041964530944824,
|
| 1070 |
+
"learning_rate": 2.888888888888889e-06,
|
| 1071 |
+
"loss": 0.2082,
|
| 1072 |
+
"step": 3700
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 1.49,
|
| 1076 |
+
"grad_norm": 10.04900074005127,
|
| 1077 |
+
"learning_rate": 2.8333333333333335e-06,
|
| 1078 |
+
"loss": 0.2298,
|
| 1079 |
+
"step": 3725
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 1.5,
|
| 1083 |
+
"grad_norm": 5.999593734741211,
|
| 1084 |
+
"learning_rate": 2.7777777777777783e-06,
|
| 1085 |
+
"loss": 0.2085,
|
| 1086 |
+
"step": 3750
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 1.51,
|
| 1090 |
+
"grad_norm": 7.189665794372559,
|
| 1091 |
+
"learning_rate": 2.7222222222222224e-06,
|
| 1092 |
+
"loss": 0.2016,
|
| 1093 |
+
"step": 3775
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 1.52,
|
| 1097 |
+
"grad_norm": 6.945767879486084,
|
| 1098 |
+
"learning_rate": 2.666666666666667e-06,
|
| 1099 |
+
"loss": 0.2196,
|
| 1100 |
+
"step": 3800
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 1.53,
|
| 1104 |
+
"grad_norm": 6.626684188842773,
|
| 1105 |
+
"learning_rate": 2.6111111111111113e-06,
|
| 1106 |
+
"loss": 0.1848,
|
| 1107 |
+
"step": 3825
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 1.54,
|
| 1111 |
+
"grad_norm": 8.809107780456543,
|
| 1112 |
+
"learning_rate": 2.5555555555555557e-06,
|
| 1113 |
+
"loss": 0.2567,
|
| 1114 |
+
"step": 3850
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 1.55,
|
| 1118 |
+
"grad_norm": 6.8933186531066895,
|
| 1119 |
+
"learning_rate": 2.5e-06,
|
| 1120 |
+
"loss": 0.2177,
|
| 1121 |
+
"step": 3875
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 1.56,
|
| 1125 |
+
"grad_norm": 5.856971740722656,
|
| 1126 |
+
"learning_rate": 2.4444444444444447e-06,
|
| 1127 |
+
"loss": 0.214,
|
| 1128 |
+
"step": 3900
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 1.5699999999999998,
|
| 1132 |
+
"grad_norm": 6.439066410064697,
|
| 1133 |
+
"learning_rate": 2.388888888888889e-06,
|
| 1134 |
+
"loss": 0.2088,
|
| 1135 |
+
"step": 3925
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 1.58,
|
| 1139 |
+
"grad_norm": 6.070130348205566,
|
| 1140 |
+
"learning_rate": 2.3333333333333336e-06,
|
| 1141 |
+
"loss": 0.2085,
|
| 1142 |
+
"step": 3950
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 1.5899999999999999,
|
| 1146 |
+
"grad_norm": 7.12228536605835,
|
| 1147 |
+
"learning_rate": 2.277777777777778e-06,
|
| 1148 |
+
"loss": 0.1959,
|
| 1149 |
+
"step": 3975
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 1.6,
|
| 1153 |
+
"grad_norm": 5.102252006530762,
|
| 1154 |
+
"learning_rate": 2.222222222222222e-06,
|
| 1155 |
+
"loss": 0.2183,
|
| 1156 |
+
"step": 4000
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 1.6,
|
| 1160 |
+
"eval_cer": 18.183709992924626,
|
| 1161 |
+
"eval_loss": 0.3092849850654602,
|
| 1162 |
+
"eval_runtime": 1753.7044,
|
| 1163 |
+
"eval_samples_per_second": 2.244,
|
| 1164 |
+
"eval_steps_per_second": 0.281,
|
| 1165 |
+
"step": 4000
|
| 1166 |
}
|
| 1167 |
],
|
| 1168 |
"logging_steps": 25,
|
|
|
|
| 1170 |
"num_input_tokens_seen": 0,
|
| 1171 |
"num_train_epochs": 2,
|
| 1172 |
"save_steps": 1000,
|
| 1173 |
+
"total_flos": 1.846946562048e+19,
|
| 1174 |
"train_batch_size": 8,
|
| 1175 |
"trial_name": null,
|
| 1176 |
"trial_params": null
|