Upload 4 files
Browse files- model.safetensors +1 -1
- trainer_state.json +3 -297
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 989717056
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:472c0afe2b754d96f0968da203ce6b1a2f7c6627739a8e76f358090fadcc8556
|
| 3 |
size 989717056
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -876,300 +876,6 @@
|
|
| 876 |
"learning_rate": 4.404992798847816e-05,
|
| 877 |
"loss": 0.1632,
|
| 878 |
"step": 12400
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 6.00096015362458,
|
| 882 |
-
"grad_norm": 0.9519465565681458,
|
| 883 |
-
"learning_rate": 4.4001920307249165e-05,
|
| 884 |
-
"loss": 0.1647,
|
| 885 |
-
"step": 12500
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 6.048967834853577,
|
| 889 |
-
"grad_norm": 1.1763761043548584,
|
| 890 |
-
"learning_rate": 4.395391262602016e-05,
|
| 891 |
-
"loss": 0.1012,
|
| 892 |
-
"step": 12600
|
| 893 |
-
},
|
| 894 |
-
{
|
| 895 |
-
"epoch": 6.096975516082574,
|
| 896 |
-
"grad_norm": 1.311137318611145,
|
| 897 |
-
"learning_rate": 4.390590494479117e-05,
|
| 898 |
-
"loss": 0.1023,
|
| 899 |
-
"step": 12700
|
| 900 |
-
},
|
| 901 |
-
{
|
| 902 |
-
"epoch": 6.14498319731157,
|
| 903 |
-
"grad_norm": 0.9711358547210693,
|
| 904 |
-
"learning_rate": 4.385789726356217e-05,
|
| 905 |
-
"loss": 0.1025,
|
| 906 |
-
"step": 12800
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 6.1929908785405665,
|
| 910 |
-
"grad_norm": 1.7737120389938354,
|
| 911 |
-
"learning_rate": 4.380988958233317e-05,
|
| 912 |
-
"loss": 0.107,
|
| 913 |
-
"step": 12900
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 6.240998559769563,
|
| 917 |
-
"grad_norm": 1.3438420295715332,
|
| 918 |
-
"learning_rate": 4.3761881901104176e-05,
|
| 919 |
-
"loss": 0.1116,
|
| 920 |
-
"step": 13000
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 6.289006240998559,
|
| 924 |
-
"grad_norm": 1.2701375484466553,
|
| 925 |
-
"learning_rate": 4.371387421987518e-05,
|
| 926 |
-
"loss": 0.1102,
|
| 927 |
-
"step": 13100
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 6.337013922227556,
|
| 931 |
-
"grad_norm": 1.3952118158340454,
|
| 932 |
-
"learning_rate": 4.366586653864618e-05,
|
| 933 |
-
"loss": 0.1114,
|
| 934 |
-
"step": 13200
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 6.385021603456553,
|
| 938 |
-
"grad_norm": 1.2892017364501953,
|
| 939 |
-
"learning_rate": 4.3617858857417185e-05,
|
| 940 |
-
"loss": 0.111,
|
| 941 |
-
"step": 13300
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 6.433029284685549,
|
| 945 |
-
"grad_norm": 1.2740648984909058,
|
| 946 |
-
"learning_rate": 4.356985117618819e-05,
|
| 947 |
-
"loss": 0.1125,
|
| 948 |
-
"step": 13400
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 6.481036965914546,
|
| 952 |
-
"grad_norm": 1.1152820587158203,
|
| 953 |
-
"learning_rate": 4.3521843494959195e-05,
|
| 954 |
-
"loss": 0.1145,
|
| 955 |
-
"step": 13500
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 6.529044647143543,
|
| 959 |
-
"grad_norm": 1.3018417358398438,
|
| 960 |
-
"learning_rate": 4.34738358137302e-05,
|
| 961 |
-
"loss": 0.1161,
|
| 962 |
-
"step": 13600
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 6.57705232837254,
|
| 966 |
-
"grad_norm": 1.3537256717681885,
|
| 967 |
-
"learning_rate": 4.34258281325012e-05,
|
| 968 |
-
"loss": 0.1127,
|
| 969 |
-
"step": 13700
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 6.625060009601536,
|
| 973 |
-
"grad_norm": 1.566361665725708,
|
| 974 |
-
"learning_rate": 4.33778204512722e-05,
|
| 975 |
-
"loss": 0.1168,
|
| 976 |
-
"step": 13800
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 6.673067690830533,
|
| 980 |
-
"grad_norm": 1.4471561908721924,
|
| 981 |
-
"learning_rate": 4.332981277004321e-05,
|
| 982 |
-
"loss": 0.1185,
|
| 983 |
-
"step": 13900
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 6.72107537205953,
|
| 987 |
-
"grad_norm": 1.2443643808364868,
|
| 988 |
-
"learning_rate": 4.328180508881421e-05,
|
| 989 |
-
"loss": 0.1211,
|
| 990 |
-
"step": 14000
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 6.769083053288526,
|
| 994 |
-
"grad_norm": 1.568220615386963,
|
| 995 |
-
"learning_rate": 4.323379740758521e-05,
|
| 996 |
-
"loss": 0.1181,
|
| 997 |
-
"step": 14100
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 6.817090734517523,
|
| 1001 |
-
"grad_norm": 1.3633244037628174,
|
| 1002 |
-
"learning_rate": 4.3185789726356216e-05,
|
| 1003 |
-
"loss": 0.1221,
|
| 1004 |
-
"step": 14200
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 6.86509841574652,
|
| 1008 |
-
"grad_norm": 1.2666867971420288,
|
| 1009 |
-
"learning_rate": 4.313778204512722e-05,
|
| 1010 |
-
"loss": 0.1223,
|
| 1011 |
-
"step": 14300
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 6.9131060969755165,
|
| 1015 |
-
"grad_norm": 1.2681658267974854,
|
| 1016 |
-
"learning_rate": 4.3089774363898227e-05,
|
| 1017 |
-
"loss": 0.1231,
|
| 1018 |
-
"step": 14400
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 6.9611137782045125,
|
| 1022 |
-
"grad_norm": 1.0958737134933472,
|
| 1023 |
-
"learning_rate": 4.3041766682669225e-05,
|
| 1024 |
-
"loss": 0.123,
|
| 1025 |
-
"step": 14500
|
| 1026 |
-
},
|
| 1027 |
-
{
|
| 1028 |
-
"epoch": 7.009121459433509,
|
| 1029 |
-
"grad_norm": 0.9932379126548767,
|
| 1030 |
-
"learning_rate": 4.299375900144023e-05,
|
| 1031 |
-
"loss": 0.1141,
|
| 1032 |
-
"step": 14600
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 7.057129140662506,
|
| 1036 |
-
"grad_norm": 1.1502649784088135,
|
| 1037 |
-
"learning_rate": 4.2945751320211235e-05,
|
| 1038 |
-
"loss": 0.0742,
|
| 1039 |
-
"step": 14700
|
| 1040 |
-
},
|
| 1041 |
-
{
|
| 1042 |
-
"epoch": 7.105136821891502,
|
| 1043 |
-
"grad_norm": 1.1436489820480347,
|
| 1044 |
-
"learning_rate": 4.289774363898224e-05,
|
| 1045 |
-
"loss": 0.0751,
|
| 1046 |
-
"step": 14800
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 7.153144503120499,
|
| 1050 |
-
"grad_norm": 1.4604742527008057,
|
| 1051 |
-
"learning_rate": 4.2849735957753245e-05,
|
| 1052 |
-
"loss": 0.0766,
|
| 1053 |
-
"step": 14900
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 7.201152184349496,
|
| 1057 |
-
"grad_norm": 1.307935118675232,
|
| 1058 |
-
"learning_rate": 4.280172827652424e-05,
|
| 1059 |
-
"loss": 0.0784,
|
| 1060 |
-
"step": 15000
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 7.249159865578493,
|
| 1064 |
-
"grad_norm": 1.0527863502502441,
|
| 1065 |
-
"learning_rate": 4.275372059529525e-05,
|
| 1066 |
-
"loss": 0.0796,
|
| 1067 |
-
"step": 15100
|
| 1068 |
-
},
|
| 1069 |
-
{
|
| 1070 |
-
"epoch": 7.297167546807489,
|
| 1071 |
-
"grad_norm": 1.2827943563461304,
|
| 1072 |
-
"learning_rate": 4.270571291406625e-05,
|
| 1073 |
-
"loss": 0.0783,
|
| 1074 |
-
"step": 15200
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 7.345175228036486,
|
| 1078 |
-
"grad_norm": 1.3460379838943481,
|
| 1079 |
-
"learning_rate": 4.265770523283726e-05,
|
| 1080 |
-
"loss": 0.0819,
|
| 1081 |
-
"step": 15300
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 7.393182909265483,
|
| 1085 |
-
"grad_norm": 1.2280845642089844,
|
| 1086 |
-
"learning_rate": 4.260969755160826e-05,
|
| 1087 |
-
"loss": 0.0843,
|
| 1088 |
-
"step": 15400
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 7.441190590494479,
|
| 1092 |
-
"grad_norm": 1.3223485946655273,
|
| 1093 |
-
"learning_rate": 4.256168987037926e-05,
|
| 1094 |
-
"loss": 0.0817,
|
| 1095 |
-
"step": 15500
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 7.489198271723476,
|
| 1099 |
-
"grad_norm": 1.296045184135437,
|
| 1100 |
-
"learning_rate": 4.251368218915027e-05,
|
| 1101 |
-
"loss": 0.0832,
|
| 1102 |
-
"step": 15600
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"epoch": 7.537205952952473,
|
| 1106 |
-
"grad_norm": 1.3680839538574219,
|
| 1107 |
-
"learning_rate": 4.246567450792127e-05,
|
| 1108 |
-
"loss": 0.0861,
|
| 1109 |
-
"step": 15700
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 7.585213634181469,
|
| 1113 |
-
"grad_norm": 1.1178674697875977,
|
| 1114 |
-
"learning_rate": 4.241766682669227e-05,
|
| 1115 |
-
"loss": 0.0838,
|
| 1116 |
-
"step": 15800
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 7.633221315410466,
|
| 1120 |
-
"grad_norm": 1.034132719039917,
|
| 1121 |
-
"learning_rate": 4.2369659145463275e-05,
|
| 1122 |
-
"loss": 0.0884,
|
| 1123 |
-
"step": 15900
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 7.6812289966394625,
|
| 1127 |
-
"grad_norm": 1.2711660861968994,
|
| 1128 |
-
"learning_rate": 4.232165146423428e-05,
|
| 1129 |
-
"loss": 0.0881,
|
| 1130 |
-
"step": 16000
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 7.7292366778684585,
|
| 1134 |
-
"grad_norm": 1.1275447607040405,
|
| 1135 |
-
"learning_rate": 4.2273643783005285e-05,
|
| 1136 |
-
"loss": 0.0891,
|
| 1137 |
-
"step": 16100
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 7.777244359097455,
|
| 1141 |
-
"grad_norm": 1.1659106016159058,
|
| 1142 |
-
"learning_rate": 4.2225636101776283e-05,
|
| 1143 |
-
"loss": 0.0905,
|
| 1144 |
-
"step": 16200
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 7.825252040326452,
|
| 1148 |
-
"grad_norm": 1.3686105012893677,
|
| 1149 |
-
"learning_rate": 4.217762842054729e-05,
|
| 1150 |
-
"loss": 0.0911,
|
| 1151 |
-
"step": 16300
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 7.873259721555449,
|
| 1155 |
-
"grad_norm": 1.3593393564224243,
|
| 1156 |
-
"learning_rate": 4.2129620739318294e-05,
|
| 1157 |
-
"loss": 0.0917,
|
| 1158 |
-
"step": 16400
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 7.921267402784445,
|
| 1162 |
-
"grad_norm": 1.4164599180221558,
|
| 1163 |
-
"learning_rate": 4.20816130580893e-05,
|
| 1164 |
-
"loss": 0.092,
|
| 1165 |
-
"step": 16500
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 7.969275084013442,
|
| 1169 |
-
"grad_norm": 1.3200361728668213,
|
| 1170 |
-
"learning_rate": 4.2033605376860304e-05,
|
| 1171 |
-
"loss": 0.0926,
|
| 1172 |
-
"step": 16600
|
| 1173 |
}
|
| 1174 |
],
|
| 1175 |
"logging_steps": 100,
|
|
@@ -1189,7 +895,7 @@
|
|
| 1189 |
"attributes": {}
|
| 1190 |
}
|
| 1191 |
},
|
| 1192 |
-
"total_flos": 1.
|
| 1193 |
"train_batch_size": 16,
|
| 1194 |
"trial_name": null,
|
| 1195 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 6.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 12498,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 876 |
"learning_rate": 4.404992798847816e-05,
|
| 877 |
"loss": 0.1632,
|
| 878 |
"step": 12400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 879 |
}
|
| 880 |
],
|
| 881 |
"logging_steps": 100,
|
|
|
|
| 895 |
"attributes": {}
|
| 896 |
}
|
| 897 |
},
|
| 898 |
+
"total_flos": 1.186514987204802e+20,
|
| 899 |
"train_batch_size": 16,
|
| 900 |
"trial_name": null,
|
| 901 |
"trial_params": null
|