Training in progress, step 2000, 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 1917255968
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:92ecbb21d1e0fe04a76374b42b85859839cd5847c3b922def8d9c835efea99e0
|
| 3 |
size 1917255968
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2479129381
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eb40c6114ec922d7714527c3e12b5ccaf476fde4ca857ba014ddc6cfb4ede0c4
|
| 3 |
size 2479129381
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c800b778fa7e115e4c34de8529902de8b61c9a1b4bab3eb8295d06dafff030e
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:946649cc0ec301f9f67d287ff0bc2472a821330c5cd88309b298943469bb0e90
|
| 3 |
size 1465
|
last-checkpoint/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,
|
|
@@ -1065,6 +1065,356 @@
|
|
| 1065 |
"learning_rate": 2.8464381903616438e-05,
|
| 1066 |
"loss": 0.3255,
|
| 1067 |
"step": 1500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1068 |
}
|
| 1069 |
],
|
| 1070 |
"logging_steps": 10,
|
|
@@ -1084,7 +1434,7 @@
|
|
| 1084 |
"attributes": {}
|
| 1085 |
}
|
| 1086 |
},
|
| 1087 |
-
"total_flos": 1.
|
| 1088 |
"train_batch_size": 4,
|
| 1089 |
"trial_name": null,
|
| 1090 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.2878531755614586,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 2000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1065 |
"learning_rate": 2.8464381903616438e-05,
|
| 1066 |
"loss": 0.3255,
|
| 1067 |
"step": 1500
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 0.9723899219190212,
|
| 1071 |
+
"grad_norm": 10.125,
|
| 1072 |
+
"learning_rate": 2.8200432913515235e-05,
|
| 1073 |
+
"loss": 0.3507,
|
| 1074 |
+
"step": 1510
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 0.9788295902761008,
|
| 1078 |
+
"grad_norm": 17.0,
|
| 1079 |
+
"learning_rate": 2.7936120715902957e-05,
|
| 1080 |
+
"loss": 0.4079,
|
| 1081 |
+
"step": 1520
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.9852692586331804,
|
| 1085 |
+
"grad_norm": 11.6875,
|
| 1086 |
+
"learning_rate": 2.7671475306776396e-05,
|
| 1087 |
+
"loss": 0.4972,
|
| 1088 |
+
"step": 1530
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.99170892699026,
|
| 1092 |
+
"grad_norm": 10.4375,
|
| 1093 |
+
"learning_rate": 2.740652671994754e-05,
|
| 1094 |
+
"loss": 0.3757,
|
| 1095 |
+
"step": 1540
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.9981485953473396,
|
| 1099 |
+
"grad_norm": 8.0625,
|
| 1100 |
+
"learning_rate": 2.7141305023635083e-05,
|
| 1101 |
+
"loss": 0.3437,
|
| 1102 |
+
"step": 1550
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.0045077678499557,
|
| 1106 |
+
"grad_norm": 11.25,
|
| 1107 |
+
"learning_rate": 2.687584031705211e-05,
|
| 1108 |
+
"loss": 0.2757,
|
| 1109 |
+
"step": 1560
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.0109474362070354,
|
| 1113 |
+
"grad_norm": 7.03125,
|
| 1114 |
+
"learning_rate": 2.6610162726990185e-05,
|
| 1115 |
+
"loss": 0.2032,
|
| 1116 |
+
"step": 1570
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.017387104564115,
|
| 1120 |
+
"grad_norm": 10.4375,
|
| 1121 |
+
"learning_rate": 2.6344302404400417e-05,
|
| 1122 |
+
"loss": 0.2548,
|
| 1123 |
+
"step": 1580
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.0238267729211945,
|
| 1127 |
+
"grad_norm": 8.75,
|
| 1128 |
+
"learning_rate": 2.607828952097165e-05,
|
| 1129 |
+
"loss": 0.2578,
|
| 1130 |
+
"step": 1590
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.0302664412782743,
|
| 1134 |
+
"grad_norm": 10.75,
|
| 1135 |
+
"learning_rate": 2.5812154265706413e-05,
|
| 1136 |
+
"loss": 0.2078,
|
| 1137 |
+
"step": 1600
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.0367061096353538,
|
| 1141 |
+
"grad_norm": 11.375,
|
| 1142 |
+
"learning_rate": 2.5545926841494826e-05,
|
| 1143 |
+
"loss": 0.1952,
|
| 1144 |
+
"step": 1610
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.0431457779924334,
|
| 1148 |
+
"grad_norm": 7.6875,
|
| 1149 |
+
"learning_rate": 2.5279637461687026e-05,
|
| 1150 |
+
"loss": 0.1958,
|
| 1151 |
+
"step": 1620
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.049585446349513,
|
| 1155 |
+
"grad_norm": 9.9375,
|
| 1156 |
+
"learning_rate": 2.5013316346664273e-05,
|
| 1157 |
+
"loss": 0.2133,
|
| 1158 |
+
"step": 1630
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.0560251147065927,
|
| 1162 |
+
"grad_norm": 8.75,
|
| 1163 |
+
"learning_rate": 2.474699372040936e-05,
|
| 1164 |
+
"loss": 0.2355,
|
| 1165 |
+
"step": 1640
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.0624647830636722,
|
| 1169 |
+
"grad_norm": 9.9375,
|
| 1170 |
+
"learning_rate": 2.4480699807076592e-05,
|
| 1171 |
+
"loss": 0.225,
|
| 1172 |
+
"step": 1650
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.0689044514207517,
|
| 1176 |
+
"grad_norm": 9.125,
|
| 1177 |
+
"learning_rate": 2.4214464827561752e-05,
|
| 1178 |
+
"loss": 0.272,
|
| 1179 |
+
"step": 1660
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 1.0753441197778315,
|
| 1183 |
+
"grad_norm": 12.4375,
|
| 1184 |
+
"learning_rate": 2.3948318996072363e-05,
|
| 1185 |
+
"loss": 0.2001,
|
| 1186 |
+
"step": 1670
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.081783788134911,
|
| 1190 |
+
"grad_norm": 7.375,
|
| 1191 |
+
"learning_rate": 2.3682292516698832e-05,
|
| 1192 |
+
"loss": 0.183,
|
| 1193 |
+
"step": 1680
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 1.0882234564919906,
|
| 1197 |
+
"grad_norm": 7.71875,
|
| 1198 |
+
"learning_rate": 2.3416415579986643e-05,
|
| 1199 |
+
"loss": 0.1954,
|
| 1200 |
+
"step": 1690
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 1.0946631248490704,
|
| 1204 |
+
"grad_norm": 6.625,
|
| 1205 |
+
"learning_rate": 2.3150718359510142e-05,
|
| 1206 |
+
"loss": 0.1848,
|
| 1207 |
+
"step": 1700
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.10110279320615,
|
| 1211 |
+
"grad_norm": 8.5,
|
| 1212 |
+
"learning_rate": 2.2885231008448216e-05,
|
| 1213 |
+
"loss": 0.1769,
|
| 1214 |
+
"step": 1710
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 1.1075424615632294,
|
| 1218 |
+
"grad_norm": 17.875,
|
| 1219 |
+
"learning_rate": 2.261998365616228e-05,
|
| 1220 |
+
"loss": 0.2485,
|
| 1221 |
+
"step": 1720
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1.1139821299203092,
|
| 1225 |
+
"grad_norm": 9.9375,
|
| 1226 |
+
"learning_rate": 2.2355006404777057e-05,
|
| 1227 |
+
"loss": 0.1795,
|
| 1228 |
+
"step": 1730
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.1204217982773887,
|
| 1232 |
+
"grad_norm": 8.375,
|
| 1233 |
+
"learning_rate": 2.2090329325764294e-05,
|
| 1234 |
+
"loss": 0.2424,
|
| 1235 |
+
"step": 1740
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.1268614666344683,
|
| 1239 |
+
"grad_norm": 8.5,
|
| 1240 |
+
"learning_rate": 2.1825982456530086e-05,
|
| 1241 |
+
"loss": 0.203,
|
| 1242 |
+
"step": 1750
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.133301134991548,
|
| 1246 |
+
"grad_norm": 6.0,
|
| 1247 |
+
"learning_rate": 2.1561995797006017e-05,
|
| 1248 |
+
"loss": 0.2003,
|
| 1249 |
+
"step": 1760
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 1.1397408033486276,
|
| 1253 |
+
"grad_norm": 18.5,
|
| 1254 |
+
"learning_rate": 2.129839930624454e-05,
|
| 1255 |
+
"loss": 0.1599,
|
| 1256 |
+
"step": 1770
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.1461804717057071,
|
| 1260 |
+
"grad_norm": 11.375,
|
| 1261 |
+
"learning_rate": 2.1035222899019057e-05,
|
| 1262 |
+
"loss": 0.2198,
|
| 1263 |
+
"step": 1780
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.1526201400627867,
|
| 1267 |
+
"grad_norm": 6.65625,
|
| 1268 |
+
"learning_rate": 2.0772496442428914e-05,
|
| 1269 |
+
"loss": 0.1721,
|
| 1270 |
+
"step": 1790
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.1590598084198664,
|
| 1274 |
+
"grad_norm": 6.625,
|
| 1275 |
+
"learning_rate": 2.051024975250996e-05,
|
| 1276 |
+
"loss": 0.1557,
|
| 1277 |
+
"step": 1800
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 1.165499476776946,
|
| 1281 |
+
"grad_norm": 7.96875,
|
| 1282 |
+
"learning_rate": 2.0248512590850716e-05,
|
| 1283 |
+
"loss": 0.205,
|
| 1284 |
+
"step": 1810
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 1.1719391451340255,
|
| 1288 |
+
"grad_norm": 5.53125,
|
| 1289 |
+
"learning_rate": 1.9987314661214904e-05,
|
| 1290 |
+
"loss": 0.203,
|
| 1291 |
+
"step": 1820
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 1.1783788134911053,
|
| 1295 |
+
"grad_norm": 20.5,
|
| 1296 |
+
"learning_rate": 1.9726685606170415e-05,
|
| 1297 |
+
"loss": 0.2571,
|
| 1298 |
+
"step": 1830
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 1.1848184818481848,
|
| 1302 |
+
"grad_norm": 8.25,
|
| 1303 |
+
"learning_rate": 1.9466655003725273e-05,
|
| 1304 |
+
"loss": 0.2108,
|
| 1305 |
+
"step": 1840
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 1.1912581502052644,
|
| 1309 |
+
"grad_norm": 10.375,
|
| 1310 |
+
"learning_rate": 1.920725236397091e-05,
|
| 1311 |
+
"loss": 0.2667,
|
| 1312 |
+
"step": 1850
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.197697818562344,
|
| 1316 |
+
"grad_norm": 5.71875,
|
| 1317 |
+
"learning_rate": 1.8948507125733177e-05,
|
| 1318 |
+
"loss": 0.1615,
|
| 1319 |
+
"step": 1860
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.2041374869194237,
|
| 1323 |
+
"grad_norm": 5.5625,
|
| 1324 |
+
"learning_rate": 1.869044865323138e-05,
|
| 1325 |
+
"loss": 0.2647,
|
| 1326 |
+
"step": 1870
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1.2105771552765032,
|
| 1330 |
+
"grad_norm": 5.90625,
|
| 1331 |
+
"learning_rate": 1.843310623274587e-05,
|
| 1332 |
+
"loss": 0.1771,
|
| 1333 |
+
"step": 1880
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.2170168236335828,
|
| 1337 |
+
"grad_norm": 5.28125,
|
| 1338 |
+
"learning_rate": 1.8176509069294396e-05,
|
| 1339 |
+
"loss": 0.1781,
|
| 1340 |
+
"step": 1890
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1.2234564919906625,
|
| 1344 |
+
"grad_norm": 6.78125,
|
| 1345 |
+
"learning_rate": 1.7920686283317712e-05,
|
| 1346 |
+
"loss": 0.167,
|
| 1347 |
+
"step": 1900
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1.229896160347742,
|
| 1351 |
+
"grad_norm": 8.0625,
|
| 1352 |
+
"learning_rate": 1.7665666907374822e-05,
|
| 1353 |
+
"loss": 0.1785,
|
| 1354 |
+
"step": 1910
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1.2363358287048216,
|
| 1358 |
+
"grad_norm": 6.71875,
|
| 1359 |
+
"learning_rate": 1.741147988284813e-05,
|
| 1360 |
+
"loss": 0.1816,
|
| 1361 |
+
"step": 1920
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.2427754970619014,
|
| 1365 |
+
"grad_norm": 11.25,
|
| 1366 |
+
"learning_rate": 1.715815405665897e-05,
|
| 1367 |
+
"loss": 0.1934,
|
| 1368 |
+
"step": 1930
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.249215165418981,
|
| 1372 |
+
"grad_norm": 8.125,
|
| 1373 |
+
"learning_rate": 1.6905718177993897e-05,
|
| 1374 |
+
"loss": 0.1683,
|
| 1375 |
+
"step": 1940
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.2556548337760605,
|
| 1379 |
+
"grad_norm": 8.4375,
|
| 1380 |
+
"learning_rate": 1.6654200895041962e-05,
|
| 1381 |
+
"loss": 0.1812,
|
| 1382 |
+
"step": 1950
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 1.2620945021331402,
|
| 1386 |
+
"grad_norm": 7.25,
|
| 1387 |
+
"learning_rate": 1.6403630751743576e-05,
|
| 1388 |
+
"loss": 0.1787,
|
| 1389 |
+
"step": 1960
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.2685341704902198,
|
| 1393 |
+
"grad_norm": 6.0625,
|
| 1394 |
+
"learning_rate": 1.6154036184551098e-05,
|
| 1395 |
+
"loss": 0.1483,
|
| 1396 |
+
"step": 1970
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 1.2749738388472993,
|
| 1400 |
+
"grad_norm": 9.0,
|
| 1401 |
+
"learning_rate": 1.5905445519201694e-05,
|
| 1402 |
+
"loss": 0.3355,
|
| 1403 |
+
"step": 1980
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 1.281413507204379,
|
| 1407 |
+
"grad_norm": 9.0,
|
| 1408 |
+
"learning_rate": 1.5657886967502733e-05,
|
| 1409 |
+
"loss": 0.3235,
|
| 1410 |
+
"step": 1990
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 1.2878531755614586,
|
| 1414 |
+
"grad_norm": 4.15625,
|
| 1415 |
+
"learning_rate": 1.541138862413009e-05,
|
| 1416 |
+
"loss": 0.1555,
|
| 1417 |
+
"step": 2000
|
| 1418 |
}
|
| 1419 |
],
|
| 1420 |
"logging_steps": 10,
|
|
|
|
| 1434 |
"attributes": {}
|
| 1435 |
}
|
| 1436 |
},
|
| 1437 |
+
"total_flos": 1.5616435963670323e+17,
|
| 1438 |
"train_batch_size": 4,
|
| 1439 |
"trial_name": null,
|
| 1440 |
"trial_params": null
|