Training in progress, step 200, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 639691872
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3bf6d6a2180aaa44b0684de1c1b4c44c876e3b61e90774ecf34f9d75d0e5306
|
| 3 |
size 639691872
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 325339796
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:753552200a46d3e7e4d375873d1c04cbc4b9acbcf9d4abe320d73e382fdda544
|
| 3 |
size 325339796
|
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:a7df43c4526bd9a8ecddc4847c60b296a89597a43586e7d1e215a1b5c2cbd258
|
| 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:f462ed10ef931ef3980d2aa020052a49dcadd400550d826d7ccc8a4e298b95f9
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "miner_id_24/checkpoint-
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 50,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1089,6 +1089,364 @@
|
|
| 1089 |
"eval_samples_per_second": 9.266,
|
| 1090 |
"eval_steps_per_second": 2.336,
|
| 1091 |
"step": 150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1092 |
}
|
| 1093 |
],
|
| 1094 |
"logging_steps": 1,
|
|
@@ -1117,7 +1475,7 @@
|
|
| 1117 |
"attributes": {}
|
| 1118 |
}
|
| 1119 |
},
|
| 1120 |
-
"total_flos": 2.
|
| 1121 |
"train_batch_size": 8,
|
| 1122 |
"trial_name": null,
|
| 1123 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.1042157411575317,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.9456264775413712,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 200,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1089 |
"eval_samples_per_second": 9.266,
|
| 1090 |
"eval_steps_per_second": 2.336,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.7139479905437353,
|
| 1095 |
+
"grad_norm": 0.479676216840744,
|
| 1096 |
+
"learning_rate": 2.0862948197684955e-05,
|
| 1097 |
+
"loss": 1.2254,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.7186761229314421,
|
| 1102 |
+
"grad_norm": 0.4174903631210327,
|
| 1103 |
+
"learning_rate": 2.0234556786166715e-05,
|
| 1104 |
+
"loss": 1.1599,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.723404255319149,
|
| 1109 |
+
"grad_norm": 0.4270111918449402,
|
| 1110 |
+
"learning_rate": 1.961336484597343e-05,
|
| 1111 |
+
"loss": 1.1773,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.7281323877068558,
|
| 1116 |
+
"grad_norm": 0.4060277044773102,
|
| 1117 |
+
"learning_rate": 1.899952262696125e-05,
|
| 1118 |
+
"loss": 1.2089,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.7328605200945626,
|
| 1123 |
+
"grad_norm": 0.40245646238327026,
|
| 1124 |
+
"learning_rate": 1.8393178601283683e-05,
|
| 1125 |
+
"loss": 1.1942,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.7375886524822695,
|
| 1130 |
+
"grad_norm": 0.3960253894329071,
|
| 1131 |
+
"learning_rate": 1.7794479427480117e-05,
|
| 1132 |
+
"loss": 1.2359,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.7423167848699763,
|
| 1137 |
+
"grad_norm": 0.36273330450057983,
|
| 1138 |
+
"learning_rate": 1.7203569915003005e-05,
|
| 1139 |
+
"loss": 1.1725,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.7470449172576832,
|
| 1144 |
+
"grad_norm": 0.38699132204055786,
|
| 1145 |
+
"learning_rate": 1.662059298919232e-05,
|
| 1146 |
+
"loss": 1.2018,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.75177304964539,
|
| 1151 |
+
"grad_norm": 0.3809574544429779,
|
| 1152 |
+
"learning_rate": 1.6045689656705716e-05,
|
| 1153 |
+
"loss": 1.1198,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.7565011820330969,
|
| 1158 |
+
"grad_norm": 0.37179070711135864,
|
| 1159 |
+
"learning_rate": 1.5478998971412668e-05,
|
| 1160 |
+
"loss": 1.1596,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.7612293144208038,
|
| 1165 |
+
"grad_norm": 0.3978999853134155,
|
| 1166 |
+
"learning_rate": 1.4920658000761174e-05,
|
| 1167 |
+
"loss": 1.1881,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.7659574468085106,
|
| 1172 |
+
"grad_norm": 0.37211042642593384,
|
| 1173 |
+
"learning_rate": 1.4370801792624656e-05,
|
| 1174 |
+
"loss": 1.1375,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.7706855791962175,
|
| 1179 |
+
"grad_norm": 0.36233842372894287,
|
| 1180 |
+
"learning_rate": 1.3829563342637513e-05,
|
| 1181 |
+
"loss": 1.0862,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.7754137115839244,
|
| 1186 |
+
"grad_norm": 0.3661029040813446,
|
| 1187 |
+
"learning_rate": 1.329707356202699e-05,
|
| 1188 |
+
"loss": 1.0632,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.7801418439716312,
|
| 1193 |
+
"grad_norm": 0.3897812068462372,
|
| 1194 |
+
"learning_rate": 1.2773461245949247e-05,
|
| 1195 |
+
"loss": 1.1953,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.7848699763593381,
|
| 1200 |
+
"grad_norm": 0.3424377143383026,
|
| 1201 |
+
"learning_rate": 1.225885304233716e-05,
|
| 1202 |
+
"loss": 1.0391,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.789598108747045,
|
| 1207 |
+
"grad_norm": 0.363320916891098,
|
| 1208 |
+
"learning_rate": 1.1753373421267621e-05,
|
| 1209 |
+
"loss": 1.0834,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.7943262411347518,
|
| 1214 |
+
"grad_norm": 0.36856764554977417,
|
| 1215 |
+
"learning_rate": 1.125714464485551e-05,
|
| 1216 |
+
"loss": 1.0455,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.7990543735224587,
|
| 1221 |
+
"grad_norm": 0.36050939559936523,
|
| 1222 |
+
"learning_rate": 1.0770286737681701e-05,
|
| 1223 |
+
"loss": 1.0576,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.8037825059101655,
|
| 1228 |
+
"grad_norm": 0.3826131224632263,
|
| 1229 |
+
"learning_rate": 1.0292917457762325e-05,
|
| 1230 |
+
"loss": 1.1386,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.8085106382978723,
|
| 1235 |
+
"grad_norm": 0.37651652097702026,
|
| 1236 |
+
"learning_rate": 9.825152268066213e-06,
|
| 1237 |
+
"loss": 1.118,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.8132387706855791,
|
| 1242 |
+
"grad_norm": 0.397661030292511,
|
| 1243 |
+
"learning_rate": 9.367104308587494e-06,
|
| 1244 |
+
"loss": 1.1479,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.817966903073286,
|
| 1249 |
+
"grad_norm": 0.37330368161201477,
|
| 1250 |
+
"learning_rate": 8.91888436897997e-06,
|
| 1251 |
+
"loss": 1.047,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.8226950354609929,
|
| 1256 |
+
"grad_norm": 0.3637228012084961,
|
| 1257 |
+
"learning_rate": 8.480600861760124e-06,
|
| 1258 |
+
"loss": 1.0695,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.8274231678486997,
|
| 1263 |
+
"grad_norm": 0.3934190273284912,
|
| 1264 |
+
"learning_rate": 8.052359796084951e-06,
|
| 1265 |
+
"loss": 1.0716,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.8321513002364066,
|
| 1270 |
+
"grad_norm": 0.36225423216819763,
|
| 1271 |
+
"learning_rate": 7.63426475211113e-06,
|
| 1272 |
+
"loss": 1.0293,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.8368794326241135,
|
| 1277 |
+
"grad_norm": 0.3973957598209381,
|
| 1278 |
+
"learning_rate": 7.226416855941814e-06,
|
| 1279 |
+
"loss": 1.072,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.8416075650118203,
|
| 1284 |
+
"grad_norm": 0.42150604724884033,
|
| 1285 |
+
"learning_rate": 6.828914755166827e-06,
|
| 1286 |
+
"loss": 1.0884,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.8463356973995272,
|
| 1291 |
+
"grad_norm": 0.38192811608314514,
|
| 1292 |
+
"learning_rate": 6.441854595002477e-06,
|
| 1293 |
+
"loss": 1.0185,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.851063829787234,
|
| 1298 |
+
"grad_norm": 0.3970332145690918,
|
| 1299 |
+
"learning_rate": 6.065329995036573e-06,
|
| 1300 |
+
"loss": 1.0495,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.8557919621749409,
|
| 1305 |
+
"grad_norm": 0.4090059995651245,
|
| 1306 |
+
"learning_rate": 5.699432026584267e-06,
|
| 1307 |
+
"loss": 1.0787,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.8605200945626478,
|
| 1312 |
+
"grad_norm": 0.3918428421020508,
|
| 1313 |
+
"learning_rate": 5.344249190660428e-06,
|
| 1314 |
+
"loss": 1.004,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.8652482269503546,
|
| 1319 |
+
"grad_norm": 0.3922417163848877,
|
| 1320 |
+
"learning_rate": 4.999867396573499e-06,
|
| 1321 |
+
"loss": 1.0025,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.8699763593380615,
|
| 1326 |
+
"grad_norm": 0.40068820118904114,
|
| 1327 |
+
"learning_rate": 4.666369941146376e-06,
|
| 1328 |
+
"loss": 1.0989,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.8747044917257684,
|
| 1333 |
+
"grad_norm": 0.3978284001350403,
|
| 1334 |
+
"learning_rate": 4.343837488569058e-06,
|
| 1335 |
+
"loss": 1.0513,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.8794326241134752,
|
| 1340 |
+
"grad_norm": 0.4173754155635834,
|
| 1341 |
+
"learning_rate": 4.032348050888179e-06,
|
| 1342 |
+
"loss": 1.1416,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.8841607565011821,
|
| 1347 |
+
"grad_norm": 0.4168488085269928,
|
| 1348 |
+
"learning_rate": 3.731976969137929e-06,
|
| 1349 |
+
"loss": 1.0955,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.8888888888888888,
|
| 1354 |
+
"grad_norm": 0.4095589816570282,
|
| 1355 |
+
"learning_rate": 3.4427968951170288e-06,
|
| 1356 |
+
"loss": 1.0602,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.8936170212765957,
|
| 1361 |
+
"grad_norm": 0.4083250164985657,
|
| 1362 |
+
"learning_rate": 3.1648777738162494e-06,
|
| 1363 |
+
"loss": 1.0999,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.8983451536643026,
|
| 1368 |
+
"grad_norm": 0.4310641586780548,
|
| 1369 |
+
"learning_rate": 2.8982868265005458e-06,
|
| 1370 |
+
"loss": 1.1334,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.9030732860520094,
|
| 1375 |
+
"grad_norm": 0.4352046847343445,
|
| 1376 |
+
"learning_rate": 2.6430885344499946e-06,
|
| 1377 |
+
"loss": 1.0898,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.9078014184397163,
|
| 1382 |
+
"grad_norm": 0.45793259143829346,
|
| 1383 |
+
"learning_rate": 2.399344623363503e-06,
|
| 1384 |
+
"loss": 1.053,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.9125295508274232,
|
| 1389 |
+
"grad_norm": 0.4551216959953308,
|
| 1390 |
+
"learning_rate": 2.1671140484290142e-06,
|
| 1391 |
+
"loss": 1.1926,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.91725768321513,
|
| 1396 |
+
"grad_norm": 0.4517664313316345,
|
| 1397 |
+
"learning_rate": 1.946452980063773e-06,
|
| 1398 |
+
"loss": 1.0751,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.9219858156028369,
|
| 1403 |
+
"grad_norm": 0.48062533140182495,
|
| 1404 |
+
"learning_rate": 1.7374147903282178e-06,
|
| 1405 |
+
"loss": 1.17,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.9267139479905437,
|
| 1410 |
+
"grad_norm": 0.4554964005947113,
|
| 1411 |
+
"learning_rate": 1.5400500400166939e-06,
|
| 1412 |
+
"loss": 1.129,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.9314420803782506,
|
| 1417 |
+
"grad_norm": 0.47884511947631836,
|
| 1418 |
+
"learning_rate": 1.3544064664281265e-06,
|
| 1419 |
+
"loss": 1.1257,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.9361702127659575,
|
| 1424 |
+
"grad_norm": 0.48259878158569336,
|
| 1425 |
+
"learning_rate": 1.18052897181965e-06,
|
| 1426 |
+
"loss": 1.1116,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.9408983451536643,
|
| 1431 |
+
"grad_norm": 0.4733264446258545,
|
| 1432 |
+
"learning_rate": 1.0184596125459135e-06,
|
| 1433 |
+
"loss": 1.1168,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.9456264775413712,
|
| 1438 |
+
"grad_norm": 0.6054824590682983,
|
| 1439 |
+
"learning_rate": 8.682375888868166e-07,
|
| 1440 |
+
"loss": 1.2112,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.9456264775413712,
|
| 1445 |
+
"eval_loss": 1.1042157411575317,
|
| 1446 |
+
"eval_runtime": 38.4897,
|
| 1447 |
+
"eval_samples_per_second": 9.275,
|
| 1448 |
+
"eval_steps_per_second": 2.338,
|
| 1449 |
+
"step": 200
|
| 1450 |
}
|
| 1451 |
],
|
| 1452 |
"logging_steps": 1,
|
|
|
|
| 1475 |
"attributes": {}
|
| 1476 |
}
|
| 1477 |
},
|
| 1478 |
+
"total_flos": 2.6742965062651085e+17,
|
| 1479 |
"train_batch_size": 8,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|