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 671149168
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac0fd6dcc3bfade4220a01c1cd9778a45e9059a321be9ec3b83b8842c3f99897
|
| 3 |
size 671149168
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 341314196
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f93f8126e690f89bf4ab8e628e51c2d056000d60b265affc0c554f930e522d31
|
| 3 |
size 341314196
|
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 sha256:57b6a7d4ae70a3de277300394e6f6a752ba31db45a63b7dd788f2b4b2e68b2c5
|
| 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 sha256:4ac116b8169c53ab649a7f15f2f32735f2c71ec2f803f70de8c655a513ee9cfc
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 2.
|
| 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": 11.142,
|
| 1090 |
"eval_steps_per_second": 2.785,
|
| 1091 |
"step": 150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1092 |
}
|
| 1093 |
],
|
| 1094 |
"logging_steps": 1,
|
|
@@ -1117,7 +1475,7 @@
|
|
| 1117 |
"attributes": {}
|
| 1118 |
}
|
| 1119 |
},
|
| 1120 |
-
"total_flos":
|
| 1121 |
"train_batch_size": 8,
|
| 1122 |
"trial_name": null,
|
| 1123 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 2.310253143310547,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.24837007140639553,
|
| 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": 11.142,
|
| 1090 |
"eval_steps_per_second": 2.785,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.1875194039118286,
|
| 1095 |
+
"grad_norm": 6.803307056427002,
|
| 1096 |
+
"learning_rate": 7.107005538862646e-05,
|
| 1097 |
+
"loss": 10.561,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.1887612542688606,
|
| 1102 |
+
"grad_norm": 8.229329109191895,
|
| 1103 |
+
"learning_rate": 7.07041155014006e-05,
|
| 1104 |
+
"loss": 10.2723,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.1900031046258926,
|
| 1109 |
+
"grad_norm": 8.097644805908203,
|
| 1110 |
+
"learning_rate": 7.033683215379002e-05,
|
| 1111 |
+
"loss": 10.7189,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.19124495498292457,
|
| 1116 |
+
"grad_norm": 9.763362884521484,
|
| 1117 |
+
"learning_rate": 6.996822917828477e-05,
|
| 1118 |
+
"loss": 10.8785,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.19248680533995655,
|
| 1123 |
+
"grad_norm": 8.936421394348145,
|
| 1124 |
+
"learning_rate": 6.959833049300377e-05,
|
| 1125 |
+
"loss": 10.7473,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.19372865569698852,
|
| 1130 |
+
"grad_norm": 14.159116744995117,
|
| 1131 |
+
"learning_rate": 6.922716010014255e-05,
|
| 1132 |
+
"loss": 11.4862,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.1949705060540205,
|
| 1137 |
+
"grad_norm": 10.795000076293945,
|
| 1138 |
+
"learning_rate": 6.885474208441603e-05,
|
| 1139 |
+
"loss": 10.2663,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.19621235641105247,
|
| 1144 |
+
"grad_norm": 12.574379920959473,
|
| 1145 |
+
"learning_rate": 6.848110061149556e-05,
|
| 1146 |
+
"loss": 10.6066,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.19745420676808445,
|
| 1151 |
+
"grad_norm": 13.208904266357422,
|
| 1152 |
+
"learning_rate": 6.810625992644085e-05,
|
| 1153 |
+
"loss": 10.2697,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.19869605712511643,
|
| 1158 |
+
"grad_norm": 12.457254409790039,
|
| 1159 |
+
"learning_rate": 6.773024435212678e-05,
|
| 1160 |
+
"loss": 12.1264,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.1999379074821484,
|
| 1165 |
+
"grad_norm": 11.590801239013672,
|
| 1166 |
+
"learning_rate": 6.735307828766515e-05,
|
| 1167 |
+
"loss": 11.8134,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.20117975783918038,
|
| 1172 |
+
"grad_norm": 8.898362159729004,
|
| 1173 |
+
"learning_rate": 6.697478620682137e-05,
|
| 1174 |
+
"loss": 10.6429,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.20242160819621235,
|
| 1179 |
+
"grad_norm": 9.280109405517578,
|
| 1180 |
+
"learning_rate": 6.659539265642643e-05,
|
| 1181 |
+
"loss": 10.5507,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.20366345855324433,
|
| 1186 |
+
"grad_norm": 11.889431953430176,
|
| 1187 |
+
"learning_rate": 6.621492225478414e-05,
|
| 1188 |
+
"loss": 11.0789,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.2049053089102763,
|
| 1193 |
+
"grad_norm": 5.70781946182251,
|
| 1194 |
+
"learning_rate": 6.583339969007363e-05,
|
| 1195 |
+
"loss": 10.9493,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.20614715926730828,
|
| 1200 |
+
"grad_norm": 8.63260269165039,
|
| 1201 |
+
"learning_rate": 6.545084971874738e-05,
|
| 1202 |
+
"loss": 10.3209,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.20738900962434026,
|
| 1207 |
+
"grad_norm": 7.887843132019043,
|
| 1208 |
+
"learning_rate": 6.506729716392481e-05,
|
| 1209 |
+
"loss": 8.8037,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.20863085998137224,
|
| 1214 |
+
"grad_norm": 11.230926513671875,
|
| 1215 |
+
"learning_rate": 6.468276691378155e-05,
|
| 1216 |
+
"loss": 10.7108,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.2098727103384042,
|
| 1221 |
+
"grad_norm": 11.069745063781738,
|
| 1222 |
+
"learning_rate": 6.429728391993446e-05,
|
| 1223 |
+
"loss": 10.1439,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.2111145606954362,
|
| 1228 |
+
"grad_norm": 15.257699966430664,
|
| 1229 |
+
"learning_rate": 6.391087319582264e-05,
|
| 1230 |
+
"loss": 10.3077,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.2123564110524682,
|
| 1235 |
+
"grad_norm": 13.242205619812012,
|
| 1236 |
+
"learning_rate": 6.35235598150842e-05,
|
| 1237 |
+
"loss": 10.6351,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.21359826140950017,
|
| 1242 |
+
"grad_norm": 18.31103515625,
|
| 1243 |
+
"learning_rate": 6.313536890992935e-05,
|
| 1244 |
+
"loss": 10.1206,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.21484011176653214,
|
| 1249 |
+
"grad_norm": 14.47295093536377,
|
| 1250 |
+
"learning_rate": 6.274632566950967e-05,
|
| 1251 |
+
"loss": 8.8579,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.21608196212356412,
|
| 1256 |
+
"grad_norm": 17.428049087524414,
|
| 1257 |
+
"learning_rate": 6.235645533828349e-05,
|
| 1258 |
+
"loss": 9.6196,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.2173238124805961,
|
| 1263 |
+
"grad_norm": 16.387075424194336,
|
| 1264 |
+
"learning_rate": 6.19657832143779e-05,
|
| 1265 |
+
"loss": 10.0083,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.21856566283762807,
|
| 1270 |
+
"grad_norm": 15.025436401367188,
|
| 1271 |
+
"learning_rate": 6.157433464794716e-05,
|
| 1272 |
+
"loss": 10.7442,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.21980751319466005,
|
| 1277 |
+
"grad_norm": 12.375956535339355,
|
| 1278 |
+
"learning_rate": 6.118213503952779e-05,
|
| 1279 |
+
"loss": 10.269,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.22104936355169202,
|
| 1284 |
+
"grad_norm": 10.693756103515625,
|
| 1285 |
+
"learning_rate": 6.078920983839031e-05,
|
| 1286 |
+
"loss": 9.8473,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.222291213908724,
|
| 1291 |
+
"grad_norm": 9.529464721679688,
|
| 1292 |
+
"learning_rate": 6.0395584540887963e-05,
|
| 1293 |
+
"loss": 10.1938,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.22353306426575598,
|
| 1298 |
+
"grad_norm": 8.467986106872559,
|
| 1299 |
+
"learning_rate": 6.0001284688802226e-05,
|
| 1300 |
+
"loss": 9.6416,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.22477491462278795,
|
| 1305 |
+
"grad_norm": 8.995750427246094,
|
| 1306 |
+
"learning_rate": 5.960633586768543e-05,
|
| 1307 |
+
"loss": 8.5713,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.22601676497981993,
|
| 1312 |
+
"grad_norm": 9.754181861877441,
|
| 1313 |
+
"learning_rate": 5.921076370520058e-05,
|
| 1314 |
+
"loss": 9.4782,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.2272586153368519,
|
| 1319 |
+
"grad_norm": 8.055115699768066,
|
| 1320 |
+
"learning_rate": 5.8814593869458455e-05,
|
| 1321 |
+
"loss": 9.0996,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.22850046569388388,
|
| 1326 |
+
"grad_norm": 10.501749038696289,
|
| 1327 |
+
"learning_rate": 5.841785206735192e-05,
|
| 1328 |
+
"loss": 10.3803,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.22974231605091586,
|
| 1333 |
+
"grad_norm": 13.05416488647461,
|
| 1334 |
+
"learning_rate": 5.8020564042888015e-05,
|
| 1335 |
+
"loss": 9.4394,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.23098416640794783,
|
| 1340 |
+
"grad_norm": 11.48218822479248,
|
| 1341 |
+
"learning_rate": 5.762275557551727e-05,
|
| 1342 |
+
"loss": 8.1828,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.2322260167649798,
|
| 1347 |
+
"grad_norm": 9.102746963500977,
|
| 1348 |
+
"learning_rate": 5.7224452478461064e-05,
|
| 1349 |
+
"loss": 9.1802,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.23346786712201179,
|
| 1354 |
+
"grad_norm": 13.362663269042969,
|
| 1355 |
+
"learning_rate": 5.682568059703659e-05,
|
| 1356 |
+
"loss": 10.5763,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.2347097174790438,
|
| 1361 |
+
"grad_norm": 14.221250534057617,
|
| 1362 |
+
"learning_rate": 5.642646580697973e-05,
|
| 1363 |
+
"loss": 8.3783,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.23595156783607577,
|
| 1368 |
+
"grad_norm": 17.283653259277344,
|
| 1369 |
+
"learning_rate": 5.602683401276615e-05,
|
| 1370 |
+
"loss": 8.8715,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.23719341819310774,
|
| 1375 |
+
"grad_norm": 20.920652389526367,
|
| 1376 |
+
"learning_rate": 5.562681114593028e-05,
|
| 1377 |
+
"loss": 7.1134,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.23843526855013972,
|
| 1382 |
+
"grad_norm": 20.802019119262695,
|
| 1383 |
+
"learning_rate": 5.522642316338268e-05,
|
| 1384 |
+
"loss": 8.4245,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.2396771189071717,
|
| 1389 |
+
"grad_norm": 16.49863052368164,
|
| 1390 |
+
"learning_rate": 5.482569604572576e-05,
|
| 1391 |
+
"loss": 7.6499,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.24091896926420367,
|
| 1396 |
+
"grad_norm": 12.959442138671875,
|
| 1397 |
+
"learning_rate": 5.442465579556793e-05,
|
| 1398 |
+
"loss": 8.1491,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.24216081962123565,
|
| 1403 |
+
"grad_norm": 14.34575080871582,
|
| 1404 |
+
"learning_rate": 5.402332843583631e-05,
|
| 1405 |
+
"loss": 9.1671,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.24340266997826762,
|
| 1410 |
+
"grad_norm": 14.149574279785156,
|
| 1411 |
+
"learning_rate": 5.3621740008088126e-05,
|
| 1412 |
+
"loss": 7.9398,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.2446445203352996,
|
| 1417 |
+
"grad_norm": 13.768582344055176,
|
| 1418 |
+
"learning_rate": 5.321991657082097e-05,
|
| 1419 |
+
"loss": 7.9755,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.24588637069233157,
|
| 1424 |
+
"grad_norm": 13.12028980255127,
|
| 1425 |
+
"learning_rate": 5.281788419778187e-05,
|
| 1426 |
+
"loss": 5.7906,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.24712822104936355,
|
| 1431 |
+
"grad_norm": 18.232118606567383,
|
| 1432 |
+
"learning_rate": 5.2415668976275355e-05,
|
| 1433 |
+
"loss": 6.1965,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.24837007140639553,
|
| 1438 |
+
"grad_norm": 16.104488372802734,
|
| 1439 |
+
"learning_rate": 5.201329700547076e-05,
|
| 1440 |
+
"loss": 9.2066,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.24837007140639553,
|
| 1445 |
+
"eval_loss": 2.310253143310547,
|
| 1446 |
+
"eval_runtime": 121.7061,
|
| 1447 |
+
"eval_samples_per_second": 11.142,
|
| 1448 |
+
"eval_steps_per_second": 2.785,
|
| 1449 |
+
"step": 200
|
| 1450 |
}
|
| 1451 |
],
|
| 1452 |
"logging_steps": 1,
|
|
|
|
| 1475 |
"attributes": {}
|
| 1476 |
}
|
| 1477 |
},
|
| 1478 |
+
"total_flos": 3.4987901319747994e+17,
|
| 1479 |
"train_batch_size": 8,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|