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 156926880
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:06ce627cdf978313cef135b956261249dacd1bd1df2c373f6ad783346e0af483
|
| 3 |
size 156926880
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 79968772
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18d9fdddc307660017992a02560363ce71fd150bf9e6b4a56622008485a71007
|
| 3 |
size 79968772
|
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 sha256:5a2b134c170c3861762b86730d8fb738bec4fdfede8e45ec8de681871bebc097
|
| 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 sha256:606e244bd4e377995f514a9afd58b985a53e5ab6166ef4b49183e16b2f949d47
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "miner_id_24/checkpoint-
|
| 4 |
-
"epoch": 1.
|
| 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": 28.452,
|
| 1090 |
"eval_steps_per_second": 7.148,
|
| 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": 0.6099563241004944,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 1.6618257261410787,
|
| 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": 28.452,
|
| 1090 |
"eval_steps_per_second": 7.148,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.2551867219917012,
|
| 1095 |
+
"grad_norm": 7.536319255828857,
|
| 1096 |
+
"learning_rate": 6.519525787871235e-05,
|
| 1097 |
+
"loss": 0.6467,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1.2634854771784232,
|
| 1102 |
+
"grad_norm": 7.516266345977783,
|
| 1103 |
+
"learning_rate": 6.476830129583206e-05,
|
| 1104 |
+
"loss": 0.5713,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1.2717842323651452,
|
| 1109 |
+
"grad_norm": 7.401801109313965,
|
| 1110 |
+
"learning_rate": 6.434016163555452e-05,
|
| 1111 |
+
"loss": 0.6164,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1.2800829875518671,
|
| 1116 |
+
"grad_norm": 6.253715515136719,
|
| 1117 |
+
"learning_rate": 6.391087319582264e-05,
|
| 1118 |
+
"loss": 0.5391,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.288381742738589,
|
| 1123 |
+
"grad_norm": 6.459432125091553,
|
| 1124 |
+
"learning_rate": 6.34804703666072e-05,
|
| 1125 |
+
"loss": 0.3725,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 1.2966804979253113,
|
| 1130 |
+
"grad_norm": 8.033556938171387,
|
| 1131 |
+
"learning_rate": 6.304898762715186e-05,
|
| 1132 |
+
"loss": 0.578,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 1.3049792531120332,
|
| 1137 |
+
"grad_norm": 6.300682067871094,
|
| 1138 |
+
"learning_rate": 6.261645954321109e-05,
|
| 1139 |
+
"loss": 0.4584,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 1.3132780082987552,
|
| 1144 |
+
"grad_norm": 6.387180328369141,
|
| 1145 |
+
"learning_rate": 6.21829207642811e-05,
|
| 1146 |
+
"loss": 0.4923,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 1.3215767634854771,
|
| 1151 |
+
"grad_norm": 7.160585880279541,
|
| 1152 |
+
"learning_rate": 6.174840602082412e-05,
|
| 1153 |
+
"loss": 0.5705,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 1.329875518672199,
|
| 1158 |
+
"grad_norm": 6.173452854156494,
|
| 1159 |
+
"learning_rate": 6.131295012148612e-05,
|
| 1160 |
+
"loss": 0.4962,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 1.3381742738589213,
|
| 1165 |
+
"grad_norm": 4.847928047180176,
|
| 1166 |
+
"learning_rate": 6.087658795030837e-05,
|
| 1167 |
+
"loss": 0.3703,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 1.3464730290456433,
|
| 1172 |
+
"grad_norm": 7.544972896575928,
|
| 1173 |
+
"learning_rate": 6.043935446393294e-05,
|
| 1174 |
+
"loss": 0.433,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 1.3547717842323652,
|
| 1179 |
+
"grad_norm": 6.9235358238220215,
|
| 1180 |
+
"learning_rate": 6.0001284688802226e-05,
|
| 1181 |
+
"loss": 0.6781,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 1.3630705394190872,
|
| 1186 |
+
"grad_norm": 7.10660457611084,
|
| 1187 |
+
"learning_rate": 5.956241371835312e-05,
|
| 1188 |
+
"loss": 0.5983,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 1.3713692946058091,
|
| 1193 |
+
"grad_norm": 4.3860955238342285,
|
| 1194 |
+
"learning_rate": 5.912277671020564e-05,
|
| 1195 |
+
"loss": 0.3219,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.379668049792531,
|
| 1200 |
+
"grad_norm": 6.816618919372559,
|
| 1201 |
+
"learning_rate": 5.868240888334653e-05,
|
| 1202 |
+
"loss": 0.4972,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 1.387966804979253,
|
| 1207 |
+
"grad_norm": 5.665396690368652,
|
| 1208 |
+
"learning_rate": 5.824134551530783e-05,
|
| 1209 |
+
"loss": 0.4513,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 1.396265560165975,
|
| 1214 |
+
"grad_norm": 8.053374290466309,
|
| 1215 |
+
"learning_rate": 5.7799621939340896e-05,
|
| 1216 |
+
"loss": 0.3688,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 1.404564315352697,
|
| 1221 |
+
"grad_norm": 5.979791164398193,
|
| 1222 |
+
"learning_rate": 5.735727354158581e-05,
|
| 1223 |
+
"loss": 0.3598,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 1.412863070539419,
|
| 1228 |
+
"grad_norm": 4.926863193511963,
|
| 1229 |
+
"learning_rate": 5.691433575823666e-05,
|
| 1230 |
+
"loss": 0.3195,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 1.4211618257261411,
|
| 1235 |
+
"grad_norm": 6.836885452270508,
|
| 1236 |
+
"learning_rate": 5.6470844072702764e-05,
|
| 1237 |
+
"loss": 0.6218,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.429460580912863,
|
| 1242 |
+
"grad_norm": 7.920914649963379,
|
| 1243 |
+
"learning_rate": 5.602683401276615e-05,
|
| 1244 |
+
"loss": 0.4448,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 1.437759336099585,
|
| 1249 |
+
"grad_norm": 3.9926419258117676,
|
| 1250 |
+
"learning_rate": 5.55823411477354e-05,
|
| 1251 |
+
"loss": 0.208,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 1.446058091286307,
|
| 1256 |
+
"grad_norm": 5.798760414123535,
|
| 1257 |
+
"learning_rate": 5.513740108559622e-05,
|
| 1258 |
+
"loss": 0.3248,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 1.454356846473029,
|
| 1263 |
+
"grad_norm": 7.048202991485596,
|
| 1264 |
+
"learning_rate": 5.469204947015897e-05,
|
| 1265 |
+
"loss": 0.4915,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 1.4626556016597512,
|
| 1270 |
+
"grad_norm": 4.450428485870361,
|
| 1271 |
+
"learning_rate": 5.424632197820324e-05,
|
| 1272 |
+
"loss": 0.2565,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 1.4709543568464731,
|
| 1277 |
+
"grad_norm": 7.559440612792969,
|
| 1278 |
+
"learning_rate": 5.3800254316619806e-05,
|
| 1279 |
+
"loss": 0.5286,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 1.479253112033195,
|
| 1284 |
+
"grad_norm": 7.2518086433410645,
|
| 1285 |
+
"learning_rate": 5.335388221955012e-05,
|
| 1286 |
+
"loss": 0.5451,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 1.487551867219917,
|
| 1291 |
+
"grad_norm": 5.655428886413574,
|
| 1292 |
+
"learning_rate": 5.290724144552379e-05,
|
| 1293 |
+
"loss": 0.3642,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 1.495850622406639,
|
| 1298 |
+
"grad_norm": 6.303919792175293,
|
| 1299 |
+
"learning_rate": 5.246036777459391e-05,
|
| 1300 |
+
"loss": 0.4304,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.504149377593361,
|
| 1305 |
+
"grad_norm": 9.382453918457031,
|
| 1306 |
+
"learning_rate": 5.201329700547076e-05,
|
| 1307 |
+
"loss": 0.5932,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 1.512448132780083,
|
| 1312 |
+
"grad_norm": 5.657342910766602,
|
| 1313 |
+
"learning_rate": 5.1566064952654014e-05,
|
| 1314 |
+
"loss": 0.431,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 1.520746887966805,
|
| 1319 |
+
"grad_norm": 5.608704566955566,
|
| 1320 |
+
"learning_rate": 5.1118707443563665e-05,
|
| 1321 |
+
"loss": 0.4258,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 1.5290456431535269,
|
| 1326 |
+
"grad_norm": 6.01896333694458,
|
| 1327 |
+
"learning_rate": 5.0671260315669875e-05,
|
| 1328 |
+
"loss": 0.3907,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 1.5373443983402488,
|
| 1333 |
+
"grad_norm": 8.329726219177246,
|
| 1334 |
+
"learning_rate": 5.022375941362217e-05,
|
| 1335 |
+
"loss": 0.3328,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 1.5456431535269708,
|
| 1340 |
+
"grad_norm": 6.260307788848877,
|
| 1341 |
+
"learning_rate": 4.977624058637783e-05,
|
| 1342 |
+
"loss": 0.3374,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.553941908713693,
|
| 1347 |
+
"grad_norm": 6.407407760620117,
|
| 1348 |
+
"learning_rate": 4.9328739684330137e-05,
|
| 1349 |
+
"loss": 0.432,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 1.562240663900415,
|
| 1354 |
+
"grad_norm": 5.673023700714111,
|
| 1355 |
+
"learning_rate": 4.8881292556436354e-05,
|
| 1356 |
+
"loss": 0.2581,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 1.570539419087137,
|
| 1361 |
+
"grad_norm": 6.290989875793457,
|
| 1362 |
+
"learning_rate": 4.8433935047346e-05,
|
| 1363 |
+
"loss": 0.387,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 1.578838174273859,
|
| 1368 |
+
"grad_norm": 5.269549369812012,
|
| 1369 |
+
"learning_rate": 4.798670299452926e-05,
|
| 1370 |
+
"loss": 0.3035,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 1.587136929460581,
|
| 1375 |
+
"grad_norm": 7.64688777923584,
|
| 1376 |
+
"learning_rate": 4.7539632225406095e-05,
|
| 1377 |
+
"loss": 0.3908,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 1.595435684647303,
|
| 1382 |
+
"grad_norm": 11.037405014038086,
|
| 1383 |
+
"learning_rate": 4.709275855447621e-05,
|
| 1384 |
+
"loss": 0.8535,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 1.603734439834025,
|
| 1389 |
+
"grad_norm": 5.834881782531738,
|
| 1390 |
+
"learning_rate": 4.6646117780449876e-05,
|
| 1391 |
+
"loss": 0.2963,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 1.612033195020747,
|
| 1396 |
+
"grad_norm": 5.2544403076171875,
|
| 1397 |
+
"learning_rate": 4.6199745683380206e-05,
|
| 1398 |
+
"loss": 0.2959,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 1.620331950207469,
|
| 1403 |
+
"grad_norm": 8.141179084777832,
|
| 1404 |
+
"learning_rate": 4.5753678021796755e-05,
|
| 1405 |
+
"loss": 0.634,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.6286307053941909,
|
| 1410 |
+
"grad_norm": 4.7402191162109375,
|
| 1411 |
+
"learning_rate": 4.530795052984104e-05,
|
| 1412 |
+
"loss": 0.2887,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 1.6369294605809128,
|
| 1417 |
+
"grad_norm": 6.756740093231201,
|
| 1418 |
+
"learning_rate": 4.48625989144038e-05,
|
| 1419 |
+
"loss": 0.4239,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.6452282157676348,
|
| 1424 |
+
"grad_norm": 6.466264247894287,
|
| 1425 |
+
"learning_rate": 4.4417658852264614e-05,
|
| 1426 |
+
"loss": 0.2381,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.6535269709543567,
|
| 1431 |
+
"grad_norm": 5.09955358505249,
|
| 1432 |
+
"learning_rate": 4.397316598723385e-05,
|
| 1433 |
+
"loss": 0.3166,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 1.6618257261410787,
|
| 1438 |
+
"grad_norm": 5.525445461273193,
|
| 1439 |
+
"learning_rate": 4.352915592729723e-05,
|
| 1440 |
+
"loss": 0.3016,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 1.6618257261410787,
|
| 1445 |
+
"eval_loss": 0.6099563241004944,
|
| 1446 |
+
"eval_runtime": 7.1625,
|
| 1447 |
+
"eval_samples_per_second": 28.342,
|
| 1448 |
+
"eval_steps_per_second": 7.12,
|
| 1449 |
+
"step": 200
|
| 1450 |
}
|
| 1451 |
],
|
| 1452 |
"logging_steps": 1,
|
|
|
|
| 1475 |
"attributes": {}
|
| 1476 |
}
|
| 1477 |
},
|
| 1478 |
+
"total_flos": 7.94731890081792e+16,
|
| 1479 |
"train_batch_size": 8,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|