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 335604696
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e255a3bf76f9c80de3cfe450532822ca289f621c6d171dfc4e73ced437d756e7
|
| 3 |
size 335604696
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 170920084
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fa1b7c6fb34d453805bac2195d6cccbd5a438316d5666bbd4d9e21485c319b23
|
| 3 |
size 170920084
|
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:f5ad8286ea66b701d63e50598db6aef980a8ece5be55ee5d74d7201d2703c34f
|
| 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:b60d6f1383abda4776549360effee800fe6cfe2c0604503e9e3fbaa79347f790
|
| 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": 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": 4.948,
|
| 1090 |
"eval_steps_per_second": 1.238,
|
| 1091 |
"step": 150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1092 |
}
|
| 1093 |
],
|
| 1094 |
"logging_steps": 1,
|
|
@@ -1112,12 +1470,12 @@
|
|
| 1112 |
"should_evaluate": false,
|
| 1113 |
"should_log": false,
|
| 1114 |
"should_save": true,
|
| 1115 |
-
"should_training_stop":
|
| 1116 |
},
|
| 1117 |
"attributes": {}
|
| 1118 |
}
|
| 1119 |
},
|
| 1120 |
-
"total_flos":
|
| 1121 |
"train_batch_size": 4,
|
| 1122 |
"trial_name": null,
|
| 1123 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.4560202658176422,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.11421229209793704,
|
| 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": 4.948,
|
| 1090 |
"eval_steps_per_second": 1.238,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.08623028053394247,
|
| 1095 |
+
"grad_norm": 3.791814088821411,
|
| 1096 |
+
"learning_rate": 1.339526225485725e-05,
|
| 1097 |
+
"loss": 3.4645,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.08680134199443215,
|
| 1102 |
+
"grad_norm": 3.796891212463379,
|
| 1103 |
+
"learning_rate": 1.2890114781136224e-05,
|
| 1104 |
+
"loss": 3.4475,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.08737240345492184,
|
| 1109 |
+
"grad_norm": 3.6403725147247314,
|
| 1110 |
+
"learning_rate": 1.239251781296245e-05,
|
| 1111 |
+
"loss": 3.1833,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.08794346491541152,
|
| 1116 |
+
"grad_norm": 3.6974196434020996,
|
| 1117 |
+
"learning_rate": 1.1902641279244715e-05,
|
| 1118 |
+
"loss": 3.3696,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.08851452637590121,
|
| 1123 |
+
"grad_norm": 3.578291893005371,
|
| 1124 |
+
"learning_rate": 1.1420652472370497e-05,
|
| 1125 |
+
"loss": 3.1136,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.0890855878363909,
|
| 1130 |
+
"grad_norm": 3.4341652393341064,
|
| 1131 |
+
"learning_rate": 1.0946715991075805e-05,
|
| 1132 |
+
"loss": 2.9641,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.08965664929688058,
|
| 1137 |
+
"grad_norm": 3.451934576034546,
|
| 1138 |
+
"learning_rate": 1.0480993684234815e-05,
|
| 1139 |
+
"loss": 3.1792,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.09022771075737027,
|
| 1144 |
+
"grad_norm": 3.3800251483917236,
|
| 1145 |
+
"learning_rate": 1.0023644595588671e-05,
|
| 1146 |
+
"loss": 3.3144,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.09079877221785995,
|
| 1151 |
+
"grad_norm": 3.479318618774414,
|
| 1152 |
+
"learning_rate": 9.57482490943216e-06,
|
| 1153 |
+
"loss": 3.1648,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.09136983367834964,
|
| 1158 |
+
"grad_norm": 3.491649866104126,
|
| 1159 |
+
"learning_rate": 9.134687897276934e-06,
|
| 1160 |
+
"loss": 3.3233,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.09194089513883932,
|
| 1165 |
+
"grad_norm": 3.484065055847168,
|
| 1166 |
+
"learning_rate": 8.703383865509432e-06,
|
| 1167 |
+
"loss": 3.5239,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.09251195659932901,
|
| 1172 |
+
"grad_norm": 3.6290812492370605,
|
| 1173 |
+
"learning_rate": 8.281060104061394e-06,
|
| 1174 |
+
"loss": 3.2998,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.09308301805981868,
|
| 1179 |
+
"grad_norm": 3.6152350902557373,
|
| 1180 |
+
"learning_rate": 7.867860836110453e-06,
|
| 1181 |
+
"loss": 3.7931,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.09365407952030837,
|
| 1186 |
+
"grad_norm": 3.5166842937469482,
|
| 1187 |
+
"learning_rate": 7.463927168828087e-06,
|
| 1188 |
+
"loss": 3.5036,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.09422514098079805,
|
| 1193 |
+
"grad_norm": 3.266004800796509,
|
| 1194 |
+
"learning_rate": 7.069397045191617e-06,
|
| 1195 |
+
"loss": 2.7839,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.09479620244128774,
|
| 1200 |
+
"grad_norm": 3.5984013080596924,
|
| 1201 |
+
"learning_rate": 6.684405196876842e-06,
|
| 1202 |
+
"loss": 3.7792,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.09536726390177742,
|
| 1207 |
+
"grad_norm": 3.4622771739959717,
|
| 1208 |
+
"learning_rate": 6.309083098247264e-06,
|
| 1209 |
+
"loss": 3.4876,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.09593832536226711,
|
| 1214 |
+
"grad_norm": 3.397523880004883,
|
| 1215 |
+
"learning_rate": 5.943558921455733e-06,
|
| 1216 |
+
"loss": 3.202,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.0965093868227568,
|
| 1221 |
+
"grad_norm": 3.561593770980835,
|
| 1222 |
+
"learning_rate": 5.587957492673759e-06,
|
| 1223 |
+
"loss": 3.5857,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.09708044828324648,
|
| 1228 |
+
"grad_norm": 3.7165884971618652,
|
| 1229 |
+
"learning_rate": 5.2424002494635095e-06,
|
| 1230 |
+
"loss": 3.6163,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.09765150974373617,
|
| 1235 |
+
"grad_norm": 3.710353374481201,
|
| 1236 |
+
"learning_rate": 4.9070051993069636e-06,
|
| 1237 |
+
"loss": 3.9078,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.09822257120422585,
|
| 1242 |
+
"grad_norm": 3.53379225730896,
|
| 1243 |
+
"learning_rate": 4.581886879306507e-06,
|
| 1244 |
+
"loss": 3.5316,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.09879363266471554,
|
| 1249 |
+
"grad_norm": 3.5342020988464355,
|
| 1250 |
+
"learning_rate": 4.2671563170705725e-06,
|
| 1251 |
+
"loss": 3.5556,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.09936469412520522,
|
| 1256 |
+
"grad_norm": 3.5446951389312744,
|
| 1257 |
+
"learning_rate": 3.962920992797834e-06,
|
| 1258 |
+
"loss": 3.4027,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.09993575558569491,
|
| 1263 |
+
"grad_norm": 3.7609660625457764,
|
| 1264 |
+
"learning_rate": 3.6692848025728216e-06,
|
| 1265 |
+
"loss": 4.0196,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.1005068170461846,
|
| 1270 |
+
"grad_norm": 3.5949559211730957,
|
| 1271 |
+
"learning_rate": 3.38634802288549e-06,
|
| 1272 |
+
"loss": 3.7018,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.10107787850667428,
|
| 1277 |
+
"grad_norm": 3.4899206161499023,
|
| 1278 |
+
"learning_rate": 3.1142072763869042e-06,
|
| 1279 |
+
"loss": 3.4402,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.10164893996716397,
|
| 1284 |
+
"grad_norm": 3.4958362579345703,
|
| 1285 |
+
"learning_rate": 2.852955498892694e-06,
|
| 1286 |
+
"loss": 3.6609,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.10222000142765365,
|
| 1291 |
+
"grad_norm": 3.7367665767669678,
|
| 1292 |
+
"learning_rate": 2.6026819076455325e-06,
|
| 1293 |
+
"loss": 3.9386,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.10279106288814334,
|
| 1298 |
+
"grad_norm": 3.6101346015930176,
|
| 1299 |
+
"learning_rate": 2.36347197084755e-06,
|
| 1300 |
+
"loss": 3.589,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.10336212434863302,
|
| 1305 |
+
"grad_norm": 3.714442491531372,
|
| 1306 |
+
"learning_rate": 2.1354073784730253e-06,
|
| 1307 |
+
"loss": 3.9742,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.10393318580912271,
|
| 1312 |
+
"grad_norm": 3.8775899410247803,
|
| 1313 |
+
"learning_rate": 1.9185660143713184e-06,
|
| 1314 |
+
"loss": 4.4065,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.1045042472696124,
|
| 1319 |
+
"grad_norm": 4.041444301605225,
|
| 1320 |
+
"learning_rate": 1.7130219296696263e-06,
|
| 1321 |
+
"loss": 4.5485,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.10507530873010208,
|
| 1326 |
+
"grad_norm": 3.8321800231933594,
|
| 1327 |
+
"learning_rate": 1.5188453174845743e-06,
|
| 1328 |
+
"loss": 4.2913,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.10564637019059177,
|
| 1333 |
+
"grad_norm": 3.849231481552124,
|
| 1334 |
+
"learning_rate": 1.3361024889513333e-06,
|
| 1335 |
+
"loss": 4.6328,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.10621743165108145,
|
| 1340 |
+
"grad_norm": 3.960550546646118,
|
| 1341 |
+
"learning_rate": 1.16485585057844e-06,
|
| 1342 |
+
"loss": 4.6204,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.10678849311157114,
|
| 1347 |
+
"grad_norm": 3.7694177627563477,
|
| 1348 |
+
"learning_rate": 1.0051638829360127e-06,
|
| 1349 |
+
"loss": 3.6095,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.10735955457206081,
|
| 1354 |
+
"grad_norm": 3.936204671859741,
|
| 1355 |
+
"learning_rate": 8.570811206847189e-07,
|
| 1356 |
+
"loss": 4.067,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.1079306160325505,
|
| 1361 |
+
"grad_norm": 4.248170852661133,
|
| 1362 |
+
"learning_rate": 7.206581339521939e-07,
|
| 1363 |
+
"loss": 4.3996,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.10850167749304018,
|
| 1368 |
+
"grad_norm": 4.024265766143799,
|
| 1369 |
+
"learning_rate": 5.959415110634375e-07,
|
| 1370 |
+
"loss": 3.9334,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.10907273895352987,
|
| 1375 |
+
"grad_norm": 4.193387508392334,
|
| 1376 |
+
"learning_rate": 4.829738426309099e-07,
|
| 1377 |
+
"loss": 3.9381,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.10964380041401955,
|
| 1382 |
+
"grad_norm": 4.102433204650879,
|
| 1383 |
+
"learning_rate": 3.817937070098914e-07,
|
| 1384 |
+
"loss": 3.7026,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.11021486187450924,
|
| 1389 |
+
"grad_norm": 4.2977142333984375,
|
| 1390 |
+
"learning_rate": 2.9243565712400384e-07,
|
| 1391 |
+
"loss": 3.4472,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.11078592333499893,
|
| 1396 |
+
"grad_norm": 4.350393295288086,
|
| 1397 |
+
"learning_rate": 2.1493020866542365e-07,
|
| 1398 |
+
"loss": 3.5126,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.11135698479548861,
|
| 1403 |
+
"grad_norm": 4.65367317199707,
|
| 1404 |
+
"learning_rate": 1.4930382967379363e-07,
|
| 1405 |
+
"loss": 3.6799,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.1119280462559783,
|
| 1410 |
+
"grad_norm": 4.893271446228027,
|
| 1411 |
+
"learning_rate": 9.557893149741924e-08,
|
| 1412 |
+
"loss": 3.6397,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.11249910771646798,
|
| 1417 |
+
"grad_norm": 4.931285381317139,
|
| 1418 |
+
"learning_rate": 5.377386113981197e-08,
|
| 1419 |
+
"loss": 3.6552,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.11307016917695767,
|
| 1424 |
+
"grad_norm": 5.327000141143799,
|
| 1425 |
+
"learning_rate": 2.3902894994198286e-08,
|
| 1426 |
+
"loss": 3.3647,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.11364123063744735,
|
| 1431 |
+
"grad_norm": 6.226569175720215,
|
| 1432 |
+
"learning_rate": 5.976233968155164e-09,
|
| 1433 |
+
"loss": 3.279,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.11421229209793704,
|
| 1438 |
+
"grad_norm": 8.364439010620117,
|
| 1439 |
+
"learning_rate": 0.0,
|
| 1440 |
+
"loss": 3.8711,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.11421229209793704,
|
| 1445 |
+
"eval_loss": 0.4560202658176422,
|
| 1446 |
+
"eval_runtime": 596.8378,
|
| 1447 |
+
"eval_samples_per_second": 4.943,
|
| 1448 |
+
"eval_steps_per_second": 1.237,
|
| 1449 |
+
"step": 200
|
| 1450 |
}
|
| 1451 |
],
|
| 1452 |
"logging_steps": 1,
|
|
|
|
| 1470 |
"should_evaluate": false,
|
| 1471 |
"should_log": false,
|
| 1472 |
"should_save": true,
|
| 1473 |
+
"should_training_stop": true
|
| 1474 |
},
|
| 1475 |
"attributes": {}
|
| 1476 |
}
|
| 1477 |
},
|
| 1478 |
+
"total_flos": 4.849985165910344e+17,
|
| 1479 |
"train_batch_size": 4,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|