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 sha256:3f6279195398fe54406c2705644f49563e7ebf45a4f92c96ec7e9ed26a64e355
|
| 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:a070972ec1fd65be280dc31e6493823dbbf4e7d2674229c197a49a80ff5f9697
|
| 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:c963cde87f5f69fd6101806c24b8bba742fcbb1a37eed1f3ed54a9b1d6fcaef5
|
| 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:d2d754412c61116546142914503e7369d0cc35d3c380a07e5218f595d76b6d96
|
| 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": 10.409,
|
| 1090 |
"eval_steps_per_second": 2.603,
|
| 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": 8,
|
| 1122 |
"trial_name": null,
|
| 1123 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.5162672996520996,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.03746721618583739,
|
| 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": 10.409,
|
| 1090 |
"eval_steps_per_second": 2.603,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.028287748220307232,
|
| 1095 |
+
"grad_norm": 0.22224098443984985,
|
| 1096 |
+
"learning_rate": 1.553232954407171e-05,
|
| 1097 |
+
"loss": 0.3951,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.02847508430123642,
|
| 1102 |
+
"grad_norm": 0.2531294822692871,
|
| 1103 |
+
"learning_rate": 1.4938160786375572e-05,
|
| 1104 |
+
"loss": 0.4491,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.028662420382165606,
|
| 1109 |
+
"grad_norm": 0.30002111196517944,
|
| 1110 |
+
"learning_rate": 1.435357758543015e-05,
|
| 1111 |
+
"loss": 0.5519,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.028849756463094792,
|
| 1116 |
+
"grad_norm": 0.2902084290981293,
|
| 1117 |
+
"learning_rate": 1.3778739760445552e-05,
|
| 1118 |
+
"loss": 0.3508,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.02903709254402398,
|
| 1123 |
+
"grad_norm": 0.3887433409690857,
|
| 1124 |
+
"learning_rate": 1.3213804466343421e-05,
|
| 1125 |
+
"loss": 0.4913,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.029224428624953166,
|
| 1130 |
+
"grad_norm": 0.3358180820941925,
|
| 1131 |
+
"learning_rate": 1.2658926150792322e-05,
|
| 1132 |
+
"loss": 0.4923,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.029411764705882353,
|
| 1137 |
+
"grad_norm": 0.41763749718666077,
|
| 1138 |
+
"learning_rate": 1.2114256511983274e-05,
|
| 1139 |
+
"loss": 0.4781,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.02959910078681154,
|
| 1144 |
+
"grad_norm": 0.33853206038475037,
|
| 1145 |
+
"learning_rate": 1.157994445715706e-05,
|
| 1146 |
+
"loss": 0.4668,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.029786436867740726,
|
| 1151 |
+
"grad_norm": 0.6292213797569275,
|
| 1152 |
+
"learning_rate": 1.1056136061894384e-05,
|
| 1153 |
+
"loss": 0.5098,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.029973772948669913,
|
| 1158 |
+
"grad_norm": 0.3926360309123993,
|
| 1159 |
+
"learning_rate": 1.0542974530180327e-05,
|
| 1160 |
+
"loss": 0.5681,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.0301611090295991,
|
| 1165 |
+
"grad_norm": 0.3899405002593994,
|
| 1166 |
+
"learning_rate": 1.0040600155253765e-05,
|
| 1167 |
+
"loss": 0.4905,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.030348445110528286,
|
| 1172 |
+
"grad_norm": 0.3343890905380249,
|
| 1173 |
+
"learning_rate": 9.549150281252633e-06,
|
| 1174 |
+
"loss": 0.4915,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.030535781191457476,
|
| 1179 |
+
"grad_norm": 0.34078794717788696,
|
| 1180 |
+
"learning_rate": 9.068759265665384e-06,
|
| 1181 |
+
"loss": 0.466,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.030723117272386663,
|
| 1186 |
+
"grad_norm": 0.33932462334632874,
|
| 1187 |
+
"learning_rate": 8.599558442598998e-06,
|
| 1188 |
+
"loss": 0.4713,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.03091045335331585,
|
| 1193 |
+
"grad_norm": 0.3398559093475342,
|
| 1194 |
+
"learning_rate": 8.141676086873572e-06,
|
| 1195 |
+
"loss": 0.5685,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.031097789434245036,
|
| 1200 |
+
"grad_norm": 0.3512920141220093,
|
| 1201 |
+
"learning_rate": 7.695237378953223e-06,
|
| 1202 |
+
"loss": 0.51,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.03128512551517422,
|
| 1207 |
+
"grad_norm": 0.343763530254364,
|
| 1208 |
+
"learning_rate": 7.260364370723044e-06,
|
| 1209 |
+
"loss": 0.4559,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.03147246159610341,
|
| 1214 |
+
"grad_norm": 0.37965717911720276,
|
| 1215 |
+
"learning_rate": 6.837175952121306e-06,
|
| 1216 |
+
"loss": 0.5447,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.03165979767703259,
|
| 1221 |
+
"grad_norm": 0.41954505443573,
|
| 1222 |
+
"learning_rate": 6.425787818636131e-06,
|
| 1223 |
+
"loss": 0.5379,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.03184713375796178,
|
| 1228 |
+
"grad_norm": 0.35495540499687195,
|
| 1229 |
+
"learning_rate": 6.026312439675552e-06,
|
| 1230 |
+
"loss": 0.4702,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.032034469838890973,
|
| 1235 |
+
"grad_norm": 0.3965103030204773,
|
| 1236 |
+
"learning_rate": 5.6388590278194096e-06,
|
| 1237 |
+
"loss": 0.5216,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.03222180591982016,
|
| 1242 |
+
"grad_norm": 0.3959389925003052,
|
| 1243 |
+
"learning_rate": 5.263533508961827e-06,
|
| 1244 |
+
"loss": 0.5908,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.03240914200074935,
|
| 1249 |
+
"grad_norm": 0.3949315547943115,
|
| 1250 |
+
"learning_rate": 4.900438493352055e-06,
|
| 1251 |
+
"loss": 0.594,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.03259647808167853,
|
| 1256 |
+
"grad_norm": 0.3838370442390442,
|
| 1257 |
+
"learning_rate": 4.549673247541875e-06,
|
| 1258 |
+
"loss": 0.5145,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.03278381416260772,
|
| 1263 |
+
"grad_norm": 0.4731585681438446,
|
| 1264 |
+
"learning_rate": 4.2113336672471245e-06,
|
| 1265 |
+
"loss": 0.5684,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.032971150243536904,
|
| 1270 |
+
"grad_norm": 0.43985679745674133,
|
| 1271 |
+
"learning_rate": 3.885512251130763e-06,
|
| 1272 |
+
"loss": 0.5455,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.033158486324466094,
|
| 1277 |
+
"grad_norm": 0.5088862776756287,
|
| 1278 |
+
"learning_rate": 3.5722980755146517e-06,
|
| 1279 |
+
"loss": 0.6215,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.03334582240539528,
|
| 1284 |
+
"grad_norm": 0.4820072650909424,
|
| 1285 |
+
"learning_rate": 3.271776770026963e-06,
|
| 1286 |
+
"loss": 0.5994,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.03353315848632447,
|
| 1291 |
+
"grad_norm": 0.4878690838813782,
|
| 1292 |
+
"learning_rate": 2.9840304941919415e-06,
|
| 1293 |
+
"loss": 0.5658,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.03372049456725365,
|
| 1298 |
+
"grad_norm": 0.5227633714675903,
|
| 1299 |
+
"learning_rate": 2.7091379149682685e-06,
|
| 1300 |
+
"loss": 0.6008,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.03390783064818284,
|
| 1305 |
+
"grad_norm": 0.4651011526584625,
|
| 1306 |
+
"learning_rate": 2.4471741852423237e-06,
|
| 1307 |
+
"loss": 0.6437,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.034095166729112024,
|
| 1312 |
+
"grad_norm": 0.5436872243881226,
|
| 1313 |
+
"learning_rate": 2.1982109232821178e-06,
|
| 1314 |
+
"loss": 0.6171,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.034282502810041214,
|
| 1319 |
+
"grad_norm": 0.521135687828064,
|
| 1320 |
+
"learning_rate": 1.962316193157593e-06,
|
| 1321 |
+
"loss": 0.6298,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.034469838890970404,
|
| 1326 |
+
"grad_norm": 0.5405279994010925,
|
| 1327 |
+
"learning_rate": 1.7395544861325718e-06,
|
| 1328 |
+
"loss": 0.6546,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.03465717497189959,
|
| 1333 |
+
"grad_norm": 0.47326454520225525,
|
| 1334 |
+
"learning_rate": 1.5299867030334814e-06,
|
| 1335 |
+
"loss": 0.5938,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.03484451105282878,
|
| 1340 |
+
"grad_norm": 0.5539429187774658,
|
| 1341 |
+
"learning_rate": 1.333670137599713e-06,
|
| 1342 |
+
"loss": 0.5548,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.03503184713375796,
|
| 1347 |
+
"grad_norm": 0.47998833656311035,
|
| 1348 |
+
"learning_rate": 1.1506584608200367e-06,
|
| 1349 |
+
"loss": 0.4494,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.03521918321468715,
|
| 1354 |
+
"grad_norm": 0.4977557063102722,
|
| 1355 |
+
"learning_rate": 9.810017062595322e-07,
|
| 1356 |
+
"loss": 0.5306,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.035406519295616334,
|
| 1361 |
+
"grad_norm": 0.5469347834587097,
|
| 1362 |
+
"learning_rate": 8.247462563808817e-07,
|
| 1363 |
+
"loss": 0.6215,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.035593855376545525,
|
| 1368 |
+
"grad_norm": 0.5438638925552368,
|
| 1369 |
+
"learning_rate": 6.819348298638839e-07,
|
| 1370 |
+
"loss": 0.547,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.03578119145747471,
|
| 1375 |
+
"grad_norm": 0.5327591300010681,
|
| 1376 |
+
"learning_rate": 5.526064699265753e-07,
|
| 1377 |
+
"loss": 0.4928,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.0359685275384039,
|
| 1382 |
+
"grad_norm": 0.47726142406463623,
|
| 1383 |
+
"learning_rate": 4.367965336512403e-07,
|
| 1384 |
+
"loss": 0.4826,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.03615586361933308,
|
| 1389 |
+
"grad_norm": 0.48082712292671204,
|
| 1390 |
+
"learning_rate": 3.3453668231809286e-07,
|
| 1391 |
+
"loss": 0.4501,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.03634319970026227,
|
| 1396 |
+
"grad_norm": 0.525841236114502,
|
| 1397 |
+
"learning_rate": 2.458548727494292e-07,
|
| 1398 |
+
"loss": 0.4643,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.036530535781191455,
|
| 1403 |
+
"grad_norm": 0.5812526345252991,
|
| 1404 |
+
"learning_rate": 1.7077534966650766e-07,
|
| 1405 |
+
"loss": 0.5527,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.036717871862120645,
|
| 1410 |
+
"grad_norm": 0.5962092280387878,
|
| 1411 |
+
"learning_rate": 1.0931863906127327e-07,
|
| 1412 |
+
"loss": 0.4762,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.03690520794304983,
|
| 1417 |
+
"grad_norm": 0.6215879321098328,
|
| 1418 |
+
"learning_rate": 6.150154258476315e-08,
|
| 1419 |
+
"loss": 0.5013,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.03709254402397902,
|
| 1424 |
+
"grad_norm": 0.5445265173912048,
|
| 1425 |
+
"learning_rate": 2.7337132953697554e-08,
|
| 1426 |
+
"loss": 0.5322,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.03727988010490821,
|
| 1431 |
+
"grad_norm": 0.6073286533355713,
|
| 1432 |
+
"learning_rate": 6.834750376549792e-09,
|
| 1433 |
+
"loss": 0.4223,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.03746721618583739,
|
| 1438 |
+
"grad_norm": 0.6887460350990295,
|
| 1439 |
+
"learning_rate": 0.0,
|
| 1440 |
+
"loss": 0.3864,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.03746721618583739,
|
| 1445 |
+
"eval_loss": 0.5162672996520996,
|
| 1446 |
+
"eval_runtime": 863.4541,
|
| 1447 |
+
"eval_samples_per_second": 10.412,
|
| 1448 |
+
"eval_steps_per_second": 2.603,
|
| 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": 3.763741378970911e+17,
|
| 1479 |
"train_batch_size": 8,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|