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:e22824e8d637419563622dbce6af753e8401ad3fd08251fea57a1120b0fe612c
|
| 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 sha256:714b7b370a928cd5cf4d461252c2c1acead39161ba325c5912a482e1f675677f
|
| 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:f1846127c77f897479c4ee233540552535ade54fc559876f86075c60ad002a31
|
| 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,
|
|
@@ -1081,6 +1081,364 @@
|
|
| 1081 |
"eval_samples_per_second": 18.785,
|
| 1082 |
"eval_steps_per_second": 4.696,
|
| 1083 |
"step": 150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1084 |
}
|
| 1085 |
],
|
| 1086 |
"logging_steps": 1,
|
|
@@ -1104,12 +1462,12 @@
|
|
| 1104 |
"should_evaluate": false,
|
| 1105 |
"should_log": false,
|
| 1106 |
"should_save": true,
|
| 1107 |
-
"should_training_stop":
|
| 1108 |
},
|
| 1109 |
"attributes": {}
|
| 1110 |
}
|
| 1111 |
},
|
| 1112 |
-
"total_flos": 2.
|
| 1113 |
"train_batch_size": 8,
|
| 1114 |
"trial_name": null,
|
| 1115 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.7135888934135437,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.019689399719426053,
|
| 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,
|
|
|
|
| 1081 |
"eval_samples_per_second": 18.785,
|
| 1082 |
"eval_steps_per_second": 4.696,
|
| 1083 |
"step": 150
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.014865496788166671,
|
| 1087 |
+
"grad_norm": 0.11930377781391144,
|
| 1088 |
+
"learning_rate": 1.553232954407171e-05,
|
| 1089 |
+
"loss": 0.5386,
|
| 1090 |
+
"step": 151
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.0149639437867638,
|
| 1094 |
+
"grad_norm": 0.16247321665287018,
|
| 1095 |
+
"learning_rate": 1.4938160786375572e-05,
|
| 1096 |
+
"loss": 0.6219,
|
| 1097 |
+
"step": 152
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.015062390785360931,
|
| 1101 |
+
"grad_norm": 0.20031054317951202,
|
| 1102 |
+
"learning_rate": 1.435357758543015e-05,
|
| 1103 |
+
"loss": 0.7861,
|
| 1104 |
+
"step": 153
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.015160837783958062,
|
| 1108 |
+
"grad_norm": 0.20611421763896942,
|
| 1109 |
+
"learning_rate": 1.3778739760445552e-05,
|
| 1110 |
+
"loss": 0.7678,
|
| 1111 |
+
"step": 154
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.015259284782555192,
|
| 1115 |
+
"grad_norm": 0.22807583212852478,
|
| 1116 |
+
"learning_rate": 1.3213804466343421e-05,
|
| 1117 |
+
"loss": 0.7018,
|
| 1118 |
+
"step": 155
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.015357731781152323,
|
| 1122 |
+
"grad_norm": 0.22471506893634796,
|
| 1123 |
+
"learning_rate": 1.2658926150792322e-05,
|
| 1124 |
+
"loss": 0.6784,
|
| 1125 |
+
"step": 156
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.015456178779749452,
|
| 1129 |
+
"grad_norm": 0.26132553815841675,
|
| 1130 |
+
"learning_rate": 1.2114256511983274e-05,
|
| 1131 |
+
"loss": 0.6765,
|
| 1132 |
+
"step": 157
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.015554625778346582,
|
| 1136 |
+
"grad_norm": 0.23332355916500092,
|
| 1137 |
+
"learning_rate": 1.157994445715706e-05,
|
| 1138 |
+
"loss": 0.7644,
|
| 1139 |
+
"step": 158
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 0.01565307277694371,
|
| 1143 |
+
"grad_norm": 0.24738949537277222,
|
| 1144 |
+
"learning_rate": 1.1056136061894384e-05,
|
| 1145 |
+
"loss": 0.8344,
|
| 1146 |
+
"step": 159
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.015751519775540842,
|
| 1150 |
+
"grad_norm": 0.2664797306060791,
|
| 1151 |
+
"learning_rate": 1.0542974530180327e-05,
|
| 1152 |
+
"loss": 0.7323,
|
| 1153 |
+
"step": 160
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.015849966774137973,
|
| 1157 |
+
"grad_norm": 0.27797332406044006,
|
| 1158 |
+
"learning_rate": 1.0040600155253765e-05,
|
| 1159 |
+
"loss": 0.7011,
|
| 1160 |
+
"step": 161
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.015948413772735103,
|
| 1164 |
+
"grad_norm": 0.25038793683052063,
|
| 1165 |
+
"learning_rate": 9.549150281252633e-06,
|
| 1166 |
+
"loss": 0.7564,
|
| 1167 |
+
"step": 162
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.016046860771332234,
|
| 1171 |
+
"grad_norm": 0.24009868502616882,
|
| 1172 |
+
"learning_rate": 9.068759265665384e-06,
|
| 1173 |
+
"loss": 0.7167,
|
| 1174 |
+
"step": 163
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.016145307769929364,
|
| 1178 |
+
"grad_norm": 0.28158503770828247,
|
| 1179 |
+
"learning_rate": 8.599558442598998e-06,
|
| 1180 |
+
"loss": 0.7722,
|
| 1181 |
+
"step": 164
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.016243754768526495,
|
| 1185 |
+
"grad_norm": 0.2530911862850189,
|
| 1186 |
+
"learning_rate": 8.141676086873572e-06,
|
| 1187 |
+
"loss": 0.6774,
|
| 1188 |
+
"step": 165
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.016342201767123626,
|
| 1192 |
+
"grad_norm": 0.24448774755001068,
|
| 1193 |
+
"learning_rate": 7.695237378953223e-06,
|
| 1194 |
+
"loss": 0.7117,
|
| 1195 |
+
"step": 166
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 0.016440648765720756,
|
| 1199 |
+
"grad_norm": 0.2513751685619354,
|
| 1200 |
+
"learning_rate": 7.260364370723044e-06,
|
| 1201 |
+
"loss": 0.6804,
|
| 1202 |
+
"step": 167
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.016539095764317887,
|
| 1206 |
+
"grad_norm": 0.33193540573120117,
|
| 1207 |
+
"learning_rate": 6.837175952121306e-06,
|
| 1208 |
+
"loss": 0.7996,
|
| 1209 |
+
"step": 168
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.016637542762915014,
|
| 1213 |
+
"grad_norm": 0.29664281010627747,
|
| 1214 |
+
"learning_rate": 6.425787818636131e-06,
|
| 1215 |
+
"loss": 0.7504,
|
| 1216 |
+
"step": 169
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.016735989761512145,
|
| 1220 |
+
"grad_norm": 0.27495744824409485,
|
| 1221 |
+
"learning_rate": 6.026312439675552e-06,
|
| 1222 |
+
"loss": 0.7443,
|
| 1223 |
+
"step": 170
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.016834436760109275,
|
| 1227 |
+
"grad_norm": 0.2867414355278015,
|
| 1228 |
+
"learning_rate": 5.6388590278194096e-06,
|
| 1229 |
+
"loss": 0.8144,
|
| 1230 |
+
"step": 171
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 0.016932883758706406,
|
| 1234 |
+
"grad_norm": 0.27185213565826416,
|
| 1235 |
+
"learning_rate": 5.263533508961827e-06,
|
| 1236 |
+
"loss": 0.7643,
|
| 1237 |
+
"step": 172
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.017031330757303537,
|
| 1241 |
+
"grad_norm": 0.3129732012748718,
|
| 1242 |
+
"learning_rate": 4.900438493352055e-06,
|
| 1243 |
+
"loss": 0.8846,
|
| 1244 |
+
"step": 173
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.017129777755900667,
|
| 1248 |
+
"grad_norm": 0.30969539284706116,
|
| 1249 |
+
"learning_rate": 4.549673247541875e-06,
|
| 1250 |
+
"loss": 0.761,
|
| 1251 |
+
"step": 174
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.017228224754497798,
|
| 1255 |
+
"grad_norm": 0.29353201389312744,
|
| 1256 |
+
"learning_rate": 4.2113336672471245e-06,
|
| 1257 |
+
"loss": 0.814,
|
| 1258 |
+
"step": 175
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.01732667175309493,
|
| 1262 |
+
"grad_norm": 0.2705335021018982,
|
| 1263 |
+
"learning_rate": 3.885512251130763e-06,
|
| 1264 |
+
"loss": 0.7505,
|
| 1265 |
+
"step": 176
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.01742511875169206,
|
| 1269 |
+
"grad_norm": 0.30902302265167236,
|
| 1270 |
+
"learning_rate": 3.5722980755146517e-06,
|
| 1271 |
+
"loss": 0.7311,
|
| 1272 |
+
"step": 177
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.01752356575028919,
|
| 1276 |
+
"grad_norm": 0.30058932304382324,
|
| 1277 |
+
"learning_rate": 3.271776770026963e-06,
|
| 1278 |
+
"loss": 0.7389,
|
| 1279 |
+
"step": 178
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.017622012748886317,
|
| 1283 |
+
"grad_norm": 0.3121153712272644,
|
| 1284 |
+
"learning_rate": 2.9840304941919415e-06,
|
| 1285 |
+
"loss": 0.7228,
|
| 1286 |
+
"step": 179
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.017720459747483448,
|
| 1290 |
+
"grad_norm": 0.3204442858695984,
|
| 1291 |
+
"learning_rate": 2.7091379149682685e-06,
|
| 1292 |
+
"loss": 0.6967,
|
| 1293 |
+
"step": 180
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.017818906746080578,
|
| 1297 |
+
"grad_norm": 0.3386498987674713,
|
| 1298 |
+
"learning_rate": 2.4471741852423237e-06,
|
| 1299 |
+
"loss": 0.7068,
|
| 1300 |
+
"step": 181
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.01791735374467771,
|
| 1304 |
+
"grad_norm": 0.31919077038764954,
|
| 1305 |
+
"learning_rate": 2.1982109232821178e-06,
|
| 1306 |
+
"loss": 0.7103,
|
| 1307 |
+
"step": 182
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.01801580074327484,
|
| 1311 |
+
"grad_norm": 0.36073577404022217,
|
| 1312 |
+
"learning_rate": 1.962316193157593e-06,
|
| 1313 |
+
"loss": 0.7672,
|
| 1314 |
+
"step": 183
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.01811424774187197,
|
| 1318 |
+
"grad_norm": 0.31926870346069336,
|
| 1319 |
+
"learning_rate": 1.7395544861325718e-06,
|
| 1320 |
+
"loss": 0.6566,
|
| 1321 |
+
"step": 184
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 0.0182126947404691,
|
| 1325 |
+
"grad_norm": 0.3300493359565735,
|
| 1326 |
+
"learning_rate": 1.5299867030334814e-06,
|
| 1327 |
+
"loss": 0.7093,
|
| 1328 |
+
"step": 185
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.01831114173906623,
|
| 1332 |
+
"grad_norm": 0.3521917462348938,
|
| 1333 |
+
"learning_rate": 1.333670137599713e-06,
|
| 1334 |
+
"loss": 0.7202,
|
| 1335 |
+
"step": 186
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.018409588737663362,
|
| 1339 |
+
"grad_norm": 0.4383225739002228,
|
| 1340 |
+
"learning_rate": 1.1506584608200367e-06,
|
| 1341 |
+
"loss": 0.7847,
|
| 1342 |
+
"step": 187
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.01850803573626049,
|
| 1346 |
+
"grad_norm": 0.3749508559703827,
|
| 1347 |
+
"learning_rate": 9.810017062595322e-07,
|
| 1348 |
+
"loss": 0.6671,
|
| 1349 |
+
"step": 188
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.01860648273485762,
|
| 1353 |
+
"grad_norm": 0.34106940031051636,
|
| 1354 |
+
"learning_rate": 8.247462563808817e-07,
|
| 1355 |
+
"loss": 0.7952,
|
| 1356 |
+
"step": 189
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.01870492973345475,
|
| 1360 |
+
"grad_norm": 0.38784539699554443,
|
| 1361 |
+
"learning_rate": 6.819348298638839e-07,
|
| 1362 |
+
"loss": 0.6736,
|
| 1363 |
+
"step": 190
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.01880337673205188,
|
| 1367 |
+
"grad_norm": 0.37535935640335083,
|
| 1368 |
+
"learning_rate": 5.526064699265753e-07,
|
| 1369 |
+
"loss": 0.6749,
|
| 1370 |
+
"step": 191
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.01890182373064901,
|
| 1374 |
+
"grad_norm": 0.37157732248306274,
|
| 1375 |
+
"learning_rate": 4.367965336512403e-07,
|
| 1376 |
+
"loss": 0.6325,
|
| 1377 |
+
"step": 192
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.019000270729246142,
|
| 1381 |
+
"grad_norm": 0.40247318148612976,
|
| 1382 |
+
"learning_rate": 3.3453668231809286e-07,
|
| 1383 |
+
"loss": 0.6671,
|
| 1384 |
+
"step": 193
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.019098717727843273,
|
| 1388 |
+
"grad_norm": 0.40654775500297546,
|
| 1389 |
+
"learning_rate": 2.458548727494292e-07,
|
| 1390 |
+
"loss": 0.651,
|
| 1391 |
+
"step": 194
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 0.019197164726440404,
|
| 1395 |
+
"grad_norm": 0.3884858191013336,
|
| 1396 |
+
"learning_rate": 1.7077534966650766e-07,
|
| 1397 |
+
"loss": 0.5634,
|
| 1398 |
+
"step": 195
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.019295611725037534,
|
| 1402 |
+
"grad_norm": 0.4609086215496063,
|
| 1403 |
+
"learning_rate": 1.0931863906127327e-07,
|
| 1404 |
+
"loss": 0.6149,
|
| 1405 |
+
"step": 196
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.019394058723634665,
|
| 1409 |
+
"grad_norm": 0.549687385559082,
|
| 1410 |
+
"learning_rate": 6.150154258476315e-08,
|
| 1411 |
+
"loss": 0.6001,
|
| 1412 |
+
"step": 197
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.019492505722231792,
|
| 1416 |
+
"grad_norm": 0.5061210989952087,
|
| 1417 |
+
"learning_rate": 2.7337132953697554e-08,
|
| 1418 |
+
"loss": 0.6994,
|
| 1419 |
+
"step": 198
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.019590952720828923,
|
| 1423 |
+
"grad_norm": 0.5341589450836182,
|
| 1424 |
+
"learning_rate": 6.834750376549792e-09,
|
| 1425 |
+
"loss": 0.6888,
|
| 1426 |
+
"step": 199
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 0.019689399719426053,
|
| 1430 |
+
"grad_norm": 0.6175804734230042,
|
| 1431 |
+
"learning_rate": 0.0,
|
| 1432 |
+
"loss": 0.4914,
|
| 1433 |
+
"step": 200
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.019689399719426053,
|
| 1437 |
+
"eval_loss": 0.7135888934135437,
|
| 1438 |
+
"eval_runtime": 910.6706,
|
| 1439 |
+
"eval_samples_per_second": 18.786,
|
| 1440 |
+
"eval_steps_per_second": 4.697,
|
| 1441 |
+
"step": 200
|
| 1442 |
}
|
| 1443 |
],
|
| 1444 |
"logging_steps": 1,
|
|
|
|
| 1462 |
"should_evaluate": false,
|
| 1463 |
"should_log": false,
|
| 1464 |
"should_save": true,
|
| 1465 |
+
"should_training_stop": true
|
| 1466 |
},
|
| 1467 |
"attributes": {}
|
| 1468 |
}
|
| 1469 |
},
|
| 1470 |
+
"total_flos": 2.6909537631574426e+17,
|
| 1471 |
"train_batch_size": 8,
|
| 1472 |
"trial_name": null,
|
| 1473 |
"trial_params": null
|