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 81576
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a583b8747eb17334382f8aaa2f03ee21bb91655438cbe70bad9246f230fba32
|
| 3 |
size 81576
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 173094
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:675a597eb1945bf59497ad766e676f886be84d7c1135ecdbbf982d265a485f1d
|
| 3 |
size 173094
|
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 sha256:982eb43f0b685032f5d229850fc4c6116b1560071673bb981b13063a46a6b3e1
|
| 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:ca9a25c72339c898b564e0c464a3f6fc75bbeec408008928b7ed05533156b98c
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 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": 276.357,
|
| 1090 |
"eval_steps_per_second": 138.179,
|
| 1091 |
"step": 150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1092 |
}
|
| 1093 |
],
|
| 1094 |
"logging_steps": 1,
|
|
@@ -1103,12 +1461,12 @@
|
|
| 1103 |
"should_evaluate": false,
|
| 1104 |
"should_log": false,
|
| 1105 |
"should_save": true,
|
| 1106 |
-
"should_training_stop":
|
| 1107 |
},
|
| 1108 |
"attributes": {}
|
| 1109 |
}
|
| 1110 |
},
|
| 1111 |
-
"total_flos":
|
| 1112 |
"train_batch_size": 2,
|
| 1113 |
"trial_name": null,
|
| 1114 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.13508949679162446,
|
| 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": 276.357,
|
| 1090 |
"eval_steps_per_second": 138.179,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.10199257007767647,
|
| 1095 |
+
"grad_norm": 0.1630678027868271,
|
| 1096 |
+
"learning_rate": 3.106465908814342e-05,
|
| 1097 |
+
"loss": 10.6025,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.10266801756163459,
|
| 1102 |
+
"grad_norm": 0.15854433178901672,
|
| 1103 |
+
"learning_rate": 2.9876321572751144e-05,
|
| 1104 |
+
"loss": 10.589,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.1033434650455927,
|
| 1109 |
+
"grad_norm": 0.15652941167354584,
|
| 1110 |
+
"learning_rate": 2.87071551708603e-05,
|
| 1111 |
+
"loss": 10.6022,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.10401891252955082,
|
| 1116 |
+
"grad_norm": 0.14386983215808868,
|
| 1117 |
+
"learning_rate": 2.7557479520891104e-05,
|
| 1118 |
+
"loss": 10.5974,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.10469436001350894,
|
| 1123 |
+
"grad_norm": 0.17749740183353424,
|
| 1124 |
+
"learning_rate": 2.6427608932686843e-05,
|
| 1125 |
+
"loss": 10.6112,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.10536980749746708,
|
| 1130 |
+
"grad_norm": 0.1529720574617386,
|
| 1131 |
+
"learning_rate": 2.5317852301584643e-05,
|
| 1132 |
+
"loss": 10.6075,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.1060452549814252,
|
| 1137 |
+
"grad_norm": 0.16465173661708832,
|
| 1138 |
+
"learning_rate": 2.422851302396655e-05,
|
| 1139 |
+
"loss": 10.5917,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.10672070246538332,
|
| 1144 |
+
"grad_norm": 0.148993581533432,
|
| 1145 |
+
"learning_rate": 2.315988891431412e-05,
|
| 1146 |
+
"loss": 10.5951,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.10739614994934144,
|
| 1151 |
+
"grad_norm": 0.17219752073287964,
|
| 1152 |
+
"learning_rate": 2.2112272123788768e-05,
|
| 1153 |
+
"loss": 10.5877,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.10807159743329955,
|
| 1158 |
+
"grad_norm": 0.17853744328022003,
|
| 1159 |
+
"learning_rate": 2.1085949060360654e-05,
|
| 1160 |
+
"loss": 10.5945,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.10874704491725769,
|
| 1165 |
+
"grad_norm": 0.16907915472984314,
|
| 1166 |
+
"learning_rate": 2.008120031050753e-05,
|
| 1167 |
+
"loss": 10.5967,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.1094224924012158,
|
| 1172 |
+
"grad_norm": 0.18723782896995544,
|
| 1173 |
+
"learning_rate": 1.9098300562505266e-05,
|
| 1174 |
+
"loss": 10.5955,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.11009793988517393,
|
| 1179 |
+
"grad_norm": 0.18115279078483582,
|
| 1180 |
+
"learning_rate": 1.8137518531330767e-05,
|
| 1181 |
+
"loss": 10.6003,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.11077338736913205,
|
| 1186 |
+
"grad_norm": 0.14582610130310059,
|
| 1187 |
+
"learning_rate": 1.7199116885197995e-05,
|
| 1188 |
+
"loss": 10.6084,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.11144883485309018,
|
| 1193 |
+
"grad_norm": 0.16853328049182892,
|
| 1194 |
+
"learning_rate": 1.6283352173747145e-05,
|
| 1195 |
+
"loss": 10.6109,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.1121242823370483,
|
| 1200 |
+
"grad_norm": 0.16580811142921448,
|
| 1201 |
+
"learning_rate": 1.5390474757906446e-05,
|
| 1202 |
+
"loss": 10.6031,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.11279972982100642,
|
| 1207 |
+
"grad_norm": 0.1782258301973343,
|
| 1208 |
+
"learning_rate": 1.4520728741446089e-05,
|
| 1209 |
+
"loss": 10.5978,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.11347517730496454,
|
| 1214 |
+
"grad_norm": 0.16206084191799164,
|
| 1215 |
+
"learning_rate": 1.3674351904242611e-05,
|
| 1216 |
+
"loss": 10.5997,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.11415062478892266,
|
| 1221 |
+
"grad_norm": 0.1567702740430832,
|
| 1222 |
+
"learning_rate": 1.2851575637272262e-05,
|
| 1223 |
+
"loss": 10.6037,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.11482607227288079,
|
| 1228 |
+
"grad_norm": 0.174870565533638,
|
| 1229 |
+
"learning_rate": 1.2052624879351104e-05,
|
| 1230 |
+
"loss": 10.5948,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.11550151975683891,
|
| 1235 |
+
"grad_norm": 0.15422426164150238,
|
| 1236 |
+
"learning_rate": 1.1277718055638819e-05,
|
| 1237 |
+
"loss": 10.5984,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.11617696724079703,
|
| 1242 |
+
"grad_norm": 0.17559370398521423,
|
| 1243 |
+
"learning_rate": 1.0527067017923654e-05,
|
| 1244 |
+
"loss": 10.5945,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.11685241472475515,
|
| 1249 |
+
"grad_norm": 0.1844927966594696,
|
| 1250 |
+
"learning_rate": 9.80087698670411e-06,
|
| 1251 |
+
"loss": 10.5987,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.11752786220871327,
|
| 1256 |
+
"grad_norm": 0.16079100966453552,
|
| 1257 |
+
"learning_rate": 9.09934649508375e-06,
|
| 1258 |
+
"loss": 10.6051,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.1182033096926714,
|
| 1263 |
+
"grad_norm": 0.16014216840267181,
|
| 1264 |
+
"learning_rate": 8.422667334494249e-06,
|
| 1265 |
+
"loss": 10.596,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.11887875717662952,
|
| 1270 |
+
"grad_norm": 0.16521821916103363,
|
| 1271 |
+
"learning_rate": 7.771024502261526e-06,
|
| 1272 |
+
"loss": 10.6092,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.11955420466058764,
|
| 1277 |
+
"grad_norm": 0.1629599630832672,
|
| 1278 |
+
"learning_rate": 7.144596151029303e-06,
|
| 1279 |
+
"loss": 10.5984,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.12022965214454576,
|
| 1284 |
+
"grad_norm": 0.16382640600204468,
|
| 1285 |
+
"learning_rate": 6.543553540053926e-06,
|
| 1286 |
+
"loss": 10.5851,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.12090509962850389,
|
| 1291 |
+
"grad_norm": 0.15883676707744598,
|
| 1292 |
+
"learning_rate": 5.968060988383883e-06,
|
| 1293 |
+
"loss": 10.5956,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.12158054711246201,
|
| 1298 |
+
"grad_norm": 0.184475377202034,
|
| 1299 |
+
"learning_rate": 5.418275829936537e-06,
|
| 1300 |
+
"loss": 10.589,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.12225599459642013,
|
| 1305 |
+
"grad_norm": 0.19880411028862,
|
| 1306 |
+
"learning_rate": 4.8943483704846475e-06,
|
| 1307 |
+
"loss": 10.581,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.12293144208037825,
|
| 1312 |
+
"grad_norm": 0.16972365975379944,
|
| 1313 |
+
"learning_rate": 4.3964218465642355e-06,
|
| 1314 |
+
"loss": 10.5909,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.12360688956433637,
|
| 1319 |
+
"grad_norm": 0.17554166913032532,
|
| 1320 |
+
"learning_rate": 3.924632386315186e-06,
|
| 1321 |
+
"loss": 10.5855,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.1242823370482945,
|
| 1326 |
+
"grad_norm": 0.15895400941371918,
|
| 1327 |
+
"learning_rate": 3.4791089722651436e-06,
|
| 1328 |
+
"loss": 10.6027,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.12495778453225262,
|
| 1333 |
+
"grad_norm": 0.15381865203380585,
|
| 1334 |
+
"learning_rate": 3.059973406066963e-06,
|
| 1335 |
+
"loss": 10.6043,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.12563323201621074,
|
| 1340 |
+
"grad_norm": 0.1588447093963623,
|
| 1341 |
+
"learning_rate": 2.667340275199426e-06,
|
| 1342 |
+
"loss": 10.5858,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.12630867950016886,
|
| 1347 |
+
"grad_norm": 0.170160710811615,
|
| 1348 |
+
"learning_rate": 2.3013169216400733e-06,
|
| 1349 |
+
"loss": 10.5963,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.12698412698412698,
|
| 1354 |
+
"grad_norm": 0.19386546313762665,
|
| 1355 |
+
"learning_rate": 1.9620034125190644e-06,
|
| 1356 |
+
"loss": 10.5906,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.1276595744680851,
|
| 1361 |
+
"grad_norm": 0.15733763575553894,
|
| 1362 |
+
"learning_rate": 1.6494925127617634e-06,
|
| 1363 |
+
"loss": 10.5984,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.12833502195204322,
|
| 1368 |
+
"grad_norm": 0.1530665159225464,
|
| 1369 |
+
"learning_rate": 1.3638696597277679e-06,
|
| 1370 |
+
"loss": 10.5911,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.12901046943600136,
|
| 1375 |
+
"grad_norm": 0.1876905858516693,
|
| 1376 |
+
"learning_rate": 1.1052129398531507e-06,
|
| 1377 |
+
"loss": 10.6023,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.12968591691995948,
|
| 1382 |
+
"grad_norm": 0.1700376272201538,
|
| 1383 |
+
"learning_rate": 8.735930673024806e-07,
|
| 1384 |
+
"loss": 10.5979,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.1303613644039176,
|
| 1389 |
+
"grad_norm": 0.18517427146434784,
|
| 1390 |
+
"learning_rate": 6.690733646361857e-07,
|
| 1391 |
+
"loss": 10.5894,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.13103681188787572,
|
| 1396 |
+
"grad_norm": 0.16964897513389587,
|
| 1397 |
+
"learning_rate": 4.917097454988584e-07,
|
| 1398 |
+
"loss": 10.5996,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.13171225937183384,
|
| 1403 |
+
"grad_norm": 0.1931840479373932,
|
| 1404 |
+
"learning_rate": 3.415506993330153e-07,
|
| 1405 |
+
"loss": 10.5854,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.13238770685579196,
|
| 1410 |
+
"grad_norm": 0.15033087134361267,
|
| 1411 |
+
"learning_rate": 2.1863727812254653e-07,
|
| 1412 |
+
"loss": 10.6238,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.13306315433975008,
|
| 1417 |
+
"grad_norm": 0.1763419657945633,
|
| 1418 |
+
"learning_rate": 1.230030851695263e-07,
|
| 1419 |
+
"loss": 10.5943,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.1337386018237082,
|
| 1424 |
+
"grad_norm": 0.17355754971504211,
|
| 1425 |
+
"learning_rate": 5.467426590739511e-08,
|
| 1426 |
+
"loss": 10.5912,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.13441404930766632,
|
| 1431 |
+
"grad_norm": 0.1702238917350769,
|
| 1432 |
+
"learning_rate": 1.3669500753099585e-08,
|
| 1433 |
+
"loss": 10.6278,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.13508949679162446,
|
| 1438 |
+
"grad_norm": 0.15678980946540833,
|
| 1439 |
+
"learning_rate": 0.0,
|
| 1440 |
+
"loss": 10.5993,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.13508949679162446,
|
| 1445 |
+
"eval_loss": 10.595317840576172,
|
| 1446 |
+
"eval_runtime": 2.2488,
|
| 1447 |
+
"eval_samples_per_second": 277.475,
|
| 1448 |
+
"eval_steps_per_second": 138.738,
|
| 1449 |
+
"step": 200
|
| 1450 |
}
|
| 1451 |
],
|
| 1452 |
"logging_steps": 1,
|
|
|
|
| 1461 |
"should_evaluate": false,
|
| 1462 |
"should_log": false,
|
| 1463 |
"should_save": true,
|
| 1464 |
+
"should_training_stop": true
|
| 1465 |
},
|
| 1466 |
"attributes": {}
|
| 1467 |
}
|
| 1468 |
},
|
| 1469 |
+
"total_flos": 32811830476800.0,
|
| 1470 |
"train_batch_size": 2,
|
| 1471 |
"trial_name": null,
|
| 1472 |
"trial_params": null
|