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 37762064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12810d468258b5304f33fae2ca923f932d8f2f1f9d77d00d54be78b22bf0817e
|
| 3 |
size 37762064
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 19283770
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c41ac035eb21d1f99367307de0a21b21a31efec944e3d595df3dac603b49120a
|
| 3 |
size 19283770
|
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 sha256:4fbfca1b4b6ded81771f66676f6dc6609585dd69c24c4161ef69050a6d989b68
|
| 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:ca62d85cf5423834480d3c20680f93b185c8e6574a8a14021d285e0b05c7449f
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 2.
|
| 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": 144.344,
|
| 1090 |
"eval_steps_per_second": 36.233,
|
| 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": 2.0797932147979736,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.6872852233676976,
|
| 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": 144.344,
|
| 1090 |
"eval_steps_per_second": 36.233,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.5189003436426117,
|
| 1095 |
+
"grad_norm": 2705.450439453125,
|
| 1096 |
+
"learning_rate": 9.319397726443026e-06,
|
| 1097 |
+
"loss": 7.5747,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.5223367697594502,
|
| 1102 |
+
"grad_norm": 3762.812744140625,
|
| 1103 |
+
"learning_rate": 8.962896471825342e-06,
|
| 1104 |
+
"loss": 7.5746,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.5257731958762887,
|
| 1109 |
+
"grad_norm": 1911.854736328125,
|
| 1110 |
+
"learning_rate": 8.61214655125809e-06,
|
| 1111 |
+
"loss": 7.438,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.5292096219931272,
|
| 1116 |
+
"grad_norm": 1469.72119140625,
|
| 1117 |
+
"learning_rate": 8.267243856267331e-06,
|
| 1118 |
+
"loss": 7.2063,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.5326460481099656,
|
| 1123 |
+
"grad_norm": 1573.3465576171875,
|
| 1124 |
+
"learning_rate": 7.928282679806052e-06,
|
| 1125 |
+
"loss": 7.2888,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.5360824742268041,
|
| 1130 |
+
"grad_norm": 1803.7042236328125,
|
| 1131 |
+
"learning_rate": 7.595355690475393e-06,
|
| 1132 |
+
"loss": 7.4952,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.5395189003436426,
|
| 1137 |
+
"grad_norm": 1138.683837890625,
|
| 1138 |
+
"learning_rate": 7.268553907189964e-06,
|
| 1139 |
+
"loss": 7.2513,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.5429553264604811,
|
| 1144 |
+
"grad_norm": 1677.6661376953125,
|
| 1145 |
+
"learning_rate": 6.947966674294236e-06,
|
| 1146 |
+
"loss": 7.5826,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.5463917525773195,
|
| 1151 |
+
"grad_norm": 1217.319091796875,
|
| 1152 |
+
"learning_rate": 6.6336816371366305e-06,
|
| 1153 |
+
"loss": 7.5123,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.5498281786941581,
|
| 1158 |
+
"grad_norm": 1235.499755859375,
|
| 1159 |
+
"learning_rate": 6.325784718108196e-06,
|
| 1160 |
+
"loss": 7.6381,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.5532646048109966,
|
| 1165 |
+
"grad_norm": 1441.3863525390625,
|
| 1166 |
+
"learning_rate": 6.0243600931522595e-06,
|
| 1167 |
+
"loss": 7.5427,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.5567010309278351,
|
| 1172 |
+
"grad_norm": 1629.83935546875,
|
| 1173 |
+
"learning_rate": 5.72949016875158e-06,
|
| 1174 |
+
"loss": 7.4876,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.5601374570446735,
|
| 1179 |
+
"grad_norm": 1260.6104736328125,
|
| 1180 |
+
"learning_rate": 5.44125555939923e-06,
|
| 1181 |
+
"loss": 7.7081,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.563573883161512,
|
| 1186 |
+
"grad_norm": 1765.477294921875,
|
| 1187 |
+
"learning_rate": 5.159735065559399e-06,
|
| 1188 |
+
"loss": 7.521,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.5670103092783505,
|
| 1193 |
+
"grad_norm": 2019.567626953125,
|
| 1194 |
+
"learning_rate": 4.885005652124144e-06,
|
| 1195 |
+
"loss": 7.6643,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.570446735395189,
|
| 1200 |
+
"grad_norm": 1756.0921630859375,
|
| 1201 |
+
"learning_rate": 4.617142427371934e-06,
|
| 1202 |
+
"loss": 7.7939,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.5738831615120275,
|
| 1207 |
+
"grad_norm": 1463.5172119140625,
|
| 1208 |
+
"learning_rate": 4.3562186224338265e-06,
|
| 1209 |
+
"loss": 7.6615,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.5773195876288659,
|
| 1214 |
+
"grad_norm": 1783.8548583984375,
|
| 1215 |
+
"learning_rate": 4.102305571272783e-06,
|
| 1216 |
+
"loss": 7.4967,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.5807560137457045,
|
| 1221 |
+
"grad_norm": 1875.7342529296875,
|
| 1222 |
+
"learning_rate": 3.855472691181678e-06,
|
| 1223 |
+
"loss": 7.4725,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.584192439862543,
|
| 1228 |
+
"grad_norm": 2582.930908203125,
|
| 1229 |
+
"learning_rate": 3.615787463805331e-06,
|
| 1230 |
+
"loss": 8.2118,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.5876288659793815,
|
| 1235 |
+
"grad_norm": 1897.54638671875,
|
| 1236 |
+
"learning_rate": 3.383315416691646e-06,
|
| 1237 |
+
"loss": 8.2408,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.5910652920962199,
|
| 1242 |
+
"grad_norm": 2314.038330078125,
|
| 1243 |
+
"learning_rate": 3.158120105377096e-06,
|
| 1244 |
+
"loss": 7.6892,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.5945017182130584,
|
| 1249 |
+
"grad_norm": 1786.5667724609375,
|
| 1250 |
+
"learning_rate": 2.940263096011233e-06,
|
| 1251 |
+
"loss": 8.3423,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.5979381443298969,
|
| 1256 |
+
"grad_norm": 2857.995849609375,
|
| 1257 |
+
"learning_rate": 2.729803948525125e-06,
|
| 1258 |
+
"loss": 7.616,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.6013745704467354,
|
| 1263 |
+
"grad_norm": 2287.089111328125,
|
| 1264 |
+
"learning_rate": 2.526800200348275e-06,
|
| 1265 |
+
"loss": 8.6003,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.6048109965635738,
|
| 1270 |
+
"grad_norm": 1499.41015625,
|
| 1271 |
+
"learning_rate": 2.3313073506784575e-06,
|
| 1272 |
+
"loss": 7.7733,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.6082474226804123,
|
| 1277 |
+
"grad_norm": 2719.6171875,
|
| 1278 |
+
"learning_rate": 2.143378845308791e-06,
|
| 1279 |
+
"loss": 8.3946,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.6116838487972509,
|
| 1284 |
+
"grad_norm": 2311.212158203125,
|
| 1285 |
+
"learning_rate": 1.9630660620161777e-06,
|
| 1286 |
+
"loss": 8.9659,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.6151202749140894,
|
| 1291 |
+
"grad_norm": 2324.716552734375,
|
| 1292 |
+
"learning_rate": 1.790418296515165e-06,
|
| 1293 |
+
"loss": 8.1793,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.6185567010309279,
|
| 1298 |
+
"grad_norm": 2077.054443359375,
|
| 1299 |
+
"learning_rate": 1.625482748980961e-06,
|
| 1300 |
+
"loss": 8.5296,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.6219931271477663,
|
| 1305 |
+
"grad_norm": 3390.5927734375,
|
| 1306 |
+
"learning_rate": 1.4683045111453942e-06,
|
| 1307 |
+
"loss": 8.2995,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.6254295532646048,
|
| 1312 |
+
"grad_norm": 2156.932373046875,
|
| 1313 |
+
"learning_rate": 1.3189265539692707e-06,
|
| 1314 |
+
"loss": 8.547,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.6288659793814433,
|
| 1319 |
+
"grad_norm": 2244.710205078125,
|
| 1320 |
+
"learning_rate": 1.1773897158945557e-06,
|
| 1321 |
+
"loss": 7.3688,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.6323024054982818,
|
| 1326 |
+
"grad_norm": 3128.6435546875,
|
| 1327 |
+
"learning_rate": 1.0437326916795432e-06,
|
| 1328 |
+
"loss": 9.2353,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.6357388316151202,
|
| 1333 |
+
"grad_norm": 3524.7509765625,
|
| 1334 |
+
"learning_rate": 9.179920218200888e-07,
|
| 1335 |
+
"loss": 8.7623,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.6391752577319587,
|
| 1340 |
+
"grad_norm": 4061.834716796875,
|
| 1341 |
+
"learning_rate": 8.002020825598277e-07,
|
| 1342 |
+
"loss": 8.2449,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.6426116838487973,
|
| 1347 |
+
"grad_norm": 4144.5556640625,
|
| 1348 |
+
"learning_rate": 6.90395076492022e-07,
|
| 1349 |
+
"loss": 8.4433,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.6460481099656358,
|
| 1354 |
+
"grad_norm": 3348.039306640625,
|
| 1355 |
+
"learning_rate": 5.886010237557194e-07,
|
| 1356 |
+
"loss": 7.4949,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.6494845360824743,
|
| 1361 |
+
"grad_norm": 2904.456298828125,
|
| 1362 |
+
"learning_rate": 4.94847753828529e-07,
|
| 1363 |
+
"loss": 7.7912,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.6529209621993127,
|
| 1368 |
+
"grad_norm": 3423.37353515625,
|
| 1369 |
+
"learning_rate": 4.091608979183303e-07,
|
| 1370 |
+
"loss": 8.2337,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.6563573883161512,
|
| 1375 |
+
"grad_norm": 2644.664794921875,
|
| 1376 |
+
"learning_rate": 3.315638819559452e-07,
|
| 1377 |
+
"loss": 7.876,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.6597938144329897,
|
| 1382 |
+
"grad_norm": 4483.470703125,
|
| 1383 |
+
"learning_rate": 2.6207792019074414e-07,
|
| 1384 |
+
"loss": 8.5568,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.6632302405498282,
|
| 1389 |
+
"grad_norm": 8579.669921875,
|
| 1390 |
+
"learning_rate": 2.0072200939085573e-07,
|
| 1391 |
+
"loss": 8.8126,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.6666666666666666,
|
| 1396 |
+
"grad_norm": 7409.6220703125,
|
| 1397 |
+
"learning_rate": 1.475129236496575e-07,
|
| 1398 |
+
"loss": 7.9544,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.6701030927835051,
|
| 1403 |
+
"grad_norm": 4703.0517578125,
|
| 1404 |
+
"learning_rate": 1.0246520979990459e-07,
|
| 1405 |
+
"loss": 8.26,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.6735395189003437,
|
| 1410 |
+
"grad_norm": 4779.853515625,
|
| 1411 |
+
"learning_rate": 6.559118343676396e-08,
|
| 1412 |
+
"loss": 8.2973,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.6769759450171822,
|
| 1417 |
+
"grad_norm": 6850.4453125,
|
| 1418 |
+
"learning_rate": 3.690092555085789e-08,
|
| 1419 |
+
"loss": 8.3535,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.6804123711340206,
|
| 1424 |
+
"grad_norm": 5519.74365234375,
|
| 1425 |
+
"learning_rate": 1.640227977221853e-08,
|
| 1426 |
+
"loss": 8.6194,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.6838487972508591,
|
| 1431 |
+
"grad_norm": 7148.4296875,
|
| 1432 |
+
"learning_rate": 4.1008502259298755e-09,
|
| 1433 |
+
"loss": 10.8181,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.6872852233676976,
|
| 1438 |
+
"grad_norm": 3729.072509765625,
|
| 1439 |
+
"learning_rate": 0.0,
|
| 1440 |
+
"loss": 13.0084,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.6872852233676976,
|
| 1445 |
+
"eval_loss": 2.0797932147979736,
|
| 1446 |
+
"eval_runtime": 3.4064,
|
| 1447 |
+
"eval_samples_per_second": 143.848,
|
| 1448 |
+
"eval_steps_per_second": 36.109,
|
| 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": 6628506133856256.0,
|
| 1479 |
"train_batch_size": 8,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|