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 sha256:126f4393fa4a6094e7fea5464418e519fbdfb718a160aba81c7597935f2fc1c7
|
| 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:e462264687018ece9257d0b0792529505e101ab6262d9de75004440bb35f59d7
|
| 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:d859f9cf7c22d0ae8676fc22bec1276b1c49d06ba5fdef732e833e9fe86c8c08
|
| 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 sha256:4ac116b8169c53ab649a7f15f2f32735f2c71ec2f803f70de8c655a513ee9cfc
|
| 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": 12.19,
|
| 1090 |
"eval_steps_per_second": 3.054,
|
| 1091 |
"step": 150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1092 |
}
|
| 1093 |
],
|
| 1094 |
"logging_steps": 1,
|
|
@@ -1117,7 +1475,7 @@
|
|
| 1117 |
"attributes": {}
|
| 1118 |
}
|
| 1119 |
},
|
| 1120 |
-
"total_flos":
|
| 1121 |
"train_batch_size": 8,
|
| 1122 |
"trial_name": null,
|
| 1123 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.6018708348274231,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.2250351617440225,
|
| 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": 12.19,
|
| 1090 |
"eval_steps_per_second": 3.054,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.169901547116737,
|
| 1095 |
+
"grad_norm": 0.22611109912395477,
|
| 1096 |
+
"learning_rate": 7.107005538862646e-05,
|
| 1097 |
+
"loss": 0.3608,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.1710267229254571,
|
| 1102 |
+
"grad_norm": 0.2486761510372162,
|
| 1103 |
+
"learning_rate": 7.07041155014006e-05,
|
| 1104 |
+
"loss": 0.4149,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.17215189873417722,
|
| 1109 |
+
"grad_norm": 0.22183437645435333,
|
| 1110 |
+
"learning_rate": 7.033683215379002e-05,
|
| 1111 |
+
"loss": 0.4472,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.17327707454289734,
|
| 1116 |
+
"grad_norm": 0.24657611548900604,
|
| 1117 |
+
"learning_rate": 6.996822917828477e-05,
|
| 1118 |
+
"loss": 0.5571,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.17440225035161744,
|
| 1123 |
+
"grad_norm": 0.22693145275115967,
|
| 1124 |
+
"learning_rate": 6.959833049300377e-05,
|
| 1125 |
+
"loss": 0.5807,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.17552742616033756,
|
| 1130 |
+
"grad_norm": 0.22520385682582855,
|
| 1131 |
+
"learning_rate": 6.922716010014255e-05,
|
| 1132 |
+
"loss": 0.5593,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.17665260196905766,
|
| 1137 |
+
"grad_norm": 0.24080556631088257,
|
| 1138 |
+
"learning_rate": 6.885474208441603e-05,
|
| 1139 |
+
"loss": 0.6988,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.17777777777777778,
|
| 1144 |
+
"grad_norm": 0.23336027562618256,
|
| 1145 |
+
"learning_rate": 6.848110061149556e-05,
|
| 1146 |
+
"loss": 0.6938,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.17890295358649788,
|
| 1151 |
+
"grad_norm": 0.24269425868988037,
|
| 1152 |
+
"learning_rate": 6.810625992644085e-05,
|
| 1153 |
+
"loss": 0.6492,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.180028129395218,
|
| 1158 |
+
"grad_norm": 0.257687509059906,
|
| 1159 |
+
"learning_rate": 6.773024435212678e-05,
|
| 1160 |
+
"loss": 0.7288,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.1811533052039381,
|
| 1165 |
+
"grad_norm": 0.24201682209968567,
|
| 1166 |
+
"learning_rate": 6.735307828766515e-05,
|
| 1167 |
+
"loss": 0.7097,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.18227848101265823,
|
| 1172 |
+
"grad_norm": 0.2506401538848877,
|
| 1173 |
+
"learning_rate": 6.697478620682137e-05,
|
| 1174 |
+
"loss": 0.7142,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.18340365682137835,
|
| 1179 |
+
"grad_norm": 0.2325735092163086,
|
| 1180 |
+
"learning_rate": 6.659539265642643e-05,
|
| 1181 |
+
"loss": 0.6644,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.18452883263009845,
|
| 1186 |
+
"grad_norm": 0.2502652704715729,
|
| 1187 |
+
"learning_rate": 6.621492225478414e-05,
|
| 1188 |
+
"loss": 0.7367,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.18565400843881857,
|
| 1193 |
+
"grad_norm": 0.2414039522409439,
|
| 1194 |
+
"learning_rate": 6.583339969007363e-05,
|
| 1195 |
+
"loss": 0.7728,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.18677918424753867,
|
| 1200 |
+
"grad_norm": 0.2699027359485626,
|
| 1201 |
+
"learning_rate": 6.545084971874738e-05,
|
| 1202 |
+
"loss": 0.8384,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.1879043600562588,
|
| 1207 |
+
"grad_norm": 0.2658804655075073,
|
| 1208 |
+
"learning_rate": 6.506729716392481e-05,
|
| 1209 |
+
"loss": 0.8469,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.1890295358649789,
|
| 1214 |
+
"grad_norm": 0.25071781873703003,
|
| 1215 |
+
"learning_rate": 6.468276691378155e-05,
|
| 1216 |
+
"loss": 0.6885,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.19015471167369902,
|
| 1221 |
+
"grad_norm": 0.2700413763523102,
|
| 1222 |
+
"learning_rate": 6.429728391993446e-05,
|
| 1223 |
+
"loss": 0.825,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.19127988748241911,
|
| 1228 |
+
"grad_norm": 0.27662935853004456,
|
| 1229 |
+
"learning_rate": 6.391087319582264e-05,
|
| 1230 |
+
"loss": 0.8351,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.19240506329113924,
|
| 1235 |
+
"grad_norm": 0.26488763093948364,
|
| 1236 |
+
"learning_rate": 6.35235598150842e-05,
|
| 1237 |
+
"loss": 0.7827,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.19353023909985936,
|
| 1242 |
+
"grad_norm": 0.2654742896556854,
|
| 1243 |
+
"learning_rate": 6.313536890992935e-05,
|
| 1244 |
+
"loss": 0.8177,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.19465541490857946,
|
| 1249 |
+
"grad_norm": 0.2809375822544098,
|
| 1250 |
+
"learning_rate": 6.274632566950967e-05,
|
| 1251 |
+
"loss": 0.8218,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.19578059071729959,
|
| 1256 |
+
"grad_norm": 0.26899024844169617,
|
| 1257 |
+
"learning_rate": 6.235645533828349e-05,
|
| 1258 |
+
"loss": 0.7878,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.19690576652601968,
|
| 1263 |
+
"grad_norm": 0.2818717062473297,
|
| 1264 |
+
"learning_rate": 6.19657832143779e-05,
|
| 1265 |
+
"loss": 0.785,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.1980309423347398,
|
| 1270 |
+
"grad_norm": 0.3077673316001892,
|
| 1271 |
+
"learning_rate": 6.157433464794716e-05,
|
| 1272 |
+
"loss": 0.8369,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.1991561181434599,
|
| 1277 |
+
"grad_norm": 0.277925580739975,
|
| 1278 |
+
"learning_rate": 6.118213503952779e-05,
|
| 1279 |
+
"loss": 0.6012,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.20028129395218003,
|
| 1284 |
+
"grad_norm": 0.28356897830963135,
|
| 1285 |
+
"learning_rate": 6.078920983839031e-05,
|
| 1286 |
+
"loss": 0.7352,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.20140646976090015,
|
| 1291 |
+
"grad_norm": 0.2859312891960144,
|
| 1292 |
+
"learning_rate": 6.0395584540887963e-05,
|
| 1293 |
+
"loss": 0.719,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.20253164556962025,
|
| 1298 |
+
"grad_norm": 0.31765639781951904,
|
| 1299 |
+
"learning_rate": 6.0001284688802226e-05,
|
| 1300 |
+
"loss": 0.8586,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.20365682137834037,
|
| 1305 |
+
"grad_norm": 0.2968887984752655,
|
| 1306 |
+
"learning_rate": 5.960633586768543e-05,
|
| 1307 |
+
"loss": 0.6802,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.20478199718706047,
|
| 1312 |
+
"grad_norm": 0.294272780418396,
|
| 1313 |
+
"learning_rate": 5.921076370520058e-05,
|
| 1314 |
+
"loss": 0.6895,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.2059071729957806,
|
| 1319 |
+
"grad_norm": 0.3082694709300995,
|
| 1320 |
+
"learning_rate": 5.8814593869458455e-05,
|
| 1321 |
+
"loss": 0.7071,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.2070323488045007,
|
| 1326 |
+
"grad_norm": 0.3072359561920166,
|
| 1327 |
+
"learning_rate": 5.841785206735192e-05,
|
| 1328 |
+
"loss": 0.6542,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.20815752461322082,
|
| 1333 |
+
"grad_norm": 0.28556135296821594,
|
| 1334 |
+
"learning_rate": 5.8020564042888015e-05,
|
| 1335 |
+
"loss": 0.5641,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.20928270042194091,
|
| 1340 |
+
"grad_norm": 0.3505888283252716,
|
| 1341 |
+
"learning_rate": 5.762275557551727e-05,
|
| 1342 |
+
"loss": 0.8211,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.21040787623066104,
|
| 1347 |
+
"grad_norm": 0.3167312443256378,
|
| 1348 |
+
"learning_rate": 5.7224452478461064e-05,
|
| 1349 |
+
"loss": 0.5715,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.21153305203938116,
|
| 1354 |
+
"grad_norm": 0.30357834696769714,
|
| 1355 |
+
"learning_rate": 5.682568059703659e-05,
|
| 1356 |
+
"loss": 0.5422,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.21265822784810126,
|
| 1361 |
+
"grad_norm": 0.28745442628860474,
|
| 1362 |
+
"learning_rate": 5.642646580697973e-05,
|
| 1363 |
+
"loss": 0.3673,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.21378340365682139,
|
| 1368 |
+
"grad_norm": 0.32116490602493286,
|
| 1369 |
+
"learning_rate": 5.602683401276615e-05,
|
| 1370 |
+
"loss": 0.4719,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.21490857946554148,
|
| 1375 |
+
"grad_norm": 0.31803691387176514,
|
| 1376 |
+
"learning_rate": 5.562681114593028e-05,
|
| 1377 |
+
"loss": 0.4755,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.2160337552742616,
|
| 1382 |
+
"grad_norm": 0.3275885283946991,
|
| 1383 |
+
"learning_rate": 5.522642316338268e-05,
|
| 1384 |
+
"loss": 0.4121,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.2171589310829817,
|
| 1389 |
+
"grad_norm": 0.3217777907848358,
|
| 1390 |
+
"learning_rate": 5.482569604572576e-05,
|
| 1391 |
+
"loss": 0.3525,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.21828410689170183,
|
| 1396 |
+
"grad_norm": 0.33562275767326355,
|
| 1397 |
+
"learning_rate": 5.442465579556793e-05,
|
| 1398 |
+
"loss": 0.3489,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.21940928270042195,
|
| 1403 |
+
"grad_norm": 0.3709050416946411,
|
| 1404 |
+
"learning_rate": 5.402332843583631e-05,
|
| 1405 |
+
"loss": 0.4602,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.22053445850914205,
|
| 1410 |
+
"grad_norm": 0.3461264371871948,
|
| 1411 |
+
"learning_rate": 5.3621740008088126e-05,
|
| 1412 |
+
"loss": 0.3244,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.22165963431786218,
|
| 1417 |
+
"grad_norm": 0.40534359216690063,
|
| 1418 |
+
"learning_rate": 5.321991657082097e-05,
|
| 1419 |
+
"loss": 0.4012,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.22278481012658227,
|
| 1424 |
+
"grad_norm": 0.425788015127182,
|
| 1425 |
+
"learning_rate": 5.281788419778187e-05,
|
| 1426 |
+
"loss": 0.3741,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.2239099859353024,
|
| 1431 |
+
"grad_norm": 0.47200363874435425,
|
| 1432 |
+
"learning_rate": 5.2415668976275355e-05,
|
| 1433 |
+
"loss": 0.3717,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.2250351617440225,
|
| 1438 |
+
"grad_norm": 0.5842266082763672,
|
| 1439 |
+
"learning_rate": 5.201329700547076e-05,
|
| 1440 |
+
"loss": 0.4174,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.2250351617440225,
|
| 1445 |
+
"eval_loss": 0.6018708348274231,
|
| 1446 |
+
"eval_runtime": 122.7147,
|
| 1447 |
+
"eval_samples_per_second": 12.199,
|
| 1448 |
+
"eval_steps_per_second": 3.056,
|
| 1449 |
+
"step": 200
|
| 1450 |
}
|
| 1451 |
],
|
| 1452 |
"logging_steps": 1,
|
|
|
|
| 1475 |
"attributes": {}
|
| 1476 |
}
|
| 1477 |
},
|
| 1478 |
+
"total_flos": 3.205591354834944e+17,
|
| 1479 |
"train_batch_size": 8,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|