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 985240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:75a371cce18512647bdcc08fc2249dfa5009721ce21c1564202813c2eebea9ce
|
| 3 |
size 985240
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 520860
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bfbd4d4b36053e9347cc89c60dbf30acaa28e217c78c9537551228267a2dc84e
|
| 3 |
size 520860
|
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:d084659b7e93e4d18ab319800e1fe2b411e283a38507de5b33b9979776aa9512
|
| 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:e1748c065d82adf475a51dba0ff56fd123a17810f07662718ca6d5704be8f9bd
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 8.
|
| 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": 123.331,
|
| 1090 |
"eval_steps_per_second": 30.959,
|
| 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": 4,
|
| 1122 |
"trial_name": null,
|
| 1123 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 8.784167289733887,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.1718213058419244,
|
| 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": 123.331,
|
| 1090 |
"eval_steps_per_second": 30.959,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.12972508591065293,
|
| 1095 |
+
"grad_norm": 1.4299190044403076,
|
| 1096 |
+
"learning_rate": 2.6124736842105265e-05,
|
| 1097 |
+
"loss": 17.5957,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.13058419243986255,
|
| 1102 |
+
"grad_norm": 1.4363600015640259,
|
| 1103 |
+
"learning_rate": 2.5591578947368422e-05,
|
| 1104 |
+
"loss": 17.748,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.13144329896907217,
|
| 1109 |
+
"grad_norm": 1.611876368522644,
|
| 1110 |
+
"learning_rate": 2.5058421052631576e-05,
|
| 1111 |
+
"loss": 17.7324,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.1323024054982818,
|
| 1116 |
+
"grad_norm": 1.26874840259552,
|
| 1117 |
+
"learning_rate": 2.4525263157894737e-05,
|
| 1118 |
+
"loss": 17.6484,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.1331615120274914,
|
| 1123 |
+
"grad_norm": 1.37934410572052,
|
| 1124 |
+
"learning_rate": 2.3992105263157894e-05,
|
| 1125 |
+
"loss": 17.6289,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.13402061855670103,
|
| 1130 |
+
"grad_norm": 1.488385796546936,
|
| 1131 |
+
"learning_rate": 2.345894736842105e-05,
|
| 1132 |
+
"loss": 17.5723,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.13487972508591065,
|
| 1137 |
+
"grad_norm": 1.4406384229660034,
|
| 1138 |
+
"learning_rate": 2.292578947368421e-05,
|
| 1139 |
+
"loss": 17.7324,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.13573883161512026,
|
| 1144 |
+
"grad_norm": 1.2565712928771973,
|
| 1145 |
+
"learning_rate": 2.2392631578947366e-05,
|
| 1146 |
+
"loss": 17.5039,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.13659793814432988,
|
| 1151 |
+
"grad_norm": 1.3823555707931519,
|
| 1152 |
+
"learning_rate": 2.1859473684210527e-05,
|
| 1153 |
+
"loss": 17.7734,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.13745704467353953,
|
| 1158 |
+
"grad_norm": 1.2780786752700806,
|
| 1159 |
+
"learning_rate": 2.132631578947368e-05,
|
| 1160 |
+
"loss": 17.6191,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.13831615120274915,
|
| 1165 |
+
"grad_norm": 1.2663692235946655,
|
| 1166 |
+
"learning_rate": 2.0793157894736842e-05,
|
| 1167 |
+
"loss": 17.6738,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.13917525773195877,
|
| 1172 |
+
"grad_norm": 1.2461869716644287,
|
| 1173 |
+
"learning_rate": 2.026e-05,
|
| 1174 |
+
"loss": 17.5723,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.1400343642611684,
|
| 1179 |
+
"grad_norm": 1.497506022453308,
|
| 1180 |
+
"learning_rate": 1.9726842105263157e-05,
|
| 1181 |
+
"loss": 17.8398,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.140893470790378,
|
| 1186 |
+
"grad_norm": 1.3680994510650635,
|
| 1187 |
+
"learning_rate": 1.9193684210526314e-05,
|
| 1188 |
+
"loss": 17.75,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.14175257731958762,
|
| 1193 |
+
"grad_norm": 1.8215138912200928,
|
| 1194 |
+
"learning_rate": 1.866052631578947e-05,
|
| 1195 |
+
"loss": 17.6582,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.14261168384879724,
|
| 1200 |
+
"grad_norm": 1.5240401029586792,
|
| 1201 |
+
"learning_rate": 1.8127368421052632e-05,
|
| 1202 |
+
"loss": 17.5254,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.14347079037800686,
|
| 1207 |
+
"grad_norm": 1.3903863430023193,
|
| 1208 |
+
"learning_rate": 1.759421052631579e-05,
|
| 1209 |
+
"loss": 17.7422,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.14432989690721648,
|
| 1214 |
+
"grad_norm": 1.1435747146606445,
|
| 1215 |
+
"learning_rate": 1.7061052631578947e-05,
|
| 1216 |
+
"loss": 17.5332,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.14518900343642613,
|
| 1221 |
+
"grad_norm": 0.8812254071235657,
|
| 1222 |
+
"learning_rate": 1.6527894736842104e-05,
|
| 1223 |
+
"loss": 17.6309,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.14604810996563575,
|
| 1228 |
+
"grad_norm": 1.0723109245300293,
|
| 1229 |
+
"learning_rate": 1.599473684210526e-05,
|
| 1230 |
+
"loss": 17.3887,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.14690721649484537,
|
| 1235 |
+
"grad_norm": 0.7917116284370422,
|
| 1236 |
+
"learning_rate": 1.546157894736842e-05,
|
| 1237 |
+
"loss": 17.4902,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.14776632302405499,
|
| 1242 |
+
"grad_norm": 0.7812902331352234,
|
| 1243 |
+
"learning_rate": 1.4928421052631576e-05,
|
| 1244 |
+
"loss": 17.5938,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.1486254295532646,
|
| 1249 |
+
"grad_norm": 0.6956250667572021,
|
| 1250 |
+
"learning_rate": 1.4395263157894735e-05,
|
| 1251 |
+
"loss": 17.4863,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.14948453608247422,
|
| 1256 |
+
"grad_norm": 0.5251582264900208,
|
| 1257 |
+
"learning_rate": 1.3862105263157895e-05,
|
| 1258 |
+
"loss": 17.4844,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.15034364261168384,
|
| 1263 |
+
"grad_norm": 0.7197800874710083,
|
| 1264 |
+
"learning_rate": 1.3328947368421052e-05,
|
| 1265 |
+
"loss": 17.5898,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.15120274914089346,
|
| 1270 |
+
"grad_norm": 0.5786798596382141,
|
| 1271 |
+
"learning_rate": 1.2795789473684211e-05,
|
| 1272 |
+
"loss": 17.4844,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.15206185567010308,
|
| 1277 |
+
"grad_norm": 0.6620001196861267,
|
| 1278 |
+
"learning_rate": 1.2262631578947368e-05,
|
| 1279 |
+
"loss": 17.5879,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.15292096219931273,
|
| 1284 |
+
"grad_norm": 0.81223064661026,
|
| 1285 |
+
"learning_rate": 1.1729473684210526e-05,
|
| 1286 |
+
"loss": 17.377,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.15378006872852235,
|
| 1291 |
+
"grad_norm": 0.8172386884689331,
|
| 1292 |
+
"learning_rate": 1.1196315789473683e-05,
|
| 1293 |
+
"loss": 17.375,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.15463917525773196,
|
| 1298 |
+
"grad_norm": 0.4085826575756073,
|
| 1299 |
+
"learning_rate": 1.066315789473684e-05,
|
| 1300 |
+
"loss": 17.4844,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.15549828178694158,
|
| 1305 |
+
"grad_norm": 1.567670464515686,
|
| 1306 |
+
"learning_rate": 1.013e-05,
|
| 1307 |
+
"loss": 17.8047,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.1563573883161512,
|
| 1312 |
+
"grad_norm": 1.0551185607910156,
|
| 1313 |
+
"learning_rate": 9.596842105263157e-06,
|
| 1314 |
+
"loss": 17.6973,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.15721649484536082,
|
| 1319 |
+
"grad_norm": 0.39254966378211975,
|
| 1320 |
+
"learning_rate": 9.063684210526316e-06,
|
| 1321 |
+
"loss": 17.4805,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.15807560137457044,
|
| 1326 |
+
"grad_norm": 0.4392406940460205,
|
| 1327 |
+
"learning_rate": 8.530526315789473e-06,
|
| 1328 |
+
"loss": 17.4883,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.15893470790378006,
|
| 1333 |
+
"grad_norm": 1.0405336618423462,
|
| 1334 |
+
"learning_rate": 7.99736842105263e-06,
|
| 1335 |
+
"loss": 17.6973,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.15979381443298968,
|
| 1340 |
+
"grad_norm": 0.7305321097373962,
|
| 1341 |
+
"learning_rate": 7.464210526315788e-06,
|
| 1342 |
+
"loss": 17.375,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.16065292096219932,
|
| 1347 |
+
"grad_norm": 1.0625568628311157,
|
| 1348 |
+
"learning_rate": 6.931052631578947e-06,
|
| 1349 |
+
"loss": 17.6992,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.16151202749140894,
|
| 1354 |
+
"grad_norm": 0.3679310381412506,
|
| 1355 |
+
"learning_rate": 6.3978947368421055e-06,
|
| 1356 |
+
"loss": 17.4805,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.16237113402061856,
|
| 1361 |
+
"grad_norm": 0.7215369343757629,
|
| 1362 |
+
"learning_rate": 5.864736842105263e-06,
|
| 1363 |
+
"loss": 17.377,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.16323024054982818,
|
| 1368 |
+
"grad_norm": 0.4060981571674347,
|
| 1369 |
+
"learning_rate": 5.33157894736842e-06,
|
| 1370 |
+
"loss": 17.4824,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.1640893470790378,
|
| 1375 |
+
"grad_norm": 0.753779411315918,
|
| 1376 |
+
"learning_rate": 4.7984210526315785e-06,
|
| 1377 |
+
"loss": 17.373,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.16494845360824742,
|
| 1382 |
+
"grad_norm": 0.7565664052963257,
|
| 1383 |
+
"learning_rate": 4.265263157894737e-06,
|
| 1384 |
+
"loss": 17.373,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.16580756013745704,
|
| 1389 |
+
"grad_norm": 0.818123996257782,
|
| 1390 |
+
"learning_rate": 3.732105263157894e-06,
|
| 1391 |
+
"loss": 17.375,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.16666666666666666,
|
| 1396 |
+
"grad_norm": 0.4565421938896179,
|
| 1397 |
+
"learning_rate": 3.1989473684210527e-06,
|
| 1398 |
+
"loss": 17.4844,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.16752577319587628,
|
| 1403 |
+
"grad_norm": 1.5401556491851807,
|
| 1404 |
+
"learning_rate": 2.66578947368421e-06,
|
| 1405 |
+
"loss": 17.8027,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.16838487972508592,
|
| 1410 |
+
"grad_norm": 0.5871375203132629,
|
| 1411 |
+
"learning_rate": 2.1326315789473684e-06,
|
| 1412 |
+
"loss": 17.4844,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.16924398625429554,
|
| 1417 |
+
"grad_norm": 0.7226822972297668,
|
| 1418 |
+
"learning_rate": 1.5994736842105264e-06,
|
| 1419 |
+
"loss": 17.4941,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.17010309278350516,
|
| 1424 |
+
"grad_norm": 1.3671441078186035,
|
| 1425 |
+
"learning_rate": 1.0663157894736842e-06,
|
| 1426 |
+
"loss": 17.7051,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.17096219931271478,
|
| 1431 |
+
"grad_norm": 2.1927340030670166,
|
| 1432 |
+
"learning_rate": 5.331578947368421e-07,
|
| 1433 |
+
"loss": 17.9121,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.1718213058419244,
|
| 1438 |
+
"grad_norm": 3.7176849842071533,
|
| 1439 |
+
"learning_rate": 0.0,
|
| 1440 |
+
"loss": 18.2168,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.1718213058419244,
|
| 1445 |
+
"eval_loss": 8.784167289733887,
|
| 1446 |
+
"eval_runtime": 4.0091,
|
| 1447 |
+
"eval_samples_per_second": 122.221,
|
| 1448 |
+
"eval_steps_per_second": 30.68,
|
| 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": 8181434548224.0,
|
| 1479 |
"train_batch_size": 4,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|