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 628216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c93368f0c9f8b881cd65130eee22f684530e3aaf36bef1aa7ed26f00067fdc8d
|
| 3 |
size 628216
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 352122
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c2b701936911f2e005f464c36463b1c3087d57702b237cb6f248d53a0a782b9
|
| 3 |
size 352122
|
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 sha256:2d7bbf2c6a262584ee3cf7d118b7f0ae18043db13669286f517bb185ea18aa3a
|
| 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 sha256:7df65c8f4b426598b0abc4173b1983dcf7411aee63ea4061d980eae7a1af2363
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 10.
|
| 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": 71.913,
|
| 1090 |
"eval_steps_per_second": 18.006,
|
| 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": 10.490081787109375,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.176056338028169,
|
| 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": 71.913,
|
| 1090 |
"eval_steps_per_second": 18.006,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.1329225352112676,
|
| 1095 |
+
"grad_norm": 0.655456006526947,
|
| 1096 |
+
"learning_rate": 1.7085562498478883e-05,
|
| 1097 |
+
"loss": 10.4826,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.13380281690140844,
|
| 1102 |
+
"grad_norm": 0.4750436246395111,
|
| 1103 |
+
"learning_rate": 1.6431976865013128e-05,
|
| 1104 |
+
"loss": 10.5067,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.13468309859154928,
|
| 1109 |
+
"grad_norm": 0.5088787078857422,
|
| 1110 |
+
"learning_rate": 1.5788935343973164e-05,
|
| 1111 |
+
"loss": 10.4799,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.13556338028169015,
|
| 1116 |
+
"grad_norm": 0.4471653401851654,
|
| 1117 |
+
"learning_rate": 1.5156613736490108e-05,
|
| 1118 |
+
"loss": 10.5016,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.136443661971831,
|
| 1123 |
+
"grad_norm": 0.4595582187175751,
|
| 1124 |
+
"learning_rate": 1.4535184912977763e-05,
|
| 1125 |
+
"loss": 10.4885,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.13732394366197184,
|
| 1130 |
+
"grad_norm": 0.5350455641746521,
|
| 1131 |
+
"learning_rate": 1.3924818765871553e-05,
|
| 1132 |
+
"loss": 10.4808,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.1382042253521127,
|
| 1137 |
+
"grad_norm": 0.5852507948875427,
|
| 1138 |
+
"learning_rate": 1.3325682163181601e-05,
|
| 1139 |
+
"loss": 10.4695,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.13908450704225353,
|
| 1144 |
+
"grad_norm": 0.5205310583114624,
|
| 1145 |
+
"learning_rate": 1.2737938902872767e-05,
|
| 1146 |
+
"loss": 10.4588,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.13996478873239437,
|
| 1151 |
+
"grad_norm": 0.5743169784545898,
|
| 1152 |
+
"learning_rate": 1.2161749668083823e-05,
|
| 1153 |
+
"loss": 10.4531,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.14084507042253522,
|
| 1158 |
+
"grad_norm": 0.4958171844482422,
|
| 1159 |
+
"learning_rate": 1.159727198319836e-05,
|
| 1160 |
+
"loss": 10.4593,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.14172535211267606,
|
| 1165 |
+
"grad_norm": 0.43265965580940247,
|
| 1166 |
+
"learning_rate": 1.1044660170779142e-05,
|
| 1167 |
+
"loss": 10.4656,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.1426056338028169,
|
| 1172 |
+
"grad_norm": 0.5412198901176453,
|
| 1173 |
+
"learning_rate": 1.0504065309377897e-05,
|
| 1174 |
+
"loss": 10.4804,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.14348591549295775,
|
| 1179 |
+
"grad_norm": 0.4394540786743164,
|
| 1180 |
+
"learning_rate": 9.97563519223192e-06,
|
| 1181 |
+
"loss": 10.4911,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.1443661971830986,
|
| 1186 |
+
"grad_norm": 0.4882407784461975,
|
| 1187 |
+
"learning_rate": 9.459514286858898e-06,
|
| 1188 |
+
"loss": 10.5067,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.14524647887323944,
|
| 1193 |
+
"grad_norm": 0.3944062888622284,
|
| 1194 |
+
"learning_rate": 8.95584369556093e-06,
|
| 1195 |
+
"loss": 10.4899,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.14612676056338028,
|
| 1200 |
+
"grad_norm": 0.3780195415019989,
|
| 1201 |
+
"learning_rate": 8.464761116848546e-06,
|
| 1202 |
+
"loss": 10.4926,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.14700704225352113,
|
| 1207 |
+
"grad_norm": 0.31584465503692627,
|
| 1208 |
+
"learning_rate": 7.986400807795349e-06,
|
| 1209 |
+
"loss": 10.4902,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.14788732394366197,
|
| 1214 |
+
"grad_norm": 0.2965972125530243,
|
| 1215 |
+
"learning_rate": 7.520893547333436e-06,
|
| 1216 |
+
"loss": 10.4713,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.1487676056338028,
|
| 1221 |
+
"grad_norm": 0.3396422266960144,
|
| 1222 |
+
"learning_rate": 7.068366600499744e-06,
|
| 1223 |
+
"loss": 10.5113,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.14964788732394366,
|
| 1228 |
+
"grad_norm": 0.27547672390937805,
|
| 1229 |
+
"learning_rate": 6.6289436836431076e-06,
|
| 1230 |
+
"loss": 10.4863,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.1505281690140845,
|
| 1235 |
+
"grad_norm": 0.21695564687252045,
|
| 1236 |
+
"learning_rate": 6.20274493060135e-06,
|
| 1237 |
+
"loss": 10.5045,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.15140845070422534,
|
| 1242 |
+
"grad_norm": 0.2346537858247757,
|
| 1243 |
+
"learning_rate": 5.789886859858009e-06,
|
| 1244 |
+
"loss": 10.5259,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.1522887323943662,
|
| 1249 |
+
"grad_norm": 0.23305346071720123,
|
| 1250 |
+
"learning_rate": 5.3904823426872605e-06,
|
| 1251 |
+
"loss": 10.5179,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.15316901408450703,
|
| 1256 |
+
"grad_norm": 0.3447682559490204,
|
| 1257 |
+
"learning_rate": 5.004640572296062e-06,
|
| 1258 |
+
"loss": 10.5116,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.15404929577464788,
|
| 1263 |
+
"grad_norm": 0.2727600038051605,
|
| 1264 |
+
"learning_rate": 4.632467033971838e-06,
|
| 1265 |
+
"loss": 10.5402,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.15492957746478872,
|
| 1270 |
+
"grad_norm": 0.3000151813030243,
|
| 1271 |
+
"learning_rate": 4.274063476243839e-06,
|
| 1272 |
+
"loss": 10.5252,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.15580985915492956,
|
| 1277 |
+
"grad_norm": 0.24301907420158386,
|
| 1278 |
+
"learning_rate": 3.929527883066117e-06,
|
| 1279 |
+
"loss": 10.5167,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.15669014084507044,
|
| 1284 |
+
"grad_norm": 0.29026785492897034,
|
| 1285 |
+
"learning_rate": 3.5989544470296595e-06,
|
| 1286 |
+
"loss": 10.5109,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.15757042253521128,
|
| 1291 |
+
"grad_norm": 0.30544570088386536,
|
| 1292 |
+
"learning_rate": 3.282433543611136e-06,
|
| 1293 |
+
"loss": 10.4965,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.15845070422535212,
|
| 1298 |
+
"grad_norm": 0.3145160973072052,
|
| 1299 |
+
"learning_rate": 2.980051706465095e-06,
|
| 1300 |
+
"loss": 10.4923,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.15933098591549297,
|
| 1305 |
+
"grad_norm": 0.29084667563438416,
|
| 1306 |
+
"learning_rate": 2.691891603766556e-06,
|
| 1307 |
+
"loss": 10.5384,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.1602112676056338,
|
| 1312 |
+
"grad_norm": 0.2924867570400238,
|
| 1313 |
+
"learning_rate": 2.4180320156103298e-06,
|
| 1314 |
+
"loss": 10.5349,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.16109154929577466,
|
| 1319 |
+
"grad_norm": 0.31382471323013306,
|
| 1320 |
+
"learning_rate": 2.158547812473352e-06,
|
| 1321 |
+
"loss": 10.5328,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.1619718309859155,
|
| 1326 |
+
"grad_norm": 0.2887389361858368,
|
| 1327 |
+
"learning_rate": 1.9135099347458293e-06,
|
| 1328 |
+
"loss": 10.5209,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.16285211267605634,
|
| 1333 |
+
"grad_norm": 0.35892584919929504,
|
| 1334 |
+
"learning_rate": 1.6829853733368294e-06,
|
| 1335 |
+
"loss": 10.5239,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.1637323943661972,
|
| 1340 |
+
"grad_norm": 0.24830038845539093,
|
| 1341 |
+
"learning_rate": 1.4670371513596842e-06,
|
| 1342 |
+
"loss": 10.5013,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.16461267605633803,
|
| 1347 |
+
"grad_norm": 0.33427461981773376,
|
| 1348 |
+
"learning_rate": 1.2657243069020402e-06,
|
| 1349 |
+
"loss": 10.4891,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.16549295774647887,
|
| 1354 |
+
"grad_norm": 0.3475538492202759,
|
| 1355 |
+
"learning_rate": 1.0791018768854855e-06,
|
| 1356 |
+
"loss": 10.4896,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.16637323943661972,
|
| 1361 |
+
"grad_norm": 0.3785645365715027,
|
| 1362 |
+
"learning_rate": 9.072208820189698e-07,
|
| 1363 |
+
"loss": 10.5247,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.16725352112676056,
|
| 1368 |
+
"grad_norm": 0.3462846279144287,
|
| 1369 |
+
"learning_rate": 7.501283128502722e-07,
|
| 1370 |
+
"loss": 10.5091,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.1681338028169014,
|
| 1375 |
+
"grad_norm": 0.3662095367908478,
|
| 1376 |
+
"learning_rate": 6.07867116919233e-07,
|
| 1377 |
+
"loss": 10.465,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.16901408450704225,
|
| 1382 |
+
"grad_norm": 0.46672433614730835,
|
| 1383 |
+
"learning_rate": 4.804761870163643e-07,
|
| 1384 |
+
"loss": 10.4353,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.1698943661971831,
|
| 1389 |
+
"grad_norm": 0.3511195778846741,
|
| 1390 |
+
"learning_rate": 3.6799035054990215e-07,
|
| 1391 |
+
"loss": 10.4936,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.17077464788732394,
|
| 1396 |
+
"grad_norm": 0.3752671778202057,
|
| 1397 |
+
"learning_rate": 2.704403600243721e-07,
|
| 1398 |
+
"loss": 10.4624,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.17165492957746478,
|
| 1403 |
+
"grad_norm": 0.3640790283679962,
|
| 1404 |
+
"learning_rate": 1.878528846331584e-07,
|
| 1405 |
+
"loss": 10.495,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.17253521126760563,
|
| 1410 |
+
"grad_norm": 0.37574926018714905,
|
| 1411 |
+
"learning_rate": 1.202505029674006e-07,
|
| 1412 |
+
"loss": 10.4919,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.17341549295774647,
|
| 1417 |
+
"grad_norm": 0.32661178708076477,
|
| 1418 |
+
"learning_rate": 6.765169684323947e-08,
|
| 1419 |
+
"loss": 10.4605,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.1742957746478873,
|
| 1424 |
+
"grad_norm": 0.415720671415329,
|
| 1425 |
+
"learning_rate": 3.007084624906731e-08,
|
| 1426 |
+
"loss": 10.4822,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.17517605633802816,
|
| 1431 |
+
"grad_norm": 0.4906235337257385,
|
| 1432 |
+
"learning_rate": 7.518225414204771e-09,
|
| 1433 |
+
"loss": 10.5168,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.176056338028169,
|
| 1438 |
+
"grad_norm": 0.49893417954444885,
|
| 1439 |
+
"learning_rate": 0.0,
|
| 1440 |
+
"loss": 10.4814,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.176056338028169,
|
| 1445 |
+
"eval_loss": 10.490081787109375,
|
| 1446 |
+
"eval_runtime": 30.7966,
|
| 1447 |
+
"eval_samples_per_second": 62.117,
|
| 1448 |
+
"eval_steps_per_second": 15.554,
|
| 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": 136620225331200.0,
|
| 1479 |
"train_batch_size": 8,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|