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 191968
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ee5ebc4f961724343ceb4d060d045611be3b35954276a6e0d1d4094201e1149
|
| 3 |
size 191968
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 253144
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2e02d419842c13e4a0acae0f6cbd4b8e8a6f912721789027a75da4256d99321
|
| 3 |
size 253144
|
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 sha256:2845fe51bf3e15f38609d75bbf20eecb59bfdd377018560c2d0532697d9d63be
|
| 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:a3e88ef6a2716260516e17223973d6a3b0a4c88bf12c72ed47e80e6f2a6782fd
|
| 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": 174.224,
|
| 1090 |
"eval_steps_per_second": 43.567,
|
| 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": 10.344128608703613,
|
| 3 |
+
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 0.007340257643043271,
|
| 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": 174.224,
|
| 1090 |
"eval_steps_per_second": 43.567,
|
| 1091 |
"step": 150
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.00554189452049767,
|
| 1095 |
+
"grad_norm": 0.15762516856193542,
|
| 1096 |
+
"learning_rate": 2.589263157894737e-05,
|
| 1097 |
+
"loss": 10.3516,
|
| 1098 |
+
"step": 151
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.005578595808712886,
|
| 1102 |
+
"grad_norm": 0.1376420110464096,
|
| 1103 |
+
"learning_rate": 2.536421052631579e-05,
|
| 1104 |
+
"loss": 10.3499,
|
| 1105 |
+
"step": 152
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.005615297096928102,
|
| 1109 |
+
"grad_norm": 0.12220832705497742,
|
| 1110 |
+
"learning_rate": 2.483578947368421e-05,
|
| 1111 |
+
"loss": 10.3486,
|
| 1112 |
+
"step": 153
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.005651998385143318,
|
| 1116 |
+
"grad_norm": 0.13591831922531128,
|
| 1117 |
+
"learning_rate": 2.430736842105263e-05,
|
| 1118 |
+
"loss": 10.3398,
|
| 1119 |
+
"step": 154
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 0.005688699673358535,
|
| 1123 |
+
"grad_norm": 0.15385285019874573,
|
| 1124 |
+
"learning_rate": 2.3778947368421052e-05,
|
| 1125 |
+
"loss": 10.3572,
|
| 1126 |
+
"step": 155
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.0057254009615737515,
|
| 1130 |
+
"grad_norm": 0.18380072712898254,
|
| 1131 |
+
"learning_rate": 2.3250526315789473e-05,
|
| 1132 |
+
"loss": 10.3438,
|
| 1133 |
+
"step": 156
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 0.005762102249788968,
|
| 1137 |
+
"grad_norm": 0.15198898315429688,
|
| 1138 |
+
"learning_rate": 2.2722105263157894e-05,
|
| 1139 |
+
"loss": 10.3463,
|
| 1140 |
+
"step": 157
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 0.005798803538004184,
|
| 1144 |
+
"grad_norm": 0.0870397686958313,
|
| 1145 |
+
"learning_rate": 2.2193684210526316e-05,
|
| 1146 |
+
"loss": 10.3472,
|
| 1147 |
+
"step": 158
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 0.0058355048262194,
|
| 1151 |
+
"grad_norm": 0.10544353723526001,
|
| 1152 |
+
"learning_rate": 2.1665263157894737e-05,
|
| 1153 |
+
"loss": 10.3577,
|
| 1154 |
+
"step": 159
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.005872206114434617,
|
| 1158 |
+
"grad_norm": 0.10530359297990799,
|
| 1159 |
+
"learning_rate": 2.1136842105263158e-05,
|
| 1160 |
+
"loss": 10.349,
|
| 1161 |
+
"step": 160
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.005908907402649833,
|
| 1165 |
+
"grad_norm": 0.10943751037120819,
|
| 1166 |
+
"learning_rate": 2.060842105263158e-05,
|
| 1167 |
+
"loss": 10.3478,
|
| 1168 |
+
"step": 161
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.005945608690865049,
|
| 1172 |
+
"grad_norm": 0.11259687691926956,
|
| 1173 |
+
"learning_rate": 2.008e-05,
|
| 1174 |
+
"loss": 10.3527,
|
| 1175 |
+
"step": 162
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 0.0059823099790802655,
|
| 1179 |
+
"grad_norm": 0.11968999356031418,
|
| 1180 |
+
"learning_rate": 1.9551578947368422e-05,
|
| 1181 |
+
"loss": 10.3526,
|
| 1182 |
+
"step": 163
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.006019011267295482,
|
| 1186 |
+
"grad_norm": 0.10259483009576797,
|
| 1187 |
+
"learning_rate": 1.9023157894736843e-05,
|
| 1188 |
+
"loss": 10.3468,
|
| 1189 |
+
"step": 164
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.006055712555510699,
|
| 1193 |
+
"grad_norm": 0.12962016463279724,
|
| 1194 |
+
"learning_rate": 1.849473684210526e-05,
|
| 1195 |
+
"loss": 10.3491,
|
| 1196 |
+
"step": 165
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.006092413843725915,
|
| 1200 |
+
"grad_norm": 0.0966765359044075,
|
| 1201 |
+
"learning_rate": 1.7966315789473686e-05,
|
| 1202 |
+
"loss": 10.3486,
|
| 1203 |
+
"step": 166
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.006129115131941131,
|
| 1207 |
+
"grad_norm": 0.17388637363910675,
|
| 1208 |
+
"learning_rate": 1.7437894736842107e-05,
|
| 1209 |
+
"loss": 10.3503,
|
| 1210 |
+
"step": 167
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.006165816420156347,
|
| 1214 |
+
"grad_norm": 0.15526233613491058,
|
| 1215 |
+
"learning_rate": 1.6909473684210525e-05,
|
| 1216 |
+
"loss": 10.3443,
|
| 1217 |
+
"step": 168
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.006202517708371564,
|
| 1221 |
+
"grad_norm": 0.15173742175102234,
|
| 1222 |
+
"learning_rate": 1.638105263157895e-05,
|
| 1223 |
+
"loss": 10.3393,
|
| 1224 |
+
"step": 169
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.0062392189965867805,
|
| 1228 |
+
"grad_norm": 0.1370978206396103,
|
| 1229 |
+
"learning_rate": 1.5852631578947368e-05,
|
| 1230 |
+
"loss": 10.3411,
|
| 1231 |
+
"step": 170
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.006275920284801996,
|
| 1235 |
+
"grad_norm": 0.12809091806411743,
|
| 1236 |
+
"learning_rate": 1.532421052631579e-05,
|
| 1237 |
+
"loss": 10.3389,
|
| 1238 |
+
"step": 171
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.006312621573017213,
|
| 1242 |
+
"grad_norm": 0.11044250428676605,
|
| 1243 |
+
"learning_rate": 1.4795789473684209e-05,
|
| 1244 |
+
"loss": 10.3528,
|
| 1245 |
+
"step": 172
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.006349322861232429,
|
| 1249 |
+
"grad_norm": 0.09055240452289581,
|
| 1250 |
+
"learning_rate": 1.4267368421052632e-05,
|
| 1251 |
+
"loss": 10.3456,
|
| 1252 |
+
"step": 173
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.006386024149447646,
|
| 1256 |
+
"grad_norm": 0.13818956911563873,
|
| 1257 |
+
"learning_rate": 1.3738947368421053e-05,
|
| 1258 |
+
"loss": 10.3314,
|
| 1259 |
+
"step": 174
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.006422725437662862,
|
| 1263 |
+
"grad_norm": 0.14228907227516174,
|
| 1264 |
+
"learning_rate": 1.3210526315789473e-05,
|
| 1265 |
+
"loss": 10.3522,
|
| 1266 |
+
"step": 175
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.006459426725878078,
|
| 1270 |
+
"grad_norm": 0.14903728663921356,
|
| 1271 |
+
"learning_rate": 1.2682105263157896e-05,
|
| 1272 |
+
"loss": 10.349,
|
| 1273 |
+
"step": 176
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.0064961280140932946,
|
| 1277 |
+
"grad_norm": 0.1314869523048401,
|
| 1278 |
+
"learning_rate": 1.2153684210526315e-05,
|
| 1279 |
+
"loss": 10.3435,
|
| 1280 |
+
"step": 177
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.006532829302308511,
|
| 1284 |
+
"grad_norm": 0.17975281178951263,
|
| 1285 |
+
"learning_rate": 1.1625263157894737e-05,
|
| 1286 |
+
"loss": 10.3414,
|
| 1287 |
+
"step": 178
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.006569530590523728,
|
| 1291 |
+
"grad_norm": 0.15279750525951385,
|
| 1292 |
+
"learning_rate": 1.1096842105263158e-05,
|
| 1293 |
+
"loss": 10.3484,
|
| 1294 |
+
"step": 179
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.006606231878738943,
|
| 1298 |
+
"grad_norm": 0.14017164707183838,
|
| 1299 |
+
"learning_rate": 1.0568421052631579e-05,
|
| 1300 |
+
"loss": 10.3335,
|
| 1301 |
+
"step": 180
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.00664293316695416,
|
| 1305 |
+
"grad_norm": 0.12534311413764954,
|
| 1306 |
+
"learning_rate": 1.004e-05,
|
| 1307 |
+
"loss": 10.3396,
|
| 1308 |
+
"step": 181
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.006679634455169376,
|
| 1312 |
+
"grad_norm": 0.1480255424976349,
|
| 1313 |
+
"learning_rate": 9.511578947368422e-06,
|
| 1314 |
+
"loss": 10.3398,
|
| 1315 |
+
"step": 182
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.006716335743384593,
|
| 1319 |
+
"grad_norm": 0.15242981910705566,
|
| 1320 |
+
"learning_rate": 8.983157894736843e-06,
|
| 1321 |
+
"loss": 10.3556,
|
| 1322 |
+
"step": 183
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.0067530370315998095,
|
| 1326 |
+
"grad_norm": 0.15913036465644836,
|
| 1327 |
+
"learning_rate": 8.454736842105263e-06,
|
| 1328 |
+
"loss": 10.3392,
|
| 1329 |
+
"step": 184
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.006789738319815025,
|
| 1333 |
+
"grad_norm": 0.1393824964761734,
|
| 1334 |
+
"learning_rate": 7.926315789473684e-06,
|
| 1335 |
+
"loss": 10.3467,
|
| 1336 |
+
"step": 185
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.006826439608030242,
|
| 1340 |
+
"grad_norm": 0.1997556835412979,
|
| 1341 |
+
"learning_rate": 7.397894736842104e-06,
|
| 1342 |
+
"loss": 10.3353,
|
| 1343 |
+
"step": 186
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.006863140896245458,
|
| 1347 |
+
"grad_norm": 0.17985405027866364,
|
| 1348 |
+
"learning_rate": 6.8694736842105265e-06,
|
| 1349 |
+
"loss": 10.3442,
|
| 1350 |
+
"step": 187
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.006899842184460675,
|
| 1354 |
+
"grad_norm": 0.19826477766036987,
|
| 1355 |
+
"learning_rate": 6.341052631578948e-06,
|
| 1356 |
+
"loss": 10.3395,
|
| 1357 |
+
"step": 188
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.0069365434726758905,
|
| 1361 |
+
"grad_norm": 0.2310490906238556,
|
| 1362 |
+
"learning_rate": 5.812631578947368e-06,
|
| 1363 |
+
"loss": 10.345,
|
| 1364 |
+
"step": 189
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.006973244760891107,
|
| 1368 |
+
"grad_norm": 0.16586144268512726,
|
| 1369 |
+
"learning_rate": 5.2842105263157896e-06,
|
| 1370 |
+
"loss": 10.3328,
|
| 1371 |
+
"step": 190
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.007009946049106324,
|
| 1375 |
+
"grad_norm": 0.1728939265012741,
|
| 1376 |
+
"learning_rate": 4.755789473684211e-06,
|
| 1377 |
+
"loss": 10.345,
|
| 1378 |
+
"step": 191
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.00704664733732154,
|
| 1382 |
+
"grad_norm": 0.16085892915725708,
|
| 1383 |
+
"learning_rate": 4.227368421052631e-06,
|
| 1384 |
+
"loss": 10.3423,
|
| 1385 |
+
"step": 192
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.007083348625536757,
|
| 1389 |
+
"grad_norm": 0.16840451955795288,
|
| 1390 |
+
"learning_rate": 3.698947368421052e-06,
|
| 1391 |
+
"loss": 10.344,
|
| 1392 |
+
"step": 193
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.007120049913751972,
|
| 1396 |
+
"grad_norm": 0.19459696114063263,
|
| 1397 |
+
"learning_rate": 3.170526315789474e-06,
|
| 1398 |
+
"loss": 10.3336,
|
| 1399 |
+
"step": 194
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.007156751201967189,
|
| 1403 |
+
"grad_norm": 0.20487509667873383,
|
| 1404 |
+
"learning_rate": 2.6421052631578948e-06,
|
| 1405 |
+
"loss": 10.3365,
|
| 1406 |
+
"step": 195
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.0071934524901824054,
|
| 1410 |
+
"grad_norm": 0.21706953644752502,
|
| 1411 |
+
"learning_rate": 2.1136842105263157e-06,
|
| 1412 |
+
"loss": 10.3362,
|
| 1413 |
+
"step": 196
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.007230153778397622,
|
| 1417 |
+
"grad_norm": 0.22209207713603973,
|
| 1418 |
+
"learning_rate": 1.585263157894737e-06,
|
| 1419 |
+
"loss": 10.3437,
|
| 1420 |
+
"step": 197
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.0072668550666128385,
|
| 1424 |
+
"grad_norm": 0.2134910374879837,
|
| 1425 |
+
"learning_rate": 1.0568421052631578e-06,
|
| 1426 |
+
"loss": 10.3488,
|
| 1427 |
+
"step": 198
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.007303556354828054,
|
| 1431 |
+
"grad_norm": 0.24191489815711975,
|
| 1432 |
+
"learning_rate": 5.284210526315789e-07,
|
| 1433 |
+
"loss": 10.3286,
|
| 1434 |
+
"step": 199
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.007340257643043271,
|
| 1438 |
+
"grad_norm": 0.5283753275871277,
|
| 1439 |
+
"learning_rate": 0.0,
|
| 1440 |
+
"loss": 10.3668,
|
| 1441 |
+
"step": 200
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.007340257643043271,
|
| 1445 |
+
"eval_loss": 10.344128608703613,
|
| 1446 |
+
"eval_runtime": 65.8798,
|
| 1447 |
+
"eval_samples_per_second": 174.151,
|
| 1448 |
+
"eval_steps_per_second": 43.549,
|
| 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": 5591476862976.0,
|
| 1479 |
"train_batch_size": 4,
|
| 1480 |
"trial_name": null,
|
| 1481 |
"trial_params": null
|