Training in progress, step 204, 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 9823216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f41af54ef48385f1e9240a01ed0e4e9778b1bbdcbbeaa3976f0744a97cee781a
|
| 3 |
size 9823216
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5962860
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d30bc5cfe87f818a6b3f472e279714bdad1dca3906fc14f6922585bbcdc9e13
|
| 3 |
size 5962860
|
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:d1d1251a5e12d0ab0e0cece07fc1165a0b3630a00ab4bccd575f3d646ed3d1cd
|
| 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:e783401ecc99df9e68aa18c0fa2df70c08a04fbed0cb3e6ae60e028c1074e54a
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1078,6 +1078,363 @@
|
|
| 1078 |
"learning_rate": 1.5349188304533413e-05,
|
| 1079 |
"loss": 1.9898,
|
| 1080 |
"step": 153
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1081 |
}
|
| 1082 |
],
|
| 1083 |
"logging_steps": 1,
|
|
@@ -1092,12 +1449,12 @@
|
|
| 1092 |
"should_evaluate": false,
|
| 1093 |
"should_log": false,
|
| 1094 |
"should_save": true,
|
| 1095 |
-
"should_training_stop":
|
| 1096 |
},
|
| 1097 |
"attributes": {}
|
| 1098 |
}
|
| 1099 |
},
|
| 1100 |
-
"total_flos":
|
| 1101 |
"train_batch_size": 4,
|
| 1102 |
"trial_name": null,
|
| 1103 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 204,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1078 |
"learning_rate": 1.5349188304533413e-05,
|
| 1079 |
"loss": 1.9898,
|
| 1080 |
"step": 153
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.7549019607843137,
|
| 1084 |
+
"grad_norm": 0.6448200941085815,
|
| 1085 |
+
"learning_rate": 1.4784473717366387e-05,
|
| 1086 |
+
"loss": 2.0231,
|
| 1087 |
+
"step": 154
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.7598039215686274,
|
| 1091 |
+
"grad_norm": 0.6143677234649658,
|
| 1092 |
+
"learning_rate": 1.4228535577631442e-05,
|
| 1093 |
+
"loss": 1.8083,
|
| 1094 |
+
"step": 155
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.7647058823529411,
|
| 1098 |
+
"grad_norm": 0.6792232990264893,
|
| 1099 |
+
"learning_rate": 1.3681512436768045e-05,
|
| 1100 |
+
"loss": 2.2576,
|
| 1101 |
+
"step": 156
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.7696078431372549,
|
| 1105 |
+
"grad_norm": 0.6397396922111511,
|
| 1106 |
+
"learning_rate": 1.314354062441106e-05,
|
| 1107 |
+
"loss": 1.9228,
|
| 1108 |
+
"step": 157
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.7745098039215687,
|
| 1112 |
+
"grad_norm": 0.6527466773986816,
|
| 1113 |
+
"learning_rate": 1.2614754214414548e-05,
|
| 1114 |
+
"loss": 1.9549,
|
| 1115 |
+
"step": 158
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.7794117647058824,
|
| 1119 |
+
"grad_norm": 0.558052659034729,
|
| 1120 |
+
"learning_rate": 1.2095284991437733e-05,
|
| 1121 |
+
"loss": 1.6198,
|
| 1122 |
+
"step": 159
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.7843137254901961,
|
| 1126 |
+
"grad_norm": 0.5108990669250488,
|
| 1127 |
+
"learning_rate": 1.1585262418101467e-05,
|
| 1128 |
+
"loss": 1.7669,
|
| 1129 |
+
"step": 160
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.7892156862745098,
|
| 1133 |
+
"grad_norm": 0.5814594030380249,
|
| 1134 |
+
"learning_rate": 1.1084813602723515e-05,
|
| 1135 |
+
"loss": 1.8603,
|
| 1136 |
+
"step": 161
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.7941176470588235,
|
| 1140 |
+
"grad_norm": 0.5403014421463013,
|
| 1141 |
+
"learning_rate": 1.0594063267640386e-05,
|
| 1142 |
+
"loss": 1.6909,
|
| 1143 |
+
"step": 162
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.7990196078431373,
|
| 1147 |
+
"grad_norm": 0.5353929400444031,
|
| 1148 |
+
"learning_rate": 1.0113133718124035e-05,
|
| 1149 |
+
"loss": 1.8579,
|
| 1150 |
+
"step": 163
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.803921568627451,
|
| 1154 |
+
"grad_norm": 0.6234535574913025,
|
| 1155 |
+
"learning_rate": 9.642144811900739e-06,
|
| 1156 |
+
"loss": 2.0002,
|
| 1157 |
+
"step": 164
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.8088235294117647,
|
| 1161 |
+
"grad_norm": 0.5333160161972046,
|
| 1162 |
+
"learning_rate": 9.181213929280046e-06,
|
| 1163 |
+
"loss": 1.7953,
|
| 1164 |
+
"step": 165
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.8137254901960784,
|
| 1168 |
+
"grad_norm": 0.5747147798538208,
|
| 1169 |
+
"learning_rate": 8.7304559439012e-06,
|
| 1170 |
+
"loss": 2.2429,
|
| 1171 |
+
"step": 166
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.8186274509803921,
|
| 1175 |
+
"grad_norm": 0.5337631702423096,
|
| 1176 |
+
"learning_rate": 8.28998319410413e-06,
|
| 1177 |
+
"loss": 1.7715,
|
| 1178 |
+
"step": 167
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.8235294117647058,
|
| 1182 |
+
"grad_norm": 0.6177130341529846,
|
| 1183 |
+
"learning_rate": 7.859905454932471e-06,
|
| 1184 |
+
"loss": 1.684,
|
| 1185 |
+
"step": 168
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.8284313725490197,
|
| 1189 |
+
"grad_norm": 0.5750278830528259,
|
| 1190 |
+
"learning_rate": 7.440329910775273e-06,
|
| 1191 |
+
"loss": 2.0278,
|
| 1192 |
+
"step": 169
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.8333333333333334,
|
| 1196 |
+
"grad_norm": 0.5838117599487305,
|
| 1197 |
+
"learning_rate": 7.031361128654401e-06,
|
| 1198 |
+
"loss": 1.9058,
|
| 1199 |
+
"step": 170
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.8382352941176471,
|
| 1203 |
+
"grad_norm": 0.573087215423584,
|
| 1204 |
+
"learning_rate": 6.633101032164274e-06,
|
| 1205 |
+
"loss": 2.153,
|
| 1206 |
+
"step": 171
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.8431372549019608,
|
| 1210 |
+
"grad_norm": 0.5336370468139648,
|
| 1211 |
+
"learning_rate": 6.2456488760703205e-06,
|
| 1212 |
+
"loss": 1.773,
|
| 1213 |
+
"step": 172
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.8480392156862745,
|
| 1217 |
+
"grad_norm": 0.628755509853363,
|
| 1218 |
+
"learning_rate": 5.869101221572654e-06,
|
| 1219 |
+
"loss": 1.949,
|
| 1220 |
+
"step": 173
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.8529411764705882,
|
| 1224 |
+
"grad_norm": 0.5883850455284119,
|
| 1225 |
+
"learning_rate": 5.5035519122409895e-06,
|
| 1226 |
+
"loss": 1.8569,
|
| 1227 |
+
"step": 174
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.8578431372549019,
|
| 1231 |
+
"grad_norm": 0.5533519983291626,
|
| 1232 |
+
"learning_rate": 5.149092050626825e-06,
|
| 1233 |
+
"loss": 1.9951,
|
| 1234 |
+
"step": 175
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.8627450980392157,
|
| 1238 |
+
"grad_norm": 0.5445258021354675,
|
| 1239 |
+
"learning_rate": 4.805809975558828e-06,
|
| 1240 |
+
"loss": 1.8357,
|
| 1241 |
+
"step": 176
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.8676470588235294,
|
| 1245 |
+
"grad_norm": 0.7583712339401245,
|
| 1246 |
+
"learning_rate": 4.47379124012689e-06,
|
| 1247 |
+
"loss": 1.9937,
|
| 1248 |
+
"step": 177
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.8725490196078431,
|
| 1252 |
+
"grad_norm": 0.550308883190155,
|
| 1253 |
+
"learning_rate": 4.153118590360561e-06,
|
| 1254 |
+
"loss": 1.7592,
|
| 1255 |
+
"step": 178
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.8774509803921569,
|
| 1259 |
+
"grad_norm": 0.5869921445846558,
|
| 1260 |
+
"learning_rate": 3.843871944606969e-06,
|
| 1261 |
+
"loss": 1.8233,
|
| 1262 |
+
"step": 179
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.8823529411764706,
|
| 1266 |
+
"grad_norm": 0.624453067779541,
|
| 1267 |
+
"learning_rate": 3.5461283736134722e-06,
|
| 1268 |
+
"loss": 1.7236,
|
| 1269 |
+
"step": 180
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.8872549019607843,
|
| 1273 |
+
"grad_norm": 0.5934841632843018,
|
| 1274 |
+
"learning_rate": 3.2599620813200837e-06,
|
| 1275 |
+
"loss": 2.2794,
|
| 1276 |
+
"step": 181
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.8921568627450981,
|
| 1280 |
+
"grad_norm": 0.6147695183753967,
|
| 1281 |
+
"learning_rate": 2.9854443863662262e-06,
|
| 1282 |
+
"loss": 2.0399,
|
| 1283 |
+
"step": 182
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.8970588235294118,
|
| 1287 |
+
"grad_norm": 0.6072641611099243,
|
| 1288 |
+
"learning_rate": 2.722643704316652e-06,
|
| 1289 |
+
"loss": 2.0926,
|
| 1290 |
+
"step": 183
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.9019607843137255,
|
| 1294 |
+
"grad_norm": 0.6701170206069946,
|
| 1295 |
+
"learning_rate": 2.4716255306108605e-06,
|
| 1296 |
+
"loss": 2.0332,
|
| 1297 |
+
"step": 184
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.9068627450980392,
|
| 1301 |
+
"grad_norm": 0.6028889417648315,
|
| 1302 |
+
"learning_rate": 2.2324524242402613e-06,
|
| 1303 |
+
"loss": 2.066,
|
| 1304 |
+
"step": 185
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.9117647058823529,
|
| 1308 |
+
"grad_norm": 0.5733740329742432,
|
| 1309 |
+
"learning_rate": 2.0051839921571448e-06,
|
| 1310 |
+
"loss": 2.119,
|
| 1311 |
+
"step": 186
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.9166666666666666,
|
| 1315 |
+
"grad_norm": 0.6689417958259583,
|
| 1316 |
+
"learning_rate": 1.7898768744194162e-06,
|
| 1317 |
+
"loss": 2.1784,
|
| 1318 |
+
"step": 187
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.9215686274509803,
|
| 1322 |
+
"grad_norm": 0.6812567114830017,
|
| 1323 |
+
"learning_rate": 1.5865847300746417e-06,
|
| 1324 |
+
"loss": 1.7479,
|
| 1325 |
+
"step": 188
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.9264705882352942,
|
| 1329 |
+
"grad_norm": 0.6046686172485352,
|
| 1330 |
+
"learning_rate": 1.3953582237871521e-06,
|
| 1331 |
+
"loss": 1.942,
|
| 1332 |
+
"step": 189
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.9313725490196079,
|
| 1336 |
+
"grad_norm": 0.6390747427940369,
|
| 1337 |
+
"learning_rate": 1.2162450132113201e-06,
|
| 1338 |
+
"loss": 2.1545,
|
| 1339 |
+
"step": 190
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.9362745098039216,
|
| 1343 |
+
"grad_norm": 0.6433112621307373,
|
| 1344 |
+
"learning_rate": 1.049289737114273e-06,
|
| 1345 |
+
"loss": 1.9614,
|
| 1346 |
+
"step": 191
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.9411764705882353,
|
| 1350 |
+
"grad_norm": 0.6422625780105591,
|
| 1351 |
+
"learning_rate": 8.945340042509797e-07,
|
| 1352 |
+
"loss": 2.0157,
|
| 1353 |
+
"step": 192
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.946078431372549,
|
| 1357 |
+
"grad_norm": 0.696194052696228,
|
| 1358 |
+
"learning_rate": 7.520163829944804e-07,
|
| 1359 |
+
"loss": 2.3783,
|
| 1360 |
+
"step": 193
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.9509803921568627,
|
| 1364 |
+
"grad_norm": 0.6186797618865967,
|
| 1365 |
+
"learning_rate": 6.217723917238128e-07,
|
| 1366 |
+
"loss": 2.0641,
|
| 1367 |
+
"step": 194
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.9558823529411765,
|
| 1371 |
+
"grad_norm": 0.7474207878112793,
|
| 1372 |
+
"learning_rate": 5.038344899721436e-07,
|
| 1373 |
+
"loss": 2.3323,
|
| 1374 |
+
"step": 195
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.9607843137254902,
|
| 1378 |
+
"grad_norm": 0.6354583501815796,
|
| 1379 |
+
"learning_rate": 3.9823207033710676e-07,
|
| 1380 |
+
"loss": 2.0189,
|
| 1381 |
+
"step": 196
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.9656862745098039,
|
| 1385 |
+
"grad_norm": 0.6712233424186707,
|
| 1386 |
+
"learning_rate": 3.0499145115561176e-07,
|
| 1387 |
+
"loss": 2.1329,
|
| 1388 |
+
"step": 197
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.9705882352941176,
|
| 1392 |
+
"grad_norm": 0.6209347248077393,
|
| 1393 |
+
"learning_rate": 2.2413586994470825e-07,
|
| 1394 |
+
"loss": 2.3842,
|
| 1395 |
+
"step": 198
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.9754901960784313,
|
| 1399 |
+
"grad_norm": 0.6732835173606873,
|
| 1400 |
+
"learning_rate": 1.5568547761034004e-07,
|
| 1401 |
+
"loss": 2.4314,
|
| 1402 |
+
"step": 199
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.9803921568627451,
|
| 1406 |
+
"grad_norm": 0.8073440194129944,
|
| 1407 |
+
"learning_rate": 9.965733342532924e-08,
|
| 1408 |
+
"loss": 2.5627,
|
| 1409 |
+
"step": 200
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.9852941176470589,
|
| 1413 |
+
"grad_norm": 0.6623374223709106,
|
| 1414 |
+
"learning_rate": 5.606540077782163e-08,
|
| 1415 |
+
"loss": 1.9204,
|
| 1416 |
+
"step": 201
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.9901960784313726,
|
| 1420 |
+
"grad_norm": 0.5475188493728638,
|
| 1421 |
+
"learning_rate": 2.4920543691309138e-08,
|
| 1422 |
+
"loss": 1.7823,
|
| 1423 |
+
"step": 202
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.9950980392156863,
|
| 1427 |
+
"grad_norm": 0.5792528390884399,
|
| 1428 |
+
"learning_rate": 6.2305241171345395e-09,
|
| 1429 |
+
"loss": 1.8127,
|
| 1430 |
+
"step": 203
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.0,
|
| 1434 |
+
"grad_norm": 0.741917610168457,
|
| 1435 |
+
"learning_rate": 0.0,
|
| 1436 |
+
"loss": 2.1431,
|
| 1437 |
+
"step": 204
|
| 1438 |
}
|
| 1439 |
],
|
| 1440 |
"logging_steps": 1,
|
|
|
|
| 1449 |
"should_evaluate": false,
|
| 1450 |
"should_log": false,
|
| 1451 |
"should_save": true,
|
| 1452 |
+
"should_training_stop": true
|
| 1453 |
},
|
| 1454 |
"attributes": {}
|
| 1455 |
}
|
| 1456 |
},
|
| 1457 |
+
"total_flos": 2652053607677952.0,
|
| 1458 |
"train_batch_size": 4,
|
| 1459 |
"trial_name": null,
|
| 1460 |
"trial_params": null
|