shulijia commited on
Commit
f4225fe
·
verified ·
1 Parent(s): f3c184d

Training in progress, step 1500, checkpoint

Browse files
last-checkpoint/model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6fa45215f37f017ea114f5d41b72d95e062c31d219651ae1208803b3cc29ec1d
3
  size 2384234968
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ecb1c43c9a6caae32eedb74afa839b6db8df1604949ac103d65018afa63a05ce
3
  size 2384234968
last-checkpoint/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:779732aa59c1ab3aa00d1a1a4bafcd133473f97fb57c9527840c5270b6084f09
3
  size 4768662910
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d22ddd6eb04ea57bbea8812aaf6fcc69925e5324bb385f88fbe59698add4850d
3
  size 4768662910
last-checkpoint/rng_state.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5edb34d031c0c2b447f3eaadb401a4c1e7e7e6d8c096e28b7092e01a8bd48c92
3
  size 14244
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52cca5856c568bc52c683b690919168fa27bfbdfefc6e0a62355afa6011157c3
3
  size 14244
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:39e70346aa38efc4a7628079f4588786c86aea2dc16a2dfc23ef706d27a775bd
3
  size 1064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7cc13ac7175e5452c0a51656649ea1f0782593abf479b3fbadc42f21b535fdb6
3
  size 1064
last-checkpoint/trainer_state.json CHANGED
@@ -2,9 +2,9 @@
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
- "epoch": 0.5180676078228209,
6
  "eval_steps": 100,
7
- "global_step": 1000,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
@@ -988,6 +988,496 @@
988
  "eval_samples_per_second": 9.665,
989
  "eval_steps_per_second": 1.211,
990
  "step": 1000
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
991
  }
992
  ],
993
  "logging_steps": 10,
@@ -1007,7 +1497,7 @@
1007
  "attributes": {}
1008
  }
1009
  },
1010
- "total_flos": 1.0824928198656e+16,
1011
  "train_batch_size": 1,
1012
  "trial_name": null,
1013
  "trial_params": null
 
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
+ "epoch": 0.7771014117342313,
6
  "eval_steps": 100,
7
+ "global_step": 1500,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
 
988
  "eval_samples_per_second": 9.665,
989
  "eval_steps_per_second": 1.211,
990
  "step": 1000
991
+ },
992
+ {
993
+ "epoch": 0.523248283901049,
994
+ "grad_norm": 0.665536105632782,
995
+ "learning_rate": 2.6511226252158895e-05,
996
+ "loss": 0.1562,
997
+ "mean_token_accuracy": 0.9715508781373501,
998
+ "num_tokens": 4136960.0,
999
+ "step": 1010
1000
+ },
1001
+ {
1002
+ "epoch": 0.5284289599792773,
1003
+ "grad_norm": 0.7251952886581421,
1004
+ "learning_rate": 2.622337363270006e-05,
1005
+ "loss": 0.1752,
1006
+ "mean_token_accuracy": 0.9673434413969517,
1007
+ "num_tokens": 4177920.0,
1008
+ "step": 1020
1009
+ },
1010
+ {
1011
+ "epoch": 0.5336096360575056,
1012
+ "grad_norm": 0.8343415260314941,
1013
+ "learning_rate": 2.5935521013241222e-05,
1014
+ "loss": 0.1601,
1015
+ "mean_token_accuracy": 0.9714774891734124,
1016
+ "num_tokens": 4218880.0,
1017
+ "step": 1030
1018
+ },
1019
+ {
1020
+ "epoch": 0.5387903121357337,
1021
+ "grad_norm": 0.7332581281661987,
1022
+ "learning_rate": 2.5647668393782386e-05,
1023
+ "loss": 0.1772,
1024
+ "mean_token_accuracy": 0.9684197597205639,
1025
+ "num_tokens": 4259840.0,
1026
+ "step": 1040
1027
+ },
1028
+ {
1029
+ "epoch": 0.543970988213962,
1030
+ "grad_norm": 0.7279286980628967,
1031
+ "learning_rate": 2.5359815774323546e-05,
1032
+ "loss": 0.1666,
1033
+ "mean_token_accuracy": 0.9698630087077618,
1034
+ "num_tokens": 4300800.0,
1035
+ "step": 1050
1036
+ },
1037
+ {
1038
+ "epoch": 0.5491516642921901,
1039
+ "grad_norm": 0.7851588726043701,
1040
+ "learning_rate": 2.507196315486471e-05,
1041
+ "loss": 0.1549,
1042
+ "mean_token_accuracy": 0.9704011715948582,
1043
+ "num_tokens": 4341760.0,
1044
+ "step": 1060
1045
+ },
1046
+ {
1047
+ "epoch": 0.5543323403704183,
1048
+ "grad_norm": 1.1083506345748901,
1049
+ "learning_rate": 2.4784110535405873e-05,
1050
+ "loss": 0.1634,
1051
+ "mean_token_accuracy": 0.9700831651687623,
1052
+ "num_tokens": 4382720.0,
1053
+ "step": 1070
1054
+ },
1055
+ {
1056
+ "epoch": 0.5595130164486466,
1057
+ "grad_norm": 1.1186856031417847,
1058
+ "learning_rate": 2.4496257915947037e-05,
1059
+ "loss": 0.1647,
1060
+ "mean_token_accuracy": 0.9702544026076794,
1061
+ "num_tokens": 4423680.0,
1062
+ "step": 1080
1063
+ },
1064
+ {
1065
+ "epoch": 0.5646936925268747,
1066
+ "grad_norm": 0.7933104038238525,
1067
+ "learning_rate": 2.42084052964882e-05,
1068
+ "loss": 0.1477,
1069
+ "mean_token_accuracy": 0.9730185873806476,
1070
+ "num_tokens": 4464640.0,
1071
+ "step": 1090
1072
+ },
1073
+ {
1074
+ "epoch": 0.569874368605103,
1075
+ "grad_norm": 0.7880426049232483,
1076
+ "learning_rate": 2.3920552677029364e-05,
1077
+ "loss": 0.1615,
1078
+ "step": 1100
1079
+ },
1080
+ {
1081
+ "epoch": 0.569874368605103,
1082
+ "eval_loss": 0.1579154133796692,
1083
+ "eval_mean_token_accuracy": 0.9709939410520154,
1084
+ "eval_num_tokens": 4505600.0,
1085
+ "eval_runtime": 177.4807,
1086
+ "eval_samples_per_second": 9.669,
1087
+ "eval_steps_per_second": 1.211,
1088
+ "step": 1100
1089
+ },
1090
+ {
1091
+ "epoch": 0.5750550446833311,
1092
+ "grad_norm": 0.8081638216972351,
1093
+ "learning_rate": 2.3632700057570527e-05,
1094
+ "loss": 0.1659,
1095
+ "mean_token_accuracy": 0.9697529342025518,
1096
+ "num_tokens": 4546560.0,
1097
+ "step": 1110
1098
+ },
1099
+ {
1100
+ "epoch": 0.5802357207615594,
1101
+ "grad_norm": 0.8306415677070618,
1102
+ "learning_rate": 2.3344847438111687e-05,
1103
+ "loss": 0.1633,
1104
+ "mean_token_accuracy": 0.9700342454016209,
1105
+ "num_tokens": 4587520.0,
1106
+ "step": 1120
1107
+ },
1108
+ {
1109
+ "epoch": 0.5854163968397876,
1110
+ "grad_norm": 0.7340885996818542,
1111
+ "learning_rate": 2.305699481865285e-05,
1112
+ "loss": 0.1564,
1113
+ "mean_token_accuracy": 0.9713551819324493,
1114
+ "num_tokens": 4628480.0,
1115
+ "step": 1130
1116
+ },
1117
+ {
1118
+ "epoch": 0.5905970729180158,
1119
+ "grad_norm": 0.7812165021896362,
1120
+ "learning_rate": 2.2769142199194014e-05,
1121
+ "loss": 0.163,
1122
+ "mean_token_accuracy": 0.97013209015131,
1123
+ "num_tokens": 4669440.0,
1124
+ "step": 1140
1125
+ },
1126
+ {
1127
+ "epoch": 0.595777748996244,
1128
+ "grad_norm": 0.8122887015342712,
1129
+ "learning_rate": 2.2481289579735178e-05,
1130
+ "loss": 0.1554,
1131
+ "mean_token_accuracy": 0.971697648614645,
1132
+ "num_tokens": 4710400.0,
1133
+ "step": 1150
1134
+ },
1135
+ {
1136
+ "epoch": 0.6009584250744722,
1137
+ "grad_norm": 0.9578766226768494,
1138
+ "learning_rate": 2.219343696027634e-05,
1139
+ "loss": 0.1534,
1140
+ "mean_token_accuracy": 0.9720401145517826,
1141
+ "num_tokens": 4751360.0,
1142
+ "step": 1160
1143
+ },
1144
+ {
1145
+ "epoch": 0.6061391011527004,
1146
+ "grad_norm": 0.7298418879508972,
1147
+ "learning_rate": 2.1905584340817505e-05,
1148
+ "loss": 0.1596,
1149
+ "mean_token_accuracy": 0.9701810151338577,
1150
+ "num_tokens": 4792320.0,
1151
+ "step": 1170
1152
+ },
1153
+ {
1154
+ "epoch": 0.6113197772309287,
1155
+ "grad_norm": 1.0097841024398804,
1156
+ "learning_rate": 2.1617731721358665e-05,
1157
+ "loss": 0.1729,
1158
+ "mean_token_accuracy": 0.9681262224912643,
1159
+ "num_tokens": 4833280.0,
1160
+ "step": 1180
1161
+ },
1162
+ {
1163
+ "epoch": 0.6165004533091568,
1164
+ "grad_norm": 0.7758947610855103,
1165
+ "learning_rate": 2.132987910189983e-05,
1166
+ "loss": 0.1513,
1167
+ "mean_token_accuracy": 0.9727984316647053,
1168
+ "num_tokens": 4874240.0,
1169
+ "step": 1190
1170
+ },
1171
+ {
1172
+ "epoch": 0.6216811293873851,
1173
+ "grad_norm": 0.899966835975647,
1174
+ "learning_rate": 2.1042026482440992e-05,
1175
+ "loss": 0.1532,
1176
+ "step": 1200
1177
+ },
1178
+ {
1179
+ "epoch": 0.6216811293873851,
1180
+ "eval_loss": 0.15650857985019684,
1181
+ "eval_mean_token_accuracy": 0.9711953251860862,
1182
+ "eval_num_tokens": 4915200.0,
1183
+ "eval_runtime": 177.4213,
1184
+ "eval_samples_per_second": 9.672,
1185
+ "eval_steps_per_second": 1.212,
1186
+ "step": 1200
1187
+ },
1188
+ {
1189
+ "epoch": 0.6268618054656132,
1190
+ "grad_norm": 0.9350607395172119,
1191
+ "learning_rate": 2.0754173862982156e-05,
1192
+ "loss": 0.1547,
1193
+ "mean_token_accuracy": 0.9720890387892723,
1194
+ "num_tokens": 4956160.0,
1195
+ "step": 1210
1196
+ },
1197
+ {
1198
+ "epoch": 0.6320424815438415,
1199
+ "grad_norm": 0.8335353136062622,
1200
+ "learning_rate": 2.046632124352332e-05,
1201
+ "loss": 0.1657,
1202
+ "mean_token_accuracy": 0.9699608609080315,
1203
+ "num_tokens": 4997120.0,
1204
+ "step": 1220
1205
+ },
1206
+ {
1207
+ "epoch": 0.6372231576220697,
1208
+ "grad_norm": 0.9029703736305237,
1209
+ "learning_rate": 2.0178468624064483e-05,
1210
+ "loss": 0.1506,
1211
+ "mean_token_accuracy": 0.9732142850756645,
1212
+ "num_tokens": 5038080.0,
1213
+ "step": 1230
1214
+ },
1215
+ {
1216
+ "epoch": 0.6424038337002979,
1217
+ "grad_norm": 0.6102480888366699,
1218
+ "learning_rate": 1.9890616004605643e-05,
1219
+ "loss": 0.1536,
1220
+ "mean_token_accuracy": 0.9712817966938019,
1221
+ "num_tokens": 5079040.0,
1222
+ "step": 1240
1223
+ },
1224
+ {
1225
+ "epoch": 0.6475845097785261,
1226
+ "grad_norm": 1.0168322324752808,
1227
+ "learning_rate": 1.9602763385146806e-05,
1228
+ "loss": 0.156,
1229
+ "mean_token_accuracy": 0.9712817974388599,
1230
+ "num_tokens": 5120000.0,
1231
+ "step": 1250
1232
+ },
1233
+ {
1234
+ "epoch": 0.6527651858567544,
1235
+ "grad_norm": 0.8640413284301758,
1236
+ "learning_rate": 1.931491076568797e-05,
1237
+ "loss": 0.1678,
1238
+ "mean_token_accuracy": 0.9678816020488739,
1239
+ "num_tokens": 5160960.0,
1240
+ "step": 1260
1241
+ },
1242
+ {
1243
+ "epoch": 0.6579458619349825,
1244
+ "grad_norm": 1.144752025604248,
1245
+ "learning_rate": 1.9027058146229133e-05,
1246
+ "loss": 0.1526,
1247
+ "mean_token_accuracy": 0.9730675160884857,
1248
+ "num_tokens": 5201920.0,
1249
+ "step": 1270
1250
+ },
1251
+ {
1252
+ "epoch": 0.6631265380132108,
1253
+ "grad_norm": 0.7036670446395874,
1254
+ "learning_rate": 1.8739205526770297e-05,
1255
+ "loss": 0.1448,
1256
+ "mean_token_accuracy": 0.9732632048428058,
1257
+ "num_tokens": 5242880.0,
1258
+ "step": 1280
1259
+ },
1260
+ {
1261
+ "epoch": 0.6683072140914389,
1262
+ "grad_norm": 0.8692848086357117,
1263
+ "learning_rate": 1.8451352907311457e-05,
1264
+ "loss": 0.1623,
1265
+ "mean_token_accuracy": 0.9700587041676044,
1266
+ "num_tokens": 5283840.0,
1267
+ "step": 1290
1268
+ },
1269
+ {
1270
+ "epoch": 0.6734878901696671,
1271
+ "grad_norm": 0.7516316175460815,
1272
+ "learning_rate": 1.816350028785262e-05,
1273
+ "loss": 0.1547,
1274
+ "step": 1300
1275
+ },
1276
+ {
1277
+ "epoch": 0.6734878901696671,
1278
+ "eval_loss": 0.1546616405248642,
1279
+ "eval_mean_token_accuracy": 0.9714706614960071,
1280
+ "eval_num_tokens": 5324800.0,
1281
+ "eval_runtime": 177.4261,
1282
+ "eval_samples_per_second": 9.672,
1283
+ "eval_steps_per_second": 1.212,
1284
+ "step": 1300
1285
+ },
1286
+ {
1287
+ "epoch": 0.6786685662478954,
1288
+ "grad_norm": 0.7789427638053894,
1289
+ "learning_rate": 1.7875647668393784e-05,
1290
+ "loss": 0.1611,
1291
+ "mean_token_accuracy": 0.9712328743189573,
1292
+ "num_tokens": 5365760.0,
1293
+ "step": 1310
1294
+ },
1295
+ {
1296
+ "epoch": 0.6838492423261235,
1297
+ "grad_norm": 0.7356511950492859,
1298
+ "learning_rate": 1.7587795048934948e-05,
1299
+ "loss": 0.1437,
1300
+ "mean_token_accuracy": 0.9725293546915055,
1301
+ "num_tokens": 5406720.0,
1302
+ "step": 1320
1303
+ },
1304
+ {
1305
+ "epoch": 0.6890299184043518,
1306
+ "grad_norm": 0.8623766303062439,
1307
+ "learning_rate": 1.729994242947611e-05,
1308
+ "loss": 0.1631,
1309
+ "mean_token_accuracy": 0.9693003863096237,
1310
+ "num_tokens": 5447680.0,
1311
+ "step": 1330
1312
+ },
1313
+ {
1314
+ "epoch": 0.6942105944825799,
1315
+ "grad_norm": 0.9133246541023254,
1316
+ "learning_rate": 1.7012089810017275e-05,
1317
+ "loss": 0.1614,
1318
+ "mean_token_accuracy": 0.9698874779045582,
1319
+ "num_tokens": 5488640.0,
1320
+ "step": 1340
1321
+ },
1322
+ {
1323
+ "epoch": 0.6993912705608082,
1324
+ "grad_norm": 1.3479807376861572,
1325
+ "learning_rate": 1.6724237190558435e-05,
1326
+ "loss": 0.16,
1327
+ "mean_token_accuracy": 0.970132089406252,
1328
+ "num_tokens": 5529600.0,
1329
+ "step": 1350
1330
+ },
1331
+ {
1332
+ "epoch": 0.7045719466390364,
1333
+ "grad_norm": 0.6969886422157288,
1334
+ "learning_rate": 1.6436384571099598e-05,
1335
+ "loss": 0.1543,
1336
+ "mean_token_accuracy": 0.9720156520605088,
1337
+ "num_tokens": 5570560.0,
1338
+ "step": 1360
1339
+ },
1340
+ {
1341
+ "epoch": 0.7097526227172646,
1342
+ "grad_norm": 0.7935067415237427,
1343
+ "learning_rate": 1.6148531951640762e-05,
1344
+ "loss": 0.176,
1345
+ "mean_token_accuracy": 0.9679060645401478,
1346
+ "num_tokens": 5611520.0,
1347
+ "step": 1370
1348
+ },
1349
+ {
1350
+ "epoch": 0.7149332987954928,
1351
+ "grad_norm": 0.7809213995933533,
1352
+ "learning_rate": 1.5860679332181925e-05,
1353
+ "loss": 0.148,
1354
+ "mean_token_accuracy": 0.9727739706635475,
1355
+ "num_tokens": 5652480.0,
1356
+ "step": 1380
1357
+ },
1358
+ {
1359
+ "epoch": 0.720113974873721,
1360
+ "grad_norm": 0.7228508591651917,
1361
+ "learning_rate": 1.557282671272309e-05,
1362
+ "loss": 0.1504,
1363
+ "mean_token_accuracy": 0.9726516619324684,
1364
+ "num_tokens": 5693440.0,
1365
+ "step": 1390
1366
+ },
1367
+ {
1368
+ "epoch": 0.7252946509519492,
1369
+ "grad_norm": 0.7751489281654358,
1370
+ "learning_rate": 1.528497409326425e-05,
1371
+ "loss": 0.1643,
1372
+ "step": 1400
1373
+ },
1374
+ {
1375
+ "epoch": 0.7252946509519492,
1376
+ "eval_loss": 0.1535310000181198,
1377
+ "eval_mean_token_accuracy": 0.9715741989224456,
1378
+ "eval_num_tokens": 5734400.0,
1379
+ "eval_runtime": 177.4294,
1380
+ "eval_samples_per_second": 9.671,
1381
+ "eval_steps_per_second": 1.212,
1382
+ "step": 1400
1383
+ },
1384
+ {
1385
+ "epoch": 0.7304753270301775,
1386
+ "grad_norm": 0.7995248436927795,
1387
+ "learning_rate": 1.499712147380541e-05,
1388
+ "loss": 0.1504,
1389
+ "mean_token_accuracy": 0.971832187101245,
1390
+ "num_tokens": 5775360.0,
1391
+ "step": 1410
1392
+ },
1393
+ {
1394
+ "epoch": 0.7356560031084056,
1395
+ "grad_norm": 0.7627185583114624,
1396
+ "learning_rate": 1.4709268854346574e-05,
1397
+ "loss": 0.1526,
1398
+ "mean_token_accuracy": 0.9707681015133858,
1399
+ "num_tokens": 5816320.0,
1400
+ "step": 1420
1401
+ },
1402
+ {
1403
+ "epoch": 0.7408366791866339,
1404
+ "grad_norm": 0.7853283286094666,
1405
+ "learning_rate": 1.4421416234887738e-05,
1406
+ "loss": 0.1653,
1407
+ "mean_token_accuracy": 0.9694716237485409,
1408
+ "num_tokens": 5857280.0,
1409
+ "step": 1430
1410
+ },
1411
+ {
1412
+ "epoch": 0.746017355264862,
1413
+ "grad_norm": 0.7680171132087708,
1414
+ "learning_rate": 1.4133563615428901e-05,
1415
+ "loss": 0.1402,
1416
+ "mean_token_accuracy": 0.9738992147147656,
1417
+ "num_tokens": 5898240.0,
1418
+ "step": 1440
1419
+ },
1420
+ {
1421
+ "epoch": 0.7511980313430903,
1422
+ "grad_norm": 0.7696517705917358,
1423
+ "learning_rate": 1.3845710995970063e-05,
1424
+ "loss": 0.1524,
1425
+ "mean_token_accuracy": 0.9718199595808983,
1426
+ "num_tokens": 5939200.0,
1427
+ "step": 1450
1428
+ },
1429
+ {
1430
+ "epoch": 0.7563787074213185,
1431
+ "grad_norm": 0.6429949402809143,
1432
+ "learning_rate": 1.3557858376511227e-05,
1433
+ "loss": 0.153,
1434
+ "mean_token_accuracy": 0.9708659470081329,
1435
+ "num_tokens": 5980160.0,
1436
+ "step": 1460
1437
+ },
1438
+ {
1439
+ "epoch": 0.7615593834995467,
1440
+ "grad_norm": 0.8178024888038635,
1441
+ "learning_rate": 1.327000575705239e-05,
1442
+ "loss": 0.147,
1443
+ "mean_token_accuracy": 0.971819955855608,
1444
+ "num_tokens": 6021120.0,
1445
+ "step": 1470
1446
+ },
1447
+ {
1448
+ "epoch": 0.7667400595777749,
1449
+ "grad_norm": 0.7854350209236145,
1450
+ "learning_rate": 1.2982153137593552e-05,
1451
+ "loss": 0.1534,
1452
+ "mean_token_accuracy": 0.9718933448195457,
1453
+ "num_tokens": 6062080.0,
1454
+ "step": 1480
1455
+ },
1456
+ {
1457
+ "epoch": 0.771920735656003,
1458
+ "grad_norm": 0.89176344871521,
1459
+ "learning_rate": 1.2694300518134716e-05,
1460
+ "loss": 0.1443,
1461
+ "mean_token_accuracy": 0.9731408931314945,
1462
+ "num_tokens": 6103040.0,
1463
+ "step": 1490
1464
+ },
1465
+ {
1466
+ "epoch": 0.7771014117342313,
1467
+ "grad_norm": 0.8288910984992981,
1468
+ "learning_rate": 1.240644789867588e-05,
1469
+ "loss": 0.1674,
1470
+ "step": 1500
1471
+ },
1472
+ {
1473
+ "epoch": 0.7771014117342313,
1474
+ "eval_loss": 0.15236081182956696,
1475
+ "eval_mean_token_accuracy": 0.9717971990274828,
1476
+ "eval_num_tokens": 6144000.0,
1477
+ "eval_runtime": 177.4234,
1478
+ "eval_samples_per_second": 9.672,
1479
+ "eval_steps_per_second": 1.212,
1480
+ "step": 1500
1481
  }
1482
  ],
1483
  "logging_steps": 10,
 
1497
  "attributes": {}
1498
  }
1499
  },
1500
+ "total_flos": 1.6237392297984e+16,
1501
  "train_batch_size": 1,
1502
  "trial_name": null,
1503
  "trial_params": null