Training in progress, step 1200, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 847599616
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d69eef6973f05a4ab208d92ce5c41ab5bad8f593947419444dd50e17b344c4ed
|
| 3 |
size 847599616
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 350603
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf14baaa9e1a00e7bed306c374f87311b812cf88c400368167158eef59997419
|
| 3 |
size 350603
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a04205b1d94836c1922e370df1e4800bd7ea54979ba8076642c732faa6296f9c
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7edabb139266eb2c4deb6ba34aaf61f1d05a108bf136a179d89f7f89db6e2e3f
|
| 3 |
size 1465
|
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.
|
| 6 |
"eval_steps": 150,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1322,6 +1322,444 @@
|
|
| 1322 |
"eval_samples_per_second": 9.638,
|
| 1323 |
"eval_steps_per_second": 1.638,
|
| 1324 |
"step": 900
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1325 |
}
|
| 1326 |
],
|
| 1327 |
"logging_steps": 5,
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.24,
|
| 6 |
"eval_steps": 150,
|
| 7 |
+
"global_step": 1200,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1322 |
"eval_samples_per_second": 9.638,
|
| 1323 |
"eval_steps_per_second": 1.638,
|
| 1324 |
"step": 900
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 0.181,
|
| 1328 |
+
"grad_norm": 0.068506620824337,
|
| 1329 |
+
"learning_rate": 3.8158014746627e-06,
|
| 1330 |
+
"loss": 6.659144592285156,
|
| 1331 |
+
"step": 905
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 0.182,
|
| 1335 |
+
"grad_norm": 0.06771203130483627,
|
| 1336 |
+
"learning_rate": 3.8130191106020916e-06,
|
| 1337 |
+
"loss": 6.752458190917968,
|
| 1338 |
+
"step": 910
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 0.183,
|
| 1342 |
+
"grad_norm": 0.07365380972623825,
|
| 1343 |
+
"learning_rate": 3.8102169196549606e-06,
|
| 1344 |
+
"loss": 6.792449951171875,
|
| 1345 |
+
"step": 915
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 0.184,
|
| 1349 |
+
"grad_norm": 0.09057911485433578,
|
| 1350 |
+
"learning_rate": 3.8073949324656187e-06,
|
| 1351 |
+
"loss": 6.750059509277344,
|
| 1352 |
+
"step": 920
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.185,
|
| 1356 |
+
"grad_norm": 0.1145537868142128,
|
| 1357 |
+
"learning_rate": 3.8045531798948662e-06,
|
| 1358 |
+
"loss": 6.659918212890625,
|
| 1359 |
+
"step": 925
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.186,
|
| 1363 |
+
"grad_norm": 0.07206426560878754,
|
| 1364 |
+
"learning_rate": 3.8016916930196535e-06,
|
| 1365 |
+
"loss": 6.64384765625,
|
| 1366 |
+
"step": 930
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 0.187,
|
| 1370 |
+
"grad_norm": 0.08511226624250412,
|
| 1371 |
+
"learning_rate": 3.798810503132742e-06,
|
| 1372 |
+
"loss": 6.741066741943359,
|
| 1373 |
+
"step": 935
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 0.188,
|
| 1377 |
+
"grad_norm": 0.07708951085805893,
|
| 1378 |
+
"learning_rate": 3.795909641742363e-06,
|
| 1379 |
+
"loss": 6.7345428466796875,
|
| 1380 |
+
"step": 940
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 0.189,
|
| 1384 |
+
"grad_norm": 0.10781801491975784,
|
| 1385 |
+
"learning_rate": 3.7929891405718725e-06,
|
| 1386 |
+
"loss": 6.796640014648437,
|
| 1387 |
+
"step": 945
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 0.19,
|
| 1391 |
+
"grad_norm": 0.07551534473896027,
|
| 1392 |
+
"learning_rate": 3.790049031559403e-06,
|
| 1393 |
+
"loss": 6.666252899169922,
|
| 1394 |
+
"step": 950
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.191,
|
| 1398 |
+
"grad_norm": 0.0841158851981163,
|
| 1399 |
+
"learning_rate": 3.787089346857515e-06,
|
| 1400 |
+
"loss": 6.701302337646484,
|
| 1401 |
+
"step": 955
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.192,
|
| 1405 |
+
"grad_norm": 0.0862930417060852,
|
| 1406 |
+
"learning_rate": 3.7841101188328477e-06,
|
| 1407 |
+
"loss": 6.614484405517578,
|
| 1408 |
+
"step": 960
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 0.193,
|
| 1412 |
+
"grad_norm": 0.07389580458402634,
|
| 1413 |
+
"learning_rate": 3.7811113800657597e-06,
|
| 1414 |
+
"loss": 6.6798454284667965,
|
| 1415 |
+
"step": 965
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 0.194,
|
| 1419 |
+
"grad_norm": 0.07656016200780869,
|
| 1420 |
+
"learning_rate": 3.778093163349979e-06,
|
| 1421 |
+
"loss": 6.734007263183594,
|
| 1422 |
+
"step": 970
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 0.195,
|
| 1426 |
+
"grad_norm": 0.06616536527872086,
|
| 1427 |
+
"learning_rate": 3.7750555016922397e-06,
|
| 1428 |
+
"loss": 6.699456787109375,
|
| 1429 |
+
"step": 975
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 0.196,
|
| 1433 |
+
"grad_norm": 0.0683082565665245,
|
| 1434 |
+
"learning_rate": 3.7719984283119227e-06,
|
| 1435 |
+
"loss": 6.68050537109375,
|
| 1436 |
+
"step": 980
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 0.197,
|
| 1440 |
+
"grad_norm": 0.1232413798570633,
|
| 1441 |
+
"learning_rate": 3.768921976640693e-06,
|
| 1442 |
+
"loss": 6.699308776855469,
|
| 1443 |
+
"step": 985
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 0.198,
|
| 1447 |
+
"grad_norm": 0.07576856017112732,
|
| 1448 |
+
"learning_rate": 3.7658261803221327e-06,
|
| 1449 |
+
"loss": 6.67384033203125,
|
| 1450 |
+
"step": 990
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 0.199,
|
| 1454 |
+
"grad_norm": 0.11441260576248169,
|
| 1455 |
+
"learning_rate": 3.7627110732113748e-06,
|
| 1456 |
+
"loss": 6.6964164733886715,
|
| 1457 |
+
"step": 995
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 0.2,
|
| 1461 |
+
"grad_norm": 0.16522112488746643,
|
| 1462 |
+
"learning_rate": 3.7595766893747325e-06,
|
| 1463 |
+
"loss": 6.660438537597656,
|
| 1464 |
+
"step": 1000
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 0.201,
|
| 1468 |
+
"grad_norm": 0.08984258025884628,
|
| 1469 |
+
"learning_rate": 3.7564230630893235e-06,
|
| 1470 |
+
"loss": 6.695003509521484,
|
| 1471 |
+
"step": 1005
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 0.202,
|
| 1475 |
+
"grad_norm": 0.1421244740486145,
|
| 1476 |
+
"learning_rate": 3.753250228842701e-06,
|
| 1477 |
+
"loss": 6.6853515625,
|
| 1478 |
+
"step": 1010
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 0.203,
|
| 1482 |
+
"grad_norm": 0.07362934201955795,
|
| 1483 |
+
"learning_rate": 3.7500582213324704e-06,
|
| 1484 |
+
"loss": 6.669375610351563,
|
| 1485 |
+
"step": 1015
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 0.204,
|
| 1489 |
+
"grad_norm": 0.08078701794147491,
|
| 1490 |
+
"learning_rate": 3.7468470754659164e-06,
|
| 1491 |
+
"loss": 6.731100463867188,
|
| 1492 |
+
"step": 1020
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 0.205,
|
| 1496 |
+
"grad_norm": 0.07613933831453323,
|
| 1497 |
+
"learning_rate": 3.743616826359614e-06,
|
| 1498 |
+
"loss": 6.699374389648438,
|
| 1499 |
+
"step": 1025
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 0.206,
|
| 1503 |
+
"grad_norm": 0.0956382229924202,
|
| 1504 |
+
"learning_rate": 3.740367509339051e-06,
|
| 1505 |
+
"loss": 6.7087150573730465,
|
| 1506 |
+
"step": 1030
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.207,
|
| 1510 |
+
"grad_norm": 0.06961136311292648,
|
| 1511 |
+
"learning_rate": 3.7370991599382374e-06,
|
| 1512 |
+
"loss": 6.619333648681641,
|
| 1513 |
+
"step": 1035
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 0.208,
|
| 1517 |
+
"grad_norm": 0.08904080837965012,
|
| 1518 |
+
"learning_rate": 3.7338118138993187e-06,
|
| 1519 |
+
"loss": 6.6525390625,
|
| 1520 |
+
"step": 1040
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 0.209,
|
| 1524 |
+
"grad_norm": 0.08716033399105072,
|
| 1525 |
+
"learning_rate": 3.7305055071721843e-06,
|
| 1526 |
+
"loss": 6.656343078613281,
|
| 1527 |
+
"step": 1045
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 0.21,
|
| 1531 |
+
"grad_norm": 0.06856364011764526,
|
| 1532 |
+
"learning_rate": 3.727180275914075e-06,
|
| 1533 |
+
"loss": 6.584860992431641,
|
| 1534 |
+
"step": 1050
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 0.21,
|
| 1538 |
+
"eval_accuracy": 0.13902075702075703,
|
| 1539 |
+
"eval_loss": 6.554867744445801,
|
| 1540 |
+
"eval_runtime": 10.4147,
|
| 1541 |
+
"eval_samples_per_second": 9.602,
|
| 1542 |
+
"eval_steps_per_second": 1.632,
|
| 1543 |
+
"step": 1050
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 0.211,
|
| 1547 |
+
"grad_norm": 0.13052572309970856,
|
| 1548 |
+
"learning_rate": 3.723836156489187e-06,
|
| 1549 |
+
"loss": 6.7243499755859375,
|
| 1550 |
+
"step": 1055
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.212,
|
| 1554 |
+
"grad_norm": 0.09087914228439331,
|
| 1555 |
+
"learning_rate": 3.7204731854682744e-06,
|
| 1556 |
+
"loss": 6.679405975341797,
|
| 1557 |
+
"step": 1060
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.213,
|
| 1561 |
+
"grad_norm": 0.08263049274682999,
|
| 1562 |
+
"learning_rate": 3.71709139962825e-06,
|
| 1563 |
+
"loss": 6.654142761230469,
|
| 1564 |
+
"step": 1065
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 0.214,
|
| 1568 |
+
"grad_norm": 0.07467084378004074,
|
| 1569 |
+
"learning_rate": 3.713690835951782e-06,
|
| 1570 |
+
"loss": 6.614359283447266,
|
| 1571 |
+
"step": 1070
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.215,
|
| 1575 |
+
"grad_norm": 0.073064424097538,
|
| 1576 |
+
"learning_rate": 3.710271531626889e-06,
|
| 1577 |
+
"loss": 6.567655944824219,
|
| 1578 |
+
"step": 1075
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.216,
|
| 1582 |
+
"grad_norm": 0.06931977719068527,
|
| 1583 |
+
"learning_rate": 3.706833524046536e-06,
|
| 1584 |
+
"loss": 6.589354705810547,
|
| 1585 |
+
"step": 1080
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.217,
|
| 1589 |
+
"grad_norm": 0.07509584724903107,
|
| 1590 |
+
"learning_rate": 3.703376850808223e-06,
|
| 1591 |
+
"loss": 6.710356903076172,
|
| 1592 |
+
"step": 1085
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.218,
|
| 1596 |
+
"grad_norm": 0.09575289487838745,
|
| 1597 |
+
"learning_rate": 3.699901549713575e-06,
|
| 1598 |
+
"loss": 6.6289726257324215,
|
| 1599 |
+
"step": 1090
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 0.219,
|
| 1603 |
+
"grad_norm": 0.09864190220832825,
|
| 1604 |
+
"learning_rate": 3.6964076587679268e-06,
|
| 1605 |
+
"loss": 6.5806831359863285,
|
| 1606 |
+
"step": 1095
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 0.22,
|
| 1610 |
+
"grad_norm": 0.23898720741271973,
|
| 1611 |
+
"learning_rate": 3.6928952161799113e-06,
|
| 1612 |
+
"loss": 6.584079742431641,
|
| 1613 |
+
"step": 1100
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.221,
|
| 1617 |
+
"grad_norm": 0.12194394320249557,
|
| 1618 |
+
"learning_rate": 3.6893642603610374e-06,
|
| 1619 |
+
"loss": 6.62684326171875,
|
| 1620 |
+
"step": 1105
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.222,
|
| 1624 |
+
"grad_norm": 0.07811807841062546,
|
| 1625 |
+
"learning_rate": 3.685814829925272e-06,
|
| 1626 |
+
"loss": 6.622457122802734,
|
| 1627 |
+
"step": 1110
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 0.223,
|
| 1631 |
+
"grad_norm": 0.0715632289648056,
|
| 1632 |
+
"learning_rate": 3.682246963688618e-06,
|
| 1633 |
+
"loss": 6.651722717285156,
|
| 1634 |
+
"step": 1115
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.224,
|
| 1638 |
+
"grad_norm": 0.10769006609916687,
|
| 1639 |
+
"learning_rate": 3.6786607006686895e-06,
|
| 1640 |
+
"loss": 6.6332145690917965,
|
| 1641 |
+
"step": 1120
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.225,
|
| 1645 |
+
"grad_norm": 0.06909330189228058,
|
| 1646 |
+
"learning_rate": 3.6750560800842837e-06,
|
| 1647 |
+
"loss": 6.65299072265625,
|
| 1648 |
+
"step": 1125
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.226,
|
| 1652 |
+
"grad_norm": 0.07310458272695541,
|
| 1653 |
+
"learning_rate": 3.671433141354953e-06,
|
| 1654 |
+
"loss": 6.617790985107422,
|
| 1655 |
+
"step": 1130
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.227,
|
| 1659 |
+
"grad_norm": 0.0946388989686966,
|
| 1660 |
+
"learning_rate": 3.667791924100576e-06,
|
| 1661 |
+
"loss": 6.5729728698730465,
|
| 1662 |
+
"step": 1135
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.228,
|
| 1666 |
+
"grad_norm": 0.10393473505973816,
|
| 1667 |
+
"learning_rate": 3.6641324681409215e-06,
|
| 1668 |
+
"loss": 6.568301391601563,
|
| 1669 |
+
"step": 1140
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.229,
|
| 1673 |
+
"grad_norm": 0.06355820596218109,
|
| 1674 |
+
"learning_rate": 3.6604548134952125e-06,
|
| 1675 |
+
"loss": 6.632942199707031,
|
| 1676 |
+
"step": 1145
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.23,
|
| 1680 |
+
"grad_norm": 0.06845368444919586,
|
| 1681 |
+
"learning_rate": 3.656759000381691e-06,
|
| 1682 |
+
"loss": 6.537193298339844,
|
| 1683 |
+
"step": 1150
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.231,
|
| 1687 |
+
"grad_norm": 0.0652310699224472,
|
| 1688 |
+
"learning_rate": 3.6530450692171787e-06,
|
| 1689 |
+
"loss": 6.58575668334961,
|
| 1690 |
+
"step": 1155
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.232,
|
| 1694 |
+
"grad_norm": 0.07378797978162766,
|
| 1695 |
+
"learning_rate": 3.64931306061663e-06,
|
| 1696 |
+
"loss": 6.643540954589843,
|
| 1697 |
+
"step": 1160
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.233,
|
| 1701 |
+
"grad_norm": 0.07988571375608444,
|
| 1702 |
+
"learning_rate": 3.6455630153926943e-06,
|
| 1703 |
+
"loss": 6.519971466064453,
|
| 1704 |
+
"step": 1165
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.234,
|
| 1708 |
+
"grad_norm": 0.06847741454839706,
|
| 1709 |
+
"learning_rate": 3.6417949745552638e-06,
|
| 1710 |
+
"loss": 6.616749572753906,
|
| 1711 |
+
"step": 1170
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.235,
|
| 1715 |
+
"grad_norm": 0.09600138664245605,
|
| 1716 |
+
"learning_rate": 3.6380089793110307e-06,
|
| 1717 |
+
"loss": 6.599732971191406,
|
| 1718 |
+
"step": 1175
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.236,
|
| 1722 |
+
"grad_norm": 0.09037528187036514,
|
| 1723 |
+
"learning_rate": 3.634205071063032e-06,
|
| 1724 |
+
"loss": 6.518192291259766,
|
| 1725 |
+
"step": 1180
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.237,
|
| 1729 |
+
"grad_norm": 0.12540659308433533,
|
| 1730 |
+
"learning_rate": 3.6303832914102e-06,
|
| 1731 |
+
"loss": 6.674017333984375,
|
| 1732 |
+
"step": 1185
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.238,
|
| 1736 |
+
"grad_norm": 0.07415056228637695,
|
| 1737 |
+
"learning_rate": 3.626543682146903e-06,
|
| 1738 |
+
"loss": 6.6067352294921875,
|
| 1739 |
+
"step": 1190
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.239,
|
| 1743 |
+
"grad_norm": 0.084800586104393,
|
| 1744 |
+
"learning_rate": 3.6226862852624927e-06,
|
| 1745 |
+
"loss": 6.528235626220703,
|
| 1746 |
+
"step": 1195
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 0.24,
|
| 1750 |
+
"grad_norm": 0.0605861060321331,
|
| 1751 |
+
"learning_rate": 3.6188111429408448e-06,
|
| 1752 |
+
"loss": 6.609767913818359,
|
| 1753 |
+
"step": 1200
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.24,
|
| 1757 |
+
"eval_accuracy": 0.13638827838827838,
|
| 1758 |
+
"eval_loss": 6.499367713928223,
|
| 1759 |
+
"eval_runtime": 10.4681,
|
| 1760 |
+
"eval_samples_per_second": 9.553,
|
| 1761 |
+
"eval_steps_per_second": 1.624,
|
| 1762 |
+
"step": 1200
|
| 1763 |
}
|
| 1764 |
],
|
| 1765 |
"logging_steps": 5,
|