Training in progress, step 2000, 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 536223056
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:07298ce60c7d8cb3c6e42c406800f94fe196f562e78c3b7b1b241dde9eb2a84a
|
| 3 |
size 536223056
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1072594443
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:27ee21889421d66393ba1ec9978f870de383bd198e67267eebcf4f41f26ae0d1
|
| 3 |
size 1072594443
|
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:ff72c43fe4ca3fcf70d83ca8f9e37fa2293fdb34dea33c1c2460b564dd80a06f
|
| 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": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1358,6 +1358,456 @@
|
|
| 1358 |
"mean_token_accuracy": 0.7567296206951142,
|
| 1359 |
"num_tokens": 1663193.0,
|
| 1360 |
"step": 1500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1361 |
}
|
| 1362 |
],
|
| 1363 |
"logging_steps": 10,
|
|
@@ -1377,7 +1827,7 @@
|
|
| 1377 |
"attributes": {}
|
| 1378 |
}
|
| 1379 |
},
|
| 1380 |
-
"total_flos":
|
| 1381 |
"train_batch_size": 8,
|
| 1382 |
"trial_name": null,
|
| 1383 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.40298206729800523,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 2000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1358 |
"mean_token_accuracy": 0.7567296206951142,
|
| 1359 |
"num_tokens": 1663193.0,
|
| 1360 |
"step": 1500
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.30425146080999393,
|
| 1364 |
+
"grad_norm": 11.75,
|
| 1365 |
+
"learning_rate": 1.7973000201491034e-05,
|
| 1366 |
+
"loss": 0.9706,
|
| 1367 |
+
"mean_token_accuracy": 0.7680864214897156,
|
| 1368 |
+
"num_tokens": 1674150.0,
|
| 1369 |
+
"step": 1510
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 0.306266371146484,
|
| 1373 |
+
"grad_norm": 11.75,
|
| 1374 |
+
"learning_rate": 1.7959567465914436e-05,
|
| 1375 |
+
"loss": 0.8763,
|
| 1376 |
+
"mean_token_accuracy": 0.7913140594959259,
|
| 1377 |
+
"num_tokens": 1683177.0,
|
| 1378 |
+
"step": 1520
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.308281281482974,
|
| 1382 |
+
"grad_norm": 13.25,
|
| 1383 |
+
"learning_rate": 1.7946134730337835e-05,
|
| 1384 |
+
"loss": 1.0059,
|
| 1385 |
+
"mean_token_accuracy": 0.7636462986469269,
|
| 1386 |
+
"num_tokens": 1694685.0,
|
| 1387 |
+
"step": 1530
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 0.31029619181946405,
|
| 1391 |
+
"grad_norm": 10.375,
|
| 1392 |
+
"learning_rate": 1.7932701994761234e-05,
|
| 1393 |
+
"loss": 0.9025,
|
| 1394 |
+
"mean_token_accuracy": 0.7839828193187713,
|
| 1395 |
+
"num_tokens": 1706205.0,
|
| 1396 |
+
"step": 1540
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 0.31231110215595403,
|
| 1400 |
+
"grad_norm": 11.6875,
|
| 1401 |
+
"learning_rate": 1.7919269259184636e-05,
|
| 1402 |
+
"loss": 0.8492,
|
| 1403 |
+
"mean_token_accuracy": 0.7961820960044861,
|
| 1404 |
+
"num_tokens": 1717490.0,
|
| 1405 |
+
"step": 1550
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.3143260124924441,
|
| 1409 |
+
"grad_norm": 12.625,
|
| 1410 |
+
"learning_rate": 1.7905836523608032e-05,
|
| 1411 |
+
"loss": 0.9433,
|
| 1412 |
+
"mean_token_accuracy": 0.7731126248836517,
|
| 1413 |
+
"num_tokens": 1728246.0,
|
| 1414 |
+
"step": 1560
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 0.3163409228289341,
|
| 1418 |
+
"grad_norm": 11.4375,
|
| 1419 |
+
"learning_rate": 1.7892403788031434e-05,
|
| 1420 |
+
"loss": 0.9186,
|
| 1421 |
+
"mean_token_accuracy": 0.7792913258075714,
|
| 1422 |
+
"num_tokens": 1739564.0,
|
| 1423 |
+
"step": 1570
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.31835583316542415,
|
| 1427 |
+
"grad_norm": 10.5625,
|
| 1428 |
+
"learning_rate": 1.7878971052454833e-05,
|
| 1429 |
+
"loss": 1.03,
|
| 1430 |
+
"mean_token_accuracy": 0.7519110560417175,
|
| 1431 |
+
"num_tokens": 1750523.0,
|
| 1432 |
+
"step": 1580
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 0.3203707435019142,
|
| 1436 |
+
"grad_norm": 14.125,
|
| 1437 |
+
"learning_rate": 1.7865538316878235e-05,
|
| 1438 |
+
"loss": 0.8591,
|
| 1439 |
+
"mean_token_accuracy": 0.7900331735610961,
|
| 1440 |
+
"num_tokens": 1761179.0,
|
| 1441 |
+
"step": 1590
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.3223856538384042,
|
| 1445 |
+
"grad_norm": 11.875,
|
| 1446 |
+
"learning_rate": 1.7852105581301634e-05,
|
| 1447 |
+
"loss": 0.9531,
|
| 1448 |
+
"mean_token_accuracy": 0.7734063506126404,
|
| 1449 |
+
"num_tokens": 1771529.0,
|
| 1450 |
+
"step": 1600
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 0.3244005641748942,
|
| 1454 |
+
"grad_norm": 12.125,
|
| 1455 |
+
"learning_rate": 1.7838672845725033e-05,
|
| 1456 |
+
"loss": 0.9979,
|
| 1457 |
+
"mean_token_accuracy": 0.7627780497074127,
|
| 1458 |
+
"num_tokens": 1783491.0,
|
| 1459 |
+
"step": 1610
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.32641547451138425,
|
| 1463 |
+
"grad_norm": 13.125,
|
| 1464 |
+
"learning_rate": 1.7825240110148435e-05,
|
| 1465 |
+
"loss": 0.9718,
|
| 1466 |
+
"mean_token_accuracy": 0.7701032817363739,
|
| 1467 |
+
"num_tokens": 1793982.0,
|
| 1468 |
+
"step": 1620
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 0.3284303848478743,
|
| 1472 |
+
"grad_norm": 10.875,
|
| 1473 |
+
"learning_rate": 1.7811807374571834e-05,
|
| 1474 |
+
"loss": 0.8964,
|
| 1475 |
+
"mean_token_accuracy": 0.778260862827301,
|
| 1476 |
+
"num_tokens": 1804365.0,
|
| 1477 |
+
"step": 1630
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 0.3304452951843643,
|
| 1481 |
+
"grad_norm": 12.25,
|
| 1482 |
+
"learning_rate": 1.7798374638995233e-05,
|
| 1483 |
+
"loss": 0.8359,
|
| 1484 |
+
"mean_token_accuracy": 0.7983898043632507,
|
| 1485 |
+
"num_tokens": 1815367.0,
|
| 1486 |
+
"step": 1640
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.3324602055208543,
|
| 1490 |
+
"grad_norm": 10.375,
|
| 1491 |
+
"learning_rate": 1.7784941903418635e-05,
|
| 1492 |
+
"loss": 0.9085,
|
| 1493 |
+
"mean_token_accuracy": 0.7792557597160339,
|
| 1494 |
+
"num_tokens": 1827485.0,
|
| 1495 |
+
"step": 1650
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.33447511585734435,
|
| 1499 |
+
"grad_norm": 11.0,
|
| 1500 |
+
"learning_rate": 1.777150916784203e-05,
|
| 1501 |
+
"loss": 0.9841,
|
| 1502 |
+
"mean_token_accuracy": 0.7674819171428681,
|
| 1503 |
+
"num_tokens": 1839427.0,
|
| 1504 |
+
"step": 1660
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 0.3364900261938344,
|
| 1508 |
+
"grad_norm": 11.9375,
|
| 1509 |
+
"learning_rate": 1.7758076432265433e-05,
|
| 1510 |
+
"loss": 0.9128,
|
| 1511 |
+
"mean_token_accuracy": 0.7775863528251648,
|
| 1512 |
+
"num_tokens": 1850272.0,
|
| 1513 |
+
"step": 1670
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 0.3385049365303244,
|
| 1517 |
+
"grad_norm": 12.1875,
|
| 1518 |
+
"learning_rate": 1.774464369668883e-05,
|
| 1519 |
+
"loss": 0.8176,
|
| 1520 |
+
"mean_token_accuracy": 0.7950083971023559,
|
| 1521 |
+
"num_tokens": 1861066.0,
|
| 1522 |
+
"step": 1680
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 0.3405198468668144,
|
| 1526 |
+
"grad_norm": 11.8125,
|
| 1527 |
+
"learning_rate": 1.7731210961112234e-05,
|
| 1528 |
+
"loss": 0.8512,
|
| 1529 |
+
"mean_token_accuracy": 0.7908262252807617,
|
| 1530 |
+
"num_tokens": 1873013.0,
|
| 1531 |
+
"step": 1690
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 0.34253475720330445,
|
| 1535 |
+
"grad_norm": 12.25,
|
| 1536 |
+
"learning_rate": 1.7717778225535632e-05,
|
| 1537 |
+
"loss": 0.9421,
|
| 1538 |
+
"mean_token_accuracy": 0.7760106027126312,
|
| 1539 |
+
"num_tokens": 1884754.0,
|
| 1540 |
+
"step": 1700
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.3445496675397945,
|
| 1544 |
+
"grad_norm": 10.3125,
|
| 1545 |
+
"learning_rate": 1.770434548995903e-05,
|
| 1546 |
+
"loss": 0.9274,
|
| 1547 |
+
"mean_token_accuracy": 0.7754071950912476,
|
| 1548 |
+
"num_tokens": 1894900.0,
|
| 1549 |
+
"step": 1710
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.34656457787628453,
|
| 1553 |
+
"grad_norm": 12.3125,
|
| 1554 |
+
"learning_rate": 1.7690912754382433e-05,
|
| 1555 |
+
"loss": 0.8674,
|
| 1556 |
+
"mean_token_accuracy": 0.7867493867874146,
|
| 1557 |
+
"num_tokens": 1904852.0,
|
| 1558 |
+
"step": 1720
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 0.3485794882127745,
|
| 1562 |
+
"grad_norm": 13.875,
|
| 1563 |
+
"learning_rate": 1.767748001880583e-05,
|
| 1564 |
+
"loss": 0.89,
|
| 1565 |
+
"mean_token_accuracy": 0.7835441708564759,
|
| 1566 |
+
"num_tokens": 1914856.0,
|
| 1567 |
+
"step": 1730
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 0.35059439854926455,
|
| 1571 |
+
"grad_norm": 17.375,
|
| 1572 |
+
"learning_rate": 1.766404728322923e-05,
|
| 1573 |
+
"loss": 0.895,
|
| 1574 |
+
"mean_token_accuracy": 0.777972149848938,
|
| 1575 |
+
"num_tokens": 1926675.0,
|
| 1576 |
+
"step": 1740
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 0.3526093088857546,
|
| 1580 |
+
"grad_norm": 13.375,
|
| 1581 |
+
"learning_rate": 1.765061454765263e-05,
|
| 1582 |
+
"loss": 0.9383,
|
| 1583 |
+
"mean_token_accuracy": 0.7738365948200225,
|
| 1584 |
+
"num_tokens": 1938008.0,
|
| 1585 |
+
"step": 1750
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 0.35462421922224463,
|
| 1589 |
+
"grad_norm": 12.4375,
|
| 1590 |
+
"learning_rate": 1.763718181207603e-05,
|
| 1591 |
+
"loss": 0.9798,
|
| 1592 |
+
"mean_token_accuracy": 0.7683565199375153,
|
| 1593 |
+
"num_tokens": 1947741.0,
|
| 1594 |
+
"step": 1760
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.3566391295587346,
|
| 1598 |
+
"grad_norm": 11.1875,
|
| 1599 |
+
"learning_rate": 1.762374907649943e-05,
|
| 1600 |
+
"loss": 0.9872,
|
| 1601 |
+
"mean_token_accuracy": 0.7626317620277405,
|
| 1602 |
+
"num_tokens": 1958393.0,
|
| 1603 |
+
"step": 1770
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 0.35865403989522465,
|
| 1607 |
+
"grad_norm": 9.625,
|
| 1608 |
+
"learning_rate": 1.761031634092283e-05,
|
| 1609 |
+
"loss": 0.8591,
|
| 1610 |
+
"mean_token_accuracy": 0.7906990587711334,
|
| 1611 |
+
"num_tokens": 1970077.0,
|
| 1612 |
+
"step": 1780
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.3606689502317147,
|
| 1616 |
+
"grad_norm": 10.5625,
|
| 1617 |
+
"learning_rate": 1.7596883605346232e-05,
|
| 1618 |
+
"loss": 0.8462,
|
| 1619 |
+
"mean_token_accuracy": 0.7935640037059783,
|
| 1620 |
+
"num_tokens": 1980357.0,
|
| 1621 |
+
"step": 1790
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 0.36268386056820473,
|
| 1625 |
+
"grad_norm": 11.6875,
|
| 1626 |
+
"learning_rate": 1.758345086976963e-05,
|
| 1627 |
+
"loss": 1.0495,
|
| 1628 |
+
"mean_token_accuracy": 0.7528632760047913,
|
| 1629 |
+
"num_tokens": 1991123.0,
|
| 1630 |
+
"step": 1800
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 0.3646987709046947,
|
| 1634 |
+
"grad_norm": 10.625,
|
| 1635 |
+
"learning_rate": 1.757001813419303e-05,
|
| 1636 |
+
"loss": 0.9717,
|
| 1637 |
+
"mean_token_accuracy": 0.7630142509937287,
|
| 1638 |
+
"num_tokens": 2001627.0,
|
| 1639 |
+
"step": 1810
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 0.36671368124118475,
|
| 1643 |
+
"grad_norm": 11.3125,
|
| 1644 |
+
"learning_rate": 1.7556585398616432e-05,
|
| 1645 |
+
"loss": 0.8413,
|
| 1646 |
+
"mean_token_accuracy": 0.792235940694809,
|
| 1647 |
+
"num_tokens": 2013933.0,
|
| 1648 |
+
"step": 1820
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.3687285915776748,
|
| 1652 |
+
"grad_norm": 12.9375,
|
| 1653 |
+
"learning_rate": 1.7543152663039827e-05,
|
| 1654 |
+
"loss": 0.9425,
|
| 1655 |
+
"mean_token_accuracy": 0.7698959350585938,
|
| 1656 |
+
"num_tokens": 2024301.0,
|
| 1657 |
+
"step": 1830
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.37074350191416483,
|
| 1661 |
+
"grad_norm": 11.4375,
|
| 1662 |
+
"learning_rate": 1.752971992746323e-05,
|
| 1663 |
+
"loss": 0.8502,
|
| 1664 |
+
"mean_token_accuracy": 0.781799453496933,
|
| 1665 |
+
"num_tokens": 2034610.0,
|
| 1666 |
+
"step": 1840
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 0.37275841225065487,
|
| 1670 |
+
"grad_norm": 10.3125,
|
| 1671 |
+
"learning_rate": 1.751628719188663e-05,
|
| 1672 |
+
"loss": 1.0261,
|
| 1673 |
+
"mean_token_accuracy": 0.7564146995544434,
|
| 1674 |
+
"num_tokens": 2046053.0,
|
| 1675 |
+
"step": 1850
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.37477332258714485,
|
| 1679 |
+
"grad_norm": 15.625,
|
| 1680 |
+
"learning_rate": 1.750285445631003e-05,
|
| 1681 |
+
"loss": 1.0081,
|
| 1682 |
+
"mean_token_accuracy": 0.760506784915924,
|
| 1683 |
+
"num_tokens": 2056818.0,
|
| 1684 |
+
"step": 1860
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.3767882329236349,
|
| 1688 |
+
"grad_norm": 13.25,
|
| 1689 |
+
"learning_rate": 1.748942172073343e-05,
|
| 1690 |
+
"loss": 1.0131,
|
| 1691 |
+
"mean_token_accuracy": 0.7612803816795349,
|
| 1692 |
+
"num_tokens": 2067615.0,
|
| 1693 |
+
"step": 1870
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 0.37880314326012493,
|
| 1697 |
+
"grad_norm": 9.8125,
|
| 1698 |
+
"learning_rate": 1.747598898515683e-05,
|
| 1699 |
+
"loss": 0.8831,
|
| 1700 |
+
"mean_token_accuracy": 0.7784870386123657,
|
| 1701 |
+
"num_tokens": 2079047.0,
|
| 1702 |
+
"step": 1880
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 0.38081805359661497,
|
| 1706 |
+
"grad_norm": 12.3125,
|
| 1707 |
+
"learning_rate": 1.746255624958023e-05,
|
| 1708 |
+
"loss": 1.0057,
|
| 1709 |
+
"mean_token_accuracy": 0.7611044764518737,
|
| 1710 |
+
"num_tokens": 2090465.0,
|
| 1711 |
+
"step": 1890
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.38283296393310495,
|
| 1715 |
+
"grad_norm": 12.25,
|
| 1716 |
+
"learning_rate": 1.7449123514003626e-05,
|
| 1717 |
+
"loss": 0.9935,
|
| 1718 |
+
"mean_token_accuracy": 0.7548797488212585,
|
| 1719 |
+
"num_tokens": 2102421.0,
|
| 1720 |
+
"step": 1900
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 0.384847874269595,
|
| 1724 |
+
"grad_norm": 12.0625,
|
| 1725 |
+
"learning_rate": 1.7435690778427028e-05,
|
| 1726 |
+
"loss": 0.8325,
|
| 1727 |
+
"mean_token_accuracy": 0.7887258946895599,
|
| 1728 |
+
"num_tokens": 2112546.0,
|
| 1729 |
+
"step": 1910
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 0.38686278460608503,
|
| 1733 |
+
"grad_norm": 20.25,
|
| 1734 |
+
"learning_rate": 1.7422258042850427e-05,
|
| 1735 |
+
"loss": 0.886,
|
| 1736 |
+
"mean_token_accuracy": 0.7784853160381318,
|
| 1737 |
+
"num_tokens": 2123698.0,
|
| 1738 |
+
"step": 1920
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.38887769494257507,
|
| 1742 |
+
"grad_norm": 14.25,
|
| 1743 |
+
"learning_rate": 1.7408825307273826e-05,
|
| 1744 |
+
"loss": 0.9794,
|
| 1745 |
+
"mean_token_accuracy": 0.7646930754184723,
|
| 1746 |
+
"num_tokens": 2134235.0,
|
| 1747 |
+
"step": 1930
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.3908926052790651,
|
| 1751 |
+
"grad_norm": 13.125,
|
| 1752 |
+
"learning_rate": 1.7395392571697228e-05,
|
| 1753 |
+
"loss": 0.8836,
|
| 1754 |
+
"mean_token_accuracy": 0.7887324035167694,
|
| 1755 |
+
"num_tokens": 2144224.0,
|
| 1756 |
+
"step": 1940
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 0.3929075156155551,
|
| 1760 |
+
"grad_norm": 13.125,
|
| 1761 |
+
"learning_rate": 1.7381959836120627e-05,
|
| 1762 |
+
"loss": 0.8652,
|
| 1763 |
+
"mean_token_accuracy": 0.7922823071479798,
|
| 1764 |
+
"num_tokens": 2154554.0,
|
| 1765 |
+
"step": 1950
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 0.39492242595204513,
|
| 1769 |
+
"grad_norm": 12.375,
|
| 1770 |
+
"learning_rate": 1.736852710054403e-05,
|
| 1771 |
+
"loss": 0.9197,
|
| 1772 |
+
"mean_token_accuracy": 0.7768184185028076,
|
| 1773 |
+
"num_tokens": 2165026.0,
|
| 1774 |
+
"step": 1960
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.39693733628853517,
|
| 1778 |
+
"grad_norm": 12.125,
|
| 1779 |
+
"learning_rate": 1.7355094364967428e-05,
|
| 1780 |
+
"loss": 0.9668,
|
| 1781 |
+
"mean_token_accuracy": 0.7621938049793243,
|
| 1782 |
+
"num_tokens": 2176032.0,
|
| 1783 |
+
"step": 1970
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.3989522466250252,
|
| 1787 |
+
"grad_norm": 13.0,
|
| 1788 |
+
"learning_rate": 1.7341661629390827e-05,
|
| 1789 |
+
"loss": 0.8779,
|
| 1790 |
+
"mean_token_accuracy": 0.7878279089927673,
|
| 1791 |
+
"num_tokens": 2186771.0,
|
| 1792 |
+
"step": 1980
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 0.4009671569615152,
|
| 1796 |
+
"grad_norm": 11.8125,
|
| 1797 |
+
"learning_rate": 1.7328228893814226e-05,
|
| 1798 |
+
"loss": 0.8448,
|
| 1799 |
+
"mean_token_accuracy": 0.7939679741859436,
|
| 1800 |
+
"num_tokens": 2197173.0,
|
| 1801 |
+
"step": 1990
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.40298206729800523,
|
| 1805 |
+
"grad_norm": 10.6875,
|
| 1806 |
+
"learning_rate": 1.7314796158237625e-05,
|
| 1807 |
+
"loss": 0.821,
|
| 1808 |
+
"mean_token_accuracy": 0.7935750424861908,
|
| 1809 |
+
"num_tokens": 2208982.0,
|
| 1810 |
+
"step": 2000
|
| 1811 |
}
|
| 1812 |
],
|
| 1813 |
"logging_steps": 10,
|
|
|
|
| 1827 |
"attributes": {}
|
| 1828 |
}
|
| 1829 |
},
|
| 1830 |
+
"total_flos": 2679210270320640.0,
|
| 1831 |
"train_batch_size": 8,
|
| 1832 |
"trial_name": null,
|
| 1833 |
"trial_params": null
|