Training in progress, step 2500, 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 328277848
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:35a25838e882b8b8d3abc2c30fca06831b937a101be8b7eca174b157589ba0d0
|
| 3 |
size 328277848
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 318646859
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08742255bcc023f4a34f8f2a127617bb854ab5fb96a4602c10f2895bfc656f64
|
| 3 |
size 318646859
|
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:a04575953c998a8fd3197b1b8249c8e72c33f4bb7c27b036788a4d9e537cf3cd
|
| 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 sha256:5c5f5054d1fb89b5c064db193ff9ee8b30b57ffe17a11e00d28cfa91ea00081e
|
| 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,
|
|
@@ -1447,6 +1447,364 @@
|
|
| 1447 |
"eval_samples_per_second": 255.761,
|
| 1448 |
"eval_steps_per_second": 5.371,
|
| 1449 |
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1450 |
}
|
| 1451 |
],
|
| 1452 |
"logging_steps": 10,
|
|
@@ -1466,7 +1824,7 @@
|
|
| 1466 |
"attributes": {}
|
| 1467 |
}
|
| 1468 |
},
|
| 1469 |
-
"total_flos":
|
| 1470 |
"train_batch_size": 48,
|
| 1471 |
"trial_name": null,
|
| 1472 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.42236864335191754,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 2500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1447 |
"eval_samples_per_second": 255.761,
|
| 1448 |
"eval_steps_per_second": 5.371,
|
| 1449 |
"step": 2000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.3395843892549417,
|
| 1453 |
+
"grad_norm": 1.0625675916671753,
|
| 1454 |
+
"learning_rate": 0.0002999993805131495,
|
| 1455 |
+
"loss": 5.482983016967774,
|
| 1456 |
+
"step": 2010
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 0.3412738638283494,
|
| 1460 |
+
"grad_norm": 0.9310702681541443,
|
| 1461 |
+
"learning_rate": 0.00029999723908369233,
|
| 1462 |
+
"loss": 5.477756500244141,
|
| 1463 |
+
"step": 2020
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.3429633384017571,
|
| 1467 |
+
"grad_norm": 0.8275931477546692,
|
| 1468 |
+
"learning_rate": 0.0002999935680854744,
|
| 1469 |
+
"loss": 5.4467018127441404,
|
| 1470 |
+
"step": 2030
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.3446528129751647,
|
| 1474 |
+
"grad_norm": 0.8972215056419373,
|
| 1475 |
+
"learning_rate": 0.00029998836755593,
|
| 1476 |
+
"loss": 5.415990829467773,
|
| 1477 |
+
"step": 2040
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 0.3463422875485724,
|
| 1481 |
+
"grad_norm": 1.0727229118347168,
|
| 1482 |
+
"learning_rate": 0.00029998163754809044,
|
| 1483 |
+
"loss": 5.403407287597656,
|
| 1484 |
+
"step": 2050
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.34803176212198006,
|
| 1488 |
+
"grad_norm": 1.0068520307540894,
|
| 1489 |
+
"learning_rate": 0.0002999733781305839,
|
| 1490 |
+
"loss": 5.4188987731933596,
|
| 1491 |
+
"step": 2060
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.34972123669538774,
|
| 1495 |
+
"grad_norm": 0.9327341914176941,
|
| 1496 |
+
"learning_rate": 0.00029996358938763406,
|
| 1497 |
+
"loss": 5.406315612792969,
|
| 1498 |
+
"step": 2070
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 0.3514107112687954,
|
| 1502 |
+
"grad_norm": 1.022828221321106,
|
| 1503 |
+
"learning_rate": 0.0002999522714190599,
|
| 1504 |
+
"loss": 5.410961532592774,
|
| 1505 |
+
"step": 2080
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.3531001858422031,
|
| 1509 |
+
"grad_norm": 0.8379955887794495,
|
| 1510 |
+
"learning_rate": 0.0002999394243402743,
|
| 1511 |
+
"loss": 5.411350250244141,
|
| 1512 |
+
"step": 2090
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 0.3547896604156107,
|
| 1516 |
+
"grad_norm": 0.8905497193336487,
|
| 1517 |
+
"learning_rate": 0.00029992504828228283,
|
| 1518 |
+
"loss": 5.384899520874024,
|
| 1519 |
+
"step": 2100
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 0.3564791349890184,
|
| 1523 |
+
"grad_norm": 0.7869957685470581,
|
| 1524 |
+
"learning_rate": 0.00029990914339168286,
|
| 1525 |
+
"loss": 5.391331481933594,
|
| 1526 |
+
"step": 2110
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 0.3581686095624261,
|
| 1530 |
+
"grad_norm": 0.7781967520713806,
|
| 1531 |
+
"learning_rate": 0.00029989170983066126,
|
| 1532 |
+
"loss": 5.365080261230469,
|
| 1533 |
+
"step": 2120
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.35985808413583376,
|
| 1537 |
+
"grad_norm": 0.8611620664596558,
|
| 1538 |
+
"learning_rate": 0.0002998727477769937,
|
| 1539 |
+
"loss": 5.367116546630859,
|
| 1540 |
+
"step": 2130
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.36154755870924143,
|
| 1544 |
+
"grad_norm": 0.8369846940040588,
|
| 1545 |
+
"learning_rate": 0.0002998522574240421,
|
| 1546 |
+
"loss": 5.361904525756836,
|
| 1547 |
+
"step": 2140
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 0.3632370332826491,
|
| 1551 |
+
"grad_norm": 0.893395721912384,
|
| 1552 |
+
"learning_rate": 0.00029983023898075305,
|
| 1553 |
+
"loss": 5.338259887695313,
|
| 1554 |
+
"step": 2150
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.3649265078560568,
|
| 1558 |
+
"grad_norm": 0.9806540012359619,
|
| 1559 |
+
"learning_rate": 0.00029980669267165545,
|
| 1560 |
+
"loss": 5.33393440246582,
|
| 1561 |
+
"step": 2160
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 0.3666159824294644,
|
| 1565 |
+
"grad_norm": 0.789153516292572,
|
| 1566 |
+
"learning_rate": 0.0002997816187368584,
|
| 1567 |
+
"loss": 5.347314834594727,
|
| 1568 |
+
"step": 2170
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.3683054570028721,
|
| 1572 |
+
"grad_norm": 0.731369137763977,
|
| 1573 |
+
"learning_rate": 0.00029975501743204866,
|
| 1574 |
+
"loss": 5.322664260864258,
|
| 1575 |
+
"step": 2180
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 0.3699949315762798,
|
| 1579 |
+
"grad_norm": 0.6811886429786682,
|
| 1580 |
+
"learning_rate": 0.00029972688902848803,
|
| 1581 |
+
"loss": 5.326079177856445,
|
| 1582 |
+
"step": 2190
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 0.37168440614968745,
|
| 1586 |
+
"grad_norm": 0.8143295645713806,
|
| 1587 |
+
"learning_rate": 0.0002996972338130106,
|
| 1588 |
+
"loss": 5.30379638671875,
|
| 1589 |
+
"step": 2200
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 0.37337388072309513,
|
| 1593 |
+
"grad_norm": 0.8854978680610657,
|
| 1594 |
+
"learning_rate": 0.00029966605208801996,
|
| 1595 |
+
"loss": 5.301242828369141,
|
| 1596 |
+
"step": 2210
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 0.3750633552965028,
|
| 1600 |
+
"grad_norm": 0.77631014585495,
|
| 1601 |
+
"learning_rate": 0.0002996333441714859,
|
| 1602 |
+
"loss": 5.294522476196289,
|
| 1603 |
+
"step": 2220
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 0.37675282986991043,
|
| 1607 |
+
"grad_norm": 0.7743359208106995,
|
| 1608 |
+
"learning_rate": 0.00029959911039694127,
|
| 1609 |
+
"loss": 5.313030624389649,
|
| 1610 |
+
"step": 2230
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.3784423044433181,
|
| 1614 |
+
"grad_norm": 0.8531479239463806,
|
| 1615 |
+
"learning_rate": 0.00029956335111347855,
|
| 1616 |
+
"loss": 5.275916671752929,
|
| 1617 |
+
"step": 2240
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 0.3801317790167258,
|
| 1621 |
+
"grad_norm": 0.722363293170929,
|
| 1622 |
+
"learning_rate": 0.0002995260666857463,
|
| 1623 |
+
"loss": 5.2906639099121096,
|
| 1624 |
+
"step": 2250
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 0.38182125359013347,
|
| 1628 |
+
"grad_norm": 0.7797225713729858,
|
| 1629 |
+
"learning_rate": 0.00029948725749394563,
|
| 1630 |
+
"loss": 5.2658641815185545,
|
| 1631 |
+
"step": 2260
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 0.38351072816354115,
|
| 1635 |
+
"grad_norm": 0.8231165409088135,
|
| 1636 |
+
"learning_rate": 0.00029944692393382586,
|
| 1637 |
+
"loss": 5.2770263671875,
|
| 1638 |
+
"step": 2270
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 0.3852002027369488,
|
| 1642 |
+
"grad_norm": 0.8083261847496033,
|
| 1643 |
+
"learning_rate": 0.000299405066416681,
|
| 1644 |
+
"loss": 5.277169799804687,
|
| 1645 |
+
"step": 2280
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 0.3868896773103565,
|
| 1649 |
+
"grad_norm": 0.8675849437713623,
|
| 1650 |
+
"learning_rate": 0.0002993616853693452,
|
| 1651 |
+
"loss": 5.258210754394531,
|
| 1652 |
+
"step": 2290
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.38857915188376413,
|
| 1656 |
+
"grad_norm": 0.7585932016372681,
|
| 1657 |
+
"learning_rate": 0.0002993167812341886,
|
| 1658 |
+
"loss": 5.252765655517578,
|
| 1659 |
+
"step": 2300
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 0.3902686264571718,
|
| 1663 |
+
"grad_norm": 0.8213605284690857,
|
| 1664 |
+
"learning_rate": 0.0002992703544691127,
|
| 1665 |
+
"loss": 5.222419357299804,
|
| 1666 |
+
"step": 2310
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 0.3919581010305795,
|
| 1670 |
+
"grad_norm": 0.7984234690666199,
|
| 1671 |
+
"learning_rate": 0.00029922240554754577,
|
| 1672 |
+
"loss": 5.227847671508789,
|
| 1673 |
+
"step": 2320
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 0.39364757560398717,
|
| 1677 |
+
"grad_norm": 0.8216149806976318,
|
| 1678 |
+
"learning_rate": 0.00029917293495843793,
|
| 1679 |
+
"loss": 5.215268325805664,
|
| 1680 |
+
"step": 2330
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 0.39533705017739484,
|
| 1684 |
+
"grad_norm": 0.7992113828659058,
|
| 1685 |
+
"learning_rate": 0.0002991219432062562,
|
| 1686 |
+
"loss": 5.251160049438477,
|
| 1687 |
+
"step": 2340
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 0.3970265247508025,
|
| 1691 |
+
"grad_norm": 0.7669650316238403,
|
| 1692 |
+
"learning_rate": 0.0002990694308109795,
|
| 1693 |
+
"loss": 5.255714797973633,
|
| 1694 |
+
"step": 2350
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 0.39871599932421015,
|
| 1698 |
+
"grad_norm": 0.7685340046882629,
|
| 1699 |
+
"learning_rate": 0.0002990153983080932,
|
| 1700 |
+
"loss": 5.2186332702636715,
|
| 1701 |
+
"step": 2360
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 0.4004054738976178,
|
| 1705 |
+
"grad_norm": 0.8289806246757507,
|
| 1706 |
+
"learning_rate": 0.0002989598462485835,
|
| 1707 |
+
"loss": 5.2316020965576175,
|
| 1708 |
+
"step": 2370
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 0.4020949484710255,
|
| 1712 |
+
"grad_norm": 0.7260857224464417,
|
| 1713 |
+
"learning_rate": 0.00029890277519893215,
|
| 1714 |
+
"loss": 5.210884857177734,
|
| 1715 |
+
"step": 2380
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 0.4037844230444332,
|
| 1719 |
+
"grad_norm": 0.6450658440589905,
|
| 1720 |
+
"learning_rate": 0.0002988441857411106,
|
| 1721 |
+
"loss": 5.194115066528321,
|
| 1722 |
+
"step": 2390
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 0.40547389761784086,
|
| 1726 |
+
"grad_norm": 0.723818838596344,
|
| 1727 |
+
"learning_rate": 0.0002987840784725737,
|
| 1728 |
+
"loss": 5.197711563110351,
|
| 1729 |
+
"step": 2400
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 0.40716337219124854,
|
| 1733 |
+
"grad_norm": 0.8113153576850891,
|
| 1734 |
+
"learning_rate": 0.0002987224540062542,
|
| 1735 |
+
"loss": 5.196290588378906,
|
| 1736 |
+
"step": 2410
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 0.4088528467646562,
|
| 1740 |
+
"grad_norm": 0.8224965929985046,
|
| 1741 |
+
"learning_rate": 0.00029865931297055605,
|
| 1742 |
+
"loss": 5.174480819702149,
|
| 1743 |
+
"step": 2420
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 0.41054232133806384,
|
| 1747 |
+
"grad_norm": 0.9786369204521179,
|
| 1748 |
+
"learning_rate": 0.00029859465600934814,
|
| 1749 |
+
"loss": 5.19611701965332,
|
| 1750 |
+
"step": 2430
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 0.4122317959114715,
|
| 1754 |
+
"grad_norm": 0.8020685911178589,
|
| 1755 |
+
"learning_rate": 0.0002985284837819577,
|
| 1756 |
+
"loss": 5.181368637084961,
|
| 1757 |
+
"step": 2440
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 0.4139212704848792,
|
| 1761 |
+
"grad_norm": 0.7282792329788208,
|
| 1762 |
+
"learning_rate": 0.0002984607969631636,
|
| 1763 |
+
"loss": 5.1728168487548825,
|
| 1764 |
+
"step": 2450
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 0.4156107450582869,
|
| 1768 |
+
"grad_norm": 0.6869542598724365,
|
| 1769 |
+
"learning_rate": 0.00029839159624318954,
|
| 1770 |
+
"loss": 5.172641372680664,
|
| 1771 |
+
"step": 2460
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 0.41730021963169456,
|
| 1775 |
+
"grad_norm": 0.8235262632369995,
|
| 1776 |
+
"learning_rate": 0.00029832088232769694,
|
| 1777 |
+
"loss": 5.165771484375,
|
| 1778 |
+
"step": 2470
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 0.41898969420510224,
|
| 1782 |
+
"grad_norm": 0.7626176476478577,
|
| 1783 |
+
"learning_rate": 0.0002982486559377776,
|
| 1784 |
+
"loss": 5.175928115844727,
|
| 1785 |
+
"step": 2480
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 0.42067916877850986,
|
| 1789 |
+
"grad_norm": 0.636053740978241,
|
| 1790 |
+
"learning_rate": 0.0002981749178099467,
|
| 1791 |
+
"loss": 5.135253143310547,
|
| 1792 |
+
"step": 2490
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 0.42236864335191754,
|
| 1796 |
+
"grad_norm": 0.6814470291137695,
|
| 1797 |
+
"learning_rate": 0.000298099668696135,
|
| 1798 |
+
"loss": 5.177354049682617,
|
| 1799 |
+
"step": 2500
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 0.42236864335191754,
|
| 1803 |
+
"eval_loss": 5.138686656951904,
|
| 1804 |
+
"eval_runtime": 3.9981,
|
| 1805 |
+
"eval_samples_per_second": 250.119,
|
| 1806 |
+
"eval_steps_per_second": 5.253,
|
| 1807 |
+
"step": 2500
|
| 1808 |
}
|
| 1809 |
],
|
| 1810 |
"logging_steps": 10,
|
|
|
|
| 1824 |
"attributes": {}
|
| 1825 |
}
|
| 1826 |
},
|
| 1827 |
+
"total_flos": 8.361420521472e+16,
|
| 1828 |
"train_batch_size": 48,
|
| 1829 |
"trial_name": null,
|
| 1830 |
"trial_params": null
|