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 737580392
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d26418e85327bf4072ba76aa01dde18a1716eaa86660c26fb299ceaf6d71c5cd
|
| 3 |
size 737580392
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1475248442
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6df5d87deb87ff09bd2a9c94e587bd25dd1604f3de6ca09e11fd8c4b22cb9b3a
|
| 3 |
size 1475248442
|
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 sha256:7c733f603eddfbf8aea2db46dfb96d2d44052ea3c8d772ba82a9011002700581
|
| 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 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05888e792c6fd0be576ebc92f377eec054b707510467523fa4c87d94e18c0540
|
| 3 |
size 1000
|
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": 2.0,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1407,6 +1407,356 @@
|
|
| 1407 |
"learning_rate": 2.8828491096532332e-05,
|
| 1408 |
"loss": 0.3261,
|
| 1409 |
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1410 |
}
|
| 1411 |
],
|
| 1412 |
"logging_steps": 10,
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.1715089034676662,
|
| 5 |
"eval_steps": 2.0,
|
| 6 |
+
"global_step": 2500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1407 |
"learning_rate": 2.8828491096532332e-05,
|
| 1408 |
"loss": 0.3261,
|
| 1409 |
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.9418931583880038,
|
| 1413 |
+
"grad_norm": 1.3600877523422241,
|
| 1414 |
+
"learning_rate": 2.8822633552014998e-05,
|
| 1415 |
+
"loss": 0.2581,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.9465791940018744,
|
| 1420 |
+
"grad_norm": 2.0142452716827393,
|
| 1421 |
+
"learning_rate": 2.8816776007497657e-05,
|
| 1422 |
+
"loss": 0.2926,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.9512652296157451,
|
| 1427 |
+
"grad_norm": 2.2205429077148438,
|
| 1428 |
+
"learning_rate": 2.881091846298032e-05,
|
| 1429 |
+
"loss": 0.2978,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.9559512652296157,
|
| 1434 |
+
"grad_norm": 1.5943437814712524,
|
| 1435 |
+
"learning_rate": 2.880506091846298e-05,
|
| 1436 |
+
"loss": 0.2801,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.9606373008434864,
|
| 1441 |
+
"grad_norm": 1.8343549966812134,
|
| 1442 |
+
"learning_rate": 2.8799203373945644e-05,
|
| 1443 |
+
"loss": 0.3144,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.9653233364573571,
|
| 1448 |
+
"grad_norm": 2.438775062561035,
|
| 1449 |
+
"learning_rate": 2.8793345829428303e-05,
|
| 1450 |
+
"loss": 0.3031,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.9700093720712277,
|
| 1455 |
+
"grad_norm": 1.6122407913208008,
|
| 1456 |
+
"learning_rate": 2.8787488284910966e-05,
|
| 1457 |
+
"loss": 0.3102,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.9746954076850984,
|
| 1462 |
+
"grad_norm": 1.8684320449829102,
|
| 1463 |
+
"learning_rate": 2.8781630740393625e-05,
|
| 1464 |
+
"loss": 0.3064,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.979381443298969,
|
| 1469 |
+
"grad_norm": 1.4418230056762695,
|
| 1470 |
+
"learning_rate": 2.877577319587629e-05,
|
| 1471 |
+
"loss": 0.2812,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.9840674789128397,
|
| 1476 |
+
"grad_norm": 1.684371829032898,
|
| 1477 |
+
"learning_rate": 2.8769915651358953e-05,
|
| 1478 |
+
"loss": 0.2871,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.9887535145267105,
|
| 1483 |
+
"grad_norm": 1.8952980041503906,
|
| 1484 |
+
"learning_rate": 2.8764058106841612e-05,
|
| 1485 |
+
"loss": 0.3098,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.993439550140581,
|
| 1490 |
+
"grad_norm": 1.9636595249176025,
|
| 1491 |
+
"learning_rate": 2.8758200562324274e-05,
|
| 1492 |
+
"loss": 0.2493,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.9981255857544518,
|
| 1497 |
+
"grad_norm": 1.9494707584381104,
|
| 1498 |
+
"learning_rate": 2.8752343017806937e-05,
|
| 1499 |
+
"loss": 0.304,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 1.0028116213683225,
|
| 1504 |
+
"grad_norm": 1.5540398359298706,
|
| 1505 |
+
"learning_rate": 2.87464854732896e-05,
|
| 1506 |
+
"loss": 0.2886,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 1.007497656982193,
|
| 1511 |
+
"grad_norm": 1.4713643789291382,
|
| 1512 |
+
"learning_rate": 2.8740627928772258e-05,
|
| 1513 |
+
"loss": 0.2194,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 1.0121836925960637,
|
| 1518 |
+
"grad_norm": 1.5880275964736938,
|
| 1519 |
+
"learning_rate": 2.8734770384254924e-05,
|
| 1520 |
+
"loss": 0.208,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 1.0168697282099344,
|
| 1525 |
+
"grad_norm": 1.8418326377868652,
|
| 1526 |
+
"learning_rate": 2.8728912839737583e-05,
|
| 1527 |
+
"loss": 0.2111,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.021555763823805,
|
| 1532 |
+
"grad_norm": 1.6184002161026,
|
| 1533 |
+
"learning_rate": 2.8723055295220245e-05,
|
| 1534 |
+
"loss": 0.223,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 1.0262417994376758,
|
| 1539 |
+
"grad_norm": 1.714209794998169,
|
| 1540 |
+
"learning_rate": 2.8717197750702904e-05,
|
| 1541 |
+
"loss": 0.2069,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.0309278350515463,
|
| 1546 |
+
"grad_norm": 1.788393497467041,
|
| 1547 |
+
"learning_rate": 2.871134020618557e-05,
|
| 1548 |
+
"loss": 0.1906,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 1.035613870665417,
|
| 1553 |
+
"grad_norm": 1.3916633129119873,
|
| 1554 |
+
"learning_rate": 2.870548266166823e-05,
|
| 1555 |
+
"loss": 0.2254,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 1.0402999062792877,
|
| 1560 |
+
"grad_norm": 1.4615910053253174,
|
| 1561 |
+
"learning_rate": 2.869962511715089e-05,
|
| 1562 |
+
"loss": 0.2189,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 1.0449859418931584,
|
| 1567 |
+
"grad_norm": 1.9119755029678345,
|
| 1568 |
+
"learning_rate": 2.869376757263355e-05,
|
| 1569 |
+
"loss": 0.2505,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.0496719775070291,
|
| 1574 |
+
"grad_norm": 1.8039140701293945,
|
| 1575 |
+
"learning_rate": 2.8687910028116216e-05,
|
| 1576 |
+
"loss": 0.2127,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 1.0543580131208996,
|
| 1581 |
+
"grad_norm": 1.394958734512329,
|
| 1582 |
+
"learning_rate": 2.8682052483598875e-05,
|
| 1583 |
+
"loss": 0.2105,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.0590440487347703,
|
| 1588 |
+
"grad_norm": 1.660672903060913,
|
| 1589 |
+
"learning_rate": 2.8676194939081537e-05,
|
| 1590 |
+
"loss": 0.2337,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.063730084348641,
|
| 1595 |
+
"grad_norm": 1.7532225847244263,
|
| 1596 |
+
"learning_rate": 2.8670337394564196e-05,
|
| 1597 |
+
"loss": 0.2435,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.0684161199625117,
|
| 1602 |
+
"grad_norm": 1.5861754417419434,
|
| 1603 |
+
"learning_rate": 2.8664479850046862e-05,
|
| 1604 |
+
"loss": 0.2258,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 1.0731021555763824,
|
| 1609 |
+
"grad_norm": 2.068645477294922,
|
| 1610 |
+
"learning_rate": 2.865862230552952e-05,
|
| 1611 |
+
"loss": 0.2332,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 1.077788191190253,
|
| 1616 |
+
"grad_norm": 1.5656896829605103,
|
| 1617 |
+
"learning_rate": 2.8652764761012184e-05,
|
| 1618 |
+
"loss": 0.226,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 1.0824742268041236,
|
| 1623 |
+
"grad_norm": 1.4442079067230225,
|
| 1624 |
+
"learning_rate": 2.8646907216494846e-05,
|
| 1625 |
+
"loss": 0.2371,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 1.0871602624179943,
|
| 1630 |
+
"grad_norm": 1.9563679695129395,
|
| 1631 |
+
"learning_rate": 2.864104967197751e-05,
|
| 1632 |
+
"loss": 0.1806,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 1.091846298031865,
|
| 1637 |
+
"grad_norm": 2.3428823947906494,
|
| 1638 |
+
"learning_rate": 2.863519212746017e-05,
|
| 1639 |
+
"loss": 0.2078,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 1.0965323336457358,
|
| 1644 |
+
"grad_norm": 1.921542763710022,
|
| 1645 |
+
"learning_rate": 2.862933458294283e-05,
|
| 1646 |
+
"loss": 0.1881,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.1012183692596063,
|
| 1651 |
+
"grad_norm": 2.0102312564849854,
|
| 1652 |
+
"learning_rate": 2.8623477038425496e-05,
|
| 1653 |
+
"loss": 0.2039,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.105904404873477,
|
| 1658 |
+
"grad_norm": 2.0634419918060303,
|
| 1659 |
+
"learning_rate": 2.8617619493908155e-05,
|
| 1660 |
+
"loss": 0.2375,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 1.1105904404873477,
|
| 1665 |
+
"grad_norm": 1.7795848846435547,
|
| 1666 |
+
"learning_rate": 2.8611761949390817e-05,
|
| 1667 |
+
"loss": 0.1827,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.1152764761012184,
|
| 1672 |
+
"grad_norm": 1.6190413236618042,
|
| 1673 |
+
"learning_rate": 2.8605904404873476e-05,
|
| 1674 |
+
"loss": 0.2205,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.119962511715089,
|
| 1679 |
+
"grad_norm": 1.917844295501709,
|
| 1680 |
+
"learning_rate": 2.8600046860356142e-05,
|
| 1681 |
+
"loss": 0.2144,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.1246485473289598,
|
| 1686 |
+
"grad_norm": 1.9821206331253052,
|
| 1687 |
+
"learning_rate": 2.85941893158388e-05,
|
| 1688 |
+
"loss": 0.2305,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.1293345829428303,
|
| 1693 |
+
"grad_norm": 1.897702932357788,
|
| 1694 |
+
"learning_rate": 2.8588331771321463e-05,
|
| 1695 |
+
"loss": 0.2189,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.134020618556701,
|
| 1700 |
+
"grad_norm": 1.575378179550171,
|
| 1701 |
+
"learning_rate": 2.8582474226804122e-05,
|
| 1702 |
+
"loss": 0.2072,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.1387066541705717,
|
| 1707 |
+
"grad_norm": 1.5781910419464111,
|
| 1708 |
+
"learning_rate": 2.8576616682286788e-05,
|
| 1709 |
+
"loss": 0.2167,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.1433926897844424,
|
| 1714 |
+
"grad_norm": 1.253761887550354,
|
| 1715 |
+
"learning_rate": 2.8570759137769447e-05,
|
| 1716 |
+
"loss": 0.2054,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 1.148078725398313,
|
| 1721 |
+
"grad_norm": 1.4306617975234985,
|
| 1722 |
+
"learning_rate": 2.856490159325211e-05,
|
| 1723 |
+
"loss": 0.2507,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.1527647610121836,
|
| 1728 |
+
"grad_norm": 1.815042495727539,
|
| 1729 |
+
"learning_rate": 2.855904404873477e-05,
|
| 1730 |
+
"loss": 0.1929,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 1.1574507966260543,
|
| 1735 |
+
"grad_norm": 1.771567940711975,
|
| 1736 |
+
"learning_rate": 2.8553186504217434e-05,
|
| 1737 |
+
"loss": 0.2143,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 1.162136832239925,
|
| 1742 |
+
"grad_norm": 1.0400657653808594,
|
| 1743 |
+
"learning_rate": 2.8547328959700093e-05,
|
| 1744 |
+
"loss": 0.223,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.1668228678537957,
|
| 1749 |
+
"grad_norm": 1.6412447690963745,
|
| 1750 |
+
"learning_rate": 2.8541471415182755e-05,
|
| 1751 |
+
"loss": 0.1909,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 1.1715089034676662,
|
| 1756 |
+
"grad_norm": 1.7133567333221436,
|
| 1757 |
+
"learning_rate": 2.8535613870665418e-05,
|
| 1758 |
+
"loss": 0.2286,
|
| 1759 |
+
"step": 2500
|
| 1760 |
}
|
| 1761 |
],
|
| 1762 |
"logging_steps": 10,
|