Training in progress, step 250, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3537299144
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c2a5ff429f650539cd5c6ad9ea7f9569fd24863056cad28726290ed985d9fea
|
| 3 |
size 3537299144
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1830175435
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85bb5e2364254f0b84ca558a536ce2983868014e01a90e171fbe557dd01d62f6
|
| 3 |
size 1830175435
|
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:d895ccae2b55d4ea213653ca4a80d00de131463e105716eab1b7022906f260bf
|
| 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 sha256:5b7f2a236446ef1e40ceb20dfad68baf17d74c3d4a45e7640820b9ddfc1c6c59
|
| 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": 50,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1440,6 +1440,364 @@
|
|
| 1440 |
"eval_samples_per_second": 2.575,
|
| 1441 |
"eval_steps_per_second": 0.644,
|
| 1442 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1443 |
}
|
| 1444 |
],
|
| 1445 |
"logging_steps": 1,
|
|
@@ -1459,7 +1817,7 @@
|
|
| 1459 |
"attributes": {}
|
| 1460 |
}
|
| 1461 |
},
|
| 1462 |
-
"total_flos":
|
| 1463 |
"train_batch_size": 4,
|
| 1464 |
"trial_name": null,
|
| 1465 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9389671361502347,
|
| 6 |
"eval_steps": 50,
|
| 7 |
+
"global_step": 250,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1440 |
"eval_samples_per_second": 2.575,
|
| 1441 |
"eval_steps_per_second": 0.644,
|
| 1442 |
"step": 200
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.7549295774647887,
|
| 1446 |
+
"grad_norm": 0.4975653886795044,
|
| 1447 |
+
"learning_rate": 3.0572508230355246e-05,
|
| 1448 |
+
"loss": 3.763,
|
| 1449 |
+
"step": 201
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.7586854460093897,
|
| 1453 |
+
"grad_norm": 0.5943359136581421,
|
| 1454 |
+
"learning_rate": 2.971452939326802e-05,
|
| 1455 |
+
"loss": 4.1011,
|
| 1456 |
+
"step": 202
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 0.7624413145539906,
|
| 1460 |
+
"grad_norm": 0.5947958827018738,
|
| 1461 |
+
"learning_rate": 2.8866656049429162e-05,
|
| 1462 |
+
"loss": 3.837,
|
| 1463 |
+
"step": 203
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.7661971830985915,
|
| 1467 |
+
"grad_norm": 0.55486661195755,
|
| 1468 |
+
"learning_rate": 2.8029010104237785e-05,
|
| 1469 |
+
"loss": 3.773,
|
| 1470 |
+
"step": 204
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.7699530516431925,
|
| 1474 |
+
"grad_norm": 0.6001894474029541,
|
| 1475 |
+
"learning_rate": 2.720171199261987e-05,
|
| 1476 |
+
"loss": 4.1092,
|
| 1477 |
+
"step": 205
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 0.7737089201877935,
|
| 1481 |
+
"grad_norm": 0.611171305179596,
|
| 1482 |
+
"learning_rate": 2.638488066171201e-05,
|
| 1483 |
+
"loss": 4.2872,
|
| 1484 |
+
"step": 206
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.7774647887323943,
|
| 1488 |
+
"grad_norm": 0.5929466485977173,
|
| 1489 |
+
"learning_rate": 2.5578633553759878e-05,
|
| 1490 |
+
"loss": 4.0139,
|
| 1491 |
+
"step": 207
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.7812206572769953,
|
| 1495 |
+
"grad_norm": 0.5859886407852173,
|
| 1496 |
+
"learning_rate": 2.4783086589232295e-05,
|
| 1497 |
+
"loss": 3.9495,
|
| 1498 |
+
"step": 208
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 0.7849765258215963,
|
| 1502 |
+
"grad_norm": 0.5463722348213196,
|
| 1503 |
+
"learning_rate": 2.3998354150154555e-05,
|
| 1504 |
+
"loss": 3.7008,
|
| 1505 |
+
"step": 209
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 0.7887323943661971,
|
| 1509 |
+
"grad_norm": 0.5370416045188904,
|
| 1510 |
+
"learning_rate": 2.3224549063662927e-05,
|
| 1511 |
+
"loss": 3.9123,
|
| 1512 |
+
"step": 210
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 0.7924882629107981,
|
| 1516 |
+
"grad_norm": 0.5654124021530151,
|
| 1517 |
+
"learning_rate": 2.246178258578234e-05,
|
| 1518 |
+
"loss": 3.816,
|
| 1519 |
+
"step": 211
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"epoch": 0.7962441314553991,
|
| 1523 |
+
"grad_norm": 0.5404929518699646,
|
| 1524 |
+
"learning_rate": 2.171016438543059e-05,
|
| 1525 |
+
"loss": 3.943,
|
| 1526 |
+
"step": 212
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 0.8,
|
| 1530 |
+
"grad_norm": 0.5264220237731934,
|
| 1531 |
+
"learning_rate": 2.096980252865005e-05,
|
| 1532 |
+
"loss": 3.8148,
|
| 1533 |
+
"step": 213
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.8037558685446009,
|
| 1537 |
+
"grad_norm": 0.5364089012145996,
|
| 1538 |
+
"learning_rate": 2.0240803463070425e-05,
|
| 1539 |
+
"loss": 4.0956,
|
| 1540 |
+
"step": 214
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 0.8075117370892019,
|
| 1544 |
+
"grad_norm": 0.49832502007484436,
|
| 1545 |
+
"learning_rate": 1.9523272002603742e-05,
|
| 1546 |
+
"loss": 3.5919,
|
| 1547 |
+
"step": 215
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 0.8112676056338028,
|
| 1551 |
+
"grad_norm": 0.5661212205886841,
|
| 1552 |
+
"learning_rate": 1.8817311312374564e-05,
|
| 1553 |
+
"loss": 3.9309,
|
| 1554 |
+
"step": 216
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.8150234741784037,
|
| 1558 |
+
"grad_norm": 0.6174516677856445,
|
| 1559 |
+
"learning_rate": 1.8123022893887065e-05,
|
| 1560 |
+
"loss": 4.4702,
|
| 1561 |
+
"step": 217
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 0.8187793427230047,
|
| 1565 |
+
"grad_norm": 0.5399917364120483,
|
| 1566 |
+
"learning_rate": 1.744050657043137e-05,
|
| 1567 |
+
"loss": 3.8469,
|
| 1568 |
+
"step": 218
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.8225352112676056,
|
| 1572 |
+
"grad_norm": 0.48354753851890564,
|
| 1573 |
+
"learning_rate": 1.6769860472731257e-05,
|
| 1574 |
+
"loss": 3.5587,
|
| 1575 |
+
"step": 219
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 0.8262910798122066,
|
| 1579 |
+
"grad_norm": 0.5603431463241577,
|
| 1580 |
+
"learning_rate": 1.6111181024835e-05,
|
| 1581 |
+
"loss": 4.3805,
|
| 1582 |
+
"step": 220
|
| 1583 |
+
},
|
| 1584 |
+
{
|
| 1585 |
+
"epoch": 0.8300469483568075,
|
| 1586 |
+
"grad_norm": 0.5792990326881409,
|
| 1587 |
+
"learning_rate": 1.5464562930251814e-05,
|
| 1588 |
+
"loss": 4.2204,
|
| 1589 |
+
"step": 221
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 0.8338028169014085,
|
| 1593 |
+
"grad_norm": 0.5376021862030029,
|
| 1594 |
+
"learning_rate": 1.4830099158335563e-05,
|
| 1595 |
+
"loss": 3.8365,
|
| 1596 |
+
"step": 222
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 0.8375586854460094,
|
| 1600 |
+
"grad_norm": 0.5793043971061707,
|
| 1601 |
+
"learning_rate": 1.4207880930917871e-05,
|
| 1602 |
+
"loss": 4.064,
|
| 1603 |
+
"step": 223
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 0.8413145539906103,
|
| 1607 |
+
"grad_norm": 0.5597378611564636,
|
| 1608 |
+
"learning_rate": 1.3597997709192378e-05,
|
| 1609 |
+
"loss": 3.8224,
|
| 1610 |
+
"step": 224
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.8450704225352113,
|
| 1614 |
+
"grad_norm": 0.5336353182792664,
|
| 1615 |
+
"learning_rate": 1.3000537180852212e-05,
|
| 1616 |
+
"loss": 3.7203,
|
| 1617 |
+
"step": 225
|
| 1618 |
+
},
|
| 1619 |
+
{
|
| 1620 |
+
"epoch": 0.8488262910798122,
|
| 1621 |
+
"grad_norm": 0.640953004360199,
|
| 1622 |
+
"learning_rate": 1.2415585247482498e-05,
|
| 1623 |
+
"loss": 4.3212,
|
| 1624 |
+
"step": 226
|
| 1625 |
+
},
|
| 1626 |
+
{
|
| 1627 |
+
"epoch": 0.8525821596244132,
|
| 1628 |
+
"grad_norm": 0.45982062816619873,
|
| 1629 |
+
"learning_rate": 1.1843226012209529e-05,
|
| 1630 |
+
"loss": 3.6229,
|
| 1631 |
+
"step": 227
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 0.856338028169014,
|
| 1635 |
+
"grad_norm": 0.5055301189422607,
|
| 1636 |
+
"learning_rate": 1.128354176760873e-05,
|
| 1637 |
+
"loss": 3.6906,
|
| 1638 |
+
"step": 228
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 0.860093896713615,
|
| 1642 |
+
"grad_norm": 0.4451459050178528,
|
| 1643 |
+
"learning_rate": 1.073661298387265e-05,
|
| 1644 |
+
"loss": 3.3596,
|
| 1645 |
+
"step": 229
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 0.863849765258216,
|
| 1649 |
+
"grad_norm": 0.6167091727256775,
|
| 1650 |
+
"learning_rate": 1.0202518297241237e-05,
|
| 1651 |
+
"loss": 4.6817,
|
| 1652 |
+
"step": 230
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.8676056338028169,
|
| 1656 |
+
"grad_norm": 0.5457577705383301,
|
| 1657 |
+
"learning_rate": 9.681334498695648e-06,
|
| 1658 |
+
"loss": 4.2546,
|
| 1659 |
+
"step": 231
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 0.8713615023474178,
|
| 1663 |
+
"grad_norm": 0.49405384063720703,
|
| 1664 |
+
"learning_rate": 9.173136522917457e-06,
|
| 1665 |
+
"loss": 3.7713,
|
| 1666 |
+
"step": 232
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 0.8751173708920188,
|
| 1670 |
+
"grad_norm": 0.5279140472412109,
|
| 1671 |
+
"learning_rate": 8.677997437514629e-06,
|
| 1672 |
+
"loss": 3.7468,
|
| 1673 |
+
"step": 233
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 0.8788732394366198,
|
| 1677 |
+
"grad_norm": 0.5161781311035156,
|
| 1678 |
+
"learning_rate": 8.195988432516078e-06,
|
| 1679 |
+
"loss": 4.2746,
|
| 1680 |
+
"step": 234
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 0.8826291079812206,
|
| 1684 |
+
"grad_norm": 0.5855900049209595,
|
| 1685 |
+
"learning_rate": 7.727178810136093e-06,
|
| 1686 |
+
"loss": 4.1113,
|
| 1687 |
+
"step": 235
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 0.8863849765258216,
|
| 1691 |
+
"grad_norm": 0.4686482548713684,
|
| 1692 |
+
"learning_rate": 7.27163597481022e-06,
|
| 1693 |
+
"loss": 3.3821,
|
| 1694 |
+
"step": 236
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 0.8901408450704226,
|
| 1698 |
+
"grad_norm": 0.5629131197929382,
|
| 1699 |
+
"learning_rate": 6.829425423504021e-06,
|
| 1700 |
+
"loss": 4.1901,
|
| 1701 |
+
"step": 237
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 0.8938967136150234,
|
| 1705 |
+
"grad_norm": 0.5782991647720337,
|
| 1706 |
+
"learning_rate": 6.4006107362960195e-06,
|
| 1707 |
+
"loss": 4.3302,
|
| 1708 |
+
"step": 238
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 0.8976525821596244,
|
| 1712 |
+
"grad_norm": 0.5707590579986572,
|
| 1713 |
+
"learning_rate": 5.985253567236304e-06,
|
| 1714 |
+
"loss": 3.9955,
|
| 1715 |
+
"step": 239
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 0.9014084507042254,
|
| 1719 |
+
"grad_norm": 0.4625610411167145,
|
| 1720 |
+
"learning_rate": 5.583413635482082e-06,
|
| 1721 |
+
"loss": 3.5662,
|
| 1722 |
+
"step": 240
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 0.9051643192488263,
|
| 1726 |
+
"grad_norm": 0.6621753573417664,
|
| 1727 |
+
"learning_rate": 5.19514871671134e-06,
|
| 1728 |
+
"loss": 4.5634,
|
| 1729 |
+
"step": 241
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 0.9089201877934272,
|
| 1733 |
+
"grad_norm": 0.4976242482662201,
|
| 1734 |
+
"learning_rate": 4.82051463481602e-06,
|
| 1735 |
+
"loss": 3.5897,
|
| 1736 |
+
"step": 242
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 0.9126760563380282,
|
| 1740 |
+
"grad_norm": 0.51161789894104,
|
| 1741 |
+
"learning_rate": 4.45956525387573e-06,
|
| 1742 |
+
"loss": 3.6594,
|
| 1743 |
+
"step": 243
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 0.9164319248826291,
|
| 1747 |
+
"grad_norm": 0.5785262584686279,
|
| 1748 |
+
"learning_rate": 4.112352470413328e-06,
|
| 1749 |
+
"loss": 4.031,
|
| 1750 |
+
"step": 244
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 0.92018779342723,
|
| 1754 |
+
"grad_norm": 0.5122177004814148,
|
| 1755 |
+
"learning_rate": 3.778926205933342e-06,
|
| 1756 |
+
"loss": 3.6733,
|
| 1757 |
+
"step": 245
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 0.923943661971831,
|
| 1761 |
+
"grad_norm": 0.5668466687202454,
|
| 1762 |
+
"learning_rate": 3.459334399744374e-06,
|
| 1763 |
+
"loss": 3.8761,
|
| 1764 |
+
"step": 246
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 0.927699530516432,
|
| 1768 |
+
"grad_norm": 0.5304160714149475,
|
| 1769 |
+
"learning_rate": 3.1536230020664417e-06,
|
| 1770 |
+
"loss": 3.3638,
|
| 1771 |
+
"step": 247
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 0.9314553990610329,
|
| 1775 |
+
"grad_norm": 0.5929594039916992,
|
| 1776 |
+
"learning_rate": 2.861835967424409e-06,
|
| 1777 |
+
"loss": 4.1158,
|
| 1778 |
+
"step": 248
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 0.9352112676056338,
|
| 1782 |
+
"grad_norm": 0.5661305785179138,
|
| 1783 |
+
"learning_rate": 2.5840152483282752e-06,
|
| 1784 |
+
"loss": 3.8846,
|
| 1785 |
+
"step": 249
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 0.9389671361502347,
|
| 1789 |
+
"grad_norm": 0.5555335879325867,
|
| 1790 |
+
"learning_rate": 2.3202007892413447e-06,
|
| 1791 |
+
"loss": 3.9409,
|
| 1792 |
+
"step": 250
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 0.9389671361502347,
|
| 1796 |
+
"eval_loss": 0.4938514232635498,
|
| 1797 |
+
"eval_runtime": 365.2814,
|
| 1798 |
+
"eval_samples_per_second": 2.593,
|
| 1799 |
+
"eval_steps_per_second": 0.649,
|
| 1800 |
+
"step": 250
|
| 1801 |
}
|
| 1802 |
],
|
| 1803 |
"logging_steps": 1,
|
|
|
|
| 1817 |
"attributes": {}
|
| 1818 |
}
|
| 1819 |
},
|
| 1820 |
+
"total_flos": 3.688311494350195e+18,
|
| 1821 |
"train_batch_size": 4,
|
| 1822 |
"trial_name": null,
|
| 1823 |
"trial_params": null
|