update
Browse files- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +693 -3
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1585984129
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:641c85a58e4cda476ffb7fb5c7db5fc4127f3ae4e0e86f2f8596e568eec42bb4
|
| 3 |
size 1585984129
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1561347007
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:795bd227bc8849bbff4aa95b4e34cb546e42243310b55d6d27498891fd16ede9
|
| 3 |
size 1561347007
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d554d225b4de5166bd769415b56e27db6938660c20b187cbf1afe1eaf730132
|
| 3 |
size 14503
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58dc44076a97aa8560be589b10abfd196fdd4df11329f0a9f0516806ffc7c4e9
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -1416,11 +1416,701 @@
|
|
| 1416 |
"learning_rate": 2.9185119574845e-05,
|
| 1417 |
"loss": 2.191,
|
| 1418 |
"step": 23500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1419 |
}
|
| 1420 |
],
|
| 1421 |
"max_steps": 56450,
|
| 1422 |
"num_train_epochs": 10,
|
| 1423 |
-
"total_flos":
|
| 1424 |
"trial_name": null,
|
| 1425 |
"trial_params": null
|
| 1426 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.2001682830698375,
|
| 5 |
+
"global_step": 35000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 1416 |
"learning_rate": 2.9185119574845e-05,
|
| 1417 |
"loss": 2.191,
|
| 1418 |
"step": 23500
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 4.18,
|
| 1422 |
+
"learning_rate": 2.9096545615589016e-05,
|
| 1423 |
+
"loss": 2.1673,
|
| 1424 |
+
"step": 23600
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 4.2,
|
| 1428 |
+
"learning_rate": 2.900797165633304e-05,
|
| 1429 |
+
"loss": 2.1964,
|
| 1430 |
+
"step": 23700
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 4.22,
|
| 1434 |
+
"learning_rate": 2.891939769707706e-05,
|
| 1435 |
+
"loss": 2.21,
|
| 1436 |
+
"step": 23800
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 4.23,
|
| 1440 |
+
"learning_rate": 2.8830823737821084e-05,
|
| 1441 |
+
"loss": 2.2209,
|
| 1442 |
+
"step": 23900
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 4.25,
|
| 1446 |
+
"learning_rate": 2.8742249778565107e-05,
|
| 1447 |
+
"loss": 2.1892,
|
| 1448 |
+
"step": 24000
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 4.27,
|
| 1452 |
+
"learning_rate": 2.8653675819309123e-05,
|
| 1453 |
+
"loss": 2.2222,
|
| 1454 |
+
"step": 24100
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 4.29,
|
| 1458 |
+
"learning_rate": 2.8565101860053146e-05,
|
| 1459 |
+
"loss": 2.2133,
|
| 1460 |
+
"step": 24200
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 4.3,
|
| 1464 |
+
"learning_rate": 2.847652790079717e-05,
|
| 1465 |
+
"loss": 2.2059,
|
| 1466 |
+
"step": 24300
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 4.32,
|
| 1470 |
+
"learning_rate": 2.838795394154119e-05,
|
| 1471 |
+
"loss": 2.2117,
|
| 1472 |
+
"step": 24400
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 4.34,
|
| 1476 |
+
"learning_rate": 2.8299379982285208e-05,
|
| 1477 |
+
"loss": 2.1632,
|
| 1478 |
+
"step": 24500
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 4.36,
|
| 1482 |
+
"learning_rate": 2.821080602302923e-05,
|
| 1483 |
+
"loss": 2.2157,
|
| 1484 |
+
"step": 24600
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 4.38,
|
| 1488 |
+
"learning_rate": 2.812223206377325e-05,
|
| 1489 |
+
"loss": 2.2001,
|
| 1490 |
+
"step": 24700
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 4.39,
|
| 1494 |
+
"learning_rate": 2.8033658104517273e-05,
|
| 1495 |
+
"loss": 2.186,
|
| 1496 |
+
"step": 24800
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 4.41,
|
| 1500 |
+
"learning_rate": 2.7945084145261296e-05,
|
| 1501 |
+
"loss": 2.1866,
|
| 1502 |
+
"step": 24900
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 4.43,
|
| 1506 |
+
"learning_rate": 2.7856510186005312e-05,
|
| 1507 |
+
"loss": 2.1672,
|
| 1508 |
+
"step": 25000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 4.45,
|
| 1512 |
+
"learning_rate": 2.7767936226749335e-05,
|
| 1513 |
+
"loss": 2.1915,
|
| 1514 |
+
"step": 25100
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 4.46,
|
| 1518 |
+
"learning_rate": 2.7679362267493358e-05,
|
| 1519 |
+
"loss": 2.1873,
|
| 1520 |
+
"step": 25200
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 4.48,
|
| 1524 |
+
"learning_rate": 2.759078830823738e-05,
|
| 1525 |
+
"loss": 2.1836,
|
| 1526 |
+
"step": 25300
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 4.5,
|
| 1530 |
+
"learning_rate": 2.7502214348981404e-05,
|
| 1531 |
+
"loss": 2.1446,
|
| 1532 |
+
"step": 25400
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 4.52,
|
| 1536 |
+
"learning_rate": 2.741364038972542e-05,
|
| 1537 |
+
"loss": 2.1815,
|
| 1538 |
+
"step": 25500
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 4.53,
|
| 1542 |
+
"learning_rate": 2.7325066430469442e-05,
|
| 1543 |
+
"loss": 2.1676,
|
| 1544 |
+
"step": 25600
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 4.55,
|
| 1548 |
+
"learning_rate": 2.7236492471213465e-05,
|
| 1549 |
+
"loss": 2.1986,
|
| 1550 |
+
"step": 25700
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.57,
|
| 1554 |
+
"learning_rate": 2.7147918511957488e-05,
|
| 1555 |
+
"loss": 2.1716,
|
| 1556 |
+
"step": 25800
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 4.59,
|
| 1560 |
+
"learning_rate": 2.7059344552701504e-05,
|
| 1561 |
+
"loss": 2.1595,
|
| 1562 |
+
"step": 25900
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 4.61,
|
| 1566 |
+
"learning_rate": 2.6970770593445527e-05,
|
| 1567 |
+
"loss": 2.1486,
|
| 1568 |
+
"step": 26000
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 4.62,
|
| 1572 |
+
"learning_rate": 2.688219663418955e-05,
|
| 1573 |
+
"loss": 2.2041,
|
| 1574 |
+
"step": 26100
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 4.64,
|
| 1578 |
+
"learning_rate": 2.6793622674933573e-05,
|
| 1579 |
+
"loss": 2.184,
|
| 1580 |
+
"step": 26200
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 4.66,
|
| 1584 |
+
"learning_rate": 2.6705048715677596e-05,
|
| 1585 |
+
"loss": 2.1567,
|
| 1586 |
+
"step": 26300
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 4.68,
|
| 1590 |
+
"learning_rate": 2.6616474756421612e-05,
|
| 1591 |
+
"loss": 2.1839,
|
| 1592 |
+
"step": 26400
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 4.69,
|
| 1596 |
+
"learning_rate": 2.6527900797165635e-05,
|
| 1597 |
+
"loss": 2.1784,
|
| 1598 |
+
"step": 26500
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 4.71,
|
| 1602 |
+
"learning_rate": 2.6439326837909658e-05,
|
| 1603 |
+
"loss": 2.1683,
|
| 1604 |
+
"step": 26600
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 4.73,
|
| 1608 |
+
"learning_rate": 2.6350752878653677e-05,
|
| 1609 |
+
"loss": 2.197,
|
| 1610 |
+
"step": 26700
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 4.75,
|
| 1614 |
+
"learning_rate": 2.62621789193977e-05,
|
| 1615 |
+
"loss": 2.2058,
|
| 1616 |
+
"step": 26800
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 4.77,
|
| 1620 |
+
"learning_rate": 2.6173604960141716e-05,
|
| 1621 |
+
"loss": 2.1386,
|
| 1622 |
+
"step": 26900
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 4.78,
|
| 1626 |
+
"learning_rate": 2.608503100088574e-05,
|
| 1627 |
+
"loss": 2.1782,
|
| 1628 |
+
"step": 27000
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 4.8,
|
| 1632 |
+
"learning_rate": 2.5996457041629762e-05,
|
| 1633 |
+
"loss": 2.2116,
|
| 1634 |
+
"step": 27100
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 4.82,
|
| 1638 |
+
"learning_rate": 2.5907883082373785e-05,
|
| 1639 |
+
"loss": 2.164,
|
| 1640 |
+
"step": 27200
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 4.84,
|
| 1644 |
+
"learning_rate": 2.58193091231178e-05,
|
| 1645 |
+
"loss": 2.1668,
|
| 1646 |
+
"step": 27300
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 4.85,
|
| 1650 |
+
"learning_rate": 2.5730735163861824e-05,
|
| 1651 |
+
"loss": 2.1444,
|
| 1652 |
+
"step": 27400
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 4.87,
|
| 1656 |
+
"learning_rate": 2.5642161204605846e-05,
|
| 1657 |
+
"loss": 2.1801,
|
| 1658 |
+
"step": 27500
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 4.89,
|
| 1662 |
+
"learning_rate": 2.555358724534987e-05,
|
| 1663 |
+
"loss": 2.1782,
|
| 1664 |
+
"step": 27600
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 4.91,
|
| 1668 |
+
"learning_rate": 2.5465013286093892e-05,
|
| 1669 |
+
"loss": 2.1644,
|
| 1670 |
+
"step": 27700
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 4.92,
|
| 1674 |
+
"learning_rate": 2.5376439326837908e-05,
|
| 1675 |
+
"loss": 2.1015,
|
| 1676 |
+
"step": 27800
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 4.94,
|
| 1680 |
+
"learning_rate": 2.528786536758193e-05,
|
| 1681 |
+
"loss": 2.1643,
|
| 1682 |
+
"step": 27900
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 4.96,
|
| 1686 |
+
"learning_rate": 2.5199291408325954e-05,
|
| 1687 |
+
"loss": 2.1822,
|
| 1688 |
+
"step": 28000
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 4.98,
|
| 1692 |
+
"learning_rate": 2.5110717449069977e-05,
|
| 1693 |
+
"loss": 2.195,
|
| 1694 |
+
"step": 28100
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 5.0,
|
| 1698 |
+
"learning_rate": 2.5022143489814e-05,
|
| 1699 |
+
"loss": 2.1692,
|
| 1700 |
+
"step": 28200
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 5.01,
|
| 1704 |
+
"learning_rate": 2.493356953055802e-05,
|
| 1705 |
+
"loss": 2.1791,
|
| 1706 |
+
"step": 28300
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 5.03,
|
| 1710 |
+
"learning_rate": 2.484499557130204e-05,
|
| 1711 |
+
"loss": 2.1332,
|
| 1712 |
+
"step": 28400
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 5.05,
|
| 1716 |
+
"learning_rate": 2.475642161204606e-05,
|
| 1717 |
+
"loss": 2.1833,
|
| 1718 |
+
"step": 28500
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 5.07,
|
| 1722 |
+
"learning_rate": 2.466784765279008e-05,
|
| 1723 |
+
"loss": 2.1712,
|
| 1724 |
+
"step": 28600
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 5.08,
|
| 1728 |
+
"learning_rate": 2.4579273693534104e-05,
|
| 1729 |
+
"loss": 2.1017,
|
| 1730 |
+
"step": 28700
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 5.1,
|
| 1734 |
+
"learning_rate": 2.4490699734278123e-05,
|
| 1735 |
+
"loss": 2.1789,
|
| 1736 |
+
"step": 28800
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 5.12,
|
| 1740 |
+
"learning_rate": 2.4402125775022143e-05,
|
| 1741 |
+
"loss": 2.1292,
|
| 1742 |
+
"step": 28900
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 5.14,
|
| 1746 |
+
"learning_rate": 2.4313551815766166e-05,
|
| 1747 |
+
"loss": 2.1539,
|
| 1748 |
+
"step": 29000
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 5.15,
|
| 1752 |
+
"learning_rate": 2.4224977856510185e-05,
|
| 1753 |
+
"loss": 2.163,
|
| 1754 |
+
"step": 29100
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 5.17,
|
| 1758 |
+
"learning_rate": 2.4136403897254208e-05,
|
| 1759 |
+
"loss": 2.1536,
|
| 1760 |
+
"step": 29200
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 5.19,
|
| 1764 |
+
"learning_rate": 2.4047829937998228e-05,
|
| 1765 |
+
"loss": 2.1545,
|
| 1766 |
+
"step": 29300
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 5.21,
|
| 1770 |
+
"learning_rate": 2.395925597874225e-05,
|
| 1771 |
+
"loss": 2.1721,
|
| 1772 |
+
"step": 29400
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 5.23,
|
| 1776 |
+
"learning_rate": 2.3870682019486273e-05,
|
| 1777 |
+
"loss": 2.1453,
|
| 1778 |
+
"step": 29500
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 5.24,
|
| 1782 |
+
"learning_rate": 2.3782108060230293e-05,
|
| 1783 |
+
"loss": 2.1464,
|
| 1784 |
+
"step": 29600
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 5.26,
|
| 1788 |
+
"learning_rate": 2.3693534100974316e-05,
|
| 1789 |
+
"loss": 2.1423,
|
| 1790 |
+
"step": 29700
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 5.28,
|
| 1794 |
+
"learning_rate": 2.3604960141718335e-05,
|
| 1795 |
+
"loss": 2.1569,
|
| 1796 |
+
"step": 29800
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 5.3,
|
| 1800 |
+
"learning_rate": 2.3516386182462358e-05,
|
| 1801 |
+
"loss": 2.1246,
|
| 1802 |
+
"step": 29900
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 5.31,
|
| 1806 |
+
"learning_rate": 2.3427812223206377e-05,
|
| 1807 |
+
"loss": 2.1579,
|
| 1808 |
+
"step": 30000
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 5.33,
|
| 1812 |
+
"learning_rate": 2.33392382639504e-05,
|
| 1813 |
+
"loss": 2.1465,
|
| 1814 |
+
"step": 30100
|
| 1815 |
+
},
|
| 1816 |
+
{
|
| 1817 |
+
"epoch": 5.35,
|
| 1818 |
+
"learning_rate": 2.3250664304694423e-05,
|
| 1819 |
+
"loss": 2.1828,
|
| 1820 |
+
"step": 30200
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 5.37,
|
| 1824 |
+
"learning_rate": 2.3162090345438443e-05,
|
| 1825 |
+
"loss": 2.1064,
|
| 1826 |
+
"step": 30300
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 5.39,
|
| 1830 |
+
"learning_rate": 2.3073516386182465e-05,
|
| 1831 |
+
"loss": 2.1668,
|
| 1832 |
+
"step": 30400
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 5.4,
|
| 1836 |
+
"learning_rate": 2.2984942426926485e-05,
|
| 1837 |
+
"loss": 2.1365,
|
| 1838 |
+
"step": 30500
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 5.42,
|
| 1842 |
+
"learning_rate": 2.2896368467670508e-05,
|
| 1843 |
+
"loss": 2.1518,
|
| 1844 |
+
"step": 30600
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 5.44,
|
| 1848 |
+
"learning_rate": 2.2807794508414527e-05,
|
| 1849 |
+
"loss": 2.0826,
|
| 1850 |
+
"step": 30700
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 5.46,
|
| 1854 |
+
"learning_rate": 2.271922054915855e-05,
|
| 1855 |
+
"loss": 2.1492,
|
| 1856 |
+
"step": 30800
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 5.47,
|
| 1860 |
+
"learning_rate": 2.263064658990257e-05,
|
| 1861 |
+
"loss": 2.1666,
|
| 1862 |
+
"step": 30900
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 5.49,
|
| 1866 |
+
"learning_rate": 2.254207263064659e-05,
|
| 1867 |
+
"loss": 2.137,
|
| 1868 |
+
"step": 31000
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 5.51,
|
| 1872 |
+
"learning_rate": 2.2453498671390612e-05,
|
| 1873 |
+
"loss": 2.1304,
|
| 1874 |
+
"step": 31100
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 5.53,
|
| 1878 |
+
"learning_rate": 2.236492471213463e-05,
|
| 1879 |
+
"loss": 2.0874,
|
| 1880 |
+
"step": 31200
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 5.54,
|
| 1884 |
+
"learning_rate": 2.2276350752878654e-05,
|
| 1885 |
+
"loss": 2.1058,
|
| 1886 |
+
"step": 31300
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 5.56,
|
| 1890 |
+
"learning_rate": 2.2187776793622674e-05,
|
| 1891 |
+
"loss": 2.1319,
|
| 1892 |
+
"step": 31400
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 5.58,
|
| 1896 |
+
"learning_rate": 2.2099202834366697e-05,
|
| 1897 |
+
"loss": 2.168,
|
| 1898 |
+
"step": 31500
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 5.6,
|
| 1902 |
+
"learning_rate": 2.201062887511072e-05,
|
| 1903 |
+
"loss": 2.1371,
|
| 1904 |
+
"step": 31600
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 5.62,
|
| 1908 |
+
"learning_rate": 2.192205491585474e-05,
|
| 1909 |
+
"loss": 2.1326,
|
| 1910 |
+
"step": 31700
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 5.63,
|
| 1914 |
+
"learning_rate": 2.1833480956598762e-05,
|
| 1915 |
+
"loss": 2.1217,
|
| 1916 |
+
"step": 31800
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 5.65,
|
| 1920 |
+
"learning_rate": 2.174490699734278e-05,
|
| 1921 |
+
"loss": 2.1494,
|
| 1922 |
+
"step": 31900
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 5.67,
|
| 1926 |
+
"learning_rate": 2.1656333038086804e-05,
|
| 1927 |
+
"loss": 2.1289,
|
| 1928 |
+
"step": 32000
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 5.69,
|
| 1932 |
+
"learning_rate": 2.1567759078830824e-05,
|
| 1933 |
+
"loss": 2.0962,
|
| 1934 |
+
"step": 32100
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 5.7,
|
| 1938 |
+
"learning_rate": 2.1479185119574847e-05,
|
| 1939 |
+
"loss": 2.1188,
|
| 1940 |
+
"step": 32200
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 5.72,
|
| 1944 |
+
"learning_rate": 2.139061116031887e-05,
|
| 1945 |
+
"loss": 2.1163,
|
| 1946 |
+
"step": 32300
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 5.74,
|
| 1950 |
+
"learning_rate": 2.130203720106289e-05,
|
| 1951 |
+
"loss": 2.1361,
|
| 1952 |
+
"step": 32400
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 5.76,
|
| 1956 |
+
"learning_rate": 2.1213463241806912e-05,
|
| 1957 |
+
"loss": 2.0793,
|
| 1958 |
+
"step": 32500
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 5.77,
|
| 1962 |
+
"learning_rate": 2.112488928255093e-05,
|
| 1963 |
+
"loss": 2.0944,
|
| 1964 |
+
"step": 32600
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 5.79,
|
| 1968 |
+
"learning_rate": 2.1036315323294954e-05,
|
| 1969 |
+
"loss": 2.145,
|
| 1970 |
+
"step": 32700
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 5.81,
|
| 1974 |
+
"learning_rate": 2.0947741364038974e-05,
|
| 1975 |
+
"loss": 2.1493,
|
| 1976 |
+
"step": 32800
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 5.83,
|
| 1980 |
+
"learning_rate": 2.0859167404782996e-05,
|
| 1981 |
+
"loss": 2.1309,
|
| 1982 |
+
"step": 32900
|
| 1983 |
+
},
|
| 1984 |
+
{
|
| 1985 |
+
"epoch": 5.85,
|
| 1986 |
+
"learning_rate": 2.0770593445527016e-05,
|
| 1987 |
+
"loss": 2.1422,
|
| 1988 |
+
"step": 33000
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 5.86,
|
| 1992 |
+
"learning_rate": 2.0682019486271035e-05,
|
| 1993 |
+
"loss": 2.1383,
|
| 1994 |
+
"step": 33100
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 5.88,
|
| 1998 |
+
"learning_rate": 2.0593445527015058e-05,
|
| 1999 |
+
"loss": 2.1247,
|
| 2000 |
+
"step": 33200
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 5.9,
|
| 2004 |
+
"learning_rate": 2.0504871567759078e-05,
|
| 2005 |
+
"loss": 2.0899,
|
| 2006 |
+
"step": 33300
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 5.92,
|
| 2010 |
+
"learning_rate": 2.04162976085031e-05,
|
| 2011 |
+
"loss": 2.1189,
|
| 2012 |
+
"step": 33400
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 5.93,
|
| 2016 |
+
"learning_rate": 2.032772364924712e-05,
|
| 2017 |
+
"loss": 2.1178,
|
| 2018 |
+
"step": 33500
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 5.95,
|
| 2022 |
+
"learning_rate": 2.0239149689991143e-05,
|
| 2023 |
+
"loss": 2.1256,
|
| 2024 |
+
"step": 33600
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 5.97,
|
| 2028 |
+
"learning_rate": 2.0150575730735166e-05,
|
| 2029 |
+
"loss": 2.1403,
|
| 2030 |
+
"step": 33700
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 5.99,
|
| 2034 |
+
"learning_rate": 2.0062001771479185e-05,
|
| 2035 |
+
"loss": 2.1421,
|
| 2036 |
+
"step": 33800
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 6.01,
|
| 2040 |
+
"learning_rate": 1.9973427812223208e-05,
|
| 2041 |
+
"loss": 2.1434,
|
| 2042 |
+
"step": 33900
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 6.02,
|
| 2046 |
+
"learning_rate": 1.9884853852967228e-05,
|
| 2047 |
+
"loss": 2.1076,
|
| 2048 |
+
"step": 34000
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 6.04,
|
| 2052 |
+
"learning_rate": 1.979627989371125e-05,
|
| 2053 |
+
"loss": 2.0935,
|
| 2054 |
+
"step": 34100
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 6.06,
|
| 2058 |
+
"learning_rate": 1.970770593445527e-05,
|
| 2059 |
+
"loss": 2.1139,
|
| 2060 |
+
"step": 34200
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 6.08,
|
| 2064 |
+
"learning_rate": 1.9619131975199293e-05,
|
| 2065 |
+
"loss": 2.1342,
|
| 2066 |
+
"step": 34300
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 6.09,
|
| 2070 |
+
"learning_rate": 1.9530558015943316e-05,
|
| 2071 |
+
"loss": 2.1207,
|
| 2072 |
+
"step": 34400
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 6.11,
|
| 2076 |
+
"learning_rate": 1.9441984056687335e-05,
|
| 2077 |
+
"loss": 2.0827,
|
| 2078 |
+
"step": 34500
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 6.13,
|
| 2082 |
+
"learning_rate": 1.9353410097431358e-05,
|
| 2083 |
+
"loss": 2.0926,
|
| 2084 |
+
"step": 34600
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 6.15,
|
| 2088 |
+
"learning_rate": 1.9264836138175378e-05,
|
| 2089 |
+
"loss": 2.1078,
|
| 2090 |
+
"step": 34700
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 6.16,
|
| 2094 |
+
"learning_rate": 1.91762621789194e-05,
|
| 2095 |
+
"loss": 2.0997,
|
| 2096 |
+
"step": 34800
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 6.18,
|
| 2100 |
+
"learning_rate": 1.908768821966342e-05,
|
| 2101 |
+
"loss": 2.1346,
|
| 2102 |
+
"step": 34900
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 6.2,
|
| 2106 |
+
"learning_rate": 1.8999114260407443e-05,
|
| 2107 |
+
"loss": 2.108,
|
| 2108 |
+
"step": 35000
|
| 2109 |
}
|
| 2110 |
],
|
| 2111 |
"max_steps": 56450,
|
| 2112 |
"num_train_epochs": 10,
|
| 2113 |
+
"total_flos": 1.2999215903313946e+18,
|
| 2114 |
"trial_name": null,
|
| 2115 |
"trial_params": null
|
| 2116 |
}
|