Training in progress, step 7000, checkpoint
Browse files- last-checkpoint/global_step7000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step7000/mp_rank_00_model_states.pt +3 -0
- last-checkpoint/latest +1 -1
- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +585 -5
last-checkpoint/global_step7000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1b40c29506ba51102a5e99f143966af19218dad4afdf7f5a54e3c91a9ba8aaa6
|
| 3 |
+
size 5117197489
|
last-checkpoint/global_step7000/mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8d472475f1ab585d49c8a44e0b9d581931a511d04b94f5f0d50d7c89a6e10a7
|
| 3 |
+
size 859127933
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step7000
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 962205216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:856733172381a37b6de12c25512bfa5cc33814241a1986b18ae46a3c6cd69ce1
|
| 3 |
size 962205216
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14709
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56a8ea7c15005d31ade663058f08a1d5a4619da6c77df5179c75f15bb9cc3f05
|
| 3 |
size 14709
|
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 sha256:34a83060f11df4fe46a27d45e8744a4c0e7bb60df156e5d496780133906eacd7
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "./iteboshi_student_model_temp/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1457,6 +1457,586 @@
|
|
| 1457 |
"eval_steps_per_second": 1.999,
|
| 1458 |
"eval_wer": 86.11975483262611,
|
| 1459 |
"step": 5000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1460 |
}
|
| 1461 |
],
|
| 1462 |
"logging_steps": 25,
|
|
@@ -1476,7 +2056,7 @@
|
|
| 1476 |
"attributes": {}
|
| 1477 |
}
|
| 1478 |
},
|
| 1479 |
-
"total_flos":
|
| 1480 |
"train_batch_size": 4,
|
| 1481 |
"trial_name": null,
|
| 1482 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 84.13012729844414,
|
| 3 |
+
"best_model_checkpoint": "./iteboshi_student_model_temp/checkpoint-7000",
|
| 4 |
+
"epoch": 7.709251101321586,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 7000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1457 |
"eval_steps_per_second": 1.999,
|
| 1458 |
"eval_wer": 86.11975483262611,
|
| 1459 |
"step": 5000
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 5.534140969162996,
|
| 1463 |
+
"grad_norm": 0.6520294547080994,
|
| 1464 |
+
"learning_rate": 1.535897435897436e-05,
|
| 1465 |
+
"loss": 0.1639,
|
| 1466 |
+
"step": 5025
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 5.561674008810573,
|
| 1470 |
+
"grad_norm": 0.6221819519996643,
|
| 1471 |
+
"learning_rate": 1.5333333333333334e-05,
|
| 1472 |
+
"loss": 0.1851,
|
| 1473 |
+
"step": 5050
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 5.5892070484581495,
|
| 1477 |
+
"grad_norm": 0.5925490856170654,
|
| 1478 |
+
"learning_rate": 1.5307692307692308e-05,
|
| 1479 |
+
"loss": 0.1733,
|
| 1480 |
+
"step": 5075
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 5.616740088105727,
|
| 1484 |
+
"grad_norm": 1.0411094427108765,
|
| 1485 |
+
"learning_rate": 1.5282051282051282e-05,
|
| 1486 |
+
"loss": 0.1502,
|
| 1487 |
+
"step": 5100
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 5.644273127753304,
|
| 1491 |
+
"grad_norm": 0.8545799255371094,
|
| 1492 |
+
"learning_rate": 1.5256410256410257e-05,
|
| 1493 |
+
"loss": 0.1681,
|
| 1494 |
+
"step": 5125
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 5.671806167400881,
|
| 1498 |
+
"grad_norm": 0.5849359631538391,
|
| 1499 |
+
"learning_rate": 1.523076923076923e-05,
|
| 1500 |
+
"loss": 0.1778,
|
| 1501 |
+
"step": 5150
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 5.6993392070484585,
|
| 1505 |
+
"grad_norm": 0.5703755617141724,
|
| 1506 |
+
"learning_rate": 1.5205128205128206e-05,
|
| 1507 |
+
"loss": 0.1328,
|
| 1508 |
+
"step": 5175
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 5.726872246696035,
|
| 1512 |
+
"grad_norm": 0.7638615965843201,
|
| 1513 |
+
"learning_rate": 1.517948717948718e-05,
|
| 1514 |
+
"loss": 0.1929,
|
| 1515 |
+
"step": 5200
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 5.754405286343612,
|
| 1519 |
+
"grad_norm": 0.5087124109268188,
|
| 1520 |
+
"learning_rate": 1.5153846153846155e-05,
|
| 1521 |
+
"loss": 0.1545,
|
| 1522 |
+
"step": 5225
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 5.78193832599119,
|
| 1526 |
+
"grad_norm": 0.8324174284934998,
|
| 1527 |
+
"learning_rate": 1.5128205128205129e-05,
|
| 1528 |
+
"loss": 0.1975,
|
| 1529 |
+
"step": 5250
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 5.809471365638767,
|
| 1533 |
+
"grad_norm": 0.4413852095603943,
|
| 1534 |
+
"learning_rate": 1.5102564102564104e-05,
|
| 1535 |
+
"loss": 0.1663,
|
| 1536 |
+
"step": 5275
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 5.8370044052863435,
|
| 1540 |
+
"grad_norm": 0.708247184753418,
|
| 1541 |
+
"learning_rate": 1.5076923076923078e-05,
|
| 1542 |
+
"loss": 0.1543,
|
| 1543 |
+
"step": 5300
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 5.864537444933921,
|
| 1547 |
+
"grad_norm": 0.689794659614563,
|
| 1548 |
+
"learning_rate": 1.5051282051282053e-05,
|
| 1549 |
+
"loss": 0.1621,
|
| 1550 |
+
"step": 5325
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 5.892070484581498,
|
| 1554 |
+
"grad_norm": 0.5020695328712463,
|
| 1555 |
+
"learning_rate": 1.5025641025641027e-05,
|
| 1556 |
+
"loss": 0.1774,
|
| 1557 |
+
"step": 5350
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 5.919603524229075,
|
| 1561 |
+
"grad_norm": 0.4771401882171631,
|
| 1562 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 1563 |
+
"loss": 0.1442,
|
| 1564 |
+
"step": 5375
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 5.9471365638766525,
|
| 1568 |
+
"grad_norm": 1.0237714052200317,
|
| 1569 |
+
"learning_rate": 1.4974358974358976e-05,
|
| 1570 |
+
"loss": 0.1629,
|
| 1571 |
+
"step": 5400
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 5.974669603524229,
|
| 1575 |
+
"grad_norm": 0.7134143114089966,
|
| 1576 |
+
"learning_rate": 1.494871794871795e-05,
|
| 1577 |
+
"loss": 0.144,
|
| 1578 |
+
"step": 5425
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 6.002202643171806,
|
| 1582 |
+
"grad_norm": 0.30046069622039795,
|
| 1583 |
+
"learning_rate": 1.4923076923076925e-05,
|
| 1584 |
+
"loss": 0.1373,
|
| 1585 |
+
"step": 5450
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 6.029735682819383,
|
| 1589 |
+
"grad_norm": 0.4057641923427582,
|
| 1590 |
+
"learning_rate": 1.4897435897435898e-05,
|
| 1591 |
+
"loss": 0.1082,
|
| 1592 |
+
"step": 5475
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 6.057268722466961,
|
| 1596 |
+
"grad_norm": 0.486996591091156,
|
| 1597 |
+
"learning_rate": 1.4871794871794874e-05,
|
| 1598 |
+
"loss": 0.0848,
|
| 1599 |
+
"step": 5500
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 6.084801762114537,
|
| 1603 |
+
"grad_norm": 0.46409764885902405,
|
| 1604 |
+
"learning_rate": 1.4846153846153847e-05,
|
| 1605 |
+
"loss": 0.1075,
|
| 1606 |
+
"step": 5525
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 6.112334801762114,
|
| 1610 |
+
"grad_norm": 0.4308403432369232,
|
| 1611 |
+
"learning_rate": 1.4820512820512823e-05,
|
| 1612 |
+
"loss": 0.1193,
|
| 1613 |
+
"step": 5550
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 6.139867841409692,
|
| 1617 |
+
"grad_norm": 0.34751376509666443,
|
| 1618 |
+
"learning_rate": 1.4794871794871796e-05,
|
| 1619 |
+
"loss": 0.1139,
|
| 1620 |
+
"step": 5575
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 6.167400881057269,
|
| 1624 |
+
"grad_norm": 0.8365034461021423,
|
| 1625 |
+
"learning_rate": 1.4769230769230772e-05,
|
| 1626 |
+
"loss": 0.1273,
|
| 1627 |
+
"step": 5600
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 6.1949339207048455,
|
| 1631 |
+
"grad_norm": 0.34338051080703735,
|
| 1632 |
+
"learning_rate": 1.4743589743589745e-05,
|
| 1633 |
+
"loss": 0.0895,
|
| 1634 |
+
"step": 5625
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 6.222466960352423,
|
| 1638 |
+
"grad_norm": 0.6777989864349365,
|
| 1639 |
+
"learning_rate": 1.471794871794872e-05,
|
| 1640 |
+
"loss": 0.121,
|
| 1641 |
+
"step": 5650
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 6.25,
|
| 1645 |
+
"grad_norm": 0.5982616543769836,
|
| 1646 |
+
"learning_rate": 1.4692307692307694e-05,
|
| 1647 |
+
"loss": 0.1214,
|
| 1648 |
+
"step": 5675
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 6.277533039647577,
|
| 1652 |
+
"grad_norm": 0.5918659567832947,
|
| 1653 |
+
"learning_rate": 1.4666666666666666e-05,
|
| 1654 |
+
"loss": 0.0863,
|
| 1655 |
+
"step": 5700
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 6.3050660792951545,
|
| 1659 |
+
"grad_norm": 0.35085636377334595,
|
| 1660 |
+
"learning_rate": 1.4641025641025642e-05,
|
| 1661 |
+
"loss": 0.11,
|
| 1662 |
+
"step": 5725
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 6.332599118942731,
|
| 1666 |
+
"grad_norm": 0.43525975942611694,
|
| 1667 |
+
"learning_rate": 1.4615384615384615e-05,
|
| 1668 |
+
"loss": 0.0945,
|
| 1669 |
+
"step": 5750
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 6.360132158590308,
|
| 1673 |
+
"grad_norm": 0.3799566328525543,
|
| 1674 |
+
"learning_rate": 1.458974358974359e-05,
|
| 1675 |
+
"loss": 0.0984,
|
| 1676 |
+
"step": 5775
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 6.387665198237886,
|
| 1680 |
+
"grad_norm": 0.7915482521057129,
|
| 1681 |
+
"learning_rate": 1.4564102564102564e-05,
|
| 1682 |
+
"loss": 0.1154,
|
| 1683 |
+
"step": 5800
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 6.415198237885463,
|
| 1687 |
+
"grad_norm": 0.47404220700263977,
|
| 1688 |
+
"learning_rate": 1.453846153846154e-05,
|
| 1689 |
+
"loss": 0.0984,
|
| 1690 |
+
"step": 5825
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 6.442731277533039,
|
| 1694 |
+
"grad_norm": 0.3866804838180542,
|
| 1695 |
+
"learning_rate": 1.4512820512820513e-05,
|
| 1696 |
+
"loss": 0.1273,
|
| 1697 |
+
"step": 5850
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 6.470264317180617,
|
| 1701 |
+
"grad_norm": 0.4380825459957123,
|
| 1702 |
+
"learning_rate": 1.4487179487179489e-05,
|
| 1703 |
+
"loss": 0.1332,
|
| 1704 |
+
"step": 5875
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 6.497797356828194,
|
| 1708 |
+
"grad_norm": 0.4749581515789032,
|
| 1709 |
+
"learning_rate": 1.4461538461538462e-05,
|
| 1710 |
+
"loss": 0.1243,
|
| 1711 |
+
"step": 5900
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 6.525330396475771,
|
| 1715 |
+
"grad_norm": 0.820015549659729,
|
| 1716 |
+
"learning_rate": 1.4435897435897438e-05,
|
| 1717 |
+
"loss": 0.1305,
|
| 1718 |
+
"step": 5925
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 6.5528634361233475,
|
| 1722 |
+
"grad_norm": 0.5644270181655884,
|
| 1723 |
+
"learning_rate": 1.4410256410256411e-05,
|
| 1724 |
+
"loss": 0.1167,
|
| 1725 |
+
"step": 5950
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 6.580396475770925,
|
| 1729 |
+
"grad_norm": 0.4454534649848938,
|
| 1730 |
+
"learning_rate": 1.4384615384615387e-05,
|
| 1731 |
+
"loss": 0.1101,
|
| 1732 |
+
"step": 5975
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 6.607929515418502,
|
| 1736 |
+
"grad_norm": 0.5850095748901367,
|
| 1737 |
+
"learning_rate": 1.435897435897436e-05,
|
| 1738 |
+
"loss": 0.104,
|
| 1739 |
+
"step": 6000
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 6.607929515418502,
|
| 1743 |
+
"eval_cer": 24.67816078777527,
|
| 1744 |
+
"eval_loss": 0.7848840355873108,
|
| 1745 |
+
"eval_runtime": 1300.262,
|
| 1746 |
+
"eval_samples_per_second": 8.138,
|
| 1747 |
+
"eval_steps_per_second": 2.035,
|
| 1748 |
+
"eval_wer": 84.6016030174446,
|
| 1749 |
+
"step": 6000
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 6.635462555066079,
|
| 1753 |
+
"grad_norm": 0.6678868532180786,
|
| 1754 |
+
"learning_rate": 1.4333333333333334e-05,
|
| 1755 |
+
"loss": 0.0875,
|
| 1756 |
+
"step": 6025
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 6.6629955947136565,
|
| 1760 |
+
"grad_norm": 0.279801607131958,
|
| 1761 |
+
"learning_rate": 1.430769230769231e-05,
|
| 1762 |
+
"loss": 0.1333,
|
| 1763 |
+
"step": 6050
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 6.690528634361233,
|
| 1767 |
+
"grad_norm": 0.8138979077339172,
|
| 1768 |
+
"learning_rate": 1.4282051282051283e-05,
|
| 1769 |
+
"loss": 0.1458,
|
| 1770 |
+
"step": 6075
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 6.71806167400881,
|
| 1774 |
+
"grad_norm": 0.6547926068305969,
|
| 1775 |
+
"learning_rate": 1.4256410256410258e-05,
|
| 1776 |
+
"loss": 0.1421,
|
| 1777 |
+
"step": 6100
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 6.745594713656388,
|
| 1781 |
+
"grad_norm": 0.6097781658172607,
|
| 1782 |
+
"learning_rate": 1.4230769230769232e-05,
|
| 1783 |
+
"loss": 0.1285,
|
| 1784 |
+
"step": 6125
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 6.773127753303965,
|
| 1788 |
+
"grad_norm": 0.4184475839138031,
|
| 1789 |
+
"learning_rate": 1.4205128205128207e-05,
|
| 1790 |
+
"loss": 0.1139,
|
| 1791 |
+
"step": 6150
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 6.8006607929515415,
|
| 1795 |
+
"grad_norm": 0.4548538029193878,
|
| 1796 |
+
"learning_rate": 1.4179487179487181e-05,
|
| 1797 |
+
"loss": 0.1214,
|
| 1798 |
+
"step": 6175
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 6.828193832599119,
|
| 1802 |
+
"grad_norm": 0.3974076509475708,
|
| 1803 |
+
"learning_rate": 1.4153846153846156e-05,
|
| 1804 |
+
"loss": 0.1051,
|
| 1805 |
+
"step": 6200
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 6.855726872246696,
|
| 1809 |
+
"grad_norm": 0.7746002078056335,
|
| 1810 |
+
"learning_rate": 1.412820512820513e-05,
|
| 1811 |
+
"loss": 0.1349,
|
| 1812 |
+
"step": 6225
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 6.883259911894273,
|
| 1816 |
+
"grad_norm": 0.3809688687324524,
|
| 1817 |
+
"learning_rate": 1.4102564102564105e-05,
|
| 1818 |
+
"loss": 0.11,
|
| 1819 |
+
"step": 6250
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 6.9107929515418505,
|
| 1823 |
+
"grad_norm": 0.39594364166259766,
|
| 1824 |
+
"learning_rate": 1.4076923076923079e-05,
|
| 1825 |
+
"loss": 0.1006,
|
| 1826 |
+
"step": 6275
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 6.938325991189427,
|
| 1830 |
+
"grad_norm": 0.5483039617538452,
|
| 1831 |
+
"learning_rate": 1.405128205128205e-05,
|
| 1832 |
+
"loss": 0.1122,
|
| 1833 |
+
"step": 6300
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 6.965859030837004,
|
| 1837 |
+
"grad_norm": 0.6642032861709595,
|
| 1838 |
+
"learning_rate": 1.4025641025641026e-05,
|
| 1839 |
+
"loss": 0.1232,
|
| 1840 |
+
"step": 6325
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 6.993392070484582,
|
| 1844 |
+
"grad_norm": 0.42328912019729614,
|
| 1845 |
+
"learning_rate": 1.4e-05,
|
| 1846 |
+
"loss": 0.1125,
|
| 1847 |
+
"step": 6350
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 7.020925110132159,
|
| 1851 |
+
"grad_norm": 0.6559634804725647,
|
| 1852 |
+
"learning_rate": 1.3974358974358975e-05,
|
| 1853 |
+
"loss": 0.0831,
|
| 1854 |
+
"step": 6375
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 7.048458149779735,
|
| 1858 |
+
"grad_norm": 0.5906537175178528,
|
| 1859 |
+
"learning_rate": 1.3948717948717949e-05,
|
| 1860 |
+
"loss": 0.0752,
|
| 1861 |
+
"step": 6400
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 7.075991189427313,
|
| 1865 |
+
"grad_norm": 0.5048888921737671,
|
| 1866 |
+
"learning_rate": 1.3923076923076924e-05,
|
| 1867 |
+
"loss": 0.0702,
|
| 1868 |
+
"step": 6425
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 7.10352422907489,
|
| 1872 |
+
"grad_norm": 0.5171650648117065,
|
| 1873 |
+
"learning_rate": 1.3897435897435898e-05,
|
| 1874 |
+
"loss": 0.0622,
|
| 1875 |
+
"step": 6450
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 7.131057268722467,
|
| 1879 |
+
"grad_norm": 0.608253538608551,
|
| 1880 |
+
"learning_rate": 1.3871794871794873e-05,
|
| 1881 |
+
"loss": 0.0795,
|
| 1882 |
+
"step": 6475
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 7.158590308370044,
|
| 1886 |
+
"grad_norm": 0.3569038212299347,
|
| 1887 |
+
"learning_rate": 1.3846153846153847e-05,
|
| 1888 |
+
"loss": 0.0862,
|
| 1889 |
+
"step": 6500
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 7.186123348017621,
|
| 1893 |
+
"grad_norm": 0.4823140501976013,
|
| 1894 |
+
"learning_rate": 1.3820512820512822e-05,
|
| 1895 |
+
"loss": 0.0749,
|
| 1896 |
+
"step": 6525
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 7.213656387665198,
|
| 1900 |
+
"grad_norm": 0.6069055199623108,
|
| 1901 |
+
"learning_rate": 1.3794871794871796e-05,
|
| 1902 |
+
"loss": 0.0732,
|
| 1903 |
+
"step": 6550
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 7.241189427312776,
|
| 1907 |
+
"grad_norm": 0.3300100564956665,
|
| 1908 |
+
"learning_rate": 1.3769230769230771e-05,
|
| 1909 |
+
"loss": 0.0831,
|
| 1910 |
+
"step": 6575
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 7.2687224669603525,
|
| 1914 |
+
"grad_norm": 0.5404714941978455,
|
| 1915 |
+
"learning_rate": 1.3743589743589745e-05,
|
| 1916 |
+
"loss": 0.0783,
|
| 1917 |
+
"step": 6600
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 7.296255506607929,
|
| 1921 |
+
"grad_norm": 0.6272768974304199,
|
| 1922 |
+
"learning_rate": 1.3717948717948718e-05,
|
| 1923 |
+
"loss": 0.0878,
|
| 1924 |
+
"step": 6625
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 7.323788546255507,
|
| 1928 |
+
"grad_norm": 0.4452053904533386,
|
| 1929 |
+
"learning_rate": 1.3692307692307694e-05,
|
| 1930 |
+
"loss": 0.0808,
|
| 1931 |
+
"step": 6650
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 7.351321585903084,
|
| 1935 |
+
"grad_norm": 0.3930460810661316,
|
| 1936 |
+
"learning_rate": 1.3666666666666667e-05,
|
| 1937 |
+
"loss": 0.0861,
|
| 1938 |
+
"step": 6675
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 7.378854625550661,
|
| 1942 |
+
"grad_norm": 0.31089282035827637,
|
| 1943 |
+
"learning_rate": 1.3641025641025643e-05,
|
| 1944 |
+
"loss": 0.0642,
|
| 1945 |
+
"step": 6700
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 7.406387665198238,
|
| 1949 |
+
"grad_norm": 0.219461590051651,
|
| 1950 |
+
"learning_rate": 1.3615384615384616e-05,
|
| 1951 |
+
"loss": 0.0589,
|
| 1952 |
+
"step": 6725
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 7.433920704845815,
|
| 1956 |
+
"grad_norm": 0.435345321893692,
|
| 1957 |
+
"learning_rate": 1.3589743589743592e-05,
|
| 1958 |
+
"loss": 0.076,
|
| 1959 |
+
"step": 6750
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 7.461453744493392,
|
| 1963 |
+
"grad_norm": 0.5823142528533936,
|
| 1964 |
+
"learning_rate": 1.3564102564102565e-05,
|
| 1965 |
+
"loss": 0.0884,
|
| 1966 |
+
"step": 6775
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 7.48898678414097,
|
| 1970 |
+
"grad_norm": 0.6687197685241699,
|
| 1971 |
+
"learning_rate": 1.353846153846154e-05,
|
| 1972 |
+
"loss": 0.0831,
|
| 1973 |
+
"step": 6800
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 7.516519823788546,
|
| 1977 |
+
"grad_norm": 0.25778886675834656,
|
| 1978 |
+
"learning_rate": 1.3512820512820514e-05,
|
| 1979 |
+
"loss": 0.0794,
|
| 1980 |
+
"step": 6825
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 7.544052863436123,
|
| 1984 |
+
"grad_norm": 0.5225228071212769,
|
| 1985 |
+
"learning_rate": 1.348717948717949e-05,
|
| 1986 |
+
"loss": 0.0772,
|
| 1987 |
+
"step": 6850
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 7.5715859030837,
|
| 1991 |
+
"grad_norm": 0.4801703095436096,
|
| 1992 |
+
"learning_rate": 1.3461538461538463e-05,
|
| 1993 |
+
"loss": 0.0717,
|
| 1994 |
+
"step": 6875
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 7.599118942731278,
|
| 1998 |
+
"grad_norm": 0.48749440908432007,
|
| 1999 |
+
"learning_rate": 1.3435897435897435e-05,
|
| 2000 |
+
"loss": 0.0952,
|
| 2001 |
+
"step": 6900
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 7.6266519823788546,
|
| 2005 |
+
"grad_norm": 0.34919875860214233,
|
| 2006 |
+
"learning_rate": 1.341025641025641e-05,
|
| 2007 |
+
"loss": 0.0755,
|
| 2008 |
+
"step": 6925
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 7.654185022026431,
|
| 2012 |
+
"grad_norm": 0.6202211976051331,
|
| 2013 |
+
"learning_rate": 1.3384615384615384e-05,
|
| 2014 |
+
"loss": 0.095,
|
| 2015 |
+
"step": 6950
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 7.681718061674009,
|
| 2019 |
+
"grad_norm": 0.3826860189437866,
|
| 2020 |
+
"learning_rate": 1.335897435897436e-05,
|
| 2021 |
+
"loss": 0.095,
|
| 2022 |
+
"step": 6975
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 7.709251101321586,
|
| 2026 |
+
"grad_norm": 0.6930757761001587,
|
| 2027 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 2028 |
+
"loss": 0.0641,
|
| 2029 |
+
"step": 7000
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 7.709251101321586,
|
| 2033 |
+
"eval_cer": 38.884367154631846,
|
| 2034 |
+
"eval_loss": 0.8057100772857666,
|
| 2035 |
+
"eval_runtime": 1352.106,
|
| 2036 |
+
"eval_samples_per_second": 7.826,
|
| 2037 |
+
"eval_steps_per_second": 1.957,
|
| 2038 |
+
"eval_wer": 84.13012729844414,
|
| 2039 |
+
"step": 7000
|
| 2040 |
}
|
| 2041 |
],
|
| 2042 |
"logging_steps": 25,
|
|
|
|
| 2056 |
"attributes": {}
|
| 2057 |
}
|
| 2058 |
},
|
| 2059 |
+
"total_flos": 1.2027002305430684e+20,
|
| 2060 |
"train_batch_size": 4,
|
| 2061 |
"trial_name": null,
|
| 2062 |
"trial_params": null
|