Training in progress, step 300, checkpoint
Browse files
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2548d336e519eef3ca9af065a415a11b2b53d7a9b7920f4f8bc81c95c166b4df
|
| 3 |
+
size 85723732
|
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 sha256:724bedb14fc7c8992624dc14eb408852ecffce5afb2fadf4db7bb21f5d042d98
|
| 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 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1906f5f6c2f6ce8e763bf9088760948b97bf5816f1a2007492cdb224a2b43706
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": NaN,
|
| 3 |
"best_model_checkpoint": "miner_id_24/checkpoint-100",
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 100,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1431,6 +1431,714 @@
|
|
| 1431 |
"eval_samples_per_second": 6.892,
|
| 1432 |
"eval_steps_per_second": 1.724,
|
| 1433 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1434 |
}
|
| 1435 |
],
|
| 1436 |
"logging_steps": 1,
|
|
@@ -1445,7 +2153,7 @@
|
|
| 1445 |
"early_stopping_threshold": 0.0
|
| 1446 |
},
|
| 1447 |
"attributes": {
|
| 1448 |
-
"early_stopping_patience_counter":
|
| 1449 |
}
|
| 1450 |
},
|
| 1451 |
"TrainerControl": {
|
|
@@ -1454,12 +2162,12 @@
|
|
| 1454 |
"should_evaluate": false,
|
| 1455 |
"should_log": false,
|
| 1456 |
"should_save": true,
|
| 1457 |
-
"should_training_stop":
|
| 1458 |
},
|
| 1459 |
"attributes": {}
|
| 1460 |
}
|
| 1461 |
},
|
| 1462 |
-
"total_flos":
|
| 1463 |
"train_batch_size": 4,
|
| 1464 |
"trial_name": null,
|
| 1465 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": NaN,
|
| 3 |
"best_model_checkpoint": "miner_id_24/checkpoint-100",
|
| 4 |
+
"epoch": 0.1639456247011408,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 300,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1431 |
"eval_samples_per_second": 6.892,
|
| 1432 |
"eval_steps_per_second": 1.724,
|
| 1433 |
"step": 200
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.10984356854976433,
|
| 1437 |
+
"grad_norm": NaN,
|
| 1438 |
+
"learning_rate": 0.00015696624124628494,
|
| 1439 |
+
"loss": 0.0,
|
| 1440 |
+
"step": 201
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.1103900539654348,
|
| 1444 |
+
"grad_norm": NaN,
|
| 1445 |
+
"learning_rate": 0.00015655040251142784,
|
| 1446 |
+
"loss": 0.0,
|
| 1447 |
+
"step": 202
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 0.11093653938110527,
|
| 1451 |
+
"grad_norm": NaN,
|
| 1452 |
+
"learning_rate": 0.00015613312115057696,
|
| 1453 |
+
"loss": 0.0,
|
| 1454 |
+
"step": 203
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 0.11148302479677574,
|
| 1458 |
+
"grad_norm": NaN,
|
| 1459 |
+
"learning_rate": 0.00015571440780876586,
|
| 1460 |
+
"loss": 0.0,
|
| 1461 |
+
"step": 204
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 0.11202951021244621,
|
| 1465 |
+
"grad_norm": NaN,
|
| 1466 |
+
"learning_rate": 0.00015529427316755874,
|
| 1467 |
+
"loss": 0.0,
|
| 1468 |
+
"step": 205
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 0.11257599562811667,
|
| 1472 |
+
"grad_norm": NaN,
|
| 1473 |
+
"learning_rate": 0.0001548727279447777,
|
| 1474 |
+
"loss": 0.0,
|
| 1475 |
+
"step": 206
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 0.11312248104378714,
|
| 1479 |
+
"grad_norm": NaN,
|
| 1480 |
+
"learning_rate": 0.00015444978289422937,
|
| 1481 |
+
"loss": 0.0,
|
| 1482 |
+
"step": 207
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.11366896645945761,
|
| 1486 |
+
"grad_norm": NaN,
|
| 1487 |
+
"learning_rate": 0.0001540254488054307,
|
| 1488 |
+
"loss": 0.0,
|
| 1489 |
+
"step": 208
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 0.11421545187512808,
|
| 1493 |
+
"grad_norm": NaN,
|
| 1494 |
+
"learning_rate": 0.00015359973650333353,
|
| 1495 |
+
"loss": 0.0,
|
| 1496 |
+
"step": 209
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 0.11476193729079855,
|
| 1500 |
+
"grad_norm": NaN,
|
| 1501 |
+
"learning_rate": 0.00015317265684804864,
|
| 1502 |
+
"loss": 0.0,
|
| 1503 |
+
"step": 210
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 0.11530842270646902,
|
| 1507 |
+
"grad_norm": NaN,
|
| 1508 |
+
"learning_rate": 0.00015274422073456852,
|
| 1509 |
+
"loss": 0.0,
|
| 1510 |
+
"step": 211
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 0.11585490812213949,
|
| 1514 |
+
"grad_norm": NaN,
|
| 1515 |
+
"learning_rate": 0.00015231443909248955,
|
| 1516 |
+
"loss": 0.0,
|
| 1517 |
+
"step": 212
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.11640139353780996,
|
| 1521 |
+
"grad_norm": NaN,
|
| 1522 |
+
"learning_rate": 0.00015188332288573313,
|
| 1523 |
+
"loss": 0.0,
|
| 1524 |
+
"step": 213
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 0.11694787895348042,
|
| 1528 |
+
"grad_norm": NaN,
|
| 1529 |
+
"learning_rate": 0.00015145088311226598,
|
| 1530 |
+
"loss": 0.0,
|
| 1531 |
+
"step": 214
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 0.1174943643691509,
|
| 1535 |
+
"grad_norm": NaN,
|
| 1536 |
+
"learning_rate": 0.0001510171308038197,
|
| 1537 |
+
"loss": 0.0,
|
| 1538 |
+
"step": 215
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.11804084978482136,
|
| 1542 |
+
"grad_norm": NaN,
|
| 1543 |
+
"learning_rate": 0.00015058207702560908,
|
| 1544 |
+
"loss": 0.0,
|
| 1545 |
+
"step": 216
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.11858733520049183,
|
| 1549 |
+
"grad_norm": NaN,
|
| 1550 |
+
"learning_rate": 0.0001501457328760501,
|
| 1551 |
+
"loss": 0.0,
|
| 1552 |
+
"step": 217
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 0.1191338206161623,
|
| 1556 |
+
"grad_norm": NaN,
|
| 1557 |
+
"learning_rate": 0.00014970810948647663,
|
| 1558 |
+
"loss": 0.0,
|
| 1559 |
+
"step": 218
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 0.11968030603183277,
|
| 1563 |
+
"grad_norm": NaN,
|
| 1564 |
+
"learning_rate": 0.00014926921802085663,
|
| 1565 |
+
"loss": 0.0,
|
| 1566 |
+
"step": 219
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 0.12022679144750324,
|
| 1570 |
+
"grad_norm": NaN,
|
| 1571 |
+
"learning_rate": 0.00014882906967550707,
|
| 1572 |
+
"loss": 0.0,
|
| 1573 |
+
"step": 220
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 0.12077327686317371,
|
| 1577 |
+
"grad_norm": NaN,
|
| 1578 |
+
"learning_rate": 0.00014838767567880865,
|
| 1579 |
+
"loss": 0.0,
|
| 1580 |
+
"step": 221
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.12131976227884418,
|
| 1584 |
+
"grad_norm": NaN,
|
| 1585 |
+
"learning_rate": 0.00014794504729091909,
|
| 1586 |
+
"loss": 0.0,
|
| 1587 |
+
"step": 222
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.12186624769451465,
|
| 1591 |
+
"grad_norm": NaN,
|
| 1592 |
+
"learning_rate": 0.000147501195803486,
|
| 1593 |
+
"loss": 0.0,
|
| 1594 |
+
"step": 223
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.12241273311018512,
|
| 1598 |
+
"grad_norm": NaN,
|
| 1599 |
+
"learning_rate": 0.00014705613253935887,
|
| 1600 |
+
"loss": 0.0,
|
| 1601 |
+
"step": 224
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.12295921852585559,
|
| 1605 |
+
"grad_norm": NaN,
|
| 1606 |
+
"learning_rate": 0.00014660986885230002,
|
| 1607 |
+
"loss": 0.0,
|
| 1608 |
+
"step": 225
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.12350570394152606,
|
| 1612 |
+
"grad_norm": NaN,
|
| 1613 |
+
"learning_rate": 0.00014616241612669522,
|
| 1614 |
+
"loss": 0.0,
|
| 1615 |
+
"step": 226
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.12405218935719653,
|
| 1619 |
+
"grad_norm": NaN,
|
| 1620 |
+
"learning_rate": 0.00014571378577726317,
|
| 1621 |
+
"loss": 0.0,
|
| 1622 |
+
"step": 227
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.124598674772867,
|
| 1626 |
+
"grad_norm": NaN,
|
| 1627 |
+
"learning_rate": 0.00014526398924876406,
|
| 1628 |
+
"loss": 0.0,
|
| 1629 |
+
"step": 228
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.12514516018853747,
|
| 1633 |
+
"grad_norm": NaN,
|
| 1634 |
+
"learning_rate": 0.00014481303801570805,
|
| 1635 |
+
"loss": 0.0,
|
| 1636 |
+
"step": 229
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 0.12569164560420792,
|
| 1640 |
+
"grad_norm": NaN,
|
| 1641 |
+
"learning_rate": 0.00014436094358206224,
|
| 1642 |
+
"loss": 0.0,
|
| 1643 |
+
"step": 230
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.1262381310198784,
|
| 1647 |
+
"grad_norm": NaN,
|
| 1648 |
+
"learning_rate": 0.00014390771748095735,
|
| 1649 |
+
"loss": 0.0,
|
| 1650 |
+
"step": 231
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 0.12678461643554886,
|
| 1654 |
+
"grad_norm": NaN,
|
| 1655 |
+
"learning_rate": 0.00014345337127439335,
|
| 1656 |
+
"loss": 0.0,
|
| 1657 |
+
"step": 232
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.12733110185121935,
|
| 1661 |
+
"grad_norm": NaN,
|
| 1662 |
+
"learning_rate": 0.0001429979165529446,
|
| 1663 |
+
"loss": 0.0,
|
| 1664 |
+
"step": 233
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.1278775872668898,
|
| 1668 |
+
"grad_norm": NaN,
|
| 1669 |
+
"learning_rate": 0.0001425413649354643,
|
| 1670 |
+
"loss": 0.0,
|
| 1671 |
+
"step": 234
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 0.1284240726825603,
|
| 1675 |
+
"grad_norm": NaN,
|
| 1676 |
+
"learning_rate": 0.00014208372806878784,
|
| 1677 |
+
"loss": 0.0,
|
| 1678 |
+
"step": 235
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 0.12897055809823074,
|
| 1682 |
+
"grad_norm": NaN,
|
| 1683 |
+
"learning_rate": 0.0001416250176274358,
|
| 1684 |
+
"loss": 0.0,
|
| 1685 |
+
"step": 236
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 0.12951704351390123,
|
| 1689 |
+
"grad_norm": NaN,
|
| 1690 |
+
"learning_rate": 0.00014116524531331616,
|
| 1691 |
+
"loss": 0.0,
|
| 1692 |
+
"step": 237
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 0.13006352892957168,
|
| 1696 |
+
"grad_norm": NaN,
|
| 1697 |
+
"learning_rate": 0.00014070442285542576,
|
| 1698 |
+
"loss": 0.0,
|
| 1699 |
+
"step": 238
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 0.13061001434524216,
|
| 1703 |
+
"grad_norm": NaN,
|
| 1704 |
+
"learning_rate": 0.00014024256200955107,
|
| 1705 |
+
"loss": 0.0,
|
| 1706 |
+
"step": 239
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 0.13115649976091262,
|
| 1710 |
+
"grad_norm": NaN,
|
| 1711 |
+
"learning_rate": 0.00013977967455796827,
|
| 1712 |
+
"loss": 0.0,
|
| 1713 |
+
"step": 240
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.1317029851765831,
|
| 1717 |
+
"grad_norm": NaN,
|
| 1718 |
+
"learning_rate": 0.0001393157723091428,
|
| 1719 |
+
"loss": 0.0,
|
| 1720 |
+
"step": 241
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 0.13224947059225356,
|
| 1724 |
+
"grad_norm": NaN,
|
| 1725 |
+
"learning_rate": 0.00013885086709742786,
|
| 1726 |
+
"loss": 0.0,
|
| 1727 |
+
"step": 242
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.13279595600792404,
|
| 1731 |
+
"grad_norm": NaN,
|
| 1732 |
+
"learning_rate": 0.00013838497078276287,
|
| 1733 |
+
"loss": 0.0,
|
| 1734 |
+
"step": 243
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 0.1333424414235945,
|
| 1738 |
+
"grad_norm": NaN,
|
| 1739 |
+
"learning_rate": 0.00013791809525037056,
|
| 1740 |
+
"loss": 0.0,
|
| 1741 |
+
"step": 244
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 0.13388892683926498,
|
| 1745 |
+
"grad_norm": NaN,
|
| 1746 |
+
"learning_rate": 0.00013745025241045412,
|
| 1747 |
+
"loss": 0.0,
|
| 1748 |
+
"step": 245
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 0.13443541225493544,
|
| 1752 |
+
"grad_norm": NaN,
|
| 1753 |
+
"learning_rate": 0.00013698145419789302,
|
| 1754 |
+
"loss": 0.0,
|
| 1755 |
+
"step": 246
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 0.13498189767060592,
|
| 1759 |
+
"grad_norm": NaN,
|
| 1760 |
+
"learning_rate": 0.0001365117125719388,
|
| 1761 |
+
"loss": 0.0,
|
| 1762 |
+
"step": 247
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 0.13552838308627638,
|
| 1766 |
+
"grad_norm": NaN,
|
| 1767 |
+
"learning_rate": 0.0001360410395159099,
|
| 1768 |
+
"loss": 0.0,
|
| 1769 |
+
"step": 248
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.13607486850194686,
|
| 1773 |
+
"grad_norm": NaN,
|
| 1774 |
+
"learning_rate": 0.00013556944703688593,
|
| 1775 |
+
"loss": 0.0,
|
| 1776 |
+
"step": 249
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 0.13662135391761732,
|
| 1780 |
+
"grad_norm": NaN,
|
| 1781 |
+
"learning_rate": 0.00013509694716540134,
|
| 1782 |
+
"loss": 0.0,
|
| 1783 |
+
"step": 250
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.1371678393332878,
|
| 1787 |
+
"grad_norm": NaN,
|
| 1788 |
+
"learning_rate": 0.00013462355195513866,
|
| 1789 |
+
"loss": 0.0,
|
| 1790 |
+
"step": 251
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 0.13771432474895826,
|
| 1794 |
+
"grad_norm": NaN,
|
| 1795 |
+
"learning_rate": 0.0001341492734826209,
|
| 1796 |
+
"loss": 0.0,
|
| 1797 |
+
"step": 252
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 0.13826081016462874,
|
| 1801 |
+
"grad_norm": NaN,
|
| 1802 |
+
"learning_rate": 0.00013367412384690346,
|
| 1803 |
+
"loss": 0.0,
|
| 1804 |
+
"step": 253
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.1388072955802992,
|
| 1808 |
+
"grad_norm": NaN,
|
| 1809 |
+
"learning_rate": 0.0001331981151692654,
|
| 1810 |
+
"loss": 0.0,
|
| 1811 |
+
"step": 254
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.13935378099596968,
|
| 1815 |
+
"grad_norm": NaN,
|
| 1816 |
+
"learning_rate": 0.0001327212595929006,
|
| 1817 |
+
"loss": 0.0,
|
| 1818 |
+
"step": 255
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.13990026641164013,
|
| 1822 |
+
"grad_norm": NaN,
|
| 1823 |
+
"learning_rate": 0.00013224356928260734,
|
| 1824 |
+
"loss": 0.0,
|
| 1825 |
+
"step": 256
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.14044675182731062,
|
| 1829 |
+
"grad_norm": NaN,
|
| 1830 |
+
"learning_rate": 0.0001317650564244787,
|
| 1831 |
+
"loss": 0.0,
|
| 1832 |
+
"step": 257
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.14099323724298107,
|
| 1836 |
+
"grad_norm": NaN,
|
| 1837 |
+
"learning_rate": 0.00013128573322559098,
|
| 1838 |
+
"loss": 0.0,
|
| 1839 |
+
"step": 258
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.14153972265865156,
|
| 1843 |
+
"grad_norm": NaN,
|
| 1844 |
+
"learning_rate": 0.00013080561191369284,
|
| 1845 |
+
"loss": 0.0,
|
| 1846 |
+
"step": 259
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.142086208074322,
|
| 1850 |
+
"grad_norm": NaN,
|
| 1851 |
+
"learning_rate": 0.0001303247047368932,
|
| 1852 |
+
"loss": 0.0,
|
| 1853 |
+
"step": 260
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.1426326934899925,
|
| 1857 |
+
"grad_norm": NaN,
|
| 1858 |
+
"learning_rate": 0.0001298430239633486,
|
| 1859 |
+
"loss": 0.0,
|
| 1860 |
+
"step": 261
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.14317917890566295,
|
| 1864 |
+
"grad_norm": NaN,
|
| 1865 |
+
"learning_rate": 0.00012936058188095043,
|
| 1866 |
+
"loss": 0.0,
|
| 1867 |
+
"step": 262
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.14372566432133344,
|
| 1871 |
+
"grad_norm": NaN,
|
| 1872 |
+
"learning_rate": 0.00012887739079701146,
|
| 1873 |
+
"loss": 0.0,
|
| 1874 |
+
"step": 263
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.1442721497370039,
|
| 1878 |
+
"grad_norm": NaN,
|
| 1879 |
+
"learning_rate": 0.00012839346303795173,
|
| 1880 |
+
"loss": 0.0,
|
| 1881 |
+
"step": 264
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.14481863515267437,
|
| 1885 |
+
"grad_norm": NaN,
|
| 1886 |
+
"learning_rate": 0.00012790881094898427,
|
| 1887 |
+
"loss": 0.0,
|
| 1888 |
+
"step": 265
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.14536512056834483,
|
| 1892 |
+
"grad_norm": NaN,
|
| 1893 |
+
"learning_rate": 0.00012742344689380009,
|
| 1894 |
+
"loss": 0.0,
|
| 1895 |
+
"step": 266
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 0.1459116059840153,
|
| 1899 |
+
"grad_norm": NaN,
|
| 1900 |
+
"learning_rate": 0.0001269373832542527,
|
| 1901 |
+
"loss": 0.0,
|
| 1902 |
+
"step": 267
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 0.14645809139968577,
|
| 1906 |
+
"grad_norm": NaN,
|
| 1907 |
+
"learning_rate": 0.00012645063243004235,
|
| 1908 |
+
"loss": 0.0,
|
| 1909 |
+
"step": 268
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 0.14700457681535625,
|
| 1913 |
+
"grad_norm": NaN,
|
| 1914 |
+
"learning_rate": 0.00012596320683839975,
|
| 1915 |
+
"loss": 0.0,
|
| 1916 |
+
"step": 269
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.1475510622310267,
|
| 1920 |
+
"grad_norm": NaN,
|
| 1921 |
+
"learning_rate": 0.00012547511891376916,
|
| 1922 |
+
"loss": 0.0,
|
| 1923 |
+
"step": 270
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.1480975476466972,
|
| 1927 |
+
"grad_norm": NaN,
|
| 1928 |
+
"learning_rate": 0.0001249863811074912,
|
| 1929 |
+
"loss": 0.0,
|
| 1930 |
+
"step": 271
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.14864403306236765,
|
| 1934 |
+
"grad_norm": NaN,
|
| 1935 |
+
"learning_rate": 0.0001244970058874854,
|
| 1936 |
+
"loss": 0.0,
|
| 1937 |
+
"step": 272
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 0.14919051847803813,
|
| 1941 |
+
"grad_norm": NaN,
|
| 1942 |
+
"learning_rate": 0.00012400700573793192,
|
| 1943 |
+
"loss": 0.0,
|
| 1944 |
+
"step": 273
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 0.1497370038937086,
|
| 1948 |
+
"grad_norm": NaN,
|
| 1949 |
+
"learning_rate": 0.0001235163931589531,
|
| 1950 |
+
"loss": 0.0,
|
| 1951 |
+
"step": 274
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.15028348930937904,
|
| 1955 |
+
"grad_norm": NaN,
|
| 1956 |
+
"learning_rate": 0.00012302518066629466,
|
| 1957 |
+
"loss": 0.0,
|
| 1958 |
+
"step": 275
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 0.15082997472504953,
|
| 1962 |
+
"grad_norm": NaN,
|
| 1963 |
+
"learning_rate": 0.0001225333807910065,
|
| 1964 |
+
"loss": 0.0,
|
| 1965 |
+
"step": 276
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.15137646014071998,
|
| 1969 |
+
"grad_norm": NaN,
|
| 1970 |
+
"learning_rate": 0.00012204100607912277,
|
| 1971 |
+
"loss": 0.0,
|
| 1972 |
+
"step": 277
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 0.15192294555639047,
|
| 1976 |
+
"grad_norm": NaN,
|
| 1977 |
+
"learning_rate": 0.00012154806909134198,
|
| 1978 |
+
"loss": 0.0,
|
| 1979 |
+
"step": 278
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 0.15246943097206092,
|
| 1983 |
+
"grad_norm": NaN,
|
| 1984 |
+
"learning_rate": 0.0001210545824027066,
|
| 1985 |
+
"loss": 0.0,
|
| 1986 |
+
"step": 279
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 0.1530159163877314,
|
| 1990 |
+
"grad_norm": NaN,
|
| 1991 |
+
"learning_rate": 0.0001205605586022822,
|
| 1992 |
+
"loss": 0.0,
|
| 1993 |
+
"step": 280
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 0.15356240180340186,
|
| 1997 |
+
"grad_norm": NaN,
|
| 1998 |
+
"learning_rate": 0.00012006601029283628,
|
| 1999 |
+
"loss": 0.0,
|
| 2000 |
+
"step": 281
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 0.15410888721907234,
|
| 2004 |
+
"grad_norm": NaN,
|
| 2005 |
+
"learning_rate": 0.00011957095009051682,
|
| 2006 |
+
"loss": 0.0,
|
| 2007 |
+
"step": 282
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 0.1546553726347428,
|
| 2011 |
+
"grad_norm": NaN,
|
| 2012 |
+
"learning_rate": 0.00011907539062453044,
|
| 2013 |
+
"loss": 0.0,
|
| 2014 |
+
"step": 283
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 0.15520185805041328,
|
| 2018 |
+
"grad_norm": NaN,
|
| 2019 |
+
"learning_rate": 0.00011857934453682015,
|
| 2020 |
+
"loss": 0.0,
|
| 2021 |
+
"step": 284
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.15574834346608374,
|
| 2025 |
+
"grad_norm": NaN,
|
| 2026 |
+
"learning_rate": 0.00011808282448174295,
|
| 2027 |
+
"loss": 0.0,
|
| 2028 |
+
"step": 285
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 0.15629482888175422,
|
| 2032 |
+
"grad_norm": NaN,
|
| 2033 |
+
"learning_rate": 0.00011758584312574692,
|
| 2034 |
+
"loss": 0.0,
|
| 2035 |
+
"step": 286
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.15684131429742468,
|
| 2039 |
+
"grad_norm": NaN,
|
| 2040 |
+
"learning_rate": 0.00011708841314704811,
|
| 2041 |
+
"loss": 0.0,
|
| 2042 |
+
"step": 287
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 0.15738779971309516,
|
| 2046 |
+
"grad_norm": NaN,
|
| 2047 |
+
"learning_rate": 0.0001165905472353072,
|
| 2048 |
+
"loss": 0.0,
|
| 2049 |
+
"step": 288
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.15793428512876562,
|
| 2053 |
+
"grad_norm": NaN,
|
| 2054 |
+
"learning_rate": 0.00011609225809130566,
|
| 2055 |
+
"loss": 0.0,
|
| 2056 |
+
"step": 289
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 0.1584807705444361,
|
| 2060 |
+
"grad_norm": NaN,
|
| 2061 |
+
"learning_rate": 0.00011559355842662187,
|
| 2062 |
+
"loss": 0.0,
|
| 2063 |
+
"step": 290
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 0.15902725596010656,
|
| 2067 |
+
"grad_norm": NaN,
|
| 2068 |
+
"learning_rate": 0.00011509446096330669,
|
| 2069 |
+
"loss": 0.0,
|
| 2070 |
+
"step": 291
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 0.15957374137577704,
|
| 2074 |
+
"grad_norm": NaN,
|
| 2075 |
+
"learning_rate": 0.00011459497843355906,
|
| 2076 |
+
"loss": 0.0,
|
| 2077 |
+
"step": 292
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 0.1601202267914475,
|
| 2081 |
+
"grad_norm": NaN,
|
| 2082 |
+
"learning_rate": 0.00011409512357940113,
|
| 2083 |
+
"loss": 0.0,
|
| 2084 |
+
"step": 293
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 0.16066671220711798,
|
| 2088 |
+
"grad_norm": NaN,
|
| 2089 |
+
"learning_rate": 0.00011359490915235322,
|
| 2090 |
+
"loss": 0.0,
|
| 2091 |
+
"step": 294
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.16121319762278843,
|
| 2095 |
+
"grad_norm": NaN,
|
| 2096 |
+
"learning_rate": 0.00011309434791310846,
|
| 2097 |
+
"loss": 0.0,
|
| 2098 |
+
"step": 295
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 0.16175968303845892,
|
| 2102 |
+
"grad_norm": NaN,
|
| 2103 |
+
"learning_rate": 0.00011259345263120737,
|
| 2104 |
+
"loss": 0.0,
|
| 2105 |
+
"step": 296
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 0.16230616845412937,
|
| 2109 |
+
"grad_norm": NaN,
|
| 2110 |
+
"learning_rate": 0.00011209223608471202,
|
| 2111 |
+
"loss": 0.0,
|
| 2112 |
+
"step": 297
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 0.16285265386979986,
|
| 2116 |
+
"grad_norm": NaN,
|
| 2117 |
+
"learning_rate": 0.0001115907110598801,
|
| 2118 |
+
"loss": 0.0,
|
| 2119 |
+
"step": 298
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 0.1633991392854703,
|
| 2123 |
+
"grad_norm": NaN,
|
| 2124 |
+
"learning_rate": 0.00011108889035083869,
|
| 2125 |
+
"loss": 0.0,
|
| 2126 |
+
"step": 299
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.1639456247011408,
|
| 2130 |
+
"grad_norm": NaN,
|
| 2131 |
+
"learning_rate": 0.00011058678675925795,
|
| 2132 |
+
"loss": 0.0,
|
| 2133 |
+
"step": 300
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.1639456247011408,
|
| 2137 |
+
"eval_loss": NaN,
|
| 2138 |
+
"eval_runtime": 446.8533,
|
| 2139 |
+
"eval_samples_per_second": 6.897,
|
| 2140 |
+
"eval_steps_per_second": 1.725,
|
| 2141 |
+
"step": 300
|
| 2142 |
}
|
| 2143 |
],
|
| 2144 |
"logging_steps": 1,
|
|
|
|
| 2153 |
"early_stopping_threshold": 0.0
|
| 2154 |
},
|
| 2155 |
"attributes": {
|
| 2156 |
+
"early_stopping_patience_counter": 2
|
| 2157 |
}
|
| 2158 |
},
|
| 2159 |
"TrainerControl": {
|
|
|
|
| 2162 |
"should_evaluate": false,
|
| 2163 |
"should_log": false,
|
| 2164 |
"should_save": true,
|
| 2165 |
+
"should_training_stop": true
|
| 2166 |
},
|
| 2167 |
"attributes": {}
|
| 2168 |
}
|
| 2169 |
},
|
| 2170 |
+
"total_flos": 8.511538553275023e+17,
|
| 2171 |
"train_batch_size": 4,
|
| 2172 |
"trial_name": null,
|
| 2173 |
"trial_params": null
|