Training in progress, step 291, 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 200068512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d559c507527408151305de60bfac7ecab7fdc8cec07a0a100cf95fa50b736969
|
| 3 |
size 200068512
|
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 sha256:13ca1ecb806bdfef2a06cb71be0add3a97cea5049a108463e73c7b0210802d2a
|
| 3 |
+
size 102030580
|
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 sha256:56b62f440272694c6b2ce6984e703f343f25272dbe6369c607d93c46957b1d63
|
| 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:fb8ca612c507917e163ce09dc4e7e572d1fec912d9e349f29fc7ce0c23b3101f
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": 0.003245619358494878,
|
| 3 |
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
-
"epoch":
|
| 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 +1431,643 @@
|
|
| 1431 |
"eval_samples_per_second": 10.861,
|
| 1432 |
"eval_steps_per_second": 2.749,
|
| 1433 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1434 |
}
|
| 1435 |
],
|
| 1436 |
"logging_steps": 1,
|
|
@@ -1454,12 +2091,12 @@
|
|
| 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": 0.003245619358494878,
|
| 3 |
"best_model_checkpoint": "miner_id_24/checkpoint-200",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 291,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1431 |
"eval_samples_per_second": 10.861,
|
| 1432 |
"eval_steps_per_second": 2.749,
|
| 1433 |
"step": 200
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 1.3814432989690721,
|
| 1437 |
+
"grad_norm": 0.1602359563112259,
|
| 1438 |
+
"learning_rate": 4.649285891779327e-05,
|
| 1439 |
+
"loss": 0.0016,
|
| 1440 |
+
"step": 201
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 1.388316151202749,
|
| 1444 |
+
"grad_norm": 0.07345937192440033,
|
| 1445 |
+
"learning_rate": 4.555172473510329e-05,
|
| 1446 |
+
"loss": 0.0019,
|
| 1447 |
+
"step": 202
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 1.395189003436426,
|
| 1451 |
+
"grad_norm": 0.1161644235253334,
|
| 1452 |
+
"learning_rate": 4.461739615694929e-05,
|
| 1453 |
+
"loss": 0.0015,
|
| 1454 |
+
"step": 203
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 1.402061855670103,
|
| 1458 |
+
"grad_norm": 0.06747262179851532,
|
| 1459 |
+
"learning_rate": 4.3689989967026935e-05,
|
| 1460 |
+
"loss": 0.0039,
|
| 1461 |
+
"step": 204
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 1.40893470790378,
|
| 1465 |
+
"grad_norm": 0.14562109112739563,
|
| 1466 |
+
"learning_rate": 4.276962208378811e-05,
|
| 1467 |
+
"loss": 0.0029,
|
| 1468 |
+
"step": 205
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 1.4158075601374571,
|
| 1472 |
+
"grad_norm": 0.0071322242729365826,
|
| 1473 |
+
"learning_rate": 4.1856407545951834e-05,
|
| 1474 |
+
"loss": 0.0001,
|
| 1475 |
+
"step": 206
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 1.422680412371134,
|
| 1479 |
+
"grad_norm": 0.07651171833276749,
|
| 1480 |
+
"learning_rate": 4.095046049812545e-05,
|
| 1481 |
+
"loss": 0.0045,
|
| 1482 |
+
"step": 207
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 1.429553264604811,
|
| 1486 |
+
"grad_norm": 0.15785658359527588,
|
| 1487 |
+
"learning_rate": 4.005189417653743e-05,
|
| 1488 |
+
"loss": 0.001,
|
| 1489 |
+
"step": 208
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 1.436426116838488,
|
| 1493 |
+
"grad_norm": 0.09830784052610397,
|
| 1494 |
+
"learning_rate": 3.916082089488372e-05,
|
| 1495 |
+
"loss": 0.0036,
|
| 1496 |
+
"step": 209
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 1.443298969072165,
|
| 1500 |
+
"grad_norm": 0.18310892581939697,
|
| 1501 |
+
"learning_rate": 3.827735203028953e-05,
|
| 1502 |
+
"loss": 0.0053,
|
| 1503 |
+
"step": 210
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 1.4501718213058419,
|
| 1507 |
+
"grad_norm": 0.15072131156921387,
|
| 1508 |
+
"learning_rate": 3.740159800938784e-05,
|
| 1509 |
+
"loss": 0.0012,
|
| 1510 |
+
"step": 211
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 1.4570446735395188,
|
| 1514 |
+
"grad_norm": 0.14942792057991028,
|
| 1515 |
+
"learning_rate": 3.6533668294517155e-05,
|
| 1516 |
+
"loss": 0.0029,
|
| 1517 |
+
"step": 212
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 1.463917525773196,
|
| 1521 |
+
"grad_norm": 0.04605146124958992,
|
| 1522 |
+
"learning_rate": 3.5673671370039466e-05,
|
| 1523 |
+
"loss": 0.0019,
|
| 1524 |
+
"step": 213
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 1.470790378006873,
|
| 1528 |
+
"grad_norm": 0.09986945241689682,
|
| 1529 |
+
"learning_rate": 3.482171472878062e-05,
|
| 1530 |
+
"loss": 0.0023,
|
| 1531 |
+
"step": 214
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 1.47766323024055,
|
| 1535 |
+
"grad_norm": 0.06329240649938583,
|
| 1536 |
+
"learning_rate": 3.39779048585945e-05,
|
| 1537 |
+
"loss": 0.0009,
|
| 1538 |
+
"step": 215
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 1.4845360824742269,
|
| 1542 |
+
"grad_norm": 0.04497532546520233,
|
| 1543 |
+
"learning_rate": 3.3142347229053015e-05,
|
| 1544 |
+
"loss": 0.0016,
|
| 1545 |
+
"step": 216
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 1.4914089347079038,
|
| 1549 |
+
"grad_norm": 0.05954873189330101,
|
| 1550 |
+
"learning_rate": 3.231514627826305e-05,
|
| 1551 |
+
"loss": 0.0016,
|
| 1552 |
+
"step": 217
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 1.4982817869415808,
|
| 1556 |
+
"grad_norm": 0.11368856579065323,
|
| 1557 |
+
"learning_rate": 3.149640539981267e-05,
|
| 1558 |
+
"loss": 0.0042,
|
| 1559 |
+
"step": 218
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 1.5051546391752577,
|
| 1563 |
+
"grad_norm": 0.0854845941066742,
|
| 1564 |
+
"learning_rate": 3.068622692984762e-05,
|
| 1565 |
+
"loss": 0.0048,
|
| 1566 |
+
"step": 219
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 1.5120274914089347,
|
| 1570 |
+
"grad_norm": 0.11908043920993805,
|
| 1571 |
+
"learning_rate": 2.9884712134280324e-05,
|
| 1572 |
+
"loss": 0.004,
|
| 1573 |
+
"step": 220
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 1.5189003436426116,
|
| 1577 |
+
"grad_norm": 0.13990691304206848,
|
| 1578 |
+
"learning_rate": 2.909196119613218e-05,
|
| 1579 |
+
"loss": 0.0047,
|
| 1580 |
+
"step": 221
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 1.5257731958762886,
|
| 1584 |
+
"grad_norm": 0.08153674751520157,
|
| 1585 |
+
"learning_rate": 2.8308073203011663e-05,
|
| 1586 |
+
"loss": 0.0051,
|
| 1587 |
+
"step": 222
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 1.5326460481099655,
|
| 1591 |
+
"grad_norm": 0.06916696578264236,
|
| 1592 |
+
"learning_rate": 2.753314613472906e-05,
|
| 1593 |
+
"loss": 0.0022,
|
| 1594 |
+
"step": 223
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 1.5395189003436425,
|
| 1598 |
+
"grad_norm": 0.0983443409204483,
|
| 1599 |
+
"learning_rate": 2.6767276851049816e-05,
|
| 1600 |
+
"loss": 0.0037,
|
| 1601 |
+
"step": 224
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 1.5463917525773194,
|
| 1605 |
+
"grad_norm": 0.18421703577041626,
|
| 1606 |
+
"learning_rate": 2.6010561079587813e-05,
|
| 1607 |
+
"loss": 0.0017,
|
| 1608 |
+
"step": 225
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 1.5532646048109966,
|
| 1612 |
+
"grad_norm": 0.15690971910953522,
|
| 1613 |
+
"learning_rate": 2.5263093403840142e-05,
|
| 1614 |
+
"loss": 0.0024,
|
| 1615 |
+
"step": 226
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 1.5601374570446735,
|
| 1619 |
+
"grad_norm": 0.06236126646399498,
|
| 1620 |
+
"learning_rate": 2.4524967251365026e-05,
|
| 1621 |
+
"loss": 0.0022,
|
| 1622 |
+
"step": 227
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 1.5670103092783505,
|
| 1626 |
+
"grad_norm": 0.1433926373720169,
|
| 1627 |
+
"learning_rate": 2.3796274882103964e-05,
|
| 1628 |
+
"loss": 0.0052,
|
| 1629 |
+
"step": 228
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.5738831615120275,
|
| 1633 |
+
"grad_norm": 0.15132726728916168,
|
| 1634 |
+
"learning_rate": 2.3077107376850005e-05,
|
| 1635 |
+
"loss": 0.0059,
|
| 1636 |
+
"step": 229
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 1.5807560137457046,
|
| 1640 |
+
"grad_norm": 0.06804714351892471,
|
| 1641 |
+
"learning_rate": 2.2367554625863497e-05,
|
| 1642 |
+
"loss": 0.0026,
|
| 1643 |
+
"step": 230
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 1.5876288659793816,
|
| 1647 |
+
"grad_norm": 0.12087640911340714,
|
| 1648 |
+
"learning_rate": 2.166770531763633e-05,
|
| 1649 |
+
"loss": 0.0035,
|
| 1650 |
+
"step": 231
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 1.5945017182130585,
|
| 1654 |
+
"grad_norm": 0.22184179723262787,
|
| 1655 |
+
"learning_rate": 2.0977646927806683e-05,
|
| 1656 |
+
"loss": 0.0034,
|
| 1657 |
+
"step": 232
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 1.6013745704467355,
|
| 1661 |
+
"grad_norm": 0.08892931044101715,
|
| 1662 |
+
"learning_rate": 2.0297465708225238e-05,
|
| 1663 |
+
"loss": 0.0025,
|
| 1664 |
+
"step": 233
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 1.6082474226804124,
|
| 1668 |
+
"grad_norm": 0.09320689737796783,
|
| 1669 |
+
"learning_rate": 1.962724667617436e-05,
|
| 1670 |
+
"loss": 0.0043,
|
| 1671 |
+
"step": 234
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 1.6151202749140894,
|
| 1675 |
+
"grad_norm": 0.18926142156124115,
|
| 1676 |
+
"learning_rate": 1.896707360374167e-05,
|
| 1677 |
+
"loss": 0.0077,
|
| 1678 |
+
"step": 235
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 1.6219931271477663,
|
| 1682 |
+
"grad_norm": 0.07618946582078934,
|
| 1683 |
+
"learning_rate": 1.8317029007349085e-05,
|
| 1684 |
+
"loss": 0.0024,
|
| 1685 |
+
"step": 236
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 1.6288659793814433,
|
| 1689 |
+
"grad_norm": 0.023058777675032616,
|
| 1690 |
+
"learning_rate": 1.7677194137439035e-05,
|
| 1691 |
+
"loss": 0.0005,
|
| 1692 |
+
"step": 237
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 1.6357388316151202,
|
| 1696 |
+
"grad_norm": 0.08784357458353043,
|
| 1697 |
+
"learning_rate": 1.7047648968318698e-05,
|
| 1698 |
+
"loss": 0.0064,
|
| 1699 |
+
"step": 238
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 1.6426116838487972,
|
| 1703 |
+
"grad_norm": 0.032160189002752304,
|
| 1704 |
+
"learning_rate": 1.642847218816398e-05,
|
| 1705 |
+
"loss": 0.001,
|
| 1706 |
+
"step": 239
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 1.6494845360824741,
|
| 1710 |
+
"grad_norm": 0.043870214372873306,
|
| 1711 |
+
"learning_rate": 1.58197411891839e-05,
|
| 1712 |
+
"loss": 0.0013,
|
| 1713 |
+
"step": 240
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 1.656357388316151,
|
| 1717 |
+
"grad_norm": 0.08430828154087067,
|
| 1718 |
+
"learning_rate": 1.5221532057947419e-05,
|
| 1719 |
+
"loss": 0.0008,
|
| 1720 |
+
"step": 241
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 1.663230240549828,
|
| 1724 |
+
"grad_norm": 0.0656968504190445,
|
| 1725 |
+
"learning_rate": 1.4633919565873033e-05,
|
| 1726 |
+
"loss": 0.0012,
|
| 1727 |
+
"step": 242
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 1.670103092783505,
|
| 1731 |
+
"grad_norm": 0.26261475682258606,
|
| 1732 |
+
"learning_rate": 1.4056977159883012e-05,
|
| 1733 |
+
"loss": 0.0033,
|
| 1734 |
+
"step": 243
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 1.6769759450171822,
|
| 1738 |
+
"grad_norm": 0.08948194980621338,
|
| 1739 |
+
"learning_rate": 1.3490776953223105e-05,
|
| 1740 |
+
"loss": 0.0028,
|
| 1741 |
+
"step": 244
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 1.6838487972508591,
|
| 1745 |
+
"grad_norm": 0.07044168561697006,
|
| 1746 |
+
"learning_rate": 1.2935389716448976e-05,
|
| 1747 |
+
"loss": 0.0022,
|
| 1748 |
+
"step": 245
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 1.690721649484536,
|
| 1752 |
+
"grad_norm": 0.08131295442581177,
|
| 1753 |
+
"learning_rate": 1.23908848685804e-05,
|
| 1754 |
+
"loss": 0.002,
|
| 1755 |
+
"step": 246
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 1.697594501718213,
|
| 1759 |
+
"grad_norm": 0.1506740152835846,
|
| 1760 |
+
"learning_rate": 1.1857330468424466e-05,
|
| 1761 |
+
"loss": 0.0035,
|
| 1762 |
+
"step": 247
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 1.7044673539518902,
|
| 1766 |
+
"grad_norm": 0.045038871467113495,
|
| 1767 |
+
"learning_rate": 1.133479320606874e-05,
|
| 1768 |
+
"loss": 0.0019,
|
| 1769 |
+
"step": 248
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 1.7113402061855671,
|
| 1773 |
+
"grad_norm": 0.14267544448375702,
|
| 1774 |
+
"learning_rate": 1.082333839454559e-05,
|
| 1775 |
+
"loss": 0.0047,
|
| 1776 |
+
"step": 249
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 1.718213058419244,
|
| 1780 |
+
"grad_norm": 0.11678794771432877,
|
| 1781 |
+
"learning_rate": 1.0323029961668462e-05,
|
| 1782 |
+
"loss": 0.002,
|
| 1783 |
+
"step": 250
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 1.725085910652921,
|
| 1787 |
+
"grad_norm": 0.04625638574361801,
|
| 1788 |
+
"learning_rate": 9.833930442041506e-06,
|
| 1789 |
+
"loss": 0.0022,
|
| 1790 |
+
"step": 251
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 1.731958762886598,
|
| 1794 |
+
"grad_norm": 0.13576196134090424,
|
| 1795 |
+
"learning_rate": 9.35610096924323e-06,
|
| 1796 |
+
"loss": 0.0022,
|
| 1797 |
+
"step": 252
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 1.738831615120275,
|
| 1801 |
+
"grad_norm": 0.10366480052471161,
|
| 1802 |
+
"learning_rate": 8.889601268185232e-06,
|
| 1803 |
+
"loss": 0.0055,
|
| 1804 |
+
"step": 253
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 1.745704467353952,
|
| 1808 |
+
"grad_norm": 0.04077526181936264,
|
| 1809 |
+
"learning_rate": 8.434489647647092e-06,
|
| 1810 |
+
"loss": 0.0017,
|
| 1811 |
+
"step": 254
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 1.7525773195876289,
|
| 1815 |
+
"grad_norm": 0.11092139035463333,
|
| 1816 |
+
"learning_rate": 7.990822992988267e-06,
|
| 1817 |
+
"loss": 0.0034,
|
| 1818 |
+
"step": 255
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 1.7594501718213058,
|
| 1822 |
+
"grad_norm": 0.069175124168396,
|
| 1823 |
+
"learning_rate": 7.558656759037797e-06,
|
| 1824 |
+
"loss": 0.0026,
|
| 1825 |
+
"step": 256
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 1.7663230240549828,
|
| 1829 |
+
"grad_norm": 0.12054812163114548,
|
| 1830 |
+
"learning_rate": 7.13804496316296e-06,
|
| 1831 |
+
"loss": 0.0024,
|
| 1832 |
+
"step": 257
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 1.7731958762886597,
|
| 1836 |
+
"grad_norm": 0.03802892938256264,
|
| 1837 |
+
"learning_rate": 6.729040178517454e-06,
|
| 1838 |
+
"loss": 0.0011,
|
| 1839 |
+
"step": 258
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 1.7800687285223367,
|
| 1843 |
+
"grad_norm": 0.031027084216475487,
|
| 1844 |
+
"learning_rate": 6.331693527470306e-06,
|
| 1845 |
+
"loss": 0.0013,
|
| 1846 |
+
"step": 259
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 1.7869415807560136,
|
| 1850 |
+
"grad_norm": 0.20980218052864075,
|
| 1851 |
+
"learning_rate": 5.946054675215784e-06,
|
| 1852 |
+
"loss": 0.0011,
|
| 1853 |
+
"step": 260
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 1.7938144329896906,
|
| 1857 |
+
"grad_norm": 0.1241433396935463,
|
| 1858 |
+
"learning_rate": 5.572171823565797e-06,
|
| 1859 |
+
"loss": 0.0011,
|
| 1860 |
+
"step": 261
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 1.8006872852233677,
|
| 1864 |
+
"grad_norm": 0.10405414551496506,
|
| 1865 |
+
"learning_rate": 5.210091704924946e-06,
|
| 1866 |
+
"loss": 0.0063,
|
| 1867 |
+
"step": 262
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 1.8075601374570447,
|
| 1871 |
+
"grad_norm": 0.019170010462403297,
|
| 1872 |
+
"learning_rate": 4.859859576449444e-06,
|
| 1873 |
+
"loss": 0.0002,
|
| 1874 |
+
"step": 263
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 1.8144329896907216,
|
| 1878 |
+
"grad_norm": 0.10227034986019135,
|
| 1879 |
+
"learning_rate": 4.521519214390257e-06,
|
| 1880 |
+
"loss": 0.0024,
|
| 1881 |
+
"step": 264
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 1.8213058419243986,
|
| 1885 |
+
"grad_norm": 0.06309763342142105,
|
| 1886 |
+
"learning_rate": 4.195112908621402e-06,
|
| 1887 |
+
"loss": 0.002,
|
| 1888 |
+
"step": 265
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 1.8281786941580758,
|
| 1892 |
+
"grad_norm": 0.11718452721834183,
|
| 1893 |
+
"learning_rate": 3.880681457354118e-06,
|
| 1894 |
+
"loss": 0.0011,
|
| 1895 |
+
"step": 266
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 1.8350515463917527,
|
| 1899 |
+
"grad_norm": 0.04819872975349426,
|
| 1900 |
+
"learning_rate": 3.578264162037348e-06,
|
| 1901 |
+
"loss": 0.003,
|
| 1902 |
+
"step": 267
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 1.8419243986254297,
|
| 1906 |
+
"grad_norm": 0.0733291283249855,
|
| 1907 |
+
"learning_rate": 3.2878988224454344e-06,
|
| 1908 |
+
"loss": 0.0016,
|
| 1909 |
+
"step": 268
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 1.8487972508591066,
|
| 1913 |
+
"grad_norm": 0.1850636601448059,
|
| 1914 |
+
"learning_rate": 3.0096217319533382e-06,
|
| 1915 |
+
"loss": 0.0032,
|
| 1916 |
+
"step": 269
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 1.8556701030927836,
|
| 1920 |
+
"grad_norm": 0.40491941571235657,
|
| 1921 |
+
"learning_rate": 2.7434676730003884e-06,
|
| 1922 |
+
"loss": 0.004,
|
| 1923 |
+
"step": 270
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 1.8625429553264605,
|
| 1927 |
+
"grad_norm": 0.08381707221269608,
|
| 1928 |
+
"learning_rate": 2.4894699127426367e-06,
|
| 1929 |
+
"loss": 0.0018,
|
| 1930 |
+
"step": 271
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 1.8694158075601375,
|
| 1934 |
+
"grad_norm": 0.06890317052602768,
|
| 1935 |
+
"learning_rate": 2.2476601988947966e-06,
|
| 1936 |
+
"loss": 0.0026,
|
| 1937 |
+
"step": 272
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 1.8762886597938144,
|
| 1941 |
+
"grad_norm": 0.15883168578147888,
|
| 1942 |
+
"learning_rate": 2.0180687557619816e-06,
|
| 1943 |
+
"loss": 0.0034,
|
| 1944 |
+
"step": 273
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 1.8831615120274914,
|
| 1948 |
+
"grad_norm": 0.06412586569786072,
|
| 1949 |
+
"learning_rate": 1.8007242804619628e-06,
|
| 1950 |
+
"loss": 0.0014,
|
| 1951 |
+
"step": 274
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 1.8900343642611683,
|
| 1955 |
+
"grad_norm": 0.07986130565404892,
|
| 1956 |
+
"learning_rate": 1.595653939338204e-06,
|
| 1957 |
+
"loss": 0.0015,
|
| 1958 |
+
"step": 275
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 1.8969072164948453,
|
| 1962 |
+
"grad_norm": 0.07344045490026474,
|
| 1963 |
+
"learning_rate": 1.4028833645643113e-06,
|
| 1964 |
+
"loss": 0.0011,
|
| 1965 |
+
"step": 276
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 1.9037800687285222,
|
| 1969 |
+
"grad_norm": 0.1562926471233368,
|
| 1970 |
+
"learning_rate": 1.222436650940173e-06,
|
| 1971 |
+
"loss": 0.0092,
|
| 1972 |
+
"step": 277
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 1.9106529209621992,
|
| 1976 |
+
"grad_norm": 0.11137005686759949,
|
| 1977 |
+
"learning_rate": 1.0543363528803696e-06,
|
| 1978 |
+
"loss": 0.0027,
|
| 1979 |
+
"step": 278
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 1.9175257731958761,
|
| 1983 |
+
"grad_norm": 0.1020520031452179,
|
| 1984 |
+
"learning_rate": 8.986034815950172e-07,
|
| 1985 |
+
"loss": 0.0054,
|
| 1986 |
+
"step": 279
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 1.9243986254295533,
|
| 1990 |
+
"grad_norm": 0.044514842331409454,
|
| 1991 |
+
"learning_rate": 7.552575024634689e-07,
|
| 1992 |
+
"loss": 0.0016,
|
| 1993 |
+
"step": 280
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 1.9312714776632303,
|
| 1997 |
+
"grad_norm": 0.05561070516705513,
|
| 1998 |
+
"learning_rate": 6.243163326014267e-07,
|
| 1999 |
+
"loss": 0.0022,
|
| 2000 |
+
"step": 281
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 1.9381443298969072,
|
| 2004 |
+
"grad_norm": 0.22467757761478424,
|
| 2005 |
+
"learning_rate": 5.057963386213116e-07,
|
| 2006 |
+
"loss": 0.0035,
|
| 2007 |
+
"step": 282
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 1.9450171821305842,
|
| 2011 |
+
"grad_norm": 0.08855283260345459,
|
| 2012 |
+
"learning_rate": 3.9971233458665493e-07,
|
| 2013 |
+
"loss": 0.0023,
|
| 2014 |
+
"step": 283
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 1.9518900343642611,
|
| 2018 |
+
"grad_norm": 0.14920522272586823,
|
| 2019 |
+
"learning_rate": 3.060775801604354e-07,
|
| 2020 |
+
"loss": 0.0061,
|
| 2021 |
+
"step": 284
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 1.9587628865979383,
|
| 2025 |
+
"grad_norm": 0.11901390552520752,
|
| 2026 |
+
"learning_rate": 2.2490377894768267e-07,
|
| 2027 |
+
"loss": 0.0055,
|
| 2028 |
+
"step": 285
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 1.9656357388316152,
|
| 2032 |
+
"grad_norm": 0.07598073780536652,
|
| 2033 |
+
"learning_rate": 1.562010770326916e-07,
|
| 2034 |
+
"loss": 0.0036,
|
| 2035 |
+
"step": 286
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 1.9725085910652922,
|
| 2039 |
+
"grad_norm": 0.1735997498035431,
|
| 2040 |
+
"learning_rate": 9.99780617107815e-08,
|
| 2041 |
+
"loss": 0.0038,
|
| 2042 |
+
"step": 287
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 1.9793814432989691,
|
| 2046 |
+
"grad_norm": 0.1865747720003128,
|
| 2047 |
+
"learning_rate": 5.6241760414987856e-08,
|
| 2048 |
+
"loss": 0.007,
|
| 2049 |
+
"step": 288
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 1.986254295532646,
|
| 2053 |
+
"grad_norm": 0.05057632550597191,
|
| 2054 |
+
"learning_rate": 2.4997639837687213e-08,
|
| 2055 |
+
"loss": 0.0032,
|
| 2056 |
+
"step": 289
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 1.993127147766323,
|
| 2060 |
+
"grad_norm": 0.13223591446876526,
|
| 2061 |
+
"learning_rate": 6.2496052472549304e-09,
|
| 2062 |
+
"loss": 0.0037,
|
| 2063 |
+
"step": 290
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 2.0,
|
| 2067 |
+
"grad_norm": 0.31574028730392456,
|
| 2068 |
+
"learning_rate": 0.0,
|
| 2069 |
+
"loss": 0.0051,
|
| 2070 |
+
"step": 291
|
| 2071 |
}
|
| 2072 |
],
|
| 2073 |
"logging_steps": 1,
|
|
|
|
| 2091 |
"should_evaluate": false,
|
| 2092 |
"should_log": false,
|
| 2093 |
"should_save": true,
|
| 2094 |
+
"should_training_stop": true
|
| 2095 |
},
|
| 2096 |
"attributes": {}
|
| 2097 |
}
|
| 2098 |
},
|
| 2099 |
+
"total_flos": 4.4623665662263296e+17,
|
| 2100 |
"train_batch_size": 4,
|
| 2101 |
"trial_name": null,
|
| 2102 |
"trial_params": null
|