Training in progress, step 300, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 40001880
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f8af9875dba3abc72b08d5cc40e93518bd82b3e85f63b6cad6ce774111d3dff6
|
| 3 |
size 40001880
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 40043787
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c652b40dc1376b5f1c6481c56235751d13ff2f60b4ee715e4f48755095005c2
|
| 3 |
size 40043787
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:793829d79b248c3a7b8954f2cd95073c2ba034f6ee2bb0edff8ce8fef88cb5ad
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1401
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:954462bb708bcdba3c0732881adee53bb51d512303049568372c51a54b8cb129
|
| 3 |
size 1401
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 100,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1424,6 +1424,714 @@
|
|
| 1424 |
"eval_samples_per_second": 149.98,
|
| 1425 |
"eval_steps_per_second": 18.953,
|
| 1426 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1427 |
}
|
| 1428 |
],
|
| 1429 |
"logging_steps": 1,
|
|
@@ -1443,7 +2151,7 @@
|
|
| 1443 |
"attributes": {}
|
| 1444 |
}
|
| 1445 |
},
|
| 1446 |
-
"total_flos":
|
| 1447 |
"train_batch_size": 128,
|
| 1448 |
"trial_name": null,
|
| 1449 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.21216407355021216,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 300,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1424 |
"eval_samples_per_second": 149.98,
|
| 1425 |
"eval_steps_per_second": 18.953,
|
| 1426 |
"step": 200
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 0.14214992927864215,
|
| 1430 |
+
"grad_norm": 0.76953125,
|
| 1431 |
+
"learning_rate": 0.0008796665258991866,
|
| 1432 |
+
"loss": 10.2804,
|
| 1433 |
+
"step": 201
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.14285714285714285,
|
| 1437 |
+
"grad_norm": 0.83203125,
|
| 1438 |
+
"learning_rate": 0.0008793524987925326,
|
| 1439 |
+
"loss": 10.2891,
|
| 1440 |
+
"step": 202
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.14356435643564355,
|
| 1444 |
+
"grad_norm": 0.7890625,
|
| 1445 |
+
"learning_rate": 0.0008790361222612515,
|
| 1446 |
+
"loss": 10.2691,
|
| 1447 |
+
"step": 203
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 0.14427157001414428,
|
| 1451 |
+
"grad_norm": 0.7890625,
|
| 1452 |
+
"learning_rate": 0.0008787173980365612,
|
| 1453 |
+
"loss": 10.2514,
|
| 1454 |
+
"step": 204
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 0.144978783592645,
|
| 1458 |
+
"grad_norm": 0.796875,
|
| 1459 |
+
"learning_rate": 0.0008783963278625267,
|
| 1460 |
+
"loss": 10.264,
|
| 1461 |
+
"step": 205
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 0.1456859971711457,
|
| 1465 |
+
"grad_norm": 0.7890625,
|
| 1466 |
+
"learning_rate": 0.0008780729134960495,
|
| 1467 |
+
"loss": 10.2424,
|
| 1468 |
+
"step": 206
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 0.1463932107496464,
|
| 1472 |
+
"grad_norm": 0.74609375,
|
| 1473 |
+
"learning_rate": 0.000877747156706859,
|
| 1474 |
+
"loss": 10.2754,
|
| 1475 |
+
"step": 207
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 0.1471004243281471,
|
| 1479 |
+
"grad_norm": 0.76171875,
|
| 1480 |
+
"learning_rate": 0.0008774190592775022,
|
| 1481 |
+
"loss": 10.2663,
|
| 1482 |
+
"step": 208
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.1478076379066478,
|
| 1486 |
+
"grad_norm": 0.7421875,
|
| 1487 |
+
"learning_rate": 0.0008770886230033342,
|
| 1488 |
+
"loss": 10.255,
|
| 1489 |
+
"step": 209
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 0.1485148514851485,
|
| 1493 |
+
"grad_norm": 0.78125,
|
| 1494 |
+
"learning_rate": 0.0008767558496925081,
|
| 1495 |
+
"loss": 10.2529,
|
| 1496 |
+
"step": 210
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 0.1492220650636492,
|
| 1500 |
+
"grad_norm": 0.76953125,
|
| 1501 |
+
"learning_rate": 0.0008764207411659658,
|
| 1502 |
+
"loss": 10.2576,
|
| 1503 |
+
"step": 211
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 0.14992927864214992,
|
| 1507 |
+
"grad_norm": 0.75390625,
|
| 1508 |
+
"learning_rate": 0.0008760832992574269,
|
| 1509 |
+
"loss": 10.243,
|
| 1510 |
+
"step": 212
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 0.15063649222065065,
|
| 1514 |
+
"grad_norm": 0.7578125,
|
| 1515 |
+
"learning_rate": 0.0008757435258133798,
|
| 1516 |
+
"loss": 10.2374,
|
| 1517 |
+
"step": 213
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.15134370579915135,
|
| 1521 |
+
"grad_norm": 0.7890625,
|
| 1522 |
+
"learning_rate": 0.0008754014226930707,
|
| 1523 |
+
"loss": 10.248,
|
| 1524 |
+
"step": 214
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 0.15205091937765206,
|
| 1528 |
+
"grad_norm": 0.77734375,
|
| 1529 |
+
"learning_rate": 0.0008750569917684944,
|
| 1530 |
+
"loss": 10.2509,
|
| 1531 |
+
"step": 215
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 0.15275813295615276,
|
| 1535 |
+
"grad_norm": 0.78125,
|
| 1536 |
+
"learning_rate": 0.0008747102349243827,
|
| 1537 |
+
"loss": 10.2583,
|
| 1538 |
+
"step": 216
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.15346534653465346,
|
| 1542 |
+
"grad_norm": 0.78125,
|
| 1543 |
+
"learning_rate": 0.0008743611540581957,
|
| 1544 |
+
"loss": 10.2646,
|
| 1545 |
+
"step": 217
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.15417256011315417,
|
| 1549 |
+
"grad_norm": 0.80859375,
|
| 1550 |
+
"learning_rate": 0.00087400975108011,
|
| 1551 |
+
"loss": 10.2393,
|
| 1552 |
+
"step": 218
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 0.15487977369165487,
|
| 1556 |
+
"grad_norm": 0.796875,
|
| 1557 |
+
"learning_rate": 0.0008736560279130091,
|
| 1558 |
+
"loss": 10.2518,
|
| 1559 |
+
"step": 219
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 0.15558698727015557,
|
| 1563 |
+
"grad_norm": 0.7265625,
|
| 1564 |
+
"learning_rate": 0.0008732999864924726,
|
| 1565 |
+
"loss": 10.2396,
|
| 1566 |
+
"step": 220
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 0.1562942008486563,
|
| 1570 |
+
"grad_norm": 0.7578125,
|
| 1571 |
+
"learning_rate": 0.0008729416287667654,
|
| 1572 |
+
"loss": 10.2418,
|
| 1573 |
+
"step": 221
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 0.157001414427157,
|
| 1577 |
+
"grad_norm": 0.79296875,
|
| 1578 |
+
"learning_rate": 0.0008725809566968277,
|
| 1579 |
+
"loss": 10.2397,
|
| 1580 |
+
"step": 222
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.15770862800565771,
|
| 1584 |
+
"grad_norm": 0.8046875,
|
| 1585 |
+
"learning_rate": 0.0008722179722562636,
|
| 1586 |
+
"loss": 10.2309,
|
| 1587 |
+
"step": 223
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.15841584158415842,
|
| 1591 |
+
"grad_norm": 0.78515625,
|
| 1592 |
+
"learning_rate": 0.0008718526774313301,
|
| 1593 |
+
"loss": 10.248,
|
| 1594 |
+
"step": 224
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.15912305516265912,
|
| 1598 |
+
"grad_norm": 0.796875,
|
| 1599 |
+
"learning_rate": 0.0008714850742209275,
|
| 1600 |
+
"loss": 10.232,
|
| 1601 |
+
"step": 225
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.15983026874115983,
|
| 1605 |
+
"grad_norm": 0.7890625,
|
| 1606 |
+
"learning_rate": 0.000871115164636587,
|
| 1607 |
+
"loss": 10.261,
|
| 1608 |
+
"step": 226
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.16053748231966053,
|
| 1612 |
+
"grad_norm": 0.82421875,
|
| 1613 |
+
"learning_rate": 0.0008707429507024607,
|
| 1614 |
+
"loss": 10.2448,
|
| 1615 |
+
"step": 227
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.16124469589816123,
|
| 1619 |
+
"grad_norm": 0.8125,
|
| 1620 |
+
"learning_rate": 0.0008703684344553098,
|
| 1621 |
+
"loss": 10.2303,
|
| 1622 |
+
"step": 228
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.16195190947666196,
|
| 1626 |
+
"grad_norm": 0.76171875,
|
| 1627 |
+
"learning_rate": 0.000869991617944494,
|
| 1628 |
+
"loss": 10.2336,
|
| 1629 |
+
"step": 229
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.16265912305516267,
|
| 1633 |
+
"grad_norm": 0.7890625,
|
| 1634 |
+
"learning_rate": 0.0008696125032319601,
|
| 1635 |
+
"loss": 10.2486,
|
| 1636 |
+
"step": 230
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 0.16336633663366337,
|
| 1640 |
+
"grad_norm": 0.78125,
|
| 1641 |
+
"learning_rate": 0.0008692310923922306,
|
| 1642 |
+
"loss": 10.2299,
|
| 1643 |
+
"step": 231
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.16407355021216408,
|
| 1647 |
+
"grad_norm": 0.75390625,
|
| 1648 |
+
"learning_rate": 0.0008688473875123925,
|
| 1649 |
+
"loss": 10.2432,
|
| 1650 |
+
"step": 232
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 0.16478076379066478,
|
| 1654 |
+
"grad_norm": 0.77734375,
|
| 1655 |
+
"learning_rate": 0.000868461390692086,
|
| 1656 |
+
"loss": 10.2641,
|
| 1657 |
+
"step": 233
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.16548797736916548,
|
| 1661 |
+
"grad_norm": 0.79296875,
|
| 1662 |
+
"learning_rate": 0.0008680731040434925,
|
| 1663 |
+
"loss": 10.2143,
|
| 1664 |
+
"step": 234
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.1661951909476662,
|
| 1668 |
+
"grad_norm": 0.75390625,
|
| 1669 |
+
"learning_rate": 0.0008676825296913235,
|
| 1670 |
+
"loss": 10.2482,
|
| 1671 |
+
"step": 235
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 0.1669024045261669,
|
| 1675 |
+
"grad_norm": 0.78515625,
|
| 1676 |
+
"learning_rate": 0.0008672896697728091,
|
| 1677 |
+
"loss": 10.2266,
|
| 1678 |
+
"step": 236
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 0.1676096181046676,
|
| 1682 |
+
"grad_norm": 0.8046875,
|
| 1683 |
+
"learning_rate": 0.0008668945264376857,
|
| 1684 |
+
"loss": 10.2357,
|
| 1685 |
+
"step": 237
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 0.16831683168316833,
|
| 1689 |
+
"grad_norm": 0.79296875,
|
| 1690 |
+
"learning_rate": 0.0008664971018481848,
|
| 1691 |
+
"loss": 10.2237,
|
| 1692 |
+
"step": 238
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 0.16902404526166903,
|
| 1696 |
+
"grad_norm": 0.7890625,
|
| 1697 |
+
"learning_rate": 0.000866097398179021,
|
| 1698 |
+
"loss": 10.2447,
|
| 1699 |
+
"step": 239
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 0.16973125884016974,
|
| 1703 |
+
"grad_norm": 0.77734375,
|
| 1704 |
+
"learning_rate": 0.00086569541761738,
|
| 1705 |
+
"loss": 10.2407,
|
| 1706 |
+
"step": 240
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 0.17043847241867044,
|
| 1710 |
+
"grad_norm": 0.7734375,
|
| 1711 |
+
"learning_rate": 0.0008652911623629067,
|
| 1712 |
+
"loss": 10.2269,
|
| 1713 |
+
"step": 241
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.17114568599717114,
|
| 1717 |
+
"grad_norm": 0.7890625,
|
| 1718 |
+
"learning_rate": 0.000864884634627693,
|
| 1719 |
+
"loss": 10.2356,
|
| 1720 |
+
"step": 242
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 0.17185289957567185,
|
| 1724 |
+
"grad_norm": 0.78125,
|
| 1725 |
+
"learning_rate": 0.0008644758366362661,
|
| 1726 |
+
"loss": 10.2402,
|
| 1727 |
+
"step": 243
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.17256011315417255,
|
| 1731 |
+
"grad_norm": 0.78515625,
|
| 1732 |
+
"learning_rate": 0.0008640647706255762,
|
| 1733 |
+
"loss": 10.2485,
|
| 1734 |
+
"step": 244
|
| 1735 |
+
},
|
| 1736 |
+
{
|
| 1737 |
+
"epoch": 0.17326732673267325,
|
| 1738 |
+
"grad_norm": 0.765625,
|
| 1739 |
+
"learning_rate": 0.0008636514388449835,
|
| 1740 |
+
"loss": 10.2711,
|
| 1741 |
+
"step": 245
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 0.173974540311174,
|
| 1745 |
+
"grad_norm": 0.78515625,
|
| 1746 |
+
"learning_rate": 0.0008632358435562474,
|
| 1747 |
+
"loss": 10.2419,
|
| 1748 |
+
"step": 246
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 0.1746817538896747,
|
| 1752 |
+
"grad_norm": 0.7734375,
|
| 1753 |
+
"learning_rate": 0.0008628179870335125,
|
| 1754 |
+
"loss": 10.2475,
|
| 1755 |
+
"step": 247
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 0.1753889674681754,
|
| 1759 |
+
"grad_norm": 0.78515625,
|
| 1760 |
+
"learning_rate": 0.0008623978715632973,
|
| 1761 |
+
"loss": 10.2354,
|
| 1762 |
+
"step": 248
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 0.1760961810466761,
|
| 1766 |
+
"grad_norm": 0.8046875,
|
| 1767 |
+
"learning_rate": 0.0008619754994444814,
|
| 1768 |
+
"loss": 10.2518,
|
| 1769 |
+
"step": 249
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.1768033946251768,
|
| 1773 |
+
"grad_norm": 0.84765625,
|
| 1774 |
+
"learning_rate": 0.0008615508729882928,
|
| 1775 |
+
"loss": 10.2316,
|
| 1776 |
+
"step": 250
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 0.1775106082036775,
|
| 1780 |
+
"grad_norm": 0.8046875,
|
| 1781 |
+
"learning_rate": 0.0008611239945182946,
|
| 1782 |
+
"loss": 10.2365,
|
| 1783 |
+
"step": 251
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 0.1782178217821782,
|
| 1787 |
+
"grad_norm": 0.7890625,
|
| 1788 |
+
"learning_rate": 0.0008606948663703739,
|
| 1789 |
+
"loss": 10.2641,
|
| 1790 |
+
"step": 252
|
| 1791 |
+
},
|
| 1792 |
+
{
|
| 1793 |
+
"epoch": 0.1789250353606789,
|
| 1794 |
+
"grad_norm": 0.80078125,
|
| 1795 |
+
"learning_rate": 0.0008602634908927277,
|
| 1796 |
+
"loss": 10.2601,
|
| 1797 |
+
"step": 253
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"epoch": 0.17963224893917965,
|
| 1801 |
+
"grad_norm": 0.80859375,
|
| 1802 |
+
"learning_rate": 0.0008598298704458502,
|
| 1803 |
+
"loss": 10.2662,
|
| 1804 |
+
"step": 254
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.18033946251768035,
|
| 1808 |
+
"grad_norm": 0.80859375,
|
| 1809 |
+
"learning_rate": 0.0008593940074025203,
|
| 1810 |
+
"loss": 10.2693,
|
| 1811 |
+
"step": 255
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.18104667609618105,
|
| 1815 |
+
"grad_norm": 0.78515625,
|
| 1816 |
+
"learning_rate": 0.0008589559041477887,
|
| 1817 |
+
"loss": 10.2418,
|
| 1818 |
+
"step": 256
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.18175388967468176,
|
| 1822 |
+
"grad_norm": 0.82421875,
|
| 1823 |
+
"learning_rate": 0.000858515563078964,
|
| 1824 |
+
"loss": 10.2431,
|
| 1825 |
+
"step": 257
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.18246110325318246,
|
| 1829 |
+
"grad_norm": 0.8046875,
|
| 1830 |
+
"learning_rate": 0.0008580729866056009,
|
| 1831 |
+
"loss": 10.2722,
|
| 1832 |
+
"step": 258
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.18316831683168316,
|
| 1836 |
+
"grad_norm": 0.83203125,
|
| 1837 |
+
"learning_rate": 0.0008576281771494854,
|
| 1838 |
+
"loss": 10.2565,
|
| 1839 |
+
"step": 259
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.18387553041018387,
|
| 1843 |
+
"grad_norm": 0.84765625,
|
| 1844 |
+
"learning_rate": 0.0008571811371446231,
|
| 1845 |
+
"loss": 10.2524,
|
| 1846 |
+
"step": 260
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.18458274398868457,
|
| 1850 |
+
"grad_norm": 0.80078125,
|
| 1851 |
+
"learning_rate": 0.0008567318690372251,
|
| 1852 |
+
"loss": 10.2671,
|
| 1853 |
+
"step": 261
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.18528995756718528,
|
| 1857 |
+
"grad_norm": 0.8984375,
|
| 1858 |
+
"learning_rate": 0.0008562803752856944,
|
| 1859 |
+
"loss": 10.267,
|
| 1860 |
+
"step": 262
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.185997171145686,
|
| 1864 |
+
"grad_norm": 0.81640625,
|
| 1865 |
+
"learning_rate": 0.000855826658360613,
|
| 1866 |
+
"loss": 10.2668,
|
| 1867 |
+
"step": 263
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.1867043847241867,
|
| 1871 |
+
"grad_norm": 0.80078125,
|
| 1872 |
+
"learning_rate": 0.000855370720744728,
|
| 1873 |
+
"loss": 10.2517,
|
| 1874 |
+
"step": 264
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.18741159830268742,
|
| 1878 |
+
"grad_norm": 0.85546875,
|
| 1879 |
+
"learning_rate": 0.0008549125649329386,
|
| 1880 |
+
"loss": 10.256,
|
| 1881 |
+
"step": 265
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.18811881188118812,
|
| 1885 |
+
"grad_norm": 0.83984375,
|
| 1886 |
+
"learning_rate": 0.0008544521934322814,
|
| 1887 |
+
"loss": 10.2607,
|
| 1888 |
+
"step": 266
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.18882602545968882,
|
| 1892 |
+
"grad_norm": 0.8203125,
|
| 1893 |
+
"learning_rate": 0.0008539896087619176,
|
| 1894 |
+
"loss": 10.2778,
|
| 1895 |
+
"step": 267
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 0.18953323903818953,
|
| 1899 |
+
"grad_norm": 0.859375,
|
| 1900 |
+
"learning_rate": 0.0008535248134531189,
|
| 1901 |
+
"loss": 10.2601,
|
| 1902 |
+
"step": 268
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 0.19024045261669023,
|
| 1906 |
+
"grad_norm": 0.84765625,
|
| 1907 |
+
"learning_rate": 0.0008530578100492538,
|
| 1908 |
+
"loss": 10.2399,
|
| 1909 |
+
"step": 269
|
| 1910 |
+
},
|
| 1911 |
+
{
|
| 1912 |
+
"epoch": 0.19094766619519093,
|
| 1913 |
+
"grad_norm": 0.828125,
|
| 1914 |
+
"learning_rate": 0.000852588601105773,
|
| 1915 |
+
"loss": 10.2647,
|
| 1916 |
+
"step": 270
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 0.19165487977369167,
|
| 1920 |
+
"grad_norm": 0.8203125,
|
| 1921 |
+
"learning_rate": 0.0008521171891901965,
|
| 1922 |
+
"loss": 10.265,
|
| 1923 |
+
"step": 271
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.19236209335219237,
|
| 1927 |
+
"grad_norm": 0.796875,
|
| 1928 |
+
"learning_rate": 0.000851643576882099,
|
| 1929 |
+
"loss": 10.2793,
|
| 1930 |
+
"step": 272
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.19306930693069307,
|
| 1934 |
+
"grad_norm": 0.859375,
|
| 1935 |
+
"learning_rate": 0.0008511677667730952,
|
| 1936 |
+
"loss": 10.283,
|
| 1937 |
+
"step": 273
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 0.19377652050919378,
|
| 1941 |
+
"grad_norm": 0.85546875,
|
| 1942 |
+
"learning_rate": 0.000850689761466827,
|
| 1943 |
+
"loss": 10.2648,
|
| 1944 |
+
"step": 274
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 0.19448373408769448,
|
| 1948 |
+
"grad_norm": 0.8671875,
|
| 1949 |
+
"learning_rate": 0.0008502095635789478,
|
| 1950 |
+
"loss": 10.2946,
|
| 1951 |
+
"step": 275
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"epoch": 0.19519094766619519,
|
| 1955 |
+
"grad_norm": 0.86328125,
|
| 1956 |
+
"learning_rate": 0.0008497271757371093,
|
| 1957 |
+
"loss": 10.2982,
|
| 1958 |
+
"step": 276
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 0.1958981612446959,
|
| 1962 |
+
"grad_norm": 0.8203125,
|
| 1963 |
+
"learning_rate": 0.0008492426005809464,
|
| 1964 |
+
"loss": 10.2918,
|
| 1965 |
+
"step": 277
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.1966053748231966,
|
| 1969 |
+
"grad_norm": 0.84375,
|
| 1970 |
+
"learning_rate": 0.0008487558407620629,
|
| 1971 |
+
"loss": 10.2779,
|
| 1972 |
+
"step": 278
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 0.19731258840169733,
|
| 1976 |
+
"grad_norm": 0.859375,
|
| 1977 |
+
"learning_rate": 0.0008482668989440178,
|
| 1978 |
+
"loss": 10.294,
|
| 1979 |
+
"step": 279
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 0.19801980198019803,
|
| 1983 |
+
"grad_norm": 0.859375,
|
| 1984 |
+
"learning_rate": 0.0008477757778023092,
|
| 1985 |
+
"loss": 10.2892,
|
| 1986 |
+
"step": 280
|
| 1987 |
+
},
|
| 1988 |
+
{
|
| 1989 |
+
"epoch": 0.19872701555869873,
|
| 1990 |
+
"grad_norm": 0.828125,
|
| 1991 |
+
"learning_rate": 0.0008472824800243608,
|
| 1992 |
+
"loss": 10.3013,
|
| 1993 |
+
"step": 281
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 0.19943422913719944,
|
| 1997 |
+
"grad_norm": 0.8515625,
|
| 1998 |
+
"learning_rate": 0.0008467870083095073,
|
| 1999 |
+
"loss": 10.3026,
|
| 2000 |
+
"step": 282
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 0.20014144271570014,
|
| 2004 |
+
"grad_norm": 0.84375,
|
| 2005 |
+
"learning_rate": 0.0008462893653689785,
|
| 2006 |
+
"loss": 10.3071,
|
| 2007 |
+
"step": 283
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 0.20084865629420084,
|
| 2011 |
+
"grad_norm": 0.87890625,
|
| 2012 |
+
"learning_rate": 0.0008457895539258857,
|
| 2013 |
+
"loss": 10.3179,
|
| 2014 |
+
"step": 284
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 0.20155586987270155,
|
| 2018 |
+
"grad_norm": 0.88671875,
|
| 2019 |
+
"learning_rate": 0.0008452875767152062,
|
| 2020 |
+
"loss": 10.2782,
|
| 2021 |
+
"step": 285
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.20226308345120225,
|
| 2025 |
+
"grad_norm": 0.87109375,
|
| 2026 |
+
"learning_rate": 0.0008447834364837685,
|
| 2027 |
+
"loss": 10.3004,
|
| 2028 |
+
"step": 286
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 0.20297029702970298,
|
| 2032 |
+
"grad_norm": 0.8515625,
|
| 2033 |
+
"learning_rate": 0.0008442771359902366,
|
| 2034 |
+
"loss": 10.317,
|
| 2035 |
+
"step": 287
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.2036775106082037,
|
| 2039 |
+
"grad_norm": 0.86328125,
|
| 2040 |
+
"learning_rate": 0.0008437686780050964,
|
| 2041 |
+
"loss": 10.3256,
|
| 2042 |
+
"step": 288
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 0.2043847241867044,
|
| 2046 |
+
"grad_norm": 0.859375,
|
| 2047 |
+
"learning_rate": 0.0008432580653106389,
|
| 2048 |
+
"loss": 10.3102,
|
| 2049 |
+
"step": 289
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.2050919377652051,
|
| 2053 |
+
"grad_norm": 0.84375,
|
| 2054 |
+
"learning_rate": 0.000842745300700946,
|
| 2055 |
+
"loss": 10.3125,
|
| 2056 |
+
"step": 290
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 0.2057991513437058,
|
| 2060 |
+
"grad_norm": 0.859375,
|
| 2061 |
+
"learning_rate": 0.0008422303869818752,
|
| 2062 |
+
"loss": 10.325,
|
| 2063 |
+
"step": 291
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 0.2065063649222065,
|
| 2067 |
+
"grad_norm": 0.8671875,
|
| 2068 |
+
"learning_rate": 0.0008417133269710432,
|
| 2069 |
+
"loss": 10.3293,
|
| 2070 |
+
"step": 292
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 0.2072135785007072,
|
| 2074 |
+
"grad_norm": 0.8828125,
|
| 2075 |
+
"learning_rate": 0.0008411941234978122,
|
| 2076 |
+
"loss": 10.3085,
|
| 2077 |
+
"step": 293
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 0.2079207920792079,
|
| 2081 |
+
"grad_norm": 0.86328125,
|
| 2082 |
+
"learning_rate": 0.0008406727794032725,
|
| 2083 |
+
"loss": 10.3144,
|
| 2084 |
+
"step": 294
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 0.20862800565770862,
|
| 2088 |
+
"grad_norm": 0.90234375,
|
| 2089 |
+
"learning_rate": 0.0008401492975402288,
|
| 2090 |
+
"loss": 10.3095,
|
| 2091 |
+
"step": 295
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.20933521923620935,
|
| 2095 |
+
"grad_norm": 0.85546875,
|
| 2096 |
+
"learning_rate": 0.0008396236807731831,
|
| 2097 |
+
"loss": 10.3412,
|
| 2098 |
+
"step": 296
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 0.21004243281471005,
|
| 2102 |
+
"grad_norm": 0.859375,
|
| 2103 |
+
"learning_rate": 0.00083909593197832,
|
| 2104 |
+
"loss": 10.3543,
|
| 2105 |
+
"step": 297
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 0.21074964639321075,
|
| 2109 |
+
"grad_norm": 0.86328125,
|
| 2110 |
+
"learning_rate": 0.0008385660540434904,
|
| 2111 |
+
"loss": 10.3612,
|
| 2112 |
+
"step": 298
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 0.21145685997171146,
|
| 2116 |
+
"grad_norm": 0.890625,
|
| 2117 |
+
"learning_rate": 0.0008380340498681957,
|
| 2118 |
+
"loss": 10.3614,
|
| 2119 |
+
"step": 299
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 0.21216407355021216,
|
| 2123 |
+
"grad_norm": 0.94140625,
|
| 2124 |
+
"learning_rate": 0.0008374999223635726,
|
| 2125 |
+
"loss": 10.3179,
|
| 2126 |
+
"step": 300
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.21216407355021216,
|
| 2130 |
+
"eval_loss": 10.361818313598633,
|
| 2131 |
+
"eval_runtime": 1.2522,
|
| 2132 |
+
"eval_samples_per_second": 145.349,
|
| 2133 |
+
"eval_steps_per_second": 18.368,
|
| 2134 |
+
"step": 300
|
| 2135 |
}
|
| 2136 |
],
|
| 2137 |
"logging_steps": 1,
|
|
|
|
| 2151 |
"attributes": {}
|
| 2152 |
}
|
| 2153 |
},
|
| 2154 |
+
"total_flos": 2.2277190057984e+16,
|
| 2155 |
"train_batch_size": 128,
|
| 2156 |
"trial_name": null,
|
| 2157 |
"trial_params": null
|