Commit
·
ea87049
1
Parent(s):
de5d682
Upload folder using huggingface_hub
Browse files- adapter_model.bin +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +735 -3
- training_args.bin +1 -1
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 75641741
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9055d8916d0c01135ce2a001532dec396deba87cc0a07dbf0c98d4dace46d428
|
| 3 |
size 75641741
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 151222021
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e8f78d2228b6babafffa91ef4beb07ff87afac189965b622082f7cfda273dcc
|
| 3 |
size 151222021
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6dd3a816ab8628e6038ecf426e93a907752049203fbc39b63fcde557182a866f
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:518177a2ab7cf6b410a68ee416a0234864cb6c8e1fae811ce6d16681d4149303
|
| 3 |
size 627
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 200,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1487,10 +1487,742 @@
|
|
| 1487 |
"eval_samples_per_second": 0.729,
|
| 1488 |
"eval_steps_per_second": 0.729,
|
| 1489 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1490 |
}
|
| 1491 |
],
|
| 1492 |
"logging_steps": 10,
|
| 1493 |
-
"max_steps":
|
| 1494 |
"num_train_epochs": 1,
|
| 1495 |
"save_steps": 100,
|
| 1496 |
"total_flos": 0.0,
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.2648831203231574,
|
| 5 |
"eval_steps": 200,
|
| 6 |
+
"global_step": 1500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1487 |
"eval_samples_per_second": 0.729,
|
| 1488 |
"eval_steps_per_second": 0.729,
|
| 1489 |
"step": 1000
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 0.18,
|
| 1493 |
+
"learning_rate": 0.0004917027842051741,
|
| 1494 |
+
"logits/chosen": -2.3743691444396973,
|
| 1495 |
+
"logits/rejected": -2.3645715713500977,
|
| 1496 |
+
"logps/chosen": -467.918701171875,
|
| 1497 |
+
"logps/rejected": -392.2208557128906,
|
| 1498 |
+
"loss": 13.2044,
|
| 1499 |
+
"rewards/accuracies": 0.4124999940395355,
|
| 1500 |
+
"rewards/chosen": -34.59047317504883,
|
| 1501 |
+
"rewards/margins": -6.761924743652344,
|
| 1502 |
+
"rewards/rejected": -27.82854652404785,
|
| 1503 |
+
"step": 1010
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 0.18,
|
| 1507 |
+
"learning_rate": 0.0004913043488808868,
|
| 1508 |
+
"logits/chosen": -2.2389774322509766,
|
| 1509 |
+
"logits/rejected": -2.229212760925293,
|
| 1510 |
+
"logps/chosen": -486.40362548828125,
|
| 1511 |
+
"logps/rejected": -460.124267578125,
|
| 1512 |
+
"loss": 12.247,
|
| 1513 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 1514 |
+
"rewards/chosen": -36.01793670654297,
|
| 1515 |
+
"rewards/margins": -2.4814937114715576,
|
| 1516 |
+
"rewards/rejected": -33.53643798828125,
|
| 1517 |
+
"step": 1020
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.18,
|
| 1521 |
+
"learning_rate": 0.0004909379125837757,
|
| 1522 |
+
"logits/chosen": -2.3661270141601562,
|
| 1523 |
+
"logits/rejected": -2.35339093208313,
|
| 1524 |
+
"logps/chosen": -450.0957946777344,
|
| 1525 |
+
"logps/rejected": -419.64453125,
|
| 1526 |
+
"loss": 9.5754,
|
| 1527 |
+
"rewards/accuracies": 0.4625000059604645,
|
| 1528 |
+
"rewards/chosen": -32.51788330078125,
|
| 1529 |
+
"rewards/margins": -2.8007171154022217,
|
| 1530 |
+
"rewards/rejected": -29.717166900634766,
|
| 1531 |
+
"step": 1030
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 0.18,
|
| 1535 |
+
"learning_rate": 0.000490689498823928,
|
| 1536 |
+
"logits/chosen": -2.502885341644287,
|
| 1537 |
+
"logits/rejected": -2.475445032119751,
|
| 1538 |
+
"logps/chosen": -544.1683349609375,
|
| 1539 |
+
"logps/rejected": -447.0438537597656,
|
| 1540 |
+
"loss": 14.0802,
|
| 1541 |
+
"rewards/accuracies": 0.38749998807907104,
|
| 1542 |
+
"rewards/chosen": -40.440181732177734,
|
| 1543 |
+
"rewards/margins": -7.666708946228027,
|
| 1544 |
+
"rewards/rejected": -32.773475646972656,
|
| 1545 |
+
"step": 1040
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.19,
|
| 1549 |
+
"learning_rate": 0.0004903107023416835,
|
| 1550 |
+
"logits/chosen": -2.5913939476013184,
|
| 1551 |
+
"logits/rejected": -2.5680718421936035,
|
| 1552 |
+
"logps/chosen": -486.11907958984375,
|
| 1553 |
+
"logps/rejected": -368.26251220703125,
|
| 1554 |
+
"loss": 13.0584,
|
| 1555 |
+
"rewards/accuracies": 0.32499998807907104,
|
| 1556 |
+
"rewards/chosen": -35.45328140258789,
|
| 1557 |
+
"rewards/margins": -9.003652572631836,
|
| 1558 |
+
"rewards/rejected": -26.449630737304688,
|
| 1559 |
+
"step": 1050
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 0.19,
|
| 1563 |
+
"learning_rate": 0.0004898811381307269,
|
| 1564 |
+
"logits/chosen": -2.5621819496154785,
|
| 1565 |
+
"logits/rejected": -2.5630691051483154,
|
| 1566 |
+
"logps/chosen": -415.2149353027344,
|
| 1567 |
+
"logps/rejected": -372.393798828125,
|
| 1568 |
+
"loss": 10.9069,
|
| 1569 |
+
"rewards/accuracies": 0.48750001192092896,
|
| 1570 |
+
"rewards/chosen": -30.74907875061035,
|
| 1571 |
+
"rewards/margins": -3.5560336112976074,
|
| 1572 |
+
"rewards/rejected": -27.193042755126953,
|
| 1573 |
+
"step": 1060
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 0.19,
|
| 1577 |
+
"learning_rate": 0.0004894424536834149,
|
| 1578 |
+
"logits/chosen": -2.6120645999908447,
|
| 1579 |
+
"logits/rejected": -2.604825019836426,
|
| 1580 |
+
"logps/chosen": -513.4226684570312,
|
| 1581 |
+
"logps/rejected": -459.19647216796875,
|
| 1582 |
+
"loss": 11.3672,
|
| 1583 |
+
"rewards/accuracies": 0.4375,
|
| 1584 |
+
"rewards/chosen": -39.00333023071289,
|
| 1585 |
+
"rewards/margins": -4.482884883880615,
|
| 1586 |
+
"rewards/rejected": -34.520442962646484,
|
| 1587 |
+
"step": 1070
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.19,
|
| 1591 |
+
"learning_rate": 0.000488994665678449,
|
| 1592 |
+
"logits/chosen": -2.880282163619995,
|
| 1593 |
+
"logits/rejected": -2.8795719146728516,
|
| 1594 |
+
"logps/chosen": -424.77874755859375,
|
| 1595 |
+
"logps/rejected": -367.4000244140625,
|
| 1596 |
+
"loss": 11.0388,
|
| 1597 |
+
"rewards/accuracies": 0.4375,
|
| 1598 |
+
"rewards/chosen": -31.353382110595703,
|
| 1599 |
+
"rewards/margins": -5.197685718536377,
|
| 1600 |
+
"rewards/rejected": -26.15569496154785,
|
| 1601 |
+
"step": 1080
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.19,
|
| 1605 |
+
"learning_rate": 0.0004885377911406459,
|
| 1606 |
+
"logits/chosen": -2.947252035140991,
|
| 1607 |
+
"logits/rejected": -2.940441370010376,
|
| 1608 |
+
"logps/chosen": -442.61651611328125,
|
| 1609 |
+
"logps/rejected": -374.4497985839844,
|
| 1610 |
+
"loss": 11.9975,
|
| 1611 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 1612 |
+
"rewards/chosen": -31.82815170288086,
|
| 1613 |
+
"rewards/margins": -5.211056709289551,
|
| 1614 |
+
"rewards/rejected": -26.617095947265625,
|
| 1615 |
+
"step": 1090
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.19,
|
| 1619 |
+
"learning_rate": 0.00048807184744029076,
|
| 1620 |
+
"logits/chosen": -2.9204514026641846,
|
| 1621 |
+
"logits/rejected": -2.922818660736084,
|
| 1622 |
+
"logps/chosen": -416.2978515625,
|
| 1623 |
+
"logps/rejected": -395.44378662109375,
|
| 1624 |
+
"loss": 9.0951,
|
| 1625 |
+
"rewards/accuracies": 0.4749999940395355,
|
| 1626 |
+
"rewards/chosen": -30.16245460510254,
|
| 1627 |
+
"rewards/margins": -1.954272985458374,
|
| 1628 |
+
"rewards/rejected": -28.208179473876953,
|
| 1629 |
+
"step": 1100
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.2,
|
| 1633 |
+
"learning_rate": 0.00048759685229247675,
|
| 1634 |
+
"logits/chosen": -2.950378179550171,
|
| 1635 |
+
"logits/rejected": -2.9522385597229004,
|
| 1636 |
+
"logps/chosen": -456.4190368652344,
|
| 1637 |
+
"logps/rejected": -417.6458435058594,
|
| 1638 |
+
"loss": 10.6616,
|
| 1639 |
+
"rewards/accuracies": 0.44999998807907104,
|
| 1640 |
+
"rewards/chosen": -33.50844192504883,
|
| 1641 |
+
"rewards/margins": -3.696442127227783,
|
| 1642 |
+
"rewards/rejected": -29.811996459960938,
|
| 1643 |
+
"step": 1110
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.2,
|
| 1647 |
+
"learning_rate": 0.0004872103512563103,
|
| 1648 |
+
"logits/chosen": NaN,
|
| 1649 |
+
"logits/rejected": NaN,
|
| 1650 |
+
"logps/chosen": NaN,
|
| 1651 |
+
"logps/rejected": NaN,
|
| 1652 |
+
"loss": 24.7455,
|
| 1653 |
+
"rewards/accuracies": 0.4000000059604645,
|
| 1654 |
+
"rewards/chosen": NaN,
|
| 1655 |
+
"rewards/margins": NaN,
|
| 1656 |
+
"rewards/rejected": NaN,
|
| 1657 |
+
"step": 1120
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.2,
|
| 1661 |
+
"learning_rate": 0.00048716163259071837,
|
| 1662 |
+
"logits/chosen": NaN,
|
| 1663 |
+
"logits/rejected": NaN,
|
| 1664 |
+
"logps/chosen": NaN,
|
| 1665 |
+
"logps/rejected": NaN,
|
| 1666 |
+
"loss": 41.8228,
|
| 1667 |
+
"rewards/accuracies": 0.375,
|
| 1668 |
+
"rewards/chosen": NaN,
|
| 1669 |
+
"rewards/margins": NaN,
|
| 1670 |
+
"rewards/rejected": NaN,
|
| 1671 |
+
"step": 1130
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 0.2,
|
| 1675 |
+
"learning_rate": 0.000487112823756431,
|
| 1676 |
+
"logits/chosen": NaN,
|
| 1677 |
+
"logits/rejected": NaN,
|
| 1678 |
+
"logps/chosen": NaN,
|
| 1679 |
+
"logps/rejected": NaN,
|
| 1680 |
+
"loss": 110.7265,
|
| 1681 |
+
"rewards/accuracies": 0.25,
|
| 1682 |
+
"rewards/chosen": NaN,
|
| 1683 |
+
"rewards/margins": NaN,
|
| 1684 |
+
"rewards/rejected": NaN,
|
| 1685 |
+
"step": 1140
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 0.2,
|
| 1689 |
+
"learning_rate": 0.000487112823756431,
|
| 1690 |
+
"logits/chosen": NaN,
|
| 1691 |
+
"logits/rejected": NaN,
|
| 1692 |
+
"logps/chosen": NaN,
|
| 1693 |
+
"logps/rejected": NaN,
|
| 1694 |
+
"loss": 245.3322,
|
| 1695 |
+
"rewards/accuracies": 0.2750000059604645,
|
| 1696 |
+
"rewards/chosen": NaN,
|
| 1697 |
+
"rewards/margins": NaN,
|
| 1698 |
+
"rewards/rejected": NaN,
|
| 1699 |
+
"step": 1150
|
| 1700 |
+
},
|
| 1701 |
+
{
|
| 1702 |
+
"epoch": 0.2,
|
| 1703 |
+
"learning_rate": 0.000487112823756431,
|
| 1704 |
+
"logits/chosen": NaN,
|
| 1705 |
+
"logits/rejected": NaN,
|
| 1706 |
+
"logps/chosen": NaN,
|
| 1707 |
+
"logps/rejected": NaN,
|
| 1708 |
+
"loss": 1041.9771,
|
| 1709 |
+
"rewards/accuracies": 0.32499998807907104,
|
| 1710 |
+
"rewards/chosen": NaN,
|
| 1711 |
+
"rewards/margins": NaN,
|
| 1712 |
+
"rewards/rejected": NaN,
|
| 1713 |
+
"step": 1160
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"epoch": 0.21,
|
| 1717 |
+
"learning_rate": 0.0004870639247720053,
|
| 1718 |
+
"logits/chosen": NaN,
|
| 1719 |
+
"logits/rejected": NaN,
|
| 1720 |
+
"logps/chosen": NaN,
|
| 1721 |
+
"logps/rejected": NaN,
|
| 1722 |
+
"loss": 437.3346,
|
| 1723 |
+
"rewards/accuracies": 0.3375000059604645,
|
| 1724 |
+
"rewards/chosen": NaN,
|
| 1725 |
+
"rewards/margins": NaN,
|
| 1726 |
+
"rewards/rejected": NaN,
|
| 1727 |
+
"step": 1170
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.21,
|
| 1731 |
+
"learning_rate": 0.0004870639247720053,
|
| 1732 |
+
"logits/chosen": NaN,
|
| 1733 |
+
"logits/rejected": NaN,
|
| 1734 |
+
"logps/chosen": NaN,
|
| 1735 |
+
"logps/rejected": NaN,
|
| 1736 |
+
"loss": 123.7268,
|
| 1737 |
+
"rewards/accuracies": 0.22499999403953552,
|
| 1738 |
+
"rewards/chosen": NaN,
|
| 1739 |
+
"rewards/margins": NaN,
|
| 1740 |
+
"rewards/rejected": NaN,
|
| 1741 |
+
"step": 1180
|
| 1742 |
+
},
|
| 1743 |
+
{
|
| 1744 |
+
"epoch": 0.21,
|
| 1745 |
+
"learning_rate": 0.0004870639247720053,
|
| 1746 |
+
"logits/chosen": NaN,
|
| 1747 |
+
"logits/rejected": NaN,
|
| 1748 |
+
"logps/chosen": NaN,
|
| 1749 |
+
"logps/rejected": NaN,
|
| 1750 |
+
"loss": 315.4085,
|
| 1751 |
+
"rewards/accuracies": 0.23749999701976776,
|
| 1752 |
+
"rewards/chosen": NaN,
|
| 1753 |
+
"rewards/margins": NaN,
|
| 1754 |
+
"rewards/rejected": NaN,
|
| 1755 |
+
"step": 1190
|
| 1756 |
+
},
|
| 1757 |
+
{
|
| 1758 |
+
"epoch": 0.21,
|
| 1759 |
+
"learning_rate": 0.0004870639247720053,
|
| 1760 |
+
"logits/chosen": NaN,
|
| 1761 |
+
"logits/rejected": NaN,
|
| 1762 |
+
"logps/chosen": NaN,
|
| 1763 |
+
"logps/rejected": NaN,
|
| 1764 |
+
"loss": 52542.0625,
|
| 1765 |
+
"rewards/accuracies": 0.25,
|
| 1766 |
+
"rewards/chosen": NaN,
|
| 1767 |
+
"rewards/margins": NaN,
|
| 1768 |
+
"rewards/rejected": NaN,
|
| 1769 |
+
"step": 1200
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.21,
|
| 1773 |
+
"eval_logits/chosen": NaN,
|
| 1774 |
+
"eval_logits/rejected": NaN,
|
| 1775 |
+
"eval_logps/chosen": NaN,
|
| 1776 |
+
"eval_logps/rejected": NaN,
|
| 1777 |
+
"eval_loss": NaN,
|
| 1778 |
+
"eval_rewards/accuracies": 0.2244604378938675,
|
| 1779 |
+
"eval_rewards/chosen": NaN,
|
| 1780 |
+
"eval_rewards/margins": NaN,
|
| 1781 |
+
"eval_rewards/rejected": NaN,
|
| 1782 |
+
"eval_runtime": 988.4543,
|
| 1783 |
+
"eval_samples_per_second": 0.703,
|
| 1784 |
+
"eval_steps_per_second": 0.703,
|
| 1785 |
+
"step": 1200
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 0.21,
|
| 1789 |
+
"learning_rate": 0.0004870149356560326,
|
| 1790 |
+
"logits/chosen": NaN,
|
| 1791 |
+
"logits/rejected": NaN,
|
| 1792 |
+
"logps/chosen": NaN,
|
| 1793 |
+
"logps/rejected": NaN,
|
| 1794 |
+
"loss": 197.8537,
|
| 1795 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 1796 |
+
"rewards/chosen": NaN,
|
| 1797 |
+
"rewards/margins": NaN,
|
| 1798 |
+
"rewards/rejected": NaN,
|
| 1799 |
+
"step": 1210
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 0.22,
|
| 1803 |
+
"learning_rate": 0.0004870149356560326,
|
| 1804 |
+
"logits/chosen": NaN,
|
| 1805 |
+
"logits/rejected": NaN,
|
| 1806 |
+
"logps/chosen": NaN,
|
| 1807 |
+
"logps/rejected": NaN,
|
| 1808 |
+
"loss": 4813.8465,
|
| 1809 |
+
"rewards/accuracies": 0.13750000298023224,
|
| 1810 |
+
"rewards/chosen": NaN,
|
| 1811 |
+
"rewards/margins": NaN,
|
| 1812 |
+
"rewards/rejected": NaN,
|
| 1813 |
+
"step": 1220
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 0.22,
|
| 1817 |
+
"learning_rate": 0.0004870149356560326,
|
| 1818 |
+
"logits/chosen": NaN,
|
| 1819 |
+
"logits/rejected": NaN,
|
| 1820 |
+
"logps/chosen": NaN,
|
| 1821 |
+
"logps/rejected": NaN,
|
| 1822 |
+
"loss": 2048.3082,
|
| 1823 |
+
"rewards/accuracies": 0.25,
|
| 1824 |
+
"rewards/chosen": NaN,
|
| 1825 |
+
"rewards/margins": NaN,
|
| 1826 |
+
"rewards/rejected": NaN,
|
| 1827 |
+
"step": 1230
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 0.22,
|
| 1831 |
+
"learning_rate": 0.0004870149356560326,
|
| 1832 |
+
"logits/chosen": NaN,
|
| 1833 |
+
"logits/rejected": NaN,
|
| 1834 |
+
"logps/chosen": NaN,
|
| 1835 |
+
"logps/rejected": NaN,
|
| 1836 |
+
"loss": 143.0693,
|
| 1837 |
+
"rewards/accuracies": 0.16249999403953552,
|
| 1838 |
+
"rewards/chosen": NaN,
|
| 1839 |
+
"rewards/margins": NaN,
|
| 1840 |
+
"rewards/rejected": NaN,
|
| 1841 |
+
"step": 1240
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 0.22,
|
| 1845 |
+
"learning_rate": 0.0004870149356560326,
|
| 1846 |
+
"logits/chosen": NaN,
|
| 1847 |
+
"logits/rejected": NaN,
|
| 1848 |
+
"logps/chosen": NaN,
|
| 1849 |
+
"logps/rejected": NaN,
|
| 1850 |
+
"loss": 16874.5531,
|
| 1851 |
+
"rewards/accuracies": 0.21250000596046448,
|
| 1852 |
+
"rewards/chosen": NaN,
|
| 1853 |
+
"rewards/margins": NaN,
|
| 1854 |
+
"rewards/rejected": NaN,
|
| 1855 |
+
"step": 1250
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 0.22,
|
| 1859 |
+
"learning_rate": 0.0004870149356560326,
|
| 1860 |
+
"logits/chosen": NaN,
|
| 1861 |
+
"logits/rejected": NaN,
|
| 1862 |
+
"logps/chosen": NaN,
|
| 1863 |
+
"logps/rejected": NaN,
|
| 1864 |
+
"loss": 2509.9437,
|
| 1865 |
+
"rewards/accuracies": 0.17499999701976776,
|
| 1866 |
+
"rewards/chosen": NaN,
|
| 1867 |
+
"rewards/margins": NaN,
|
| 1868 |
+
"rewards/rejected": NaN,
|
| 1869 |
+
"step": 1260
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 0.22,
|
| 1873 |
+
"learning_rate": 0.0004870149356560326,
|
| 1874 |
+
"logits/chosen": NaN,
|
| 1875 |
+
"logits/rejected": NaN,
|
| 1876 |
+
"logps/chosen": NaN,
|
| 1877 |
+
"logps/rejected": NaN,
|
| 1878 |
+
"loss": 9635.2938,
|
| 1879 |
+
"rewards/accuracies": 0.1875,
|
| 1880 |
+
"rewards/chosen": NaN,
|
| 1881 |
+
"rewards/margins": NaN,
|
| 1882 |
+
"rewards/rejected": NaN,
|
| 1883 |
+
"step": 1270
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 0.23,
|
| 1887 |
+
"learning_rate": 0.0004870149356560326,
|
| 1888 |
+
"logits/chosen": NaN,
|
| 1889 |
+
"logits/rejected": NaN,
|
| 1890 |
+
"logps/chosen": NaN,
|
| 1891 |
+
"logps/rejected": NaN,
|
| 1892 |
+
"loss": 3857.9113,
|
| 1893 |
+
"rewards/accuracies": 0.13750000298023224,
|
| 1894 |
+
"rewards/chosen": NaN,
|
| 1895 |
+
"rewards/margins": NaN,
|
| 1896 |
+
"rewards/rejected": NaN,
|
| 1897 |
+
"step": 1280
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 0.23,
|
| 1901 |
+
"learning_rate": 0.0004870149356560326,
|
| 1902 |
+
"logits/chosen": NaN,
|
| 1903 |
+
"logits/rejected": NaN,
|
| 1904 |
+
"logps/chosen": NaN,
|
| 1905 |
+
"logps/rejected": NaN,
|
| 1906 |
+
"loss": 4572.7609,
|
| 1907 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 1908 |
+
"rewards/chosen": NaN,
|
| 1909 |
+
"rewards/margins": NaN,
|
| 1910 |
+
"rewards/rejected": NaN,
|
| 1911 |
+
"step": 1290
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 0.23,
|
| 1915 |
+
"learning_rate": 0.0004870149356560326,
|
| 1916 |
+
"logits/chosen": NaN,
|
| 1917 |
+
"logits/rejected": NaN,
|
| 1918 |
+
"logps/chosen": NaN,
|
| 1919 |
+
"logps/rejected": NaN,
|
| 1920 |
+
"loss": 606.0347,
|
| 1921 |
+
"rewards/accuracies": 0.15000000596046448,
|
| 1922 |
+
"rewards/chosen": NaN,
|
| 1923 |
+
"rewards/margins": NaN,
|
| 1924 |
+
"rewards/rejected": NaN,
|
| 1925 |
+
"step": 1300
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 0.23,
|
| 1929 |
+
"learning_rate": 0.0004870149356560326,
|
| 1930 |
+
"logits/chosen": NaN,
|
| 1931 |
+
"logits/rejected": NaN,
|
| 1932 |
+
"logps/chosen": NaN,
|
| 1933 |
+
"logps/rejected": NaN,
|
| 1934 |
+
"loss": 1759.0598,
|
| 1935 |
+
"rewards/accuracies": 0.17499999701976776,
|
| 1936 |
+
"rewards/chosen": NaN,
|
| 1937 |
+
"rewards/margins": NaN,
|
| 1938 |
+
"rewards/rejected": NaN,
|
| 1939 |
+
"step": 1310
|
| 1940 |
+
},
|
| 1941 |
+
{
|
| 1942 |
+
"epoch": 0.23,
|
| 1943 |
+
"learning_rate": 0.0004870149356560326,
|
| 1944 |
+
"logits/chosen": NaN,
|
| 1945 |
+
"logits/rejected": NaN,
|
| 1946 |
+
"logps/chosen": NaN,
|
| 1947 |
+
"logps/rejected": NaN,
|
| 1948 |
+
"loss": 14436.2953,
|
| 1949 |
+
"rewards/accuracies": 0.21250000596046448,
|
| 1950 |
+
"rewards/chosen": NaN,
|
| 1951 |
+
"rewards/margins": NaN,
|
| 1952 |
+
"rewards/rejected": NaN,
|
| 1953 |
+
"step": 1320
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 0.23,
|
| 1957 |
+
"learning_rate": 0.0004870149356560326,
|
| 1958 |
+
"logits/chosen": NaN,
|
| 1959 |
+
"logits/rejected": NaN,
|
| 1960 |
+
"logps/chosen": NaN,
|
| 1961 |
+
"logps/rejected": NaN,
|
| 1962 |
+
"loss": 386264.375,
|
| 1963 |
+
"rewards/accuracies": 0.0625,
|
| 1964 |
+
"rewards/chosen": NaN,
|
| 1965 |
+
"rewards/margins": NaN,
|
| 1966 |
+
"rewards/rejected": NaN,
|
| 1967 |
+
"step": 1330
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 0.24,
|
| 1971 |
+
"learning_rate": 0.0004870149356560326,
|
| 1972 |
+
"logits/chosen": NaN,
|
| 1973 |
+
"logits/rejected": NaN,
|
| 1974 |
+
"logps/chosen": NaN,
|
| 1975 |
+
"logps/rejected": NaN,
|
| 1976 |
+
"loss": 8430.657,
|
| 1977 |
+
"rewards/accuracies": 0.1875,
|
| 1978 |
+
"rewards/chosen": NaN,
|
| 1979 |
+
"rewards/margins": NaN,
|
| 1980 |
+
"rewards/rejected": NaN,
|
| 1981 |
+
"step": 1340
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.24,
|
| 1985 |
+
"learning_rate": 0.0004870149356560326,
|
| 1986 |
+
"logits/chosen": NaN,
|
| 1987 |
+
"logits/rejected": NaN,
|
| 1988 |
+
"logps/chosen": NaN,
|
| 1989 |
+
"logps/rejected": NaN,
|
| 1990 |
+
"loss": 6939.275,
|
| 1991 |
+
"rewards/accuracies": 0.13750000298023224,
|
| 1992 |
+
"rewards/chosen": NaN,
|
| 1993 |
+
"rewards/margins": NaN,
|
| 1994 |
+
"rewards/rejected": NaN,
|
| 1995 |
+
"step": 1350
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.24,
|
| 1999 |
+
"learning_rate": 0.0004870149356560326,
|
| 2000 |
+
"logits/chosen": NaN,
|
| 2001 |
+
"logits/rejected": NaN,
|
| 2002 |
+
"logps/chosen": NaN,
|
| 2003 |
+
"logps/rejected": NaN,
|
| 2004 |
+
"loss": 4295.0949,
|
| 2005 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 2006 |
+
"rewards/chosen": NaN,
|
| 2007 |
+
"rewards/margins": NaN,
|
| 2008 |
+
"rewards/rejected": NaN,
|
| 2009 |
+
"step": 1360
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.24,
|
| 2013 |
+
"learning_rate": 0.0004870149356560326,
|
| 2014 |
+
"logits/chosen": NaN,
|
| 2015 |
+
"logits/rejected": NaN,
|
| 2016 |
+
"logps/chosen": NaN,
|
| 2017 |
+
"logps/rejected": NaN,
|
| 2018 |
+
"loss": 17283.3672,
|
| 2019 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 2020 |
+
"rewards/chosen": NaN,
|
| 2021 |
+
"rewards/margins": NaN,
|
| 2022 |
+
"rewards/rejected": NaN,
|
| 2023 |
+
"step": 1370
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.24,
|
| 2027 |
+
"learning_rate": 0.0004870149356560326,
|
| 2028 |
+
"logits/chosen": NaN,
|
| 2029 |
+
"logits/rejected": NaN,
|
| 2030 |
+
"logps/chosen": NaN,
|
| 2031 |
+
"logps/rejected": NaN,
|
| 2032 |
+
"loss": 24895.7469,
|
| 2033 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 2034 |
+
"rewards/chosen": NaN,
|
| 2035 |
+
"rewards/margins": NaN,
|
| 2036 |
+
"rewards/rejected": NaN,
|
| 2037 |
+
"step": 1380
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.25,
|
| 2041 |
+
"learning_rate": 0.0004870149356560326,
|
| 2042 |
+
"logits/chosen": NaN,
|
| 2043 |
+
"logits/rejected": NaN,
|
| 2044 |
+
"logps/chosen": NaN,
|
| 2045 |
+
"logps/rejected": NaN,
|
| 2046 |
+
"loss": 45136.4094,
|
| 2047 |
+
"rewards/accuracies": 0.17499999701976776,
|
| 2048 |
+
"rewards/chosen": NaN,
|
| 2049 |
+
"rewards/margins": NaN,
|
| 2050 |
+
"rewards/rejected": NaN,
|
| 2051 |
+
"step": 1390
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.25,
|
| 2055 |
+
"learning_rate": 0.0004870149356560326,
|
| 2056 |
+
"logits/chosen": NaN,
|
| 2057 |
+
"logits/rejected": NaN,
|
| 2058 |
+
"logps/chosen": NaN,
|
| 2059 |
+
"logps/rejected": NaN,
|
| 2060 |
+
"loss": 165.0189,
|
| 2061 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 2062 |
+
"rewards/chosen": NaN,
|
| 2063 |
+
"rewards/margins": NaN,
|
| 2064 |
+
"rewards/rejected": NaN,
|
| 2065 |
+
"step": 1400
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.25,
|
| 2069 |
+
"eval_logits/chosen": NaN,
|
| 2070 |
+
"eval_logits/rejected": NaN,
|
| 2071 |
+
"eval_logps/chosen": NaN,
|
| 2072 |
+
"eval_logps/rejected": NaN,
|
| 2073 |
+
"eval_loss": NaN,
|
| 2074 |
+
"eval_rewards/accuracies": 0.17553956806659698,
|
| 2075 |
+
"eval_rewards/chosen": NaN,
|
| 2076 |
+
"eval_rewards/margins": NaN,
|
| 2077 |
+
"eval_rewards/rejected": NaN,
|
| 2078 |
+
"eval_runtime": 987.1061,
|
| 2079 |
+
"eval_samples_per_second": 0.704,
|
| 2080 |
+
"eval_steps_per_second": 0.704,
|
| 2081 |
+
"step": 1400
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.25,
|
| 2085 |
+
"learning_rate": 0.0004870149356560326,
|
| 2086 |
+
"logits/chosen": NaN,
|
| 2087 |
+
"logits/rejected": NaN,
|
| 2088 |
+
"logps/chosen": NaN,
|
| 2089 |
+
"logps/rejected": NaN,
|
| 2090 |
+
"loss": 8688.7203,
|
| 2091 |
+
"rewards/accuracies": 0.22499999403953552,
|
| 2092 |
+
"rewards/chosen": NaN,
|
| 2093 |
+
"rewards/margins": NaN,
|
| 2094 |
+
"rewards/rejected": NaN,
|
| 2095 |
+
"step": 1410
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.25,
|
| 2099 |
+
"learning_rate": 0.0004870149356560326,
|
| 2100 |
+
"logits/chosen": NaN,
|
| 2101 |
+
"logits/rejected": NaN,
|
| 2102 |
+
"logps/chosen": NaN,
|
| 2103 |
+
"logps/rejected": NaN,
|
| 2104 |
+
"loss": 9484.9836,
|
| 2105 |
+
"rewards/accuracies": 0.1875,
|
| 2106 |
+
"rewards/chosen": NaN,
|
| 2107 |
+
"rewards/margins": NaN,
|
| 2108 |
+
"rewards/rejected": NaN,
|
| 2109 |
+
"step": 1420
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.25,
|
| 2113 |
+
"learning_rate": 0.0004870149356560326,
|
| 2114 |
+
"logits/chosen": NaN,
|
| 2115 |
+
"logits/rejected": NaN,
|
| 2116 |
+
"logps/chosen": NaN,
|
| 2117 |
+
"logps/rejected": NaN,
|
| 2118 |
+
"loss": 9308.6922,
|
| 2119 |
+
"rewards/accuracies": 0.17499999701976776,
|
| 2120 |
+
"rewards/chosen": NaN,
|
| 2121 |
+
"rewards/margins": NaN,
|
| 2122 |
+
"rewards/rejected": NaN,
|
| 2123 |
+
"step": 1430
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.25,
|
| 2127 |
+
"learning_rate": 0.0004870149356560326,
|
| 2128 |
+
"logits/chosen": NaN,
|
| 2129 |
+
"logits/rejected": NaN,
|
| 2130 |
+
"logps/chosen": NaN,
|
| 2131 |
+
"logps/rejected": NaN,
|
| 2132 |
+
"loss": 2905.9373,
|
| 2133 |
+
"rewards/accuracies": 0.22499999403953552,
|
| 2134 |
+
"rewards/chosen": NaN,
|
| 2135 |
+
"rewards/margins": NaN,
|
| 2136 |
+
"rewards/rejected": NaN,
|
| 2137 |
+
"step": 1440
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.26,
|
| 2141 |
+
"learning_rate": 0.0004870149356560326,
|
| 2142 |
+
"logits/chosen": NaN,
|
| 2143 |
+
"logits/rejected": NaN,
|
| 2144 |
+
"logps/chosen": NaN,
|
| 2145 |
+
"logps/rejected": NaN,
|
| 2146 |
+
"loss": 4738.7867,
|
| 2147 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 2148 |
+
"rewards/chosen": NaN,
|
| 2149 |
+
"rewards/margins": NaN,
|
| 2150 |
+
"rewards/rejected": NaN,
|
| 2151 |
+
"step": 1450
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.26,
|
| 2155 |
+
"learning_rate": 0.0004870149356560326,
|
| 2156 |
+
"logits/chosen": NaN,
|
| 2157 |
+
"logits/rejected": NaN,
|
| 2158 |
+
"logps/chosen": NaN,
|
| 2159 |
+
"logps/rejected": NaN,
|
| 2160 |
+
"loss": 424.0728,
|
| 2161 |
+
"rewards/accuracies": 0.20000000298023224,
|
| 2162 |
+
"rewards/chosen": NaN,
|
| 2163 |
+
"rewards/margins": NaN,
|
| 2164 |
+
"rewards/rejected": NaN,
|
| 2165 |
+
"step": 1460
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.26,
|
| 2169 |
+
"learning_rate": 0.0004870149356560326,
|
| 2170 |
+
"logits/chosen": NaN,
|
| 2171 |
+
"logits/rejected": NaN,
|
| 2172 |
+
"logps/chosen": NaN,
|
| 2173 |
+
"logps/rejected": NaN,
|
| 2174 |
+
"loss": 2181.5133,
|
| 2175 |
+
"rewards/accuracies": 0.13750000298023224,
|
| 2176 |
+
"rewards/chosen": NaN,
|
| 2177 |
+
"rewards/margins": NaN,
|
| 2178 |
+
"rewards/rejected": NaN,
|
| 2179 |
+
"step": 1470
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.26,
|
| 2183 |
+
"learning_rate": 0.0004870149356560326,
|
| 2184 |
+
"logits/chosen": NaN,
|
| 2185 |
+
"logits/rejected": NaN,
|
| 2186 |
+
"logps/chosen": NaN,
|
| 2187 |
+
"logps/rejected": NaN,
|
| 2188 |
+
"loss": 23597.5219,
|
| 2189 |
+
"rewards/accuracies": 0.16249999403953552,
|
| 2190 |
+
"rewards/chosen": NaN,
|
| 2191 |
+
"rewards/margins": NaN,
|
| 2192 |
+
"rewards/rejected": NaN,
|
| 2193 |
+
"step": 1480
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.26,
|
| 2197 |
+
"learning_rate": 0.0004870149356560326,
|
| 2198 |
+
"logits/chosen": NaN,
|
| 2199 |
+
"logits/rejected": NaN,
|
| 2200 |
+
"logps/chosen": NaN,
|
| 2201 |
+
"logps/rejected": NaN,
|
| 2202 |
+
"loss": 1259.9596,
|
| 2203 |
+
"rewards/accuracies": 0.25,
|
| 2204 |
+
"rewards/chosen": NaN,
|
| 2205 |
+
"rewards/margins": NaN,
|
| 2206 |
+
"rewards/rejected": NaN,
|
| 2207 |
+
"step": 1490
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.26,
|
| 2211 |
+
"learning_rate": 0.0004870149356560326,
|
| 2212 |
+
"logits/chosen": NaN,
|
| 2213 |
+
"logits/rejected": NaN,
|
| 2214 |
+
"logps/chosen": NaN,
|
| 2215 |
+
"logps/rejected": NaN,
|
| 2216 |
+
"loss": 7475.4719,
|
| 2217 |
+
"rewards/accuracies": 0.21250000596046448,
|
| 2218 |
+
"rewards/chosen": NaN,
|
| 2219 |
+
"rewards/margins": NaN,
|
| 2220 |
+
"rewards/rejected": NaN,
|
| 2221 |
+
"step": 1500
|
| 2222 |
}
|
| 2223 |
],
|
| 2224 |
"logging_steps": 10,
|
| 2225 |
+
"max_steps": 5662,
|
| 2226 |
"num_train_epochs": 1,
|
| 2227 |
"save_steps": 100,
|
| 2228 |
"total_flos": 0.0,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4091
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:396a28579afc547a6f46654c7646d9a3770584a235915799960ff8d28e97bcac
|
| 3 |
size 4091
|