Training in progress, epoch 3, 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 3132668808
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:025890d5356002da0350524acea1e028a4c5f1d1f7cc3bee2cc13262f0621eeb
|
| 3 |
size 3132668808
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6265677800
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7f356132a63ab1bc73486a7f9c3ebd47f727e216f286211b35fc1d1a53f3277
|
| 3 |
size 6265677800
|
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:c1ebb3e8ecd863784104abdee43036b081c962dae091b76d5c1db6131e386d91
|
| 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 sha256:2517fb96a777d41d764ad3dee01acb528730d0eb485080f169d6588b17f93ab7
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1677,13 +1677,854 @@
|
|
| 1677 |
"eval_steps_per_second": 0.491,
|
| 1678 |
"eval_translation_length": 7761920,
|
| 1679 |
"step": 136438
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1680 |
}
|
| 1681 |
],
|
| 1682 |
"logging_steps": 500,
|
| 1683 |
"max_steps": 341095,
|
| 1684 |
"num_train_epochs": 5,
|
| 1685 |
"save_steps": 500,
|
| 1686 |
-
"total_flos":
|
| 1687 |
"trial_name": null,
|
| 1688 |
"trial_params": null
|
| 1689 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 204657,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1677 |
"eval_steps_per_second": 0.491,
|
| 1678 |
"eval_translation_length": 7761920,
|
| 1679 |
"step": 136438
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 2.0,
|
| 1683 |
+
"learning_rate": 6.555520384510561e-05,
|
| 1684 |
+
"loss": 2.8434,
|
| 1685 |
+
"step": 136500
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 2.01,
|
| 1689 |
+
"learning_rate": 6.533588640408162e-05,
|
| 1690 |
+
"loss": 2.4374,
|
| 1691 |
+
"step": 137000
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 2.02,
|
| 1695 |
+
"learning_rate": 6.511624277254071e-05,
|
| 1696 |
+
"loss": 2.421,
|
| 1697 |
+
"step": 137500
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 2.02,
|
| 1701 |
+
"learning_rate": 6.48962776222487e-05,
|
| 1702 |
+
"loss": 2.4327,
|
| 1703 |
+
"step": 138000
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 2.03,
|
| 1707 |
+
"learning_rate": 6.467599563181004e-05,
|
| 1708 |
+
"loss": 2.4567,
|
| 1709 |
+
"step": 138500
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 2.04,
|
| 1713 |
+
"learning_rate": 6.445540148656825e-05,
|
| 1714 |
+
"loss": 2.3828,
|
| 1715 |
+
"step": 139000
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 2.04,
|
| 1719 |
+
"learning_rate": 6.423449987850634e-05,
|
| 1720 |
+
"loss": 2.4582,
|
| 1721 |
+
"step": 139500
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 2.05,
|
| 1725 |
+
"learning_rate": 6.401329550614694e-05,
|
| 1726 |
+
"loss": 2.4376,
|
| 1727 |
+
"step": 140000
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 2.06,
|
| 1731 |
+
"learning_rate": 6.379179307445245e-05,
|
| 1732 |
+
"loss": 2.4871,
|
| 1733 |
+
"step": 140500
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 2.07,
|
| 1737 |
+
"learning_rate": 6.35699972947249e-05,
|
| 1738 |
+
"loss": 2.4955,
|
| 1739 |
+
"step": 141000
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 2.07,
|
| 1743 |
+
"learning_rate": 6.334791288450577e-05,
|
| 1744 |
+
"loss": 2.5095,
|
| 1745 |
+
"step": 141500
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 2.08,
|
| 1749 |
+
"learning_rate": 6.312554456747558e-05,
|
| 1750 |
+
"loss": 2.4246,
|
| 1751 |
+
"step": 142000
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 2.09,
|
| 1755 |
+
"learning_rate": 6.290289707335355e-05,
|
| 1756 |
+
"loss": 2.567,
|
| 1757 |
+
"step": 142500
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 2.1,
|
| 1761 |
+
"learning_rate": 6.26799751377969e-05,
|
| 1762 |
+
"loss": 2.4974,
|
| 1763 |
+
"step": 143000
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 2.1,
|
| 1767 |
+
"learning_rate": 6.245678350230015e-05,
|
| 1768 |
+
"loss": 2.4016,
|
| 1769 |
+
"step": 143500
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 2.11,
|
| 1773 |
+
"learning_rate": 6.223332691409428e-05,
|
| 1774 |
+
"loss": 2.4861,
|
| 1775 |
+
"step": 144000
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 2.12,
|
| 1779 |
+
"learning_rate": 6.200961012604575e-05,
|
| 1780 |
+
"loss": 2.4795,
|
| 1781 |
+
"step": 144500
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 2.13,
|
| 1785 |
+
"learning_rate": 6.178563789655537e-05,
|
| 1786 |
+
"loss": 2.5037,
|
| 1787 |
+
"step": 145000
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 2.13,
|
| 1791 |
+
"learning_rate": 6.156141498945719e-05,
|
| 1792 |
+
"loss": 2.4686,
|
| 1793 |
+
"step": 145500
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 2.14,
|
| 1797 |
+
"learning_rate": 6.133694617391704e-05,
|
| 1798 |
+
"loss": 2.5575,
|
| 1799 |
+
"step": 146000
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 2.15,
|
| 1803 |
+
"learning_rate": 6.111223622433124e-05,
|
| 1804 |
+
"loss": 2.5201,
|
| 1805 |
+
"step": 146500
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 2.15,
|
| 1809 |
+
"learning_rate": 6.088728992022491e-05,
|
| 1810 |
+
"loss": 2.4335,
|
| 1811 |
+
"step": 147000
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 2.16,
|
| 1815 |
+
"learning_rate": 6.0662112046150376e-05,
|
| 1816 |
+
"loss": 2.4387,
|
| 1817 |
+
"step": 147500
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 2.17,
|
| 1821 |
+
"learning_rate": 6.043670739158542e-05,
|
| 1822 |
+
"loss": 2.5027,
|
| 1823 |
+
"step": 148000
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 2.18,
|
| 1827 |
+
"learning_rate": 6.021108075083142e-05,
|
| 1828 |
+
"loss": 2.4055,
|
| 1829 |
+
"step": 148500
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 2.18,
|
| 1833 |
+
"learning_rate": 5.998523692291127e-05,
|
| 1834 |
+
"loss": 2.478,
|
| 1835 |
+
"step": 149000
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 2.19,
|
| 1839 |
+
"learning_rate": 5.9759180711467446e-05,
|
| 1840 |
+
"loss": 2.4491,
|
| 1841 |
+
"step": 149500
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 2.2,
|
| 1845 |
+
"learning_rate": 5.953291692465974e-05,
|
| 1846 |
+
"loss": 2.5007,
|
| 1847 |
+
"step": 150000
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 2.21,
|
| 1851 |
+
"learning_rate": 5.930645037506301e-05,
|
| 1852 |
+
"loss": 2.4752,
|
| 1853 |
+
"step": 150500
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 2.21,
|
| 1857 |
+
"learning_rate": 5.9079785879564876e-05,
|
| 1858 |
+
"loss": 2.4569,
|
| 1859 |
+
"step": 151000
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 2.22,
|
| 1863 |
+
"learning_rate": 5.885292825926314e-05,
|
| 1864 |
+
"loss": 2.441,
|
| 1865 |
+
"step": 151500
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 2.23,
|
| 1869 |
+
"learning_rate": 5.862588233936341e-05,
|
| 1870 |
+
"loss": 2.4172,
|
| 1871 |
+
"step": 152000
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 2.24,
|
| 1875 |
+
"learning_rate": 5.8398652949076324e-05,
|
| 1876 |
+
"loss": 2.4053,
|
| 1877 |
+
"step": 152500
|
| 1878 |
+
},
|
| 1879 |
+
{
|
| 1880 |
+
"epoch": 2.24,
|
| 1881 |
+
"learning_rate": 5.81712449215149e-05,
|
| 1882 |
+
"loss": 2.5097,
|
| 1883 |
+
"step": 153000
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 2.25,
|
| 1887 |
+
"learning_rate": 5.794366309359174e-05,
|
| 1888 |
+
"loss": 2.4697,
|
| 1889 |
+
"step": 153500
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 2.26,
|
| 1893 |
+
"learning_rate": 5.771591230591612e-05,
|
| 1894 |
+
"loss": 2.5604,
|
| 1895 |
+
"step": 154000
|
| 1896 |
+
},
|
| 1897 |
+
{
|
| 1898 |
+
"epoch": 2.26,
|
| 1899 |
+
"learning_rate": 5.748799740269104e-05,
|
| 1900 |
+
"loss": 2.3954,
|
| 1901 |
+
"step": 154500
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 2.27,
|
| 1905 |
+
"learning_rate": 5.7259923231610236e-05,
|
| 1906 |
+
"loss": 2.4581,
|
| 1907 |
+
"step": 155000
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 2.28,
|
| 1911 |
+
"learning_rate": 5.703169464375498e-05,
|
| 1912 |
+
"loss": 2.5286,
|
| 1913 |
+
"step": 155500
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 2.29,
|
| 1917 |
+
"learning_rate": 5.680331649349101e-05,
|
| 1918 |
+
"loss": 2.3529,
|
| 1919 |
+
"step": 156000
|
| 1920 |
+
},
|
| 1921 |
+
{
|
| 1922 |
+
"epoch": 2.29,
|
| 1923 |
+
"learning_rate": 5.657479363836519e-05,
|
| 1924 |
+
"loss": 2.5152,
|
| 1925 |
+
"step": 156500
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 2.3,
|
| 1929 |
+
"learning_rate": 5.6346130939002193e-05,
|
| 1930 |
+
"loss": 2.512,
|
| 1931 |
+
"step": 157000
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 2.31,
|
| 1935 |
+
"learning_rate": 5.6117333259001183e-05,
|
| 1936 |
+
"loss": 2.5408,
|
| 1937 |
+
"step": 157500
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 2.32,
|
| 1941 |
+
"learning_rate": 5.5888405464832314e-05,
|
| 1942 |
+
"loss": 2.5028,
|
| 1943 |
+
"step": 158000
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 2.32,
|
| 1947 |
+
"learning_rate": 5.565935242573323e-05,
|
| 1948 |
+
"loss": 2.5252,
|
| 1949 |
+
"step": 158500
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 2.33,
|
| 1953 |
+
"learning_rate": 5.5430179013605544e-05,
|
| 1954 |
+
"loss": 2.4696,
|
| 1955 |
+
"step": 159000
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 2.34,
|
| 1959 |
+
"learning_rate": 5.5200890102911096e-05,
|
| 1960 |
+
"loss": 2.4254,
|
| 1961 |
+
"step": 159500
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 2.35,
|
| 1965 |
+
"learning_rate": 5.497149057056843e-05,
|
| 1966 |
+
"loss": 2.346,
|
| 1967 |
+
"step": 160000
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 2.35,
|
| 1971 |
+
"learning_rate": 5.474198529584896e-05,
|
| 1972 |
+
"loss": 2.4914,
|
| 1973 |
+
"step": 160500
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 2.36,
|
| 1977 |
+
"learning_rate": 5.451237916027319e-05,
|
| 1978 |
+
"loss": 2.5014,
|
| 1979 |
+
"step": 161000
|
| 1980 |
+
},
|
| 1981 |
+
{
|
| 1982 |
+
"epoch": 2.37,
|
| 1983 |
+
"learning_rate": 5.428267704750691e-05,
|
| 1984 |
+
"loss": 2.4718,
|
| 1985 |
+
"step": 161500
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 2.37,
|
| 1989 |
+
"learning_rate": 5.4052883843257375e-05,
|
| 1990 |
+
"loss": 2.4263,
|
| 1991 |
+
"step": 162000
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.38,
|
| 1995 |
+
"learning_rate": 5.3823004435169276e-05,
|
| 1996 |
+
"loss": 2.5898,
|
| 1997 |
+
"step": 162500
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 2.39,
|
| 2001 |
+
"learning_rate": 5.3593043712720835e-05,
|
| 2002 |
+
"loss": 2.5174,
|
| 2003 |
+
"step": 163000
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 2.4,
|
| 2007 |
+
"learning_rate": 5.336300656711989e-05,
|
| 2008 |
+
"loss": 2.5126,
|
| 2009 |
+
"step": 163500
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 2.4,
|
| 2013 |
+
"learning_rate": 5.3132897891199684e-05,
|
| 2014 |
+
"loss": 2.4919,
|
| 2015 |
+
"step": 164000
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 2.41,
|
| 2019 |
+
"learning_rate": 5.290272257931495e-05,
|
| 2020 |
+
"loss": 2.4545,
|
| 2021 |
+
"step": 164500
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 2.42,
|
| 2025 |
+
"learning_rate": 5.267248552723771e-05,
|
| 2026 |
+
"loss": 2.5104,
|
| 2027 |
+
"step": 165000
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 2.43,
|
| 2031 |
+
"learning_rate": 5.244219163205324e-05,
|
| 2032 |
+
"loss": 2.464,
|
| 2033 |
+
"step": 165500
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 2.43,
|
| 2037 |
+
"learning_rate": 5.22118457920558e-05,
|
| 2038 |
+
"loss": 2.4752,
|
| 2039 |
+
"step": 166000
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.44,
|
| 2043 |
+
"learning_rate": 5.198145290664456e-05,
|
| 2044 |
+
"loss": 2.5599,
|
| 2045 |
+
"step": 166500
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 2.45,
|
| 2049 |
+
"learning_rate": 5.1751017876219256e-05,
|
| 2050 |
+
"loss": 2.4875,
|
| 2051 |
+
"step": 167000
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 2.46,
|
| 2055 |
+
"learning_rate": 5.15205456020761e-05,
|
| 2056 |
+
"loss": 2.4102,
|
| 2057 |
+
"step": 167500
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 2.46,
|
| 2061 |
+
"learning_rate": 5.129004098630348e-05,
|
| 2062 |
+
"loss": 2.4686,
|
| 2063 |
+
"step": 168000
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 2.47,
|
| 2067 |
+
"learning_rate": 5.105950893167761e-05,
|
| 2068 |
+
"loss": 2.5395,
|
| 2069 |
+
"step": 168500
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 2.48,
|
| 2073 |
+
"learning_rate": 5.082895434155841e-05,
|
| 2074 |
+
"loss": 2.4642,
|
| 2075 |
+
"step": 169000
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 2.48,
|
| 2079 |
+
"learning_rate": 5.059838211978506e-05,
|
| 2080 |
+
"loss": 2.6105,
|
| 2081 |
+
"step": 169500
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.49,
|
| 2085 |
+
"learning_rate": 5.0367797170571776e-05,
|
| 2086 |
+
"loss": 2.5126,
|
| 2087 |
+
"step": 170000
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 2.5,
|
| 2091 |
+
"learning_rate": 5.013720439840348e-05,
|
| 2092 |
+
"loss": 2.5685,
|
| 2093 |
+
"step": 170500
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 2.51,
|
| 2097 |
+
"learning_rate": 4.9906608707931526e-05,
|
| 2098 |
+
"loss": 2.4634,
|
| 2099 |
+
"step": 171000
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 2.51,
|
| 2103 |
+
"learning_rate": 4.967601500386929e-05,
|
| 2104 |
+
"loss": 2.4449,
|
| 2105 |
+
"step": 171500
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 2.52,
|
| 2109 |
+
"learning_rate": 4.9445428190887924e-05,
|
| 2110 |
+
"loss": 2.515,
|
| 2111 |
+
"step": 172000
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 2.53,
|
| 2115 |
+
"learning_rate": 4.9214853173511975e-05,
|
| 2116 |
+
"loss": 2.5057,
|
| 2117 |
+
"step": 172500
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 2.54,
|
| 2121 |
+
"learning_rate": 4.8984294856015164e-05,
|
| 2122 |
+
"loss": 2.5329,
|
| 2123 |
+
"step": 173000
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 2.54,
|
| 2127 |
+
"learning_rate": 4.875375814231596e-05,
|
| 2128 |
+
"loss": 2.4823,
|
| 2129 |
+
"step": 173500
|
| 2130 |
+
},
|
| 2131 |
+
{
|
| 2132 |
+
"epoch": 2.55,
|
| 2133 |
+
"learning_rate": 4.852324793587333e-05,
|
| 2134 |
+
"loss": 2.4907,
|
| 2135 |
+
"step": 174000
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 2.56,
|
| 2139 |
+
"learning_rate": 4.829276913958245e-05,
|
| 2140 |
+
"loss": 2.523,
|
| 2141 |
+
"step": 174500
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 2.57,
|
| 2145 |
+
"learning_rate": 4.80623266556704e-05,
|
| 2146 |
+
"loss": 2.3798,
|
| 2147 |
+
"step": 175000
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 2.57,
|
| 2151 |
+
"learning_rate": 4.783192538559191e-05,
|
| 2152 |
+
"loss": 2.5055,
|
| 2153 |
+
"step": 175500
|
| 2154 |
+
},
|
| 2155 |
+
{
|
| 2156 |
+
"epoch": 2.58,
|
| 2157 |
+
"learning_rate": 4.760157022992511e-05,
|
| 2158 |
+
"loss": 2.542,
|
| 2159 |
+
"step": 176000
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 2.59,
|
| 2163 |
+
"learning_rate": 4.737126608826727e-05,
|
| 2164 |
+
"loss": 2.5768,
|
| 2165 |
+
"step": 176500
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 2.59,
|
| 2169 |
+
"learning_rate": 4.7141017859130605e-05,
|
| 2170 |
+
"loss": 2.5816,
|
| 2171 |
+
"step": 177000
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 2.6,
|
| 2175 |
+
"learning_rate": 4.6910830439838114e-05,
|
| 2176 |
+
"loss": 2.4947,
|
| 2177 |
+
"step": 177500
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 2.61,
|
| 2181 |
+
"learning_rate": 4.668070872641934e-05,
|
| 2182 |
+
"loss": 2.5179,
|
| 2183 |
+
"step": 178000
|
| 2184 |
+
},
|
| 2185 |
+
{
|
| 2186 |
+
"epoch": 2.62,
|
| 2187 |
+
"learning_rate": 4.645065761350634e-05,
|
| 2188 |
+
"loss": 2.5035,
|
| 2189 |
+
"step": 178500
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 2.62,
|
| 2193 |
+
"learning_rate": 4.622068199422945e-05,
|
| 2194 |
+
"loss": 2.4886,
|
| 2195 |
+
"step": 179000
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 2.63,
|
| 2199 |
+
"learning_rate": 4.59907867601133e-05,
|
| 2200 |
+
"loss": 2.434,
|
| 2201 |
+
"step": 179500
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.64,
|
| 2205 |
+
"learning_rate": 4.576097680097277e-05,
|
| 2206 |
+
"loss": 2.4869,
|
| 2207 |
+
"step": 180000
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 2.65,
|
| 2211 |
+
"learning_rate": 4.553125700480892e-05,
|
| 2212 |
+
"loss": 2.4565,
|
| 2213 |
+
"step": 180500
|
| 2214 |
+
},
|
| 2215 |
+
{
|
| 2216 |
+
"epoch": 2.65,
|
| 2217 |
+
"learning_rate": 4.530163225770513e-05,
|
| 2218 |
+
"loss": 2.496,
|
| 2219 |
+
"step": 181000
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 2.66,
|
| 2223 |
+
"learning_rate": 4.5072107443723025e-05,
|
| 2224 |
+
"loss": 2.415,
|
| 2225 |
+
"step": 181500
|
| 2226 |
+
},
|
| 2227 |
+
{
|
| 2228 |
+
"epoch": 2.67,
|
| 2229 |
+
"learning_rate": 4.484268744479875e-05,
|
| 2230 |
+
"loss": 2.4461,
|
| 2231 |
+
"step": 182000
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 2.68,
|
| 2235 |
+
"learning_rate": 4.461337714063901e-05,
|
| 2236 |
+
"loss": 2.4858,
|
| 2237 |
+
"step": 182500
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 2.68,
|
| 2241 |
+
"learning_rate": 4.438418140861738e-05,
|
| 2242 |
+
"loss": 2.5667,
|
| 2243 |
+
"step": 183000
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 2.69,
|
| 2247 |
+
"learning_rate": 4.415510512367047e-05,
|
| 2248 |
+
"loss": 2.5009,
|
| 2249 |
+
"step": 183500
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 2.7,
|
| 2253 |
+
"learning_rate": 4.392615315819427e-05,
|
| 2254 |
+
"loss": 2.575,
|
| 2255 |
+
"step": 184000
|
| 2256 |
+
},
|
| 2257 |
+
{
|
| 2258 |
+
"epoch": 2.7,
|
| 2259 |
+
"learning_rate": 4.3697330381940595e-05,
|
| 2260 |
+
"loss": 2.3703,
|
| 2261 |
+
"step": 184500
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 2.71,
|
| 2265 |
+
"learning_rate": 4.346864166191336e-05,
|
| 2266 |
+
"loss": 2.4276,
|
| 2267 |
+
"step": 185000
|
| 2268 |
+
},
|
| 2269 |
+
{
|
| 2270 |
+
"epoch": 2.72,
|
| 2271 |
+
"learning_rate": 4.3240091862265176e-05,
|
| 2272 |
+
"loss": 2.5392,
|
| 2273 |
+
"step": 185500
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 2.73,
|
| 2277 |
+
"learning_rate": 4.301168584419385e-05,
|
| 2278 |
+
"loss": 2.4175,
|
| 2279 |
+
"step": 186000
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 2.73,
|
| 2283 |
+
"learning_rate": 4.2783428465838974e-05,
|
| 2284 |
+
"loss": 2.457,
|
| 2285 |
+
"step": 186500
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 2.74,
|
| 2289 |
+
"learning_rate": 4.2555324582178655e-05,
|
| 2290 |
+
"loss": 2.5549,
|
| 2291 |
+
"step": 187000
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.75,
|
| 2295 |
+
"learning_rate": 4.232737904492613e-05,
|
| 2296 |
+
"loss": 2.5629,
|
| 2297 |
+
"step": 187500
|
| 2298 |
+
},
|
| 2299 |
+
{
|
| 2300 |
+
"epoch": 2.76,
|
| 2301 |
+
"learning_rate": 4.2099596702426705e-05,
|
| 2302 |
+
"loss": 2.529,
|
| 2303 |
+
"step": 188000
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 2.76,
|
| 2307 |
+
"learning_rate": 4.187198239955456e-05,
|
| 2308 |
+
"loss": 2.5347,
|
| 2309 |
+
"step": 188500
|
| 2310 |
+
},
|
| 2311 |
+
{
|
| 2312 |
+
"epoch": 2.77,
|
| 2313 |
+
"learning_rate": 4.164454097760973e-05,
|
| 2314 |
+
"loss": 2.4773,
|
| 2315 |
+
"step": 189000
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 2.78,
|
| 2319 |
+
"learning_rate": 4.141727727421508e-05,
|
| 2320 |
+
"loss": 2.4718,
|
| 2321 |
+
"step": 189500
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 2.79,
|
| 2325 |
+
"learning_rate": 4.119019612321346e-05,
|
| 2326 |
+
"loss": 2.4831,
|
| 2327 |
+
"step": 190000
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 2.79,
|
| 2331 |
+
"learning_rate": 4.096330235456489e-05,
|
| 2332 |
+
"loss": 2.5369,
|
| 2333 |
+
"step": 190500
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 2.8,
|
| 2337 |
+
"learning_rate": 4.073660079424377e-05,
|
| 2338 |
+
"loss": 2.5185,
|
| 2339 |
+
"step": 191000
|
| 2340 |
+
},
|
| 2341 |
+
{
|
| 2342 |
+
"epoch": 2.81,
|
| 2343 |
+
"learning_rate": 4.051009626413636e-05,
|
| 2344 |
+
"loss": 2.4736,
|
| 2345 |
+
"step": 191500
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 2.81,
|
| 2349 |
+
"learning_rate": 4.028379358193805e-05,
|
| 2350 |
+
"loss": 2.5441,
|
| 2351 |
+
"step": 192000
|
| 2352 |
+
},
|
| 2353 |
+
{
|
| 2354 |
+
"epoch": 2.82,
|
| 2355 |
+
"learning_rate": 4.005769756105101e-05,
|
| 2356 |
+
"loss": 2.4609,
|
| 2357 |
+
"step": 192500
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 2.83,
|
| 2361 |
+
"learning_rate": 3.983181301048179e-05,
|
| 2362 |
+
"loss": 2.5473,
|
| 2363 |
+
"step": 193000
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 2.84,
|
| 2367 |
+
"learning_rate": 3.960614473473898e-05,
|
| 2368 |
+
"loss": 2.5741,
|
| 2369 |
+
"step": 193500
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 2.84,
|
| 2373 |
+
"learning_rate": 3.93806975337311e-05,
|
| 2374 |
+
"loss": 2.5476,
|
| 2375 |
+
"step": 194000
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 2.85,
|
| 2379 |
+
"learning_rate": 3.9155476202664446e-05,
|
| 2380 |
+
"loss": 2.5093,
|
| 2381 |
+
"step": 194500
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 2.86,
|
| 2385 |
+
"learning_rate": 3.8930485531941084e-05,
|
| 2386 |
+
"loss": 2.5792,
|
| 2387 |
+
"step": 195000
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 2.87,
|
| 2391 |
+
"learning_rate": 3.870573030705705e-05,
|
| 2392 |
+
"loss": 2.5136,
|
| 2393 |
+
"step": 195500
|
| 2394 |
+
},
|
| 2395 |
+
{
|
| 2396 |
+
"epoch": 2.87,
|
| 2397 |
+
"learning_rate": 3.848121530850049e-05,
|
| 2398 |
+
"loss": 2.4505,
|
| 2399 |
+
"step": 196000
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 2.88,
|
| 2403 |
+
"learning_rate": 3.825694531164997e-05,
|
| 2404 |
+
"loss": 2.4591,
|
| 2405 |
+
"step": 196500
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 2.89,
|
| 2409 |
+
"learning_rate": 3.803292508667294e-05,
|
| 2410 |
+
"loss": 2.501,
|
| 2411 |
+
"step": 197000
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.9,
|
| 2415 |
+
"learning_rate": 3.780915939842428e-05,
|
| 2416 |
+
"loss": 2.5764,
|
| 2417 |
+
"step": 197500
|
| 2418 |
+
},
|
| 2419 |
+
{
|
| 2420 |
+
"epoch": 2.9,
|
| 2421 |
+
"learning_rate": 3.7585653006344886e-05,
|
| 2422 |
+
"loss": 2.4637,
|
| 2423 |
+
"step": 198000
|
| 2424 |
+
},
|
| 2425 |
+
{
|
| 2426 |
+
"epoch": 2.91,
|
| 2427 |
+
"learning_rate": 3.736241066436059e-05,
|
| 2428 |
+
"loss": 2.5644,
|
| 2429 |
+
"step": 198500
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 2.92,
|
| 2433 |
+
"learning_rate": 3.713943712078083e-05,
|
| 2434 |
+
"loss": 2.4658,
|
| 2435 |
+
"step": 199000
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 2.92,
|
| 2439 |
+
"learning_rate": 3.691673711819783e-05,
|
| 2440 |
+
"loss": 2.5032,
|
| 2441 |
+
"step": 199500
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 2.93,
|
| 2445 |
+
"learning_rate": 3.669431539338567e-05,
|
| 2446 |
+
"loss": 2.5171,
|
| 2447 |
+
"step": 200000
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 2.94,
|
| 2451 |
+
"learning_rate": 3.647217667719949e-05,
|
| 2452 |
+
"loss": 2.5642,
|
| 2453 |
+
"step": 200500
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 2.95,
|
| 2457 |
+
"learning_rate": 3.6250325694474943e-05,
|
| 2458 |
+
"loss": 2.5403,
|
| 2459 |
+
"step": 201000
|
| 2460 |
+
},
|
| 2461 |
+
{
|
| 2462 |
+
"epoch": 2.95,
|
| 2463 |
+
"learning_rate": 3.6028767163927654e-05,
|
| 2464 |
+
"loss": 2.5735,
|
| 2465 |
+
"step": 201500
|
| 2466 |
+
},
|
| 2467 |
+
{
|
| 2468 |
+
"epoch": 2.96,
|
| 2469 |
+
"learning_rate": 3.580750579805285e-05,
|
| 2470 |
+
"loss": 2.4973,
|
| 2471 |
+
"step": 202000
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 2.97,
|
| 2475 |
+
"learning_rate": 3.558654630302516e-05,
|
| 2476 |
+
"loss": 2.4614,
|
| 2477 |
+
"step": 202500
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 2.98,
|
| 2481 |
+
"learning_rate": 3.536589337859844e-05,
|
| 2482 |
+
"loss": 2.513,
|
| 2483 |
+
"step": 203000
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 2.98,
|
| 2487 |
+
"learning_rate": 3.514555171800597e-05,
|
| 2488 |
+
"loss": 2.4906,
|
| 2489 |
+
"step": 203500
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 2.99,
|
| 2493 |
+
"learning_rate": 3.492552600786041e-05,
|
| 2494 |
+
"loss": 2.5152,
|
| 2495 |
+
"step": 204000
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 3.0,
|
| 2499 |
+
"learning_rate": 3.470582092805431e-05,
|
| 2500 |
+
"loss": 2.4693,
|
| 2501 |
+
"step": 204500
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 3.0,
|
| 2505 |
+
"eval_bleu": 1.0,
|
| 2506 |
+
"eval_brevity_penalty": 1.0,
|
| 2507 |
+
"eval_length_ratio": 1.0,
|
| 2508 |
+
"eval_loss": 2.896097421646118,
|
| 2509 |
+
"eval_precisions": [
|
| 2510 |
+
1.0,
|
| 2511 |
+
1.0,
|
| 2512 |
+
1.0,
|
| 2513 |
+
1.0
|
| 2514 |
+
],
|
| 2515 |
+
"eval_reference_length": 7761920,
|
| 2516 |
+
"eval_runtime": 14493.3817,
|
| 2517 |
+
"eval_samples_per_second": 1.046,
|
| 2518 |
+
"eval_steps_per_second": 0.523,
|
| 2519 |
+
"eval_translation_length": 7761920,
|
| 2520 |
+
"step": 204657
|
| 2521 |
}
|
| 2522 |
],
|
| 2523 |
"logging_steps": 500,
|
| 2524 |
"max_steps": 341095,
|
| 2525 |
"num_train_epochs": 5,
|
| 2526 |
"save_steps": 500,
|
| 2527 |
+
"total_flos": 9.433737890333983e+17,
|
| 2528 |
"trial_name": null,
|
| 2529 |
"trial_params": null
|
| 2530 |
}
|