Training in progress, step 1788, 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 3171119736
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd9e4fe810f2ea1790522d641825620fdde9830548e2a0a32214743a814c1c04
|
| 3 |
size 3171119736
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6342379990
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8e5f59aa42c10a605f83a0c932bbcad439fbc7ac8b194f3fdad92ecc66f43faa
|
| 3 |
size 6342379990
|
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 sha256:404b871ed8ee4bbd465f4cf5598f5d497e5307c5f756348590408d5dff207b33
|
| 3 |
size 1064
|
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":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -13608,6 +13608,710 @@
|
|
| 13608 |
"loss": 1.175,
|
| 13609 |
"num_tokens": 74904934.0,
|
| 13610 |
"step": 1700
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13611 |
}
|
| 13612 |
],
|
| 13613 |
"logging_steps": 1,
|
|
@@ -13622,12 +14326,12 @@
|
|
| 13622 |
"should_evaluate": false,
|
| 13623 |
"should_log": false,
|
| 13624 |
"should_save": true,
|
| 13625 |
-
"should_training_stop":
|
| 13626 |
},
|
| 13627 |
"attributes": {}
|
| 13628 |
}
|
| 13629 |
},
|
| 13630 |
-
"total_flos": 1.
|
| 13631 |
"train_batch_size": 8,
|
| 13632 |
"trial_name": null,
|
| 13633 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 1788,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 13608 |
"loss": 1.175,
|
| 13609 |
"num_tokens": 74904934.0,
|
| 13610 |
"step": 1700
|
| 13611 |
+
},
|
| 13612 |
+
{
|
| 13613 |
+
"epoch": 3.8053691275167782,
|
| 13614 |
+
"grad_norm": 1.7265625,
|
| 13615 |
+
"learning_rate": 6.612567676012538e-08,
|
| 13616 |
+
"loss": 0.9379,
|
| 13617 |
+
"num_tokens": 74968162.0,
|
| 13618 |
+
"step": 1701
|
| 13619 |
+
},
|
| 13620 |
+
{
|
| 13621 |
+
"epoch": 3.8076062639821027,
|
| 13622 |
+
"grad_norm": 2.140625,
|
| 13623 |
+
"learning_rate": 6.463458736367111e-08,
|
| 13624 |
+
"loss": 1.1161,
|
| 13625 |
+
"num_tokens": 75007435.0,
|
| 13626 |
+
"step": 1702
|
| 13627 |
+
},
|
| 13628 |
+
{
|
| 13629 |
+
"epoch": 3.809843400447427,
|
| 13630 |
+
"grad_norm": 2.171875,
|
| 13631 |
+
"learning_rate": 6.316039239796235e-08,
|
| 13632 |
+
"loss": 1.1358,
|
| 13633 |
+
"num_tokens": 75043772.0,
|
| 13634 |
+
"step": 1703
|
| 13635 |
+
},
|
| 13636 |
+
{
|
| 13637 |
+
"epoch": 3.8120805369127515,
|
| 13638 |
+
"grad_norm": 2.09375,
|
| 13639 |
+
"learning_rate": 6.170309690937015e-08,
|
| 13640 |
+
"loss": 1.0269,
|
| 13641 |
+
"num_tokens": 75086495.0,
|
| 13642 |
+
"step": 1704
|
| 13643 |
+
},
|
| 13644 |
+
{
|
| 13645 |
+
"epoch": 3.814317673378076,
|
| 13646 |
+
"grad_norm": 2.15625,
|
| 13647 |
+
"learning_rate": 6.02627058864158e-08,
|
| 13648 |
+
"loss": 0.8462,
|
| 13649 |
+
"num_tokens": 75110601.0,
|
| 13650 |
+
"step": 1705
|
| 13651 |
+
},
|
| 13652 |
+
{
|
| 13653 |
+
"epoch": 3.8165548098434003,
|
| 13654 |
+
"grad_norm": 1.8203125,
|
| 13655 |
+
"learning_rate": 5.883922425975464e-08,
|
| 13656 |
+
"loss": 1.2619,
|
| 13657 |
+
"num_tokens": 75166428.0,
|
| 13658 |
+
"step": 1706
|
| 13659 |
+
},
|
| 13660 |
+
{
|
| 13661 |
+
"epoch": 3.8187919463087248,
|
| 13662 |
+
"grad_norm": 2.046875,
|
| 13663 |
+
"learning_rate": 5.743265690215938e-08,
|
| 13664 |
+
"loss": 1.2676,
|
| 13665 |
+
"num_tokens": 75212607.0,
|
| 13666 |
+
"step": 1707
|
| 13667 |
+
},
|
| 13668 |
+
{
|
| 13669 |
+
"epoch": 3.821029082774049,
|
| 13670 |
+
"grad_norm": 1.890625,
|
| 13671 |
+
"learning_rate": 5.604300862850187e-08,
|
| 13672 |
+
"loss": 1.2185,
|
| 13673 |
+
"num_tokens": 75262553.0,
|
| 13674 |
+
"step": 1708
|
| 13675 |
+
},
|
| 13676 |
+
{
|
| 13677 |
+
"epoch": 3.8232662192393736,
|
| 13678 |
+
"grad_norm": 2.078125,
|
| 13679 |
+
"learning_rate": 5.467028419573861e-08,
|
| 13680 |
+
"loss": 1.2147,
|
| 13681 |
+
"num_tokens": 75301345.0,
|
| 13682 |
+
"step": 1709
|
| 13683 |
+
},
|
| 13684 |
+
{
|
| 13685 |
+
"epoch": 3.825503355704698,
|
| 13686 |
+
"grad_norm": 1.640625,
|
| 13687 |
+
"learning_rate": 5.331448830289354e-08,
|
| 13688 |
+
"loss": 1.2582,
|
| 13689 |
+
"num_tokens": 75367698.0,
|
| 13690 |
+
"step": 1710
|
| 13691 |
+
},
|
| 13692 |
+
{
|
| 13693 |
+
"epoch": 3.8277404921700224,
|
| 13694 |
+
"grad_norm": 2.078125,
|
| 13695 |
+
"learning_rate": 5.19756255910403e-08,
|
| 13696 |
+
"loss": 1.1464,
|
| 13697 |
+
"num_tokens": 75403557.0,
|
| 13698 |
+
"step": 1711
|
| 13699 |
+
},
|
| 13700 |
+
{
|
| 13701 |
+
"epoch": 3.829977628635347,
|
| 13702 |
+
"grad_norm": 1.59375,
|
| 13703 |
+
"learning_rate": 5.0653700643290006e-08,
|
| 13704 |
+
"loss": 1.0514,
|
| 13705 |
+
"num_tokens": 75464418.0,
|
| 13706 |
+
"step": 1712
|
| 13707 |
+
},
|
| 13708 |
+
{
|
| 13709 |
+
"epoch": 3.8322147651006713,
|
| 13710 |
+
"grad_norm": 2.234375,
|
| 13711 |
+
"learning_rate": 4.934871798477236e-08,
|
| 13712 |
+
"loss": 1.0188,
|
| 13713 |
+
"num_tokens": 75492830.0,
|
| 13714 |
+
"step": 1713
|
| 13715 |
+
},
|
| 13716 |
+
{
|
| 13717 |
+
"epoch": 3.8344519015659957,
|
| 13718 |
+
"grad_norm": 1.4453125,
|
| 13719 |
+
"learning_rate": 4.806068208262071e-08,
|
| 13720 |
+
"loss": 1.1834,
|
| 13721 |
+
"num_tokens": 75578238.0,
|
| 13722 |
+
"step": 1714
|
| 13723 |
+
},
|
| 13724 |
+
{
|
| 13725 |
+
"epoch": 3.83668903803132,
|
| 13726 |
+
"grad_norm": 1.8671875,
|
| 13727 |
+
"learning_rate": 4.6789597345959223e-08,
|
| 13728 |
+
"loss": 1.3146,
|
| 13729 |
+
"num_tokens": 75641338.0,
|
| 13730 |
+
"step": 1715
|
| 13731 |
+
},
|
| 13732 |
+
{
|
| 13733 |
+
"epoch": 3.8389261744966445,
|
| 13734 |
+
"grad_norm": 1.65625,
|
| 13735 |
+
"learning_rate": 4.5535468125883496e-08,
|
| 13736 |
+
"loss": 1.2417,
|
| 13737 |
+
"num_tokens": 75703922.0,
|
| 13738 |
+
"step": 1716
|
| 13739 |
+
},
|
| 13740 |
+
{
|
| 13741 |
+
"epoch": 3.841163310961969,
|
| 13742 |
+
"grad_norm": 1.9609375,
|
| 13743 |
+
"learning_rate": 4.429829871545055e-08,
|
| 13744 |
+
"loss": 1.2914,
|
| 13745 |
+
"num_tokens": 75745478.0,
|
| 13746 |
+
"step": 1717
|
| 13747 |
+
},
|
| 13748 |
+
{
|
| 13749 |
+
"epoch": 3.843400447427293,
|
| 13750 |
+
"grad_norm": 1.8125,
|
| 13751 |
+
"learning_rate": 4.3078093349659955e-08,
|
| 13752 |
+
"loss": 1.1846,
|
| 13753 |
+
"num_tokens": 75793875.0,
|
| 13754 |
+
"step": 1718
|
| 13755 |
+
},
|
| 13756 |
+
{
|
| 13757 |
+
"epoch": 3.8456375838926173,
|
| 13758 |
+
"grad_norm": 1.6796875,
|
| 13759 |
+
"learning_rate": 4.187485620544163e-08,
|
| 13760 |
+
"loss": 1.0283,
|
| 13761 |
+
"num_tokens": 75842393.0,
|
| 13762 |
+
"step": 1719
|
| 13763 |
+
},
|
| 13764 |
+
{
|
| 13765 |
+
"epoch": 3.8478747203579418,
|
| 13766 |
+
"grad_norm": 1.7578125,
|
| 13767 |
+
"learning_rate": 4.068859140164083e-08,
|
| 13768 |
+
"loss": 0.7449,
|
| 13769 |
+
"num_tokens": 75875657.0,
|
| 13770 |
+
"step": 1720
|
| 13771 |
+
},
|
| 13772 |
+
{
|
| 13773 |
+
"epoch": 3.850111856823266,
|
| 13774 |
+
"grad_norm": 2.109375,
|
| 13775 |
+
"learning_rate": 3.9519302999004305e-08,
|
| 13776 |
+
"loss": 0.6057,
|
| 13777 |
+
"num_tokens": 75896467.0,
|
| 13778 |
+
"step": 1721
|
| 13779 |
+
},
|
| 13780 |
+
{
|
| 13781 |
+
"epoch": 3.8523489932885906,
|
| 13782 |
+
"grad_norm": 2.515625,
|
| 13783 |
+
"learning_rate": 3.836699500016583e-08,
|
| 13784 |
+
"loss": 1.204,
|
| 13785 |
+
"num_tokens": 75931182.0,
|
| 13786 |
+
"step": 1722
|
| 13787 |
+
},
|
| 13788 |
+
{
|
| 13789 |
+
"epoch": 3.854586129753915,
|
| 13790 |
+
"grad_norm": 2.03125,
|
| 13791 |
+
"learning_rate": 3.7231671349634015e-08,
|
| 13792 |
+
"loss": 0.6344,
|
| 13793 |
+
"num_tokens": 75954997.0,
|
| 13794 |
+
"step": 1723
|
| 13795 |
+
},
|
| 13796 |
+
{
|
| 13797 |
+
"epoch": 3.8568232662192394,
|
| 13798 |
+
"grad_norm": 2.484375,
|
| 13799 |
+
"learning_rate": 3.611333593377564e-08,
|
| 13800 |
+
"loss": 1.4252,
|
| 13801 |
+
"num_tokens": 75995886.0,
|
| 13802 |
+
"step": 1724
|
| 13803 |
+
},
|
| 13804 |
+
{
|
| 13805 |
+
"epoch": 3.859060402684564,
|
| 13806 |
+
"grad_norm": 1.859375,
|
| 13807 |
+
"learning_rate": 3.501199258080734e-08,
|
| 13808 |
+
"loss": 1.1533,
|
| 13809 |
+
"num_tokens": 76038609.0,
|
| 13810 |
+
"step": 1725
|
| 13811 |
+
},
|
| 13812 |
+
{
|
| 13813 |
+
"epoch": 3.8612975391498883,
|
| 13814 |
+
"grad_norm": 1.34375,
|
| 13815 |
+
"learning_rate": 3.3927645060776725e-08,
|
| 13816 |
+
"loss": 1.052,
|
| 13817 |
+
"num_tokens": 76119236.0,
|
| 13818 |
+
"step": 1726
|
| 13819 |
+
},
|
| 13820 |
+
{
|
| 13821 |
+
"epoch": 3.8635346756152127,
|
| 13822 |
+
"grad_norm": 1.8125,
|
| 13823 |
+
"learning_rate": 3.286029708555405e-08,
|
| 13824 |
+
"loss": 1.0873,
|
| 13825 |
+
"num_tokens": 76164531.0,
|
| 13826 |
+
"step": 1727
|
| 13827 |
+
},
|
| 13828 |
+
{
|
| 13829 |
+
"epoch": 3.8657718120805367,
|
| 13830 |
+
"grad_norm": 2.375,
|
| 13831 |
+
"learning_rate": 3.1809952308818336e-08,
|
| 13832 |
+
"loss": 0.7005,
|
| 13833 |
+
"num_tokens": 76180927.0,
|
| 13834 |
+
"step": 1728
|
| 13835 |
+
},
|
| 13836 |
+
{
|
| 13837 |
+
"epoch": 3.868008948545861,
|
| 13838 |
+
"grad_norm": 1.8359375,
|
| 13839 |
+
"learning_rate": 3.077661432604184e-08,
|
| 13840 |
+
"loss": 1.3131,
|
| 13841 |
+
"num_tokens": 76237010.0,
|
| 13842 |
+
"step": 1729
|
| 13843 |
+
},
|
| 13844 |
+
{
|
| 13845 |
+
"epoch": 3.8702460850111855,
|
| 13846 |
+
"grad_norm": 2.4375,
|
| 13847 |
+
"learning_rate": 2.976028667448283e-08,
|
| 13848 |
+
"loss": 0.7783,
|
| 13849 |
+
"num_tokens": 76257989.0,
|
| 13850 |
+
"step": 1730
|
| 13851 |
+
},
|
| 13852 |
+
{
|
| 13853 |
+
"epoch": 3.87248322147651,
|
| 13854 |
+
"grad_norm": 2.171875,
|
| 13855 |
+
"learning_rate": 2.8760972833170032e-08,
|
| 13856 |
+
"loss": 0.8357,
|
| 13857 |
+
"num_tokens": 76284577.0,
|
| 13858 |
+
"step": 1731
|
| 13859 |
+
},
|
| 13860 |
+
{
|
| 13861 |
+
"epoch": 3.8747203579418343,
|
| 13862 |
+
"grad_norm": 1.5859375,
|
| 13863 |
+
"learning_rate": 2.7778676222890433e-08,
|
| 13864 |
+
"loss": 1.0342,
|
| 13865 |
+
"num_tokens": 76343897.0,
|
| 13866 |
+
"step": 1732
|
| 13867 |
+
},
|
| 13868 |
+
{
|
| 13869 |
+
"epoch": 3.8769574944071588,
|
| 13870 |
+
"grad_norm": 1.9609375,
|
| 13871 |
+
"learning_rate": 2.6813400206180394e-08,
|
| 13872 |
+
"loss": 1.2112,
|
| 13873 |
+
"num_tokens": 76389220.0,
|
| 13874 |
+
"step": 1733
|
| 13875 |
+
},
|
| 13876 |
+
{
|
| 13877 |
+
"epoch": 3.879194630872483,
|
| 13878 |
+
"grad_norm": 2.0,
|
| 13879 |
+
"learning_rate": 2.586514808731122e-08,
|
| 13880 |
+
"loss": 0.7824,
|
| 13881 |
+
"num_tokens": 76415002.0,
|
| 13882 |
+
"step": 1734
|
| 13883 |
+
},
|
| 13884 |
+
{
|
| 13885 |
+
"epoch": 3.8814317673378076,
|
| 13886 |
+
"grad_norm": 1.8515625,
|
| 13887 |
+
"learning_rate": 2.4933923112279712e-08,
|
| 13888 |
+
"loss": 0.8511,
|
| 13889 |
+
"num_tokens": 76449834.0,
|
| 13890 |
+
"step": 1735
|
| 13891 |
+
},
|
| 13892 |
+
{
|
| 13893 |
+
"epoch": 3.883668903803132,
|
| 13894 |
+
"grad_norm": 1.671875,
|
| 13895 |
+
"learning_rate": 2.4019728468797077e-08,
|
| 13896 |
+
"loss": 1.1136,
|
| 13897 |
+
"num_tokens": 76512621.0,
|
| 13898 |
+
"step": 1736
|
| 13899 |
+
},
|
| 13900 |
+
{
|
| 13901 |
+
"epoch": 3.8859060402684564,
|
| 13902 |
+
"grad_norm": 1.6640625,
|
| 13903 |
+
"learning_rate": 2.31225672862756e-08,
|
| 13904 |
+
"loss": 1.1071,
|
| 13905 |
+
"num_tokens": 76573378.0,
|
| 13906 |
+
"step": 1737
|
| 13907 |
+
},
|
| 13908 |
+
{
|
| 13909 |
+
"epoch": 3.888143176733781,
|
| 13910 |
+
"grad_norm": 2.140625,
|
| 13911 |
+
"learning_rate": 2.224244263582087e-08,
|
| 13912 |
+
"loss": 1.0365,
|
| 13913 |
+
"num_tokens": 76605199.0,
|
| 13914 |
+
"step": 1738
|
| 13915 |
+
},
|
| 13916 |
+
{
|
| 13917 |
+
"epoch": 3.8903803131991053,
|
| 13918 |
+
"grad_norm": 2.28125,
|
| 13919 |
+
"learning_rate": 2.137935753022069e-08,
|
| 13920 |
+
"loss": 1.0554,
|
| 13921 |
+
"num_tokens": 76654107.0,
|
| 13922 |
+
"step": 1739
|
| 13923 |
+
},
|
| 13924 |
+
{
|
| 13925 |
+
"epoch": 3.8926174496644297,
|
| 13926 |
+
"grad_norm": 1.765625,
|
| 13927 |
+
"learning_rate": 2.053331492393229e-08,
|
| 13928 |
+
"loss": 1.1568,
|
| 13929 |
+
"num_tokens": 76719172.0,
|
| 13930 |
+
"step": 1740
|
| 13931 |
+
},
|
| 13932 |
+
{
|
| 13933 |
+
"epoch": 3.894854586129754,
|
| 13934 |
+
"grad_norm": 2.171875,
|
| 13935 |
+
"learning_rate": 1.9704317713076236e-08,
|
| 13936 |
+
"loss": 0.6222,
|
| 13937 |
+
"num_tokens": 76736383.0,
|
| 13938 |
+
"step": 1741
|
| 13939 |
+
},
|
| 13940 |
+
{
|
| 13941 |
+
"epoch": 3.8970917225950785,
|
| 13942 |
+
"grad_norm": 2.34375,
|
| 13943 |
+
"learning_rate": 1.8892368735422552e-08,
|
| 13944 |
+
"loss": 1.1949,
|
| 13945 |
+
"num_tokens": 76769294.0,
|
| 13946 |
+
"step": 1742
|
| 13947 |
+
},
|
| 13948 |
+
{
|
| 13949 |
+
"epoch": 3.899328859060403,
|
| 13950 |
+
"grad_norm": 1.9453125,
|
| 13951 |
+
"learning_rate": 1.8097470770384596e-08,
|
| 13952 |
+
"loss": 0.9303,
|
| 13953 |
+
"num_tokens": 76803600.0,
|
| 13954 |
+
"step": 1743
|
| 13955 |
+
},
|
| 13956 |
+
{
|
| 13957 |
+
"epoch": 3.901565995525727,
|
| 13958 |
+
"grad_norm": 2.140625,
|
| 13959 |
+
"learning_rate": 1.7319626539005762e-08,
|
| 13960 |
+
"loss": 0.9384,
|
| 13961 |
+
"num_tokens": 76834616.0,
|
| 13962 |
+
"step": 1744
|
| 13963 |
+
},
|
| 13964 |
+
{
|
| 13965 |
+
"epoch": 3.9038031319910513,
|
| 13966 |
+
"grad_norm": 2.359375,
|
| 13967 |
+
"learning_rate": 1.655883870395336e-08,
|
| 13968 |
+
"loss": 1.0708,
|
| 13969 |
+
"num_tokens": 76877338.0,
|
| 13970 |
+
"step": 1745
|
| 13971 |
+
},
|
| 13972 |
+
{
|
| 13973 |
+
"epoch": 3.9060402684563758,
|
| 13974 |
+
"grad_norm": 1.859375,
|
| 13975 |
+
"learning_rate": 1.5815109869509183e-08,
|
| 13976 |
+
"loss": 0.9379,
|
| 13977 |
+
"num_tokens": 76915339.0,
|
| 13978 |
+
"step": 1746
|
| 13979 |
+
},
|
| 13980 |
+
{
|
| 13981 |
+
"epoch": 3.9082774049217,
|
| 13982 |
+
"grad_norm": 1.765625,
|
| 13983 |
+
"learning_rate": 1.508844258155728e-08,
|
| 13984 |
+
"loss": 0.9361,
|
| 13985 |
+
"num_tokens": 76964508.0,
|
| 13986 |
+
"step": 1747
|
| 13987 |
+
},
|
| 13988 |
+
{
|
| 13989 |
+
"epoch": 3.9105145413870246,
|
| 13990 |
+
"grad_norm": 1.875,
|
| 13991 |
+
"learning_rate": 1.4378839327580663e-08,
|
| 13992 |
+
"loss": 0.9759,
|
| 13993 |
+
"num_tokens": 77001061.0,
|
| 13994 |
+
"step": 1748
|
| 13995 |
+
},
|
| 13996 |
+
{
|
| 13997 |
+
"epoch": 3.912751677852349,
|
| 13998 |
+
"grad_norm": 2.109375,
|
| 13999 |
+
"learning_rate": 1.3686302536647378e-08,
|
| 14000 |
+
"loss": 1.1656,
|
| 14001 |
+
"num_tokens": 77038112.0,
|
| 14002 |
+
"step": 1749
|
| 14003 |
+
},
|
| 14004 |
+
{
|
| 14005 |
+
"epoch": 3.9149888143176734,
|
| 14006 |
+
"grad_norm": 1.9375,
|
| 14007 |
+
"learning_rate": 1.3010834579405552e-08,
|
| 14008 |
+
"loss": 0.9305,
|
| 14009 |
+
"num_tokens": 77072561.0,
|
| 14010 |
+
"step": 1750
|
| 14011 |
+
},
|
| 14012 |
+
{
|
| 14013 |
+
"epoch": 3.917225950782998,
|
| 14014 |
+
"grad_norm": 2.109375,
|
| 14015 |
+
"learning_rate": 1.2352437768074487e-08,
|
| 14016 |
+
"loss": 1.0788,
|
| 14017 |
+
"num_tokens": 77114800.0,
|
| 14018 |
+
"step": 1751
|
| 14019 |
+
},
|
| 14020 |
+
{
|
| 14021 |
+
"epoch": 3.9194630872483223,
|
| 14022 |
+
"grad_norm": 2.03125,
|
| 14023 |
+
"learning_rate": 1.1711114356436903e-08,
|
| 14024 |
+
"loss": 1.0074,
|
| 14025 |
+
"num_tokens": 77148923.0,
|
| 14026 |
+
"step": 1752
|
| 14027 |
+
},
|
| 14028 |
+
{
|
| 14029 |
+
"epoch": 3.9217002237136467,
|
| 14030 |
+
"grad_norm": 1.7421875,
|
| 14031 |
+
"learning_rate": 1.1086866539830044e-08,
|
| 14032 |
+
"loss": 1.2444,
|
| 14033 |
+
"num_tokens": 77206267.0,
|
| 14034 |
+
"step": 1753
|
| 14035 |
+
},
|
| 14036 |
+
{
|
| 14037 |
+
"epoch": 3.9239373601789707,
|
| 14038 |
+
"grad_norm": 1.640625,
|
| 14039 |
+
"learning_rate": 1.0479696455139576e-08,
|
| 14040 |
+
"loss": 0.9213,
|
| 14041 |
+
"num_tokens": 77251539.0,
|
| 14042 |
+
"step": 1754
|
| 14043 |
+
},
|
| 14044 |
+
{
|
| 14045 |
+
"epoch": 3.926174496644295,
|
| 14046 |
+
"grad_norm": 1.84375,
|
| 14047 |
+
"learning_rate": 9.889606180792378e-09,
|
| 14048 |
+
"loss": 0.9004,
|
| 14049 |
+
"num_tokens": 77291726.0,
|
| 14050 |
+
"step": 1755
|
| 14051 |
+
},
|
| 14052 |
+
{
|
| 14053 |
+
"epoch": 3.9284116331096195,
|
| 14054 |
+
"grad_norm": 2.5625,
|
| 14055 |
+
"learning_rate": 9.316597736747091e-09,
|
| 14056 |
+
"loss": 1.0968,
|
| 14057 |
+
"num_tokens": 77332501.0,
|
| 14058 |
+
"step": 1756
|
| 14059 |
+
},
|
| 14060 |
+
{
|
| 14061 |
+
"epoch": 3.930648769574944,
|
| 14062 |
+
"grad_norm": 2.0625,
|
| 14063 |
+
"learning_rate": 8.7606730844908e-09,
|
| 14064 |
+
"loss": 1.2321,
|
| 14065 |
+
"num_tokens": 77386516.0,
|
| 14066 |
+
"step": 1757
|
| 14067 |
+
},
|
| 14068 |
+
{
|
| 14069 |
+
"epoch": 3.9328859060402683,
|
| 14070 |
+
"grad_norm": 1.8515625,
|
| 14071 |
+
"learning_rate": 8.221834127029593e-09,
|
| 14072 |
+
"loss": 1.0625,
|
| 14073 |
+
"num_tokens": 77432402.0,
|
| 14074 |
+
"step": 1758
|
| 14075 |
+
},
|
| 14076 |
+
{
|
| 14077 |
+
"epoch": 3.9351230425055927,
|
| 14078 |
+
"grad_norm": 1.859375,
|
| 14079 |
+
"learning_rate": 7.700082708883006e-09,
|
| 14080 |
+
"loss": 1.2727,
|
| 14081 |
+
"num_tokens": 77495304.0,
|
| 14082 |
+
"step": 1759
|
| 14083 |
+
},
|
| 14084 |
+
{
|
| 14085 |
+
"epoch": 3.937360178970917,
|
| 14086 |
+
"grad_norm": 1.859375,
|
| 14087 |
+
"learning_rate": 7.1954206160768096e-09,
|
| 14088 |
+
"loss": 1.3575,
|
| 14089 |
+
"num_tokens": 77554698.0,
|
| 14090 |
+
"step": 1760
|
| 14091 |
+
},
|
| 14092 |
+
{
|
| 14093 |
+
"epoch": 3.9395973154362416,
|
| 14094 |
+
"grad_norm": 1.9765625,
|
| 14095 |
+
"learning_rate": 6.7078495761385695e-09,
|
| 14096 |
+
"loss": 0.9031,
|
| 14097 |
+
"num_tokens": 77586793.0,
|
| 14098 |
+
"step": 1761
|
| 14099 |
+
},
|
| 14100 |
+
{
|
| 14101 |
+
"epoch": 3.941834451901566,
|
| 14102 |
+
"grad_norm": 2.078125,
|
| 14103 |
+
"learning_rate": 6.237371258090985e-09,
|
| 14104 |
+
"loss": 1.2859,
|
| 14105 |
+
"num_tokens": 77631284.0,
|
| 14106 |
+
"step": 1762
|
| 14107 |
+
},
|
| 14108 |
+
{
|
| 14109 |
+
"epoch": 3.9440715883668904,
|
| 14110 |
+
"grad_norm": 2.171875,
|
| 14111 |
+
"learning_rate": 5.783987272445779e-09,
|
| 14112 |
+
"loss": 0.8371,
|
| 14113 |
+
"num_tokens": 77655093.0,
|
| 14114 |
+
"step": 1763
|
| 14115 |
+
},
|
| 14116 |
+
{
|
| 14117 |
+
"epoch": 3.946308724832215,
|
| 14118 |
+
"grad_norm": 1.90625,
|
| 14119 |
+
"learning_rate": 5.347699171197595e-09,
|
| 14120 |
+
"loss": 1.3136,
|
| 14121 |
+
"num_tokens": 77708409.0,
|
| 14122 |
+
"step": 1764
|
| 14123 |
+
},
|
| 14124 |
+
{
|
| 14125 |
+
"epoch": 3.9485458612975393,
|
| 14126 |
+
"grad_norm": 1.7734375,
|
| 14127 |
+
"learning_rate": 4.928508447821223e-09,
|
| 14128 |
+
"loss": 1.1163,
|
| 14129 |
+
"num_tokens": 77756547.0,
|
| 14130 |
+
"step": 1765
|
| 14131 |
+
},
|
| 14132 |
+
{
|
| 14133 |
+
"epoch": 3.9507829977628637,
|
| 14134 |
+
"grad_norm": 2.734375,
|
| 14135 |
+
"learning_rate": 4.526416537263267e-09,
|
| 14136 |
+
"loss": 1.0242,
|
| 14137 |
+
"num_tokens": 77784004.0,
|
| 14138 |
+
"step": 1766
|
| 14139 |
+
},
|
| 14140 |
+
{
|
| 14141 |
+
"epoch": 3.953020134228188,
|
| 14142 |
+
"grad_norm": 2.125,
|
| 14143 |
+
"learning_rate": 4.141424815938822e-09,
|
| 14144 |
+
"loss": 1.0553,
|
| 14145 |
+
"num_tokens": 77815465.0,
|
| 14146 |
+
"step": 1767
|
| 14147 |
+
},
|
| 14148 |
+
{
|
| 14149 |
+
"epoch": 3.9552572706935125,
|
| 14150 |
+
"grad_norm": 1.7109375,
|
| 14151 |
+
"learning_rate": 3.77353460172869e-09,
|
| 14152 |
+
"loss": 1.2345,
|
| 14153 |
+
"num_tokens": 77867878.0,
|
| 14154 |
+
"step": 1768
|
| 14155 |
+
},
|
| 14156 |
+
{
|
| 14157 |
+
"epoch": 3.957494407158837,
|
| 14158 |
+
"grad_norm": 1.6640625,
|
| 14159 |
+
"learning_rate": 3.422747153969952e-09,
|
| 14160 |
+
"loss": 0.8246,
|
| 14161 |
+
"num_tokens": 77910885.0,
|
| 14162 |
+
"step": 1769
|
| 14163 |
+
},
|
| 14164 |
+
{
|
| 14165 |
+
"epoch": 3.959731543624161,
|
| 14166 |
+
"grad_norm": 1.984375,
|
| 14167 |
+
"learning_rate": 3.089063673456516e-09,
|
| 14168 |
+
"loss": 1.1371,
|
| 14169 |
+
"num_tokens": 77970645.0,
|
| 14170 |
+
"step": 1770
|
| 14171 |
+
},
|
| 14172 |
+
{
|
| 14173 |
+
"epoch": 3.9619686800894853,
|
| 14174 |
+
"grad_norm": 1.7421875,
|
| 14175 |
+
"learning_rate": 2.7724853024324594e-09,
|
| 14176 |
+
"loss": 0.996,
|
| 14177 |
+
"num_tokens": 78019910.0,
|
| 14178 |
+
"step": 1771
|
| 14179 |
+
},
|
| 14180 |
+
{
|
| 14181 |
+
"epoch": 3.9642058165548097,
|
| 14182 |
+
"grad_norm": 1.8046875,
|
| 14183 |
+
"learning_rate": 2.473013124589252e-09,
|
| 14184 |
+
"loss": 1.2593,
|
| 14185 |
+
"num_tokens": 78080344.0,
|
| 14186 |
+
"step": 1772
|
| 14187 |
+
},
|
| 14188 |
+
{
|
| 14189 |
+
"epoch": 3.966442953020134,
|
| 14190 |
+
"grad_norm": 1.9140625,
|
| 14191 |
+
"learning_rate": 2.1906481650613153e-09,
|
| 14192 |
+
"loss": 0.886,
|
| 14193 |
+
"num_tokens": 78109615.0,
|
| 14194 |
+
"step": 1773
|
| 14195 |
+
},
|
| 14196 |
+
{
|
| 14197 |
+
"epoch": 3.9686800894854586,
|
| 14198 |
+
"grad_norm": 2.25,
|
| 14199 |
+
"learning_rate": 1.925391390421583e-09,
|
| 14200 |
+
"loss": 1.2837,
|
| 14201 |
+
"num_tokens": 78153287.0,
|
| 14202 |
+
"step": 1774
|
| 14203 |
+
},
|
| 14204 |
+
{
|
| 14205 |
+
"epoch": 3.970917225950783,
|
| 14206 |
+
"grad_norm": 2.125,
|
| 14207 |
+
"learning_rate": 1.6772437086803873e-09,
|
| 14208 |
+
"loss": 1.4013,
|
| 14209 |
+
"num_tokens": 78213855.0,
|
| 14210 |
+
"step": 1775
|
| 14211 |
+
},
|
| 14212 |
+
{
|
| 14213 |
+
"epoch": 3.9731543624161074,
|
| 14214 |
+
"grad_norm": 2.3125,
|
| 14215 |
+
"learning_rate": 1.446205969282133e-09,
|
| 14216 |
+
"loss": 1.1417,
|
| 14217 |
+
"num_tokens": 78249035.0,
|
| 14218 |
+
"step": 1776
|
| 14219 |
+
},
|
| 14220 |
+
{
|
| 14221 |
+
"epoch": 3.975391498881432,
|
| 14222 |
+
"grad_norm": 2.15625,
|
| 14223 |
+
"learning_rate": 1.2322789630997422e-09,
|
| 14224 |
+
"loss": 0.9906,
|
| 14225 |
+
"num_tokens": 78285916.0,
|
| 14226 |
+
"step": 1777
|
| 14227 |
+
},
|
| 14228 |
+
{
|
| 14229 |
+
"epoch": 3.9776286353467563,
|
| 14230 |
+
"grad_norm": 1.7890625,
|
| 14231 |
+
"learning_rate": 1.0354634224346572e-09,
|
| 14232 |
+
"loss": 1.1599,
|
| 14233 |
+
"num_tokens": 78343837.0,
|
| 14234 |
+
"step": 1778
|
| 14235 |
+
},
|
| 14236 |
+
{
|
| 14237 |
+
"epoch": 3.9798657718120807,
|
| 14238 |
+
"grad_norm": 1.8359375,
|
| 14239 |
+
"learning_rate": 8.557600210140627e-10,
|
| 14240 |
+
"loss": 0.9641,
|
| 14241 |
+
"num_tokens": 78385564.0,
|
| 14242 |
+
"step": 1779
|
| 14243 |
+
},
|
| 14244 |
+
{
|
| 14245 |
+
"epoch": 3.9821029082774047,
|
| 14246 |
+
"grad_norm": 2.046875,
|
| 14247 |
+
"learning_rate": 6.931693739864465e-10,
|
| 14248 |
+
"loss": 1.3055,
|
| 14249 |
+
"num_tokens": 78442663.0,
|
| 14250 |
+
"step": 1780
|
| 14251 |
+
},
|
| 14252 |
+
{
|
| 14253 |
+
"epoch": 3.984340044742729,
|
| 14254 |
+
"grad_norm": 2.390625,
|
| 14255 |
+
"learning_rate": 5.476920379221539e-10,
|
| 14256 |
+
"loss": 0.9808,
|
| 14257 |
+
"num_tokens": 78478724.0,
|
| 14258 |
+
"step": 1781
|
| 14259 |
+
},
|
| 14260 |
+
{
|
| 14261 |
+
"epoch": 3.9865771812080535,
|
| 14262 |
+
"grad_norm": 1.71875,
|
| 14263 |
+
"learning_rate": 4.1932851081005753e-10,
|
| 14264 |
+
"loss": 1.1854,
|
| 14265 |
+
"num_tokens": 78535705.0,
|
| 14266 |
+
"step": 1782
|
| 14267 |
+
},
|
| 14268 |
+
{
|
| 14269 |
+
"epoch": 3.988814317673378,
|
| 14270 |
+
"grad_norm": 2.21875,
|
| 14271 |
+
"learning_rate": 3.080792320564463e-10,
|
| 14272 |
+
"loss": 1.2227,
|
| 14273 |
+
"num_tokens": 78567714.0,
|
| 14274 |
+
"step": 1783
|
| 14275 |
+
},
|
| 14276 |
+
{
|
| 14277 |
+
"epoch": 3.9910514541387023,
|
| 14278 |
+
"grad_norm": 1.828125,
|
| 14279 |
+
"learning_rate": 2.1394458248169548e-10,
|
| 14280 |
+
"loss": 0.8626,
|
| 14281 |
+
"num_tokens": 78605188.0,
|
| 14282 |
+
"step": 1784
|
| 14283 |
+
},
|
| 14284 |
+
{
|
| 14285 |
+
"epoch": 3.9932885906040267,
|
| 14286 |
+
"grad_norm": 2.25,
|
| 14287 |
+
"learning_rate": 1.3692488432304195e-10,
|
| 14288 |
+
"loss": 0.905,
|
| 14289 |
+
"num_tokens": 78634122.0,
|
| 14290 |
+
"step": 1785
|
| 14291 |
+
},
|
| 14292 |
+
{
|
| 14293 |
+
"epoch": 3.995525727069351,
|
| 14294 |
+
"grad_norm": 1.75,
|
| 14295 |
+
"learning_rate": 7.702040122847809e-11,
|
| 14296 |
+
"loss": 0.9858,
|
| 14297 |
+
"num_tokens": 78681459.0,
|
| 14298 |
+
"step": 1786
|
| 14299 |
+
},
|
| 14300 |
+
{
|
| 14301 |
+
"epoch": 3.9977628635346756,
|
| 14302 |
+
"grad_norm": 1.7578125,
|
| 14303 |
+
"learning_rate": 3.423133825897207e-11,
|
| 14304 |
+
"loss": 1.2375,
|
| 14305 |
+
"num_tokens": 78745426.0,
|
| 14306 |
+
"step": 1787
|
| 14307 |
+
},
|
| 14308 |
+
{
|
| 14309 |
+
"epoch": 4.0,
|
| 14310 |
+
"grad_norm": 1.9609375,
|
| 14311 |
+
"learning_rate": 8.557841888467977e-12,
|
| 14312 |
+
"loss": 1.073,
|
| 14313 |
+
"num_tokens": 78788096.0,
|
| 14314 |
+
"step": 1788
|
| 14315 |
}
|
| 14316 |
],
|
| 14317 |
"logging_steps": 1,
|
|
|
|
| 14326 |
"should_evaluate": false,
|
| 14327 |
"should_log": false,
|
| 14328 |
"should_save": true,
|
| 14329 |
+
"should_training_stop": true
|
| 14330 |
},
|
| 14331 |
"attributes": {}
|
| 14332 |
}
|
| 14333 |
},
|
| 14334 |
+
"total_flos": 1.6020540287814205e+18,
|
| 14335 |
"train_batch_size": 8,
|
| 14336 |
"trial_name": null,
|
| 14337 |
"trial_params": null
|