Training in progress, step 4325, 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 791869518
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4433c2138519f63d9d5aa7ff598665f22cd57b538e906f9544c0f9720c57fab2
|
| 3 |
size 791869518
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2375752250
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9062e822dc347a524608daf51b672d8000ac0bfa81cd6464bba773f938a940fc
|
| 3 |
size 2375752250
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:46d4edc1ba983c008c39d87f9a0be72b701bdd2dc74240b405a1e67990d5bd14
|
| 3 |
size 1000
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 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,
|
|
@@ -5671,6 +5671,461 @@
|
|
| 5671 |
"eval_samples_per_second": 607.208,
|
| 5672 |
"eval_steps_per_second": 37.951,
|
| 5673 |
"step": 4000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5674 |
}
|
| 5675 |
],
|
| 5676 |
"logging_steps": 5,
|
|
@@ -5685,12 +6140,12 @@
|
|
| 5685 |
"should_evaluate": false,
|
| 5686 |
"should_log": false,
|
| 5687 |
"should_save": true,
|
| 5688 |
-
"should_training_stop":
|
| 5689 |
},
|
| 5690 |
"attributes": {}
|
| 5691 |
}
|
| 5692 |
},
|
| 5693 |
-
"total_flos": 1.
|
| 5694 |
"train_batch_size": 4,
|
| 5695 |
"trial_name": null,
|
| 5696 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9998338426417576,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 4325,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 5671 |
"eval_samples_per_second": 607.208,
|
| 5672 |
"eval_steps_per_second": 37.951,
|
| 5673 |
"step": 4000
|
| 5674 |
+
},
|
| 5675 |
+
{
|
| 5676 |
+
"epoch": 0.9258576970590148,
|
| 5677 |
+
"grad_norm": 155.25,
|
| 5678 |
+
"learning_rate": 8.221993833504626e-07,
|
| 5679 |
+
"loss": 68.988,
|
| 5680 |
+
"step": 4005
|
| 5681 |
+
},
|
| 5682 |
+
{
|
| 5683 |
+
"epoch": 0.9270135743337451,
|
| 5684 |
+
"grad_norm": 146.5,
|
| 5685 |
+
"learning_rate": 8.093525179856115e-07,
|
| 5686 |
+
"loss": 67.9789,
|
| 5687 |
+
"step": 4010
|
| 5688 |
+
},
|
| 5689 |
+
{
|
| 5690 |
+
"epoch": 0.9281694516084754,
|
| 5691 |
+
"grad_norm": 147.625,
|
| 5692 |
+
"learning_rate": 7.965056526207606e-07,
|
| 5693 |
+
"loss": 68.1106,
|
| 5694 |
+
"step": 4015
|
| 5695 |
+
},
|
| 5696 |
+
{
|
| 5697 |
+
"epoch": 0.9293253288832058,
|
| 5698 |
+
"grad_norm": 141.5,
|
| 5699 |
+
"learning_rate": 7.836587872559097e-07,
|
| 5700 |
+
"loss": 68.5058,
|
| 5701 |
+
"step": 4020
|
| 5702 |
+
},
|
| 5703 |
+
{
|
| 5704 |
+
"epoch": 0.9304812061579362,
|
| 5705 |
+
"grad_norm": 136.875,
|
| 5706 |
+
"learning_rate": 7.708119218910587e-07,
|
| 5707 |
+
"loss": 66.9191,
|
| 5708 |
+
"step": 4025
|
| 5709 |
+
},
|
| 5710 |
+
{
|
| 5711 |
+
"epoch": 0.9316370834326665,
|
| 5712 |
+
"grad_norm": 154.375,
|
| 5713 |
+
"learning_rate": 7.579650565262076e-07,
|
| 5714 |
+
"loss": 68.1875,
|
| 5715 |
+
"step": 4030
|
| 5716 |
+
},
|
| 5717 |
+
{
|
| 5718 |
+
"epoch": 0.9327929607073969,
|
| 5719 |
+
"grad_norm": 149.625,
|
| 5720 |
+
"learning_rate": 7.451181911613567e-07,
|
| 5721 |
+
"loss": 69.5334,
|
| 5722 |
+
"step": 4035
|
| 5723 |
+
},
|
| 5724 |
+
{
|
| 5725 |
+
"epoch": 0.9339488379821272,
|
| 5726 |
+
"grad_norm": 147.25,
|
| 5727 |
+
"learning_rate": 7.322713257965057e-07,
|
| 5728 |
+
"loss": 68.1886,
|
| 5729 |
+
"step": 4040
|
| 5730 |
+
},
|
| 5731 |
+
{
|
| 5732 |
+
"epoch": 0.9351047152568576,
|
| 5733 |
+
"grad_norm": 160.875,
|
| 5734 |
+
"learning_rate": 7.194244604316547e-07,
|
| 5735 |
+
"loss": 68.1861,
|
| 5736 |
+
"step": 4045
|
| 5737 |
+
},
|
| 5738 |
+
{
|
| 5739 |
+
"epoch": 0.936260592531588,
|
| 5740 |
+
"grad_norm": 156.375,
|
| 5741 |
+
"learning_rate": 7.065775950668037e-07,
|
| 5742 |
+
"loss": 66.4576,
|
| 5743 |
+
"step": 4050
|
| 5744 |
+
},
|
| 5745 |
+
{
|
| 5746 |
+
"epoch": 0.9374164698063183,
|
| 5747 |
+
"grad_norm": 141.625,
|
| 5748 |
+
"learning_rate": 6.937307297019528e-07,
|
| 5749 |
+
"loss": 67.8205,
|
| 5750 |
+
"step": 4055
|
| 5751 |
+
},
|
| 5752 |
+
{
|
| 5753 |
+
"epoch": 0.9385723470810486,
|
| 5754 |
+
"grad_norm": 156.25,
|
| 5755 |
+
"learning_rate": 6.808838643371019e-07,
|
| 5756 |
+
"loss": 68.9202,
|
| 5757 |
+
"step": 4060
|
| 5758 |
+
},
|
| 5759 |
+
{
|
| 5760 |
+
"epoch": 0.9397282243557791,
|
| 5761 |
+
"grad_norm": 148.5,
|
| 5762 |
+
"learning_rate": 6.680369989722508e-07,
|
| 5763 |
+
"loss": 68.9684,
|
| 5764 |
+
"step": 4065
|
| 5765 |
+
},
|
| 5766 |
+
{
|
| 5767 |
+
"epoch": 0.9408841016305094,
|
| 5768 |
+
"grad_norm": 143.5,
|
| 5769 |
+
"learning_rate": 6.551901336073999e-07,
|
| 5770 |
+
"loss": 68.7812,
|
| 5771 |
+
"step": 4070
|
| 5772 |
+
},
|
| 5773 |
+
{
|
| 5774 |
+
"epoch": 0.9420399789052397,
|
| 5775 |
+
"grad_norm": 149.75,
|
| 5776 |
+
"learning_rate": 6.423432682425489e-07,
|
| 5777 |
+
"loss": 68.0078,
|
| 5778 |
+
"step": 4075
|
| 5779 |
+
},
|
| 5780 |
+
{
|
| 5781 |
+
"epoch": 0.94319585617997,
|
| 5782 |
+
"grad_norm": 139.5,
|
| 5783 |
+
"learning_rate": 6.294964028776979e-07,
|
| 5784 |
+
"loss": 67.9145,
|
| 5785 |
+
"step": 4080
|
| 5786 |
+
},
|
| 5787 |
+
{
|
| 5788 |
+
"epoch": 0.9443517334547005,
|
| 5789 |
+
"grad_norm": 159.625,
|
| 5790 |
+
"learning_rate": 6.16649537512847e-07,
|
| 5791 |
+
"loss": 68.2476,
|
| 5792 |
+
"step": 4085
|
| 5793 |
+
},
|
| 5794 |
+
{
|
| 5795 |
+
"epoch": 0.9455076107294308,
|
| 5796 |
+
"grad_norm": 151.0,
|
| 5797 |
+
"learning_rate": 6.038026721479959e-07,
|
| 5798 |
+
"loss": 68.8499,
|
| 5799 |
+
"step": 4090
|
| 5800 |
+
},
|
| 5801 |
+
{
|
| 5802 |
+
"epoch": 0.9466634880041611,
|
| 5803 |
+
"grad_norm": 148.25,
|
| 5804 |
+
"learning_rate": 5.90955806783145e-07,
|
| 5805 |
+
"loss": 66.8455,
|
| 5806 |
+
"step": 4095
|
| 5807 |
+
},
|
| 5808 |
+
{
|
| 5809 |
+
"epoch": 0.9478193652788915,
|
| 5810 |
+
"grad_norm": 150.375,
|
| 5811 |
+
"learning_rate": 5.78108941418294e-07,
|
| 5812 |
+
"loss": 69.0889,
|
| 5813 |
+
"step": 4100
|
| 5814 |
+
},
|
| 5815 |
+
{
|
| 5816 |
+
"epoch": 0.9489752425536219,
|
| 5817 |
+
"grad_norm": 145.625,
|
| 5818 |
+
"learning_rate": 5.65262076053443e-07,
|
| 5819 |
+
"loss": 68.773,
|
| 5820 |
+
"step": 4105
|
| 5821 |
+
},
|
| 5822 |
+
{
|
| 5823 |
+
"epoch": 0.9501311198283522,
|
| 5824 |
+
"grad_norm": 145.875,
|
| 5825 |
+
"learning_rate": 5.524152106885921e-07,
|
| 5826 |
+
"loss": 68.0571,
|
| 5827 |
+
"step": 4110
|
| 5828 |
+
},
|
| 5829 |
+
{
|
| 5830 |
+
"epoch": 0.9512869971030826,
|
| 5831 |
+
"grad_norm": 137.75,
|
| 5832 |
+
"learning_rate": 5.39568345323741e-07,
|
| 5833 |
+
"loss": 68.6678,
|
| 5834 |
+
"step": 4115
|
| 5835 |
+
},
|
| 5836 |
+
{
|
| 5837 |
+
"epoch": 0.9524428743778129,
|
| 5838 |
+
"grad_norm": 152.75,
|
| 5839 |
+
"learning_rate": 5.267214799588901e-07,
|
| 5840 |
+
"loss": 70.0085,
|
| 5841 |
+
"step": 4120
|
| 5842 |
+
},
|
| 5843 |
+
{
|
| 5844 |
+
"epoch": 0.9535987516525433,
|
| 5845 |
+
"grad_norm": 157.75,
|
| 5846 |
+
"learning_rate": 5.138746145940391e-07,
|
| 5847 |
+
"loss": 67.1869,
|
| 5848 |
+
"step": 4125
|
| 5849 |
+
},
|
| 5850 |
+
{
|
| 5851 |
+
"epoch": 0.9547546289272737,
|
| 5852 |
+
"grad_norm": 154.75,
|
| 5853 |
+
"learning_rate": 5.010277492291881e-07,
|
| 5854 |
+
"loss": 69.5982,
|
| 5855 |
+
"step": 4130
|
| 5856 |
+
},
|
| 5857 |
+
{
|
| 5858 |
+
"epoch": 0.955910506202004,
|
| 5859 |
+
"grad_norm": 153.5,
|
| 5860 |
+
"learning_rate": 4.881808838643371e-07,
|
| 5861 |
+
"loss": 68.2536,
|
| 5862 |
+
"step": 4135
|
| 5863 |
+
},
|
| 5864 |
+
{
|
| 5865 |
+
"epoch": 0.9570663834767343,
|
| 5866 |
+
"grad_norm": 142.125,
|
| 5867 |
+
"learning_rate": 4.753340184994862e-07,
|
| 5868 |
+
"loss": 67.8511,
|
| 5869 |
+
"step": 4140
|
| 5870 |
+
},
|
| 5871 |
+
{
|
| 5872 |
+
"epoch": 0.9582222607514647,
|
| 5873 |
+
"grad_norm": 152.25,
|
| 5874 |
+
"learning_rate": 4.624871531346352e-07,
|
| 5875 |
+
"loss": 68.5388,
|
| 5876 |
+
"step": 4145
|
| 5877 |
+
},
|
| 5878 |
+
{
|
| 5879 |
+
"epoch": 0.9593781380261951,
|
| 5880 |
+
"grad_norm": 142.625,
|
| 5881 |
+
"learning_rate": 4.496402877697842e-07,
|
| 5882 |
+
"loss": 69.31,
|
| 5883 |
+
"step": 4150
|
| 5884 |
+
},
|
| 5885 |
+
{
|
| 5886 |
+
"epoch": 0.9605340153009254,
|
| 5887 |
+
"grad_norm": 151.125,
|
| 5888 |
+
"learning_rate": 4.3679342240493327e-07,
|
| 5889 |
+
"loss": 68.8235,
|
| 5890 |
+
"step": 4155
|
| 5891 |
+
},
|
| 5892 |
+
{
|
| 5893 |
+
"epoch": 0.9616898925756557,
|
| 5894 |
+
"grad_norm": 137.625,
|
| 5895 |
+
"learning_rate": 4.2394655704008227e-07,
|
| 5896 |
+
"loss": 67.4872,
|
| 5897 |
+
"step": 4160
|
| 5898 |
+
},
|
| 5899 |
+
{
|
| 5900 |
+
"epoch": 0.9628457698503862,
|
| 5901 |
+
"grad_norm": 148.625,
|
| 5902 |
+
"learning_rate": 4.110996916752313e-07,
|
| 5903 |
+
"loss": 68.9608,
|
| 5904 |
+
"step": 4165
|
| 5905 |
+
},
|
| 5906 |
+
{
|
| 5907 |
+
"epoch": 0.9640016471251165,
|
| 5908 |
+
"grad_norm": 134.875,
|
| 5909 |
+
"learning_rate": 3.982528263103803e-07,
|
| 5910 |
+
"loss": 68.4842,
|
| 5911 |
+
"step": 4170
|
| 5912 |
+
},
|
| 5913 |
+
{
|
| 5914 |
+
"epoch": 0.9651575243998468,
|
| 5915 |
+
"grad_norm": 151.5,
|
| 5916 |
+
"learning_rate": 3.8540596094552934e-07,
|
| 5917 |
+
"loss": 68.2285,
|
| 5918 |
+
"step": 4175
|
| 5919 |
+
},
|
| 5920 |
+
{
|
| 5921 |
+
"epoch": 0.9663134016745772,
|
| 5922 |
+
"grad_norm": 152.625,
|
| 5923 |
+
"learning_rate": 3.7255909558067835e-07,
|
| 5924 |
+
"loss": 67.0058,
|
| 5925 |
+
"step": 4180
|
| 5926 |
+
},
|
| 5927 |
+
{
|
| 5928 |
+
"epoch": 0.9674692789493076,
|
| 5929 |
+
"grad_norm": 141.625,
|
| 5930 |
+
"learning_rate": 3.5971223021582736e-07,
|
| 5931 |
+
"loss": 67.6228,
|
| 5932 |
+
"step": 4185
|
| 5933 |
+
},
|
| 5934 |
+
{
|
| 5935 |
+
"epoch": 0.9686251562240379,
|
| 5936 |
+
"grad_norm": 139.125,
|
| 5937 |
+
"learning_rate": 3.468653648509764e-07,
|
| 5938 |
+
"loss": 67.6947,
|
| 5939 |
+
"step": 4190
|
| 5940 |
+
},
|
| 5941 |
+
{
|
| 5942 |
+
"epoch": 0.9697810334987683,
|
| 5943 |
+
"grad_norm": 152.75,
|
| 5944 |
+
"learning_rate": 3.340184994861254e-07,
|
| 5945 |
+
"loss": 67.1925,
|
| 5946 |
+
"step": 4195
|
| 5947 |
+
},
|
| 5948 |
+
{
|
| 5949 |
+
"epoch": 0.9709369107734986,
|
| 5950 |
+
"grad_norm": 156.5,
|
| 5951 |
+
"learning_rate": 3.2117163412127443e-07,
|
| 5952 |
+
"loss": 67.7996,
|
| 5953 |
+
"step": 4200
|
| 5954 |
+
},
|
| 5955 |
+
{
|
| 5956 |
+
"epoch": 0.972092788048229,
|
| 5957 |
+
"grad_norm": 152.75,
|
| 5958 |
+
"learning_rate": 3.083247687564235e-07,
|
| 5959 |
+
"loss": 67.6101,
|
| 5960 |
+
"step": 4205
|
| 5961 |
+
},
|
| 5962 |
+
{
|
| 5963 |
+
"epoch": 0.9732486653229593,
|
| 5964 |
+
"grad_norm": 143.75,
|
| 5965 |
+
"learning_rate": 2.954779033915725e-07,
|
| 5966 |
+
"loss": 67.7158,
|
| 5967 |
+
"step": 4210
|
| 5968 |
+
},
|
| 5969 |
+
{
|
| 5970 |
+
"epoch": 0.9744045425976897,
|
| 5971 |
+
"grad_norm": 149.125,
|
| 5972 |
+
"learning_rate": 2.826310380267215e-07,
|
| 5973 |
+
"loss": 67.5787,
|
| 5974 |
+
"step": 4215
|
| 5975 |
+
},
|
| 5976 |
+
{
|
| 5977 |
+
"epoch": 0.97556041987242,
|
| 5978 |
+
"grad_norm": 157.625,
|
| 5979 |
+
"learning_rate": 2.697841726618705e-07,
|
| 5980 |
+
"loss": 68.4845,
|
| 5981 |
+
"step": 4220
|
| 5982 |
+
},
|
| 5983 |
+
{
|
| 5984 |
+
"epoch": 0.9767162971471504,
|
| 5985 |
+
"grad_norm": 134.125,
|
| 5986 |
+
"learning_rate": 2.5693730729701956e-07,
|
| 5987 |
+
"loss": 69.1448,
|
| 5988 |
+
"step": 4225
|
| 5989 |
+
},
|
| 5990 |
+
{
|
| 5991 |
+
"epoch": 0.9778721744218808,
|
| 5992 |
+
"grad_norm": 139.875,
|
| 5993 |
+
"learning_rate": 2.4409044193216857e-07,
|
| 5994 |
+
"loss": 68.03,
|
| 5995 |
+
"step": 4230
|
| 5996 |
+
},
|
| 5997 |
+
{
|
| 5998 |
+
"epoch": 0.9790280516966111,
|
| 5999 |
+
"grad_norm": 147.25,
|
| 6000 |
+
"learning_rate": 2.312435765673176e-07,
|
| 6001 |
+
"loss": 68.5574,
|
| 6002 |
+
"step": 4235
|
| 6003 |
+
},
|
| 6004 |
+
{
|
| 6005 |
+
"epoch": 0.9801839289713414,
|
| 6006 |
+
"grad_norm": 154.25,
|
| 6007 |
+
"learning_rate": 2.1839671120246663e-07,
|
| 6008 |
+
"loss": 67.5709,
|
| 6009 |
+
"step": 4240
|
| 6010 |
+
},
|
| 6011 |
+
{
|
| 6012 |
+
"epoch": 0.9813398062460719,
|
| 6013 |
+
"grad_norm": 149.875,
|
| 6014 |
+
"learning_rate": 2.0554984583761564e-07,
|
| 6015 |
+
"loss": 68.7623,
|
| 6016 |
+
"step": 4245
|
| 6017 |
+
},
|
| 6018 |
+
{
|
| 6019 |
+
"epoch": 0.9824956835208022,
|
| 6020 |
+
"grad_norm": 141.625,
|
| 6021 |
+
"learning_rate": 1.9270298047276467e-07,
|
| 6022 |
+
"loss": 68.6813,
|
| 6023 |
+
"step": 4250
|
| 6024 |
+
},
|
| 6025 |
+
{
|
| 6026 |
+
"epoch": 0.9836515607955325,
|
| 6027 |
+
"grad_norm": 156.0,
|
| 6028 |
+
"learning_rate": 1.7985611510791368e-07,
|
| 6029 |
+
"loss": 68.6549,
|
| 6030 |
+
"step": 4255
|
| 6031 |
+
},
|
| 6032 |
+
{
|
| 6033 |
+
"epoch": 0.9848074380702629,
|
| 6034 |
+
"grad_norm": 147.25,
|
| 6035 |
+
"learning_rate": 1.670092497430627e-07,
|
| 6036 |
+
"loss": 68.4055,
|
| 6037 |
+
"step": 4260
|
| 6038 |
+
},
|
| 6039 |
+
{
|
| 6040 |
+
"epoch": 0.9859633153449933,
|
| 6041 |
+
"grad_norm": 143.875,
|
| 6042 |
+
"learning_rate": 1.5416238437821174e-07,
|
| 6043 |
+
"loss": 68.2161,
|
| 6044 |
+
"step": 4265
|
| 6045 |
+
},
|
| 6046 |
+
{
|
| 6047 |
+
"epoch": 0.9871191926197236,
|
| 6048 |
+
"grad_norm": 155.875,
|
| 6049 |
+
"learning_rate": 1.4131551901336075e-07,
|
| 6050 |
+
"loss": 68.8521,
|
| 6051 |
+
"step": 4270
|
| 6052 |
+
},
|
| 6053 |
+
{
|
| 6054 |
+
"epoch": 0.9882750698944539,
|
| 6055 |
+
"grad_norm": 152.375,
|
| 6056 |
+
"learning_rate": 1.2846865364850978e-07,
|
| 6057 |
+
"loss": 68.8391,
|
| 6058 |
+
"step": 4275
|
| 6059 |
+
},
|
| 6060 |
+
{
|
| 6061 |
+
"epoch": 0.9894309471691843,
|
| 6062 |
+
"grad_norm": 151.125,
|
| 6063 |
+
"learning_rate": 1.156217882836588e-07,
|
| 6064 |
+
"loss": 67.5471,
|
| 6065 |
+
"step": 4280
|
| 6066 |
+
},
|
| 6067 |
+
{
|
| 6068 |
+
"epoch": 0.9905868244439147,
|
| 6069 |
+
"grad_norm": 153.875,
|
| 6070 |
+
"learning_rate": 1.0277492291880782e-07,
|
| 6071 |
+
"loss": 68.6625,
|
| 6072 |
+
"step": 4285
|
| 6073 |
+
},
|
| 6074 |
+
{
|
| 6075 |
+
"epoch": 0.991742701718645,
|
| 6076 |
+
"grad_norm": 139.375,
|
| 6077 |
+
"learning_rate": 8.992805755395684e-08,
|
| 6078 |
+
"loss": 68.1066,
|
| 6079 |
+
"step": 4290
|
| 6080 |
+
},
|
| 6081 |
+
{
|
| 6082 |
+
"epoch": 0.9928985789933754,
|
| 6083 |
+
"grad_norm": 156.375,
|
| 6084 |
+
"learning_rate": 7.708119218910587e-08,
|
| 6085 |
+
"loss": 68.376,
|
| 6086 |
+
"step": 4295
|
| 6087 |
+
},
|
| 6088 |
+
{
|
| 6089 |
+
"epoch": 0.9940544562681057,
|
| 6090 |
+
"grad_norm": 149.375,
|
| 6091 |
+
"learning_rate": 6.423432682425489e-08,
|
| 6092 |
+
"loss": 68.4129,
|
| 6093 |
+
"step": 4300
|
| 6094 |
+
},
|
| 6095 |
+
{
|
| 6096 |
+
"epoch": 0.9952103335428361,
|
| 6097 |
+
"grad_norm": 139.0,
|
| 6098 |
+
"learning_rate": 5.138746145940391e-08,
|
| 6099 |
+
"loss": 67.743,
|
| 6100 |
+
"step": 4305
|
| 6101 |
+
},
|
| 6102 |
+
{
|
| 6103 |
+
"epoch": 0.9963662108175665,
|
| 6104 |
+
"grad_norm": 137.375,
|
| 6105 |
+
"learning_rate": 3.8540596094552936e-08,
|
| 6106 |
+
"loss": 68.2336,
|
| 6107 |
+
"step": 4310
|
| 6108 |
+
},
|
| 6109 |
+
{
|
| 6110 |
+
"epoch": 0.9975220880922968,
|
| 6111 |
+
"grad_norm": 147.125,
|
| 6112 |
+
"learning_rate": 2.5693730729701955e-08,
|
| 6113 |
+
"loss": 67.8423,
|
| 6114 |
+
"step": 4315
|
| 6115 |
+
},
|
| 6116 |
+
{
|
| 6117 |
+
"epoch": 0.9986779653670271,
|
| 6118 |
+
"grad_norm": 148.875,
|
| 6119 |
+
"learning_rate": 1.2846865364850977e-08,
|
| 6120 |
+
"loss": 68.4695,
|
| 6121 |
+
"step": 4320
|
| 6122 |
+
},
|
| 6123 |
+
{
|
| 6124 |
+
"epoch": 0.9998338426417576,
|
| 6125 |
+
"grad_norm": 151.25,
|
| 6126 |
+
"learning_rate": 0.0,
|
| 6127 |
+
"loss": 67.3571,
|
| 6128 |
+
"step": 4325
|
| 6129 |
}
|
| 6130 |
],
|
| 6131 |
"logging_steps": 5,
|
|
|
|
| 6140 |
"should_evaluate": false,
|
| 6141 |
"should_log": false,
|
| 6142 |
"should_save": true,
|
| 6143 |
+
"should_training_stop": true
|
| 6144 |
},
|
| 6145 |
"attributes": {}
|
| 6146 |
}
|
| 6147 |
},
|
| 6148 |
+
"total_flos": 1.8737448318811505e+19,
|
| 6149 |
"train_batch_size": 4,
|
| 6150 |
"trial_name": null,
|
| 6151 |
"trial_params": null
|