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:a9463a4cf726a44d6e4bb29306731b1d8aa8f15504d4ef9951107b616d11c491
|
| 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 sha256:569e80b0d79408c2eb641c58e972a9f9458142149a67b730f7051098afc3c432
|
| 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 sha256:67ef161df088b25299c08a70e28a1b0fb30a68b220323fecbd6d8252a34d1761
|
| 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": 1099.059,
|
| 5672 |
"eval_steps_per_second": 34.348,
|
| 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": 0.9998266196613304,
|
| 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": 1099.059,
|
| 5672 |
"eval_steps_per_second": 34.348,
|
| 5673 |
"step": 4000
|
| 5674 |
+
},
|
| 5675 |
+
{
|
| 5676 |
+
"epoch": 0.9258510084956366,
|
| 5677 |
+
"grad_norm": 177.875,
|
| 5678 |
+
"learning_rate": 4.110996916752313e-07,
|
| 5679 |
+
"loss": 65.9115,
|
| 5680 |
+
"step": 4005
|
| 5681 |
+
},
|
| 5682 |
+
{
|
| 5683 |
+
"epoch": 0.9270068774201006,
|
| 5684 |
+
"grad_norm": 165.5,
|
| 5685 |
+
"learning_rate": 4.0467625899280576e-07,
|
| 5686 |
+
"loss": 65.2806,
|
| 5687 |
+
"step": 4010
|
| 5688 |
+
},
|
| 5689 |
+
{
|
| 5690 |
+
"epoch": 0.9281627463445645,
|
| 5691 |
+
"grad_norm": 159.625,
|
| 5692 |
+
"learning_rate": 3.982528263103803e-07,
|
| 5693 |
+
"loss": 65.0467,
|
| 5694 |
+
"step": 4015
|
| 5695 |
+
},
|
| 5696 |
+
{
|
| 5697 |
+
"epoch": 0.9293186152690285,
|
| 5698 |
+
"grad_norm": 184.5,
|
| 5699 |
+
"learning_rate": 3.9182939362795487e-07,
|
| 5700 |
+
"loss": 66.6678,
|
| 5701 |
+
"step": 4020
|
| 5702 |
+
},
|
| 5703 |
+
{
|
| 5704 |
+
"epoch": 0.9304744841934924,
|
| 5705 |
+
"grad_norm": 167.875,
|
| 5706 |
+
"learning_rate": 3.8540596094552934e-07,
|
| 5707 |
+
"loss": 64.9519,
|
| 5708 |
+
"step": 4025
|
| 5709 |
+
},
|
| 5710 |
+
{
|
| 5711 |
+
"epoch": 0.9316303531179564,
|
| 5712 |
+
"grad_norm": 187.875,
|
| 5713 |
+
"learning_rate": 3.789825282631038e-07,
|
| 5714 |
+
"loss": 65.5718,
|
| 5715 |
+
"step": 4030
|
| 5716 |
+
},
|
| 5717 |
+
{
|
| 5718 |
+
"epoch": 0.9327862220424203,
|
| 5719 |
+
"grad_norm": 157.375,
|
| 5720 |
+
"learning_rate": 3.7255909558067835e-07,
|
| 5721 |
+
"loss": 65.731,
|
| 5722 |
+
"step": 4035
|
| 5723 |
+
},
|
| 5724 |
+
{
|
| 5725 |
+
"epoch": 0.9339420909668844,
|
| 5726 |
+
"grad_norm": 162.125,
|
| 5727 |
+
"learning_rate": 3.661356628982528e-07,
|
| 5728 |
+
"loss": 64.8896,
|
| 5729 |
+
"step": 4040
|
| 5730 |
+
},
|
| 5731 |
+
{
|
| 5732 |
+
"epoch": 0.9350979598913484,
|
| 5733 |
+
"grad_norm": 179.625,
|
| 5734 |
+
"learning_rate": 3.5971223021582736e-07,
|
| 5735 |
+
"loss": 65.7682,
|
| 5736 |
+
"step": 4045
|
| 5737 |
+
},
|
| 5738 |
+
{
|
| 5739 |
+
"epoch": 0.9362538288158123,
|
| 5740 |
+
"grad_norm": 183.125,
|
| 5741 |
+
"learning_rate": 3.5328879753340183e-07,
|
| 5742 |
+
"loss": 64.5395,
|
| 5743 |
+
"step": 4050
|
| 5744 |
+
},
|
| 5745 |
+
{
|
| 5746 |
+
"epoch": 0.9374096977402763,
|
| 5747 |
+
"grad_norm": 177.5,
|
| 5748 |
+
"learning_rate": 3.468653648509764e-07,
|
| 5749 |
+
"loss": 65.6702,
|
| 5750 |
+
"step": 4055
|
| 5751 |
+
},
|
| 5752 |
+
{
|
| 5753 |
+
"epoch": 0.9385655666647402,
|
| 5754 |
+
"grad_norm": 171.75,
|
| 5755 |
+
"learning_rate": 3.4044193216855094e-07,
|
| 5756 |
+
"loss": 65.3081,
|
| 5757 |
+
"step": 4060
|
| 5758 |
+
},
|
| 5759 |
+
{
|
| 5760 |
+
"epoch": 0.9397214355892042,
|
| 5761 |
+
"grad_norm": 164.0,
|
| 5762 |
+
"learning_rate": 3.340184994861254e-07,
|
| 5763 |
+
"loss": 66.0151,
|
| 5764 |
+
"step": 4065
|
| 5765 |
+
},
|
| 5766 |
+
{
|
| 5767 |
+
"epoch": 0.9408773045136681,
|
| 5768 |
+
"grad_norm": 185.125,
|
| 5769 |
+
"learning_rate": 3.2759506680369995e-07,
|
| 5770 |
+
"loss": 65.2362,
|
| 5771 |
+
"step": 4070
|
| 5772 |
+
},
|
| 5773 |
+
{
|
| 5774 |
+
"epoch": 0.9420331734381321,
|
| 5775 |
+
"grad_norm": 168.5,
|
| 5776 |
+
"learning_rate": 3.2117163412127443e-07,
|
| 5777 |
+
"loss": 64.6354,
|
| 5778 |
+
"step": 4075
|
| 5779 |
+
},
|
| 5780 |
+
{
|
| 5781 |
+
"epoch": 0.943189042362596,
|
| 5782 |
+
"grad_norm": 175.75,
|
| 5783 |
+
"learning_rate": 3.1474820143884896e-07,
|
| 5784 |
+
"loss": 65.7743,
|
| 5785 |
+
"step": 4080
|
| 5786 |
+
},
|
| 5787 |
+
{
|
| 5788 |
+
"epoch": 0.94434491128706,
|
| 5789 |
+
"grad_norm": 165.875,
|
| 5790 |
+
"learning_rate": 3.083247687564235e-07,
|
| 5791 |
+
"loss": 65.3118,
|
| 5792 |
+
"step": 4085
|
| 5793 |
+
},
|
| 5794 |
+
{
|
| 5795 |
+
"epoch": 0.9455007802115241,
|
| 5796 |
+
"grad_norm": 176.125,
|
| 5797 |
+
"learning_rate": 3.0190133607399796e-07,
|
| 5798 |
+
"loss": 67.3664,
|
| 5799 |
+
"step": 4090
|
| 5800 |
+
},
|
| 5801 |
+
{
|
| 5802 |
+
"epoch": 0.946656649135988,
|
| 5803 |
+
"grad_norm": 174.0,
|
| 5804 |
+
"learning_rate": 2.954779033915725e-07,
|
| 5805 |
+
"loss": 65.0957,
|
| 5806 |
+
"step": 4095
|
| 5807 |
+
},
|
| 5808 |
+
{
|
| 5809 |
+
"epoch": 0.947812518060452,
|
| 5810 |
+
"grad_norm": 171.875,
|
| 5811 |
+
"learning_rate": 2.89054470709147e-07,
|
| 5812 |
+
"loss": 67.5809,
|
| 5813 |
+
"step": 4100
|
| 5814 |
+
},
|
| 5815 |
+
{
|
| 5816 |
+
"epoch": 0.9489683869849159,
|
| 5817 |
+
"grad_norm": 188.375,
|
| 5818 |
+
"learning_rate": 2.826310380267215e-07,
|
| 5819 |
+
"loss": 65.4258,
|
| 5820 |
+
"step": 4105
|
| 5821 |
+
},
|
| 5822 |
+
{
|
| 5823 |
+
"epoch": 0.9501242559093799,
|
| 5824 |
+
"grad_norm": 154.25,
|
| 5825 |
+
"learning_rate": 2.7620760534429603e-07,
|
| 5826 |
+
"loss": 66.9282,
|
| 5827 |
+
"step": 4110
|
| 5828 |
+
},
|
| 5829 |
+
{
|
| 5830 |
+
"epoch": 0.9512801248338438,
|
| 5831 |
+
"grad_norm": 168.25,
|
| 5832 |
+
"learning_rate": 2.697841726618705e-07,
|
| 5833 |
+
"loss": 66.7261,
|
| 5834 |
+
"step": 4115
|
| 5835 |
+
},
|
| 5836 |
+
{
|
| 5837 |
+
"epoch": 0.9524359937583078,
|
| 5838 |
+
"grad_norm": 171.75,
|
| 5839 |
+
"learning_rate": 2.6336073997944503e-07,
|
| 5840 |
+
"loss": 68.1024,
|
| 5841 |
+
"step": 4120
|
| 5842 |
+
},
|
| 5843 |
+
{
|
| 5844 |
+
"epoch": 0.9535918626827717,
|
| 5845 |
+
"grad_norm": 179.75,
|
| 5846 |
+
"learning_rate": 2.5693730729701956e-07,
|
| 5847 |
+
"loss": 64.8889,
|
| 5848 |
+
"step": 4125
|
| 5849 |
+
},
|
| 5850 |
+
{
|
| 5851 |
+
"epoch": 0.9547477316072357,
|
| 5852 |
+
"grad_norm": 188.375,
|
| 5853 |
+
"learning_rate": 2.5051387461459404e-07,
|
| 5854 |
+
"loss": 67.7614,
|
| 5855 |
+
"step": 4130
|
| 5856 |
+
},
|
| 5857 |
+
{
|
| 5858 |
+
"epoch": 0.9559036005316998,
|
| 5859 |
+
"grad_norm": 195.0,
|
| 5860 |
+
"learning_rate": 2.4409044193216857e-07,
|
| 5861 |
+
"loss": 66.5597,
|
| 5862 |
+
"step": 4135
|
| 5863 |
+
},
|
| 5864 |
+
{
|
| 5865 |
+
"epoch": 0.9570594694561637,
|
| 5866 |
+
"grad_norm": 164.5,
|
| 5867 |
+
"learning_rate": 2.376670092497431e-07,
|
| 5868 |
+
"loss": 65.3393,
|
| 5869 |
+
"step": 4140
|
| 5870 |
+
},
|
| 5871 |
+
{
|
| 5872 |
+
"epoch": 0.9582153383806277,
|
| 5873 |
+
"grad_norm": 157.0,
|
| 5874 |
+
"learning_rate": 2.312435765673176e-07,
|
| 5875 |
+
"loss": 66.6879,
|
| 5876 |
+
"step": 4145
|
| 5877 |
+
},
|
| 5878 |
+
{
|
| 5879 |
+
"epoch": 0.9593712073050916,
|
| 5880 |
+
"grad_norm": 155.625,
|
| 5881 |
+
"learning_rate": 2.248201438848921e-07,
|
| 5882 |
+
"loss": 64.9449,
|
| 5883 |
+
"step": 4150
|
| 5884 |
+
},
|
| 5885 |
+
{
|
| 5886 |
+
"epoch": 0.9605270762295556,
|
| 5887 |
+
"grad_norm": 172.375,
|
| 5888 |
+
"learning_rate": 2.1839671120246663e-07,
|
| 5889 |
+
"loss": 67.0915,
|
| 5890 |
+
"step": 4155
|
| 5891 |
+
},
|
| 5892 |
+
{
|
| 5893 |
+
"epoch": 0.9616829451540195,
|
| 5894 |
+
"grad_norm": 162.0,
|
| 5895 |
+
"learning_rate": 2.1197327852004114e-07,
|
| 5896 |
+
"loss": 64.8401,
|
| 5897 |
+
"step": 4160
|
| 5898 |
+
},
|
| 5899 |
+
{
|
| 5900 |
+
"epoch": 0.9628388140784835,
|
| 5901 |
+
"grad_norm": 167.25,
|
| 5902 |
+
"learning_rate": 2.0554984583761564e-07,
|
| 5903 |
+
"loss": 66.7578,
|
| 5904 |
+
"step": 4165
|
| 5905 |
+
},
|
| 5906 |
+
{
|
| 5907 |
+
"epoch": 0.9639946830029474,
|
| 5908 |
+
"grad_norm": 162.875,
|
| 5909 |
+
"learning_rate": 1.9912641315519014e-07,
|
| 5910 |
+
"loss": 65.6989,
|
| 5911 |
+
"step": 4170
|
| 5912 |
+
},
|
| 5913 |
+
{
|
| 5914 |
+
"epoch": 0.9651505519274114,
|
| 5915 |
+
"grad_norm": 163.625,
|
| 5916 |
+
"learning_rate": 1.9270298047276467e-07,
|
| 5917 |
+
"loss": 65.0639,
|
| 5918 |
+
"step": 4175
|
| 5919 |
+
},
|
| 5920 |
+
{
|
| 5921 |
+
"epoch": 0.9663064208518753,
|
| 5922 |
+
"grad_norm": 165.125,
|
| 5923 |
+
"learning_rate": 1.8627954779033917e-07,
|
| 5924 |
+
"loss": 65.1387,
|
| 5925 |
+
"step": 4180
|
| 5926 |
+
},
|
| 5927 |
+
{
|
| 5928 |
+
"epoch": 0.9674622897763394,
|
| 5929 |
+
"grad_norm": 154.625,
|
| 5930 |
+
"learning_rate": 1.7985611510791368e-07,
|
| 5931 |
+
"loss": 64.6432,
|
| 5932 |
+
"step": 4185
|
| 5933 |
+
},
|
| 5934 |
+
{
|
| 5935 |
+
"epoch": 0.9686181587008034,
|
| 5936 |
+
"grad_norm": 167.0,
|
| 5937 |
+
"learning_rate": 1.734326824254882e-07,
|
| 5938 |
+
"loss": 65.3206,
|
| 5939 |
+
"step": 4190
|
| 5940 |
+
},
|
| 5941 |
+
{
|
| 5942 |
+
"epoch": 0.9697740276252673,
|
| 5943 |
+
"grad_norm": 168.0,
|
| 5944 |
+
"learning_rate": 1.670092497430627e-07,
|
| 5945 |
+
"loss": 65.6291,
|
| 5946 |
+
"step": 4195
|
| 5947 |
+
},
|
| 5948 |
+
{
|
| 5949 |
+
"epoch": 0.9709298965497313,
|
| 5950 |
+
"grad_norm": 165.25,
|
| 5951 |
+
"learning_rate": 1.6058581706063721e-07,
|
| 5952 |
+
"loss": 66.5355,
|
| 5953 |
+
"step": 4200
|
| 5954 |
+
},
|
| 5955 |
+
{
|
| 5956 |
+
"epoch": 0.9720857654741952,
|
| 5957 |
+
"grad_norm": 172.75,
|
| 5958 |
+
"learning_rate": 1.5416238437821174e-07,
|
| 5959 |
+
"loss": 65.4167,
|
| 5960 |
+
"step": 4205
|
| 5961 |
+
},
|
| 5962 |
+
{
|
| 5963 |
+
"epoch": 0.9732416343986592,
|
| 5964 |
+
"grad_norm": 164.5,
|
| 5965 |
+
"learning_rate": 1.4773895169578625e-07,
|
| 5966 |
+
"loss": 65.3302,
|
| 5967 |
+
"step": 4210
|
| 5968 |
+
},
|
| 5969 |
+
{
|
| 5970 |
+
"epoch": 0.9743975033231231,
|
| 5971 |
+
"grad_norm": 165.25,
|
| 5972 |
+
"learning_rate": 1.4131551901336075e-07,
|
| 5973 |
+
"loss": 64.7231,
|
| 5974 |
+
"step": 4215
|
| 5975 |
+
},
|
| 5976 |
+
{
|
| 5977 |
+
"epoch": 0.9755533722475871,
|
| 5978 |
+
"grad_norm": 165.125,
|
| 5979 |
+
"learning_rate": 1.3489208633093525e-07,
|
| 5980 |
+
"loss": 66.1741,
|
| 5981 |
+
"step": 4220
|
| 5982 |
+
},
|
| 5983 |
+
{
|
| 5984 |
+
"epoch": 0.976709241172051,
|
| 5985 |
+
"grad_norm": 163.625,
|
| 5986 |
+
"learning_rate": 1.2846865364850978e-07,
|
| 5987 |
+
"loss": 66.4611,
|
| 5988 |
+
"step": 4225
|
| 5989 |
+
},
|
| 5990 |
+
{
|
| 5991 |
+
"epoch": 0.977865110096515,
|
| 5992 |
+
"grad_norm": 166.25,
|
| 5993 |
+
"learning_rate": 1.2204522096608428e-07,
|
| 5994 |
+
"loss": 65.0315,
|
| 5995 |
+
"step": 4230
|
| 5996 |
+
},
|
| 5997 |
+
{
|
| 5998 |
+
"epoch": 0.9790209790209791,
|
| 5999 |
+
"grad_norm": 163.375,
|
| 6000 |
+
"learning_rate": 1.156217882836588e-07,
|
| 6001 |
+
"loss": 66.8005,
|
| 6002 |
+
"step": 4235
|
| 6003 |
+
},
|
| 6004 |
+
{
|
| 6005 |
+
"epoch": 0.980176847945443,
|
| 6006 |
+
"grad_norm": 187.375,
|
| 6007 |
+
"learning_rate": 1.0919835560123332e-07,
|
| 6008 |
+
"loss": 65.212,
|
| 6009 |
+
"step": 4240
|
| 6010 |
+
},
|
| 6011 |
+
{
|
| 6012 |
+
"epoch": 0.981332716869907,
|
| 6013 |
+
"grad_norm": 168.0,
|
| 6014 |
+
"learning_rate": 1.0277492291880782e-07,
|
| 6015 |
+
"loss": 66.4805,
|
| 6016 |
+
"step": 4245
|
| 6017 |
+
},
|
| 6018 |
+
{
|
| 6019 |
+
"epoch": 0.9824885857943709,
|
| 6020 |
+
"grad_norm": 162.25,
|
| 6021 |
+
"learning_rate": 9.635149023638234e-08,
|
| 6022 |
+
"loss": 66.4234,
|
| 6023 |
+
"step": 4250
|
| 6024 |
+
},
|
| 6025 |
+
{
|
| 6026 |
+
"epoch": 0.9836444547188349,
|
| 6027 |
+
"grad_norm": 178.375,
|
| 6028 |
+
"learning_rate": 8.992805755395684e-08,
|
| 6029 |
+
"loss": 66.0467,
|
| 6030 |
+
"step": 4255
|
| 6031 |
+
},
|
| 6032 |
+
{
|
| 6033 |
+
"epoch": 0.9848003236432988,
|
| 6034 |
+
"grad_norm": 171.25,
|
| 6035 |
+
"learning_rate": 8.350462487153136e-08,
|
| 6036 |
+
"loss": 66.8108,
|
| 6037 |
+
"step": 4260
|
| 6038 |
+
},
|
| 6039 |
+
{
|
| 6040 |
+
"epoch": 0.9859561925677628,
|
| 6041 |
+
"grad_norm": 191.875,
|
| 6042 |
+
"learning_rate": 7.708119218910587e-08,
|
| 6043 |
+
"loss": 66.4269,
|
| 6044 |
+
"step": 4265
|
| 6045 |
+
},
|
| 6046 |
+
{
|
| 6047 |
+
"epoch": 0.9871120614922267,
|
| 6048 |
+
"grad_norm": 161.125,
|
| 6049 |
+
"learning_rate": 7.065775950668037e-08,
|
| 6050 |
+
"loss": 66.1149,
|
| 6051 |
+
"step": 4270
|
| 6052 |
+
},
|
| 6053 |
+
{
|
| 6054 |
+
"epoch": 0.9882679304166907,
|
| 6055 |
+
"grad_norm": 166.5,
|
| 6056 |
+
"learning_rate": 6.423432682425489e-08,
|
| 6057 |
+
"loss": 66.4215,
|
| 6058 |
+
"step": 4275
|
| 6059 |
+
},
|
| 6060 |
+
{
|
| 6061 |
+
"epoch": 0.9894237993411548,
|
| 6062 |
+
"grad_norm": 162.0,
|
| 6063 |
+
"learning_rate": 5.78108941418294e-08,
|
| 6064 |
+
"loss": 65.648,
|
| 6065 |
+
"step": 4280
|
| 6066 |
+
},
|
| 6067 |
+
{
|
| 6068 |
+
"epoch": 0.9905796682656187,
|
| 6069 |
+
"grad_norm": 195.25,
|
| 6070 |
+
"learning_rate": 5.138746145940391e-08,
|
| 6071 |
+
"loss": 66.2166,
|
| 6072 |
+
"step": 4285
|
| 6073 |
+
},
|
| 6074 |
+
{
|
| 6075 |
+
"epoch": 0.9917355371900827,
|
| 6076 |
+
"grad_norm": 171.125,
|
| 6077 |
+
"learning_rate": 4.496402877697842e-08,
|
| 6078 |
+
"loss": 66.4645,
|
| 6079 |
+
"step": 4290
|
| 6080 |
+
},
|
| 6081 |
+
{
|
| 6082 |
+
"epoch": 0.9928914061145466,
|
| 6083 |
+
"grad_norm": 167.625,
|
| 6084 |
+
"learning_rate": 3.8540596094552936e-08,
|
| 6085 |
+
"loss": 64.63,
|
| 6086 |
+
"step": 4295
|
| 6087 |
+
},
|
| 6088 |
+
{
|
| 6089 |
+
"epoch": 0.9940472750390106,
|
| 6090 |
+
"grad_norm": 172.75,
|
| 6091 |
+
"learning_rate": 3.2117163412127445e-08,
|
| 6092 |
+
"loss": 66.4142,
|
| 6093 |
+
"step": 4300
|
| 6094 |
+
},
|
| 6095 |
+
{
|
| 6096 |
+
"epoch": 0.9952031439634745,
|
| 6097 |
+
"grad_norm": 157.375,
|
| 6098 |
+
"learning_rate": 2.5693730729701955e-08,
|
| 6099 |
+
"loss": 65.1951,
|
| 6100 |
+
"step": 4305
|
| 6101 |
+
},
|
| 6102 |
+
{
|
| 6103 |
+
"epoch": 0.9963590128879385,
|
| 6104 |
+
"grad_norm": 156.375,
|
| 6105 |
+
"learning_rate": 1.9270298047276468e-08,
|
| 6106 |
+
"loss": 66.1862,
|
| 6107 |
+
"step": 4310
|
| 6108 |
+
},
|
| 6109 |
+
{
|
| 6110 |
+
"epoch": 0.9975148818124024,
|
| 6111 |
+
"grad_norm": 171.5,
|
| 6112 |
+
"learning_rate": 1.2846865364850977e-08,
|
| 6113 |
+
"loss": 65.3223,
|
| 6114 |
+
"step": 4315
|
| 6115 |
+
},
|
| 6116 |
+
{
|
| 6117 |
+
"epoch": 0.9986707507368664,
|
| 6118 |
+
"grad_norm": 153.625,
|
| 6119 |
+
"learning_rate": 6.423432682425489e-09,
|
| 6120 |
+
"loss": 66.466,
|
| 6121 |
+
"step": 4320
|
| 6122 |
+
},
|
| 6123 |
+
{
|
| 6124 |
+
"epoch": 0.9998266196613304,
|
| 6125 |
+
"grad_norm": 177.5,
|
| 6126 |
+
"learning_rate": 0.0,
|
| 6127 |
+
"loss": 64.9395,
|
| 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
|