loss 1.9
Browse files- optimizer.pt +2 -2
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +693 -3
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e910b40b828bac9676f32e9027fb4bae2807a8e503cf84aee412a09b171792a6
|
| 3 |
+
size 352532857
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 688496379
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d61e3329086073dabcac16c83a166a9b77cd0b21af46d4b24fd6ac0e6ca0dede
|
| 3 |
size 688496379
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14503
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1f260b89845223cfd8cbaf75c8c03ec7736f2f1224f8eefd7e78f7e9d52b99b
|
| 3 |
size 14503
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:28815abb7485f25e0d84f8da82443e9227d6335d5ca41dcafd1766e40fab3540
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -3756,11 +3756,701 @@
|
|
| 3756 |
"learning_rate": 0.0003211686760073857,
|
| 3757 |
"loss": 2.0217,
|
| 3758 |
"step": 62500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3759 |
}
|
| 3760 |
],
|
| 3761 |
"max_steps": 92070,
|
| 3762 |
"num_train_epochs": 10,
|
| 3763 |
-
"total_flos": 1.
|
| 3764 |
"trial_name": null,
|
| 3765 |
"trial_params": null
|
| 3766 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 8.037361861576475,
|
| 5 |
+
"global_step": 74000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 3756 |
"learning_rate": 0.0003211686760073857,
|
| 3757 |
"loss": 2.0217,
|
| 3758 |
"step": 62500
|
| 3759 |
+
},
|
| 3760 |
+
{
|
| 3761 |
+
"epoch": 6.8,
|
| 3762 |
+
"learning_rate": 0.0003200825458889975,
|
| 3763 |
+
"loss": 2.1323,
|
| 3764 |
+
"step": 62600
|
| 3765 |
+
},
|
| 3766 |
+
{
|
| 3767 |
+
"epoch": 6.81,
|
| 3768 |
+
"learning_rate": 0.00031899641577060934,
|
| 3769 |
+
"loss": 2.0287,
|
| 3770 |
+
"step": 62700
|
| 3771 |
+
},
|
| 3772 |
+
{
|
| 3773 |
+
"epoch": 6.82,
|
| 3774 |
+
"learning_rate": 0.00031791028565222114,
|
| 3775 |
+
"loss": 2.0581,
|
| 3776 |
+
"step": 62800
|
| 3777 |
+
},
|
| 3778 |
+
{
|
| 3779 |
+
"epoch": 6.83,
|
| 3780 |
+
"learning_rate": 0.000316824155533833,
|
| 3781 |
+
"loss": 2.1231,
|
| 3782 |
+
"step": 62900
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 6.84,
|
| 3786 |
+
"learning_rate": 0.0003157380254154448,
|
| 3787 |
+
"loss": 2.0398,
|
| 3788 |
+
"step": 63000
|
| 3789 |
+
},
|
| 3790 |
+
{
|
| 3791 |
+
"epoch": 6.85,
|
| 3792 |
+
"learning_rate": 0.0003146518952970566,
|
| 3793 |
+
"loss": 2.0363,
|
| 3794 |
+
"step": 63100
|
| 3795 |
+
},
|
| 3796 |
+
{
|
| 3797 |
+
"epoch": 6.86,
|
| 3798 |
+
"learning_rate": 0.0003135657651786684,
|
| 3799 |
+
"loss": 1.9909,
|
| 3800 |
+
"step": 63200
|
| 3801 |
+
},
|
| 3802 |
+
{
|
| 3803 |
+
"epoch": 6.88,
|
| 3804 |
+
"learning_rate": 0.0003124796350602802,
|
| 3805 |
+
"loss": 2.1045,
|
| 3806 |
+
"step": 63300
|
| 3807 |
+
},
|
| 3808 |
+
{
|
| 3809 |
+
"epoch": 6.89,
|
| 3810 |
+
"learning_rate": 0.00031139350494189204,
|
| 3811 |
+
"loss": 2.1585,
|
| 3812 |
+
"step": 63400
|
| 3813 |
+
},
|
| 3814 |
+
{
|
| 3815 |
+
"epoch": 6.9,
|
| 3816 |
+
"learning_rate": 0.00031030737482350384,
|
| 3817 |
+
"loss": 2.008,
|
| 3818 |
+
"step": 63500
|
| 3819 |
+
},
|
| 3820 |
+
{
|
| 3821 |
+
"epoch": 6.91,
|
| 3822 |
+
"learning_rate": 0.0003092212447051157,
|
| 3823 |
+
"loss": 2.0558,
|
| 3824 |
+
"step": 63600
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 6.92,
|
| 3828 |
+
"learning_rate": 0.0003081351145867275,
|
| 3829 |
+
"loss": 2.1185,
|
| 3830 |
+
"step": 63700
|
| 3831 |
+
},
|
| 3832 |
+
{
|
| 3833 |
+
"epoch": 6.93,
|
| 3834 |
+
"learning_rate": 0.0003070489844683393,
|
| 3835 |
+
"loss": 2.0802,
|
| 3836 |
+
"step": 63800
|
| 3837 |
+
},
|
| 3838 |
+
{
|
| 3839 |
+
"epoch": 6.94,
|
| 3840 |
+
"learning_rate": 0.00030596285434995115,
|
| 3841 |
+
"loss": 2.0766,
|
| 3842 |
+
"step": 63900
|
| 3843 |
+
},
|
| 3844 |
+
{
|
| 3845 |
+
"epoch": 6.95,
|
| 3846 |
+
"learning_rate": 0.00030487672423156294,
|
| 3847 |
+
"loss": 1.9965,
|
| 3848 |
+
"step": 64000
|
| 3849 |
+
},
|
| 3850 |
+
{
|
| 3851 |
+
"epoch": 6.96,
|
| 3852 |
+
"learning_rate": 0.0003037905941131748,
|
| 3853 |
+
"loss": 1.9354,
|
| 3854 |
+
"step": 64100
|
| 3855 |
+
},
|
| 3856 |
+
{
|
| 3857 |
+
"epoch": 6.97,
|
| 3858 |
+
"learning_rate": 0.0003027044639947866,
|
| 3859 |
+
"loss": 2.0683,
|
| 3860 |
+
"step": 64200
|
| 3861 |
+
},
|
| 3862 |
+
{
|
| 3863 |
+
"epoch": 6.98,
|
| 3864 |
+
"learning_rate": 0.0003016183338763984,
|
| 3865 |
+
"loss": 2.0038,
|
| 3866 |
+
"step": 64300
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 6.99,
|
| 3870 |
+
"learning_rate": 0.0003005322037580102,
|
| 3871 |
+
"loss": 1.9823,
|
| 3872 |
+
"step": 64400
|
| 3873 |
+
},
|
| 3874 |
+
{
|
| 3875 |
+
"epoch": 7.01,
|
| 3876 |
+
"learning_rate": 0.000299446073639622,
|
| 3877 |
+
"loss": 2.0041,
|
| 3878 |
+
"step": 64500
|
| 3879 |
+
},
|
| 3880 |
+
{
|
| 3881 |
+
"epoch": 7.02,
|
| 3882 |
+
"learning_rate": 0.00029835994352123385,
|
| 3883 |
+
"loss": 1.9983,
|
| 3884 |
+
"step": 64600
|
| 3885 |
+
},
|
| 3886 |
+
{
|
| 3887 |
+
"epoch": 7.03,
|
| 3888 |
+
"learning_rate": 0.00029727381340284565,
|
| 3889 |
+
"loss": 1.9441,
|
| 3890 |
+
"step": 64700
|
| 3891 |
+
},
|
| 3892 |
+
{
|
| 3893 |
+
"epoch": 7.04,
|
| 3894 |
+
"learning_rate": 0.0002961876832844575,
|
| 3895 |
+
"loss": 1.9004,
|
| 3896 |
+
"step": 64800
|
| 3897 |
+
},
|
| 3898 |
+
{
|
| 3899 |
+
"epoch": 7.05,
|
| 3900 |
+
"learning_rate": 0.0002951015531660693,
|
| 3901 |
+
"loss": 1.9637,
|
| 3902 |
+
"step": 64900
|
| 3903 |
+
},
|
| 3904 |
+
{
|
| 3905 |
+
"epoch": 7.06,
|
| 3906 |
+
"learning_rate": 0.00029401542304768115,
|
| 3907 |
+
"loss": 2.091,
|
| 3908 |
+
"step": 65000
|
| 3909 |
+
},
|
| 3910 |
+
{
|
| 3911 |
+
"epoch": 7.07,
|
| 3912 |
+
"learning_rate": 0.00029292929292929295,
|
| 3913 |
+
"loss": 1.9929,
|
| 3914 |
+
"step": 65100
|
| 3915 |
+
},
|
| 3916 |
+
{
|
| 3917 |
+
"epoch": 7.08,
|
| 3918 |
+
"learning_rate": 0.00029184316281090475,
|
| 3919 |
+
"loss": 1.9966,
|
| 3920 |
+
"step": 65200
|
| 3921 |
+
},
|
| 3922 |
+
{
|
| 3923 |
+
"epoch": 7.09,
|
| 3924 |
+
"learning_rate": 0.0002907570326925166,
|
| 3925 |
+
"loss": 2.0103,
|
| 3926 |
+
"step": 65300
|
| 3927 |
+
},
|
| 3928 |
+
{
|
| 3929 |
+
"epoch": 7.1,
|
| 3930 |
+
"learning_rate": 0.00028967090257412835,
|
| 3931 |
+
"loss": 2.0151,
|
| 3932 |
+
"step": 65400
|
| 3933 |
+
},
|
| 3934 |
+
{
|
| 3935 |
+
"epoch": 7.11,
|
| 3936 |
+
"learning_rate": 0.0002885847724557402,
|
| 3937 |
+
"loss": 1.9294,
|
| 3938 |
+
"step": 65500
|
| 3939 |
+
},
|
| 3940 |
+
{
|
| 3941 |
+
"epoch": 7.13,
|
| 3942 |
+
"learning_rate": 0.000287498642337352,
|
| 3943 |
+
"loss": 2.0316,
|
| 3944 |
+
"step": 65600
|
| 3945 |
+
},
|
| 3946 |
+
{
|
| 3947 |
+
"epoch": 7.14,
|
| 3948 |
+
"learning_rate": 0.00028641251221896385,
|
| 3949 |
+
"loss": 1.978,
|
| 3950 |
+
"step": 65700
|
| 3951 |
+
},
|
| 3952 |
+
{
|
| 3953 |
+
"epoch": 7.15,
|
| 3954 |
+
"learning_rate": 0.00028532638210057565,
|
| 3955 |
+
"loss": 1.9587,
|
| 3956 |
+
"step": 65800
|
| 3957 |
+
},
|
| 3958 |
+
{
|
| 3959 |
+
"epoch": 7.16,
|
| 3960 |
+
"learning_rate": 0.00028424025198218745,
|
| 3961 |
+
"loss": 1.9977,
|
| 3962 |
+
"step": 65900
|
| 3963 |
+
},
|
| 3964 |
+
{
|
| 3965 |
+
"epoch": 7.17,
|
| 3966 |
+
"learning_rate": 0.0002831541218637993,
|
| 3967 |
+
"loss": 1.9621,
|
| 3968 |
+
"step": 66000
|
| 3969 |
+
},
|
| 3970 |
+
{
|
| 3971 |
+
"epoch": 7.18,
|
| 3972 |
+
"learning_rate": 0.0002820679917454111,
|
| 3973 |
+
"loss": 2.0178,
|
| 3974 |
+
"step": 66100
|
| 3975 |
+
},
|
| 3976 |
+
{
|
| 3977 |
+
"epoch": 7.19,
|
| 3978 |
+
"learning_rate": 0.00028098186162702296,
|
| 3979 |
+
"loss": 1.9199,
|
| 3980 |
+
"step": 66200
|
| 3981 |
+
},
|
| 3982 |
+
{
|
| 3983 |
+
"epoch": 7.2,
|
| 3984 |
+
"learning_rate": 0.00027989573150863476,
|
| 3985 |
+
"loss": 1.9889,
|
| 3986 |
+
"step": 66300
|
| 3987 |
+
},
|
| 3988 |
+
{
|
| 3989 |
+
"epoch": 7.21,
|
| 3990 |
+
"learning_rate": 0.00027880960139024655,
|
| 3991 |
+
"loss": 2.0201,
|
| 3992 |
+
"step": 66400
|
| 3993 |
+
},
|
| 3994 |
+
{
|
| 3995 |
+
"epoch": 7.22,
|
| 3996 |
+
"learning_rate": 0.00027772347127185835,
|
| 3997 |
+
"loss": 1.9566,
|
| 3998 |
+
"step": 66500
|
| 3999 |
+
},
|
| 4000 |
+
{
|
| 4001 |
+
"epoch": 7.23,
|
| 4002 |
+
"learning_rate": 0.00027663734115347015,
|
| 4003 |
+
"loss": 2.0375,
|
| 4004 |
+
"step": 66600
|
| 4005 |
+
},
|
| 4006 |
+
{
|
| 4007 |
+
"epoch": 7.24,
|
| 4008 |
+
"learning_rate": 0.000275551211035082,
|
| 4009 |
+
"loss": 1.9564,
|
| 4010 |
+
"step": 66700
|
| 4011 |
+
},
|
| 4012 |
+
{
|
| 4013 |
+
"epoch": 7.26,
|
| 4014 |
+
"learning_rate": 0.0002744650809166938,
|
| 4015 |
+
"loss": 2.0215,
|
| 4016 |
+
"step": 66800
|
| 4017 |
+
},
|
| 4018 |
+
{
|
| 4019 |
+
"epoch": 7.27,
|
| 4020 |
+
"learning_rate": 0.00027337895079830566,
|
| 4021 |
+
"loss": 2.0147,
|
| 4022 |
+
"step": 66900
|
| 4023 |
+
},
|
| 4024 |
+
{
|
| 4025 |
+
"epoch": 7.28,
|
| 4026 |
+
"learning_rate": 0.00027229282067991746,
|
| 4027 |
+
"loss": 1.9813,
|
| 4028 |
+
"step": 67000
|
| 4029 |
+
},
|
| 4030 |
+
{
|
| 4031 |
+
"epoch": 7.29,
|
| 4032 |
+
"learning_rate": 0.0002712066905615293,
|
| 4033 |
+
"loss": 1.968,
|
| 4034 |
+
"step": 67100
|
| 4035 |
+
},
|
| 4036 |
+
{
|
| 4037 |
+
"epoch": 7.3,
|
| 4038 |
+
"learning_rate": 0.0002701205604431411,
|
| 4039 |
+
"loss": 1.9384,
|
| 4040 |
+
"step": 67200
|
| 4041 |
+
},
|
| 4042 |
+
{
|
| 4043 |
+
"epoch": 7.31,
|
| 4044 |
+
"learning_rate": 0.0002690344303247529,
|
| 4045 |
+
"loss": 2.0004,
|
| 4046 |
+
"step": 67300
|
| 4047 |
+
},
|
| 4048 |
+
{
|
| 4049 |
+
"epoch": 7.32,
|
| 4050 |
+
"learning_rate": 0.00026794830020636476,
|
| 4051 |
+
"loss": 2.0398,
|
| 4052 |
+
"step": 67400
|
| 4053 |
+
},
|
| 4054 |
+
{
|
| 4055 |
+
"epoch": 7.33,
|
| 4056 |
+
"learning_rate": 0.0002668621700879765,
|
| 4057 |
+
"loss": 2.0169,
|
| 4058 |
+
"step": 67500
|
| 4059 |
+
},
|
| 4060 |
+
{
|
| 4061 |
+
"epoch": 7.34,
|
| 4062 |
+
"learning_rate": 0.00026577603996958836,
|
| 4063 |
+
"loss": 2.011,
|
| 4064 |
+
"step": 67600
|
| 4065 |
+
},
|
| 4066 |
+
{
|
| 4067 |
+
"epoch": 7.35,
|
| 4068 |
+
"learning_rate": 0.00026468990985120016,
|
| 4069 |
+
"loss": 1.9144,
|
| 4070 |
+
"step": 67700
|
| 4071 |
+
},
|
| 4072 |
+
{
|
| 4073 |
+
"epoch": 7.36,
|
| 4074 |
+
"learning_rate": 0.000263603779732812,
|
| 4075 |
+
"loss": 1.989,
|
| 4076 |
+
"step": 67800
|
| 4077 |
+
},
|
| 4078 |
+
{
|
| 4079 |
+
"epoch": 7.37,
|
| 4080 |
+
"learning_rate": 0.0002625176496144238,
|
| 4081 |
+
"loss": 1.9601,
|
| 4082 |
+
"step": 67900
|
| 4083 |
+
},
|
| 4084 |
+
{
|
| 4085 |
+
"epoch": 7.39,
|
| 4086 |
+
"learning_rate": 0.0002614315194960356,
|
| 4087 |
+
"loss": 2.0163,
|
| 4088 |
+
"step": 68000
|
| 4089 |
+
},
|
| 4090 |
+
{
|
| 4091 |
+
"epoch": 7.4,
|
| 4092 |
+
"learning_rate": 0.00026034538937764746,
|
| 4093 |
+
"loss": 1.9704,
|
| 4094 |
+
"step": 68100
|
| 4095 |
+
},
|
| 4096 |
+
{
|
| 4097 |
+
"epoch": 7.41,
|
| 4098 |
+
"learning_rate": 0.00025925925925925926,
|
| 4099 |
+
"loss": 2.0961,
|
| 4100 |
+
"step": 68200
|
| 4101 |
+
},
|
| 4102 |
+
{
|
| 4103 |
+
"epoch": 7.42,
|
| 4104 |
+
"learning_rate": 0.0002581731291408711,
|
| 4105 |
+
"loss": 1.9547,
|
| 4106 |
+
"step": 68300
|
| 4107 |
+
},
|
| 4108 |
+
{
|
| 4109 |
+
"epoch": 7.43,
|
| 4110 |
+
"learning_rate": 0.0002570869990224829,
|
| 4111 |
+
"loss": 1.9263,
|
| 4112 |
+
"step": 68400
|
| 4113 |
+
},
|
| 4114 |
+
{
|
| 4115 |
+
"epoch": 7.44,
|
| 4116 |
+
"learning_rate": 0.00025600086890409477,
|
| 4117 |
+
"loss": 2.0018,
|
| 4118 |
+
"step": 68500
|
| 4119 |
+
},
|
| 4120 |
+
{
|
| 4121 |
+
"epoch": 7.45,
|
| 4122 |
+
"learning_rate": 0.0002549147387857065,
|
| 4123 |
+
"loss": 1.9383,
|
| 4124 |
+
"step": 68600
|
| 4125 |
+
},
|
| 4126 |
+
{
|
| 4127 |
+
"epoch": 7.46,
|
| 4128 |
+
"learning_rate": 0.0002538286086673183,
|
| 4129 |
+
"loss": 2.0069,
|
| 4130 |
+
"step": 68700
|
| 4131 |
+
},
|
| 4132 |
+
{
|
| 4133 |
+
"epoch": 7.47,
|
| 4134 |
+
"learning_rate": 0.00025274247854893017,
|
| 4135 |
+
"loss": 1.9821,
|
| 4136 |
+
"step": 68800
|
| 4137 |
+
},
|
| 4138 |
+
{
|
| 4139 |
+
"epoch": 7.48,
|
| 4140 |
+
"learning_rate": 0.00025165634843054196,
|
| 4141 |
+
"loss": 2.011,
|
| 4142 |
+
"step": 68900
|
| 4143 |
+
},
|
| 4144 |
+
{
|
| 4145 |
+
"epoch": 7.49,
|
| 4146 |
+
"learning_rate": 0.0002505702183121538,
|
| 4147 |
+
"loss": 1.9619,
|
| 4148 |
+
"step": 69000
|
| 4149 |
+
},
|
| 4150 |
+
{
|
| 4151 |
+
"epoch": 7.51,
|
| 4152 |
+
"learning_rate": 0.0002494840881937656,
|
| 4153 |
+
"loss": 1.9261,
|
| 4154 |
+
"step": 69100
|
| 4155 |
+
},
|
| 4156 |
+
{
|
| 4157 |
+
"epoch": 7.52,
|
| 4158 |
+
"learning_rate": 0.0002483979580753774,
|
| 4159 |
+
"loss": 2.0158,
|
| 4160 |
+
"step": 69200
|
| 4161 |
+
},
|
| 4162 |
+
{
|
| 4163 |
+
"epoch": 7.53,
|
| 4164 |
+
"learning_rate": 0.00024731182795698927,
|
| 4165 |
+
"loss": 1.9946,
|
| 4166 |
+
"step": 69300
|
| 4167 |
+
},
|
| 4168 |
+
{
|
| 4169 |
+
"epoch": 7.54,
|
| 4170 |
+
"learning_rate": 0.00024622569783860107,
|
| 4171 |
+
"loss": 2.0002,
|
| 4172 |
+
"step": 69400
|
| 4173 |
+
},
|
| 4174 |
+
{
|
| 4175 |
+
"epoch": 7.55,
|
| 4176 |
+
"learning_rate": 0.0002451395677202129,
|
| 4177 |
+
"loss": 1.9934,
|
| 4178 |
+
"step": 69500
|
| 4179 |
+
},
|
| 4180 |
+
{
|
| 4181 |
+
"epoch": 7.56,
|
| 4182 |
+
"learning_rate": 0.0002440534376018247,
|
| 4183 |
+
"loss": 2.026,
|
| 4184 |
+
"step": 69600
|
| 4185 |
+
},
|
| 4186 |
+
{
|
| 4187 |
+
"epoch": 7.57,
|
| 4188 |
+
"learning_rate": 0.00024296730748343652,
|
| 4189 |
+
"loss": 1.996,
|
| 4190 |
+
"step": 69700
|
| 4191 |
+
},
|
| 4192 |
+
{
|
| 4193 |
+
"epoch": 7.58,
|
| 4194 |
+
"learning_rate": 0.00024188117736504835,
|
| 4195 |
+
"loss": 1.9758,
|
| 4196 |
+
"step": 69800
|
| 4197 |
+
},
|
| 4198 |
+
{
|
| 4199 |
+
"epoch": 7.59,
|
| 4200 |
+
"learning_rate": 0.00024079504724666014,
|
| 4201 |
+
"loss": 1.9358,
|
| 4202 |
+
"step": 69900
|
| 4203 |
+
},
|
| 4204 |
+
{
|
| 4205 |
+
"epoch": 7.6,
|
| 4206 |
+
"learning_rate": 0.00023970891712827197,
|
| 4207 |
+
"loss": 2.0799,
|
| 4208 |
+
"step": 70000
|
| 4209 |
+
},
|
| 4210 |
+
{
|
| 4211 |
+
"epoch": 7.61,
|
| 4212 |
+
"learning_rate": 0.0002386227870098838,
|
| 4213 |
+
"loss": 2.0153,
|
| 4214 |
+
"step": 70100
|
| 4215 |
+
},
|
| 4216 |
+
{
|
| 4217 |
+
"epoch": 7.62,
|
| 4218 |
+
"learning_rate": 0.00023753665689149562,
|
| 4219 |
+
"loss": 1.9276,
|
| 4220 |
+
"step": 70200
|
| 4221 |
+
},
|
| 4222 |
+
{
|
| 4223 |
+
"epoch": 7.64,
|
| 4224 |
+
"learning_rate": 0.00023645052677310742,
|
| 4225 |
+
"loss": 1.9403,
|
| 4226 |
+
"step": 70300
|
| 4227 |
+
},
|
| 4228 |
+
{
|
| 4229 |
+
"epoch": 7.65,
|
| 4230 |
+
"learning_rate": 0.00023536439665471922,
|
| 4231 |
+
"loss": 1.8964,
|
| 4232 |
+
"step": 70400
|
| 4233 |
+
},
|
| 4234 |
+
{
|
| 4235 |
+
"epoch": 7.66,
|
| 4236 |
+
"learning_rate": 0.00023427826653633105,
|
| 4237 |
+
"loss": 1.9704,
|
| 4238 |
+
"step": 70500
|
| 4239 |
+
},
|
| 4240 |
+
{
|
| 4241 |
+
"epoch": 7.67,
|
| 4242 |
+
"learning_rate": 0.00023319213641794287,
|
| 4243 |
+
"loss": 1.9923,
|
| 4244 |
+
"step": 70600
|
| 4245 |
+
},
|
| 4246 |
+
{
|
| 4247 |
+
"epoch": 7.68,
|
| 4248 |
+
"learning_rate": 0.0002321060062995547,
|
| 4249 |
+
"loss": 2.0429,
|
| 4250 |
+
"step": 70700
|
| 4251 |
+
},
|
| 4252 |
+
{
|
| 4253 |
+
"epoch": 7.69,
|
| 4254 |
+
"learning_rate": 0.00023101987618116653,
|
| 4255 |
+
"loss": 1.8684,
|
| 4256 |
+
"step": 70800
|
| 4257 |
+
},
|
| 4258 |
+
{
|
| 4259 |
+
"epoch": 7.7,
|
| 4260 |
+
"learning_rate": 0.00022993374606277832,
|
| 4261 |
+
"loss": 2.0477,
|
| 4262 |
+
"step": 70900
|
| 4263 |
+
},
|
| 4264 |
+
{
|
| 4265 |
+
"epoch": 7.71,
|
| 4266 |
+
"learning_rate": 0.00022884761594439012,
|
| 4267 |
+
"loss": 2.0213,
|
| 4268 |
+
"step": 71000
|
| 4269 |
+
},
|
| 4270 |
+
{
|
| 4271 |
+
"epoch": 7.72,
|
| 4272 |
+
"learning_rate": 0.00022776148582600195,
|
| 4273 |
+
"loss": 2.0389,
|
| 4274 |
+
"step": 71100
|
| 4275 |
+
},
|
| 4276 |
+
{
|
| 4277 |
+
"epoch": 7.73,
|
| 4278 |
+
"learning_rate": 0.00022667535570761378,
|
| 4279 |
+
"loss": 1.9886,
|
| 4280 |
+
"step": 71200
|
| 4281 |
+
},
|
| 4282 |
+
{
|
| 4283 |
+
"epoch": 7.74,
|
| 4284 |
+
"learning_rate": 0.0002255892255892256,
|
| 4285 |
+
"loss": 1.946,
|
| 4286 |
+
"step": 71300
|
| 4287 |
+
},
|
| 4288 |
+
{
|
| 4289 |
+
"epoch": 7.75,
|
| 4290 |
+
"learning_rate": 0.00022450309547083743,
|
| 4291 |
+
"loss": 1.9715,
|
| 4292 |
+
"step": 71400
|
| 4293 |
+
},
|
| 4294 |
+
{
|
| 4295 |
+
"epoch": 7.77,
|
| 4296 |
+
"learning_rate": 0.00022341696535244923,
|
| 4297 |
+
"loss": 1.9688,
|
| 4298 |
+
"step": 71500
|
| 4299 |
+
},
|
| 4300 |
+
{
|
| 4301 |
+
"epoch": 7.78,
|
| 4302 |
+
"learning_rate": 0.00022233083523406105,
|
| 4303 |
+
"loss": 2.012,
|
| 4304 |
+
"step": 71600
|
| 4305 |
+
},
|
| 4306 |
+
{
|
| 4307 |
+
"epoch": 7.79,
|
| 4308 |
+
"learning_rate": 0.00022124470511567285,
|
| 4309 |
+
"loss": 1.9587,
|
| 4310 |
+
"step": 71700
|
| 4311 |
+
},
|
| 4312 |
+
{
|
| 4313 |
+
"epoch": 7.8,
|
| 4314 |
+
"learning_rate": 0.00022015857499728468,
|
| 4315 |
+
"loss": 1.9548,
|
| 4316 |
+
"step": 71800
|
| 4317 |
+
},
|
| 4318 |
+
{
|
| 4319 |
+
"epoch": 7.81,
|
| 4320 |
+
"learning_rate": 0.0002190724448788965,
|
| 4321 |
+
"loss": 1.9489,
|
| 4322 |
+
"step": 71900
|
| 4323 |
+
},
|
| 4324 |
+
{
|
| 4325 |
+
"epoch": 7.82,
|
| 4326 |
+
"learning_rate": 0.0002179863147605083,
|
| 4327 |
+
"loss": 1.9707,
|
| 4328 |
+
"step": 72000
|
| 4329 |
+
},
|
| 4330 |
+
{
|
| 4331 |
+
"epoch": 7.83,
|
| 4332 |
+
"learning_rate": 0.00021690018464212013,
|
| 4333 |
+
"loss": 1.9287,
|
| 4334 |
+
"step": 72100
|
| 4335 |
+
},
|
| 4336 |
+
{
|
| 4337 |
+
"epoch": 7.84,
|
| 4338 |
+
"learning_rate": 0.00021581405452373196,
|
| 4339 |
+
"loss": 1.9981,
|
| 4340 |
+
"step": 72200
|
| 4341 |
+
},
|
| 4342 |
+
{
|
| 4343 |
+
"epoch": 7.85,
|
| 4344 |
+
"learning_rate": 0.00021472792440534378,
|
| 4345 |
+
"loss": 2.0738,
|
| 4346 |
+
"step": 72300
|
| 4347 |
+
},
|
| 4348 |
+
{
|
| 4349 |
+
"epoch": 7.86,
|
| 4350 |
+
"learning_rate": 0.00021364179428695558,
|
| 4351 |
+
"loss": 1.975,
|
| 4352 |
+
"step": 72400
|
| 4353 |
+
},
|
| 4354 |
+
{
|
| 4355 |
+
"epoch": 7.87,
|
| 4356 |
+
"learning_rate": 0.00021255566416856738,
|
| 4357 |
+
"loss": 1.9052,
|
| 4358 |
+
"step": 72500
|
| 4359 |
+
},
|
| 4360 |
+
{
|
| 4361 |
+
"epoch": 7.89,
|
| 4362 |
+
"learning_rate": 0.0002114695340501792,
|
| 4363 |
+
"loss": 1.9549,
|
| 4364 |
+
"step": 72600
|
| 4365 |
+
},
|
| 4366 |
+
{
|
| 4367 |
+
"epoch": 7.9,
|
| 4368 |
+
"learning_rate": 0.00021038340393179103,
|
| 4369 |
+
"loss": 1.9218,
|
| 4370 |
+
"step": 72700
|
| 4371 |
+
},
|
| 4372 |
+
{
|
| 4373 |
+
"epoch": 7.91,
|
| 4374 |
+
"learning_rate": 0.00020929727381340286,
|
| 4375 |
+
"loss": 2.0659,
|
| 4376 |
+
"step": 72800
|
| 4377 |
+
},
|
| 4378 |
+
{
|
| 4379 |
+
"epoch": 7.92,
|
| 4380 |
+
"learning_rate": 0.00020821114369501468,
|
| 4381 |
+
"loss": 1.9476,
|
| 4382 |
+
"step": 72900
|
| 4383 |
+
},
|
| 4384 |
+
{
|
| 4385 |
+
"epoch": 7.93,
|
| 4386 |
+
"learning_rate": 0.0002071250135766265,
|
| 4387 |
+
"loss": 1.953,
|
| 4388 |
+
"step": 73000
|
| 4389 |
+
},
|
| 4390 |
+
{
|
| 4391 |
+
"epoch": 7.94,
|
| 4392 |
+
"learning_rate": 0.00020603888345823828,
|
| 4393 |
+
"loss": 1.9764,
|
| 4394 |
+
"step": 73100
|
| 4395 |
+
},
|
| 4396 |
+
{
|
| 4397 |
+
"epoch": 7.95,
|
| 4398 |
+
"learning_rate": 0.0002049527533398501,
|
| 4399 |
+
"loss": 2.0075,
|
| 4400 |
+
"step": 73200
|
| 4401 |
+
},
|
| 4402 |
+
{
|
| 4403 |
+
"epoch": 7.96,
|
| 4404 |
+
"learning_rate": 0.00020386662322146193,
|
| 4405 |
+
"loss": 1.9069,
|
| 4406 |
+
"step": 73300
|
| 4407 |
+
},
|
| 4408 |
+
{
|
| 4409 |
+
"epoch": 7.97,
|
| 4410 |
+
"learning_rate": 0.00020278049310307376,
|
| 4411 |
+
"loss": 2.0327,
|
| 4412 |
+
"step": 73400
|
| 4413 |
+
},
|
| 4414 |
+
{
|
| 4415 |
+
"epoch": 7.98,
|
| 4416 |
+
"learning_rate": 0.0002016943629846856,
|
| 4417 |
+
"loss": 1.9687,
|
| 4418 |
+
"step": 73500
|
| 4419 |
+
},
|
| 4420 |
+
{
|
| 4421 |
+
"epoch": 7.99,
|
| 4422 |
+
"learning_rate": 0.00020060823286629739,
|
| 4423 |
+
"loss": 1.9978,
|
| 4424 |
+
"step": 73600
|
| 4425 |
+
},
|
| 4426 |
+
{
|
| 4427 |
+
"epoch": 8.0,
|
| 4428 |
+
"learning_rate": 0.00019952210274790918,
|
| 4429 |
+
"loss": 1.9533,
|
| 4430 |
+
"step": 73700
|
| 4431 |
+
},
|
| 4432 |
+
{
|
| 4433 |
+
"epoch": 8.02,
|
| 4434 |
+
"learning_rate": 0.000198435972629521,
|
| 4435 |
+
"loss": 1.9278,
|
| 4436 |
+
"step": 73800
|
| 4437 |
+
},
|
| 4438 |
+
{
|
| 4439 |
+
"epoch": 8.03,
|
| 4440 |
+
"learning_rate": 0.00019734984251113284,
|
| 4441 |
+
"loss": 1.9699,
|
| 4442 |
+
"step": 73900
|
| 4443 |
+
},
|
| 4444 |
+
{
|
| 4445 |
+
"epoch": 8.04,
|
| 4446 |
+
"learning_rate": 0.00019626371239274466,
|
| 4447 |
+
"loss": 1.9263,
|
| 4448 |
+
"step": 74000
|
| 4449 |
}
|
| 4450 |
],
|
| 4451 |
"max_steps": 92070,
|
| 4452 |
"num_train_epochs": 10,
|
| 4453 |
+
"total_flos": 1.517321250322514e+17,
|
| 4454 |
"trial_name": null,
|
| 4455 |
"trial_params": null
|
| 4456 |
}
|