Skip to content

da-molchanov/variance-networks

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Variance Networks

The code for our ICLR 2019 paper on Variance Networks: When Expectation Does Not Meet Your Expectations.

Talk video

Code

We actually have two version of the code:

  • TensorFlow implementation is done with python 2.7, and will help to reproduce CIFAR results i.e. training variance layers via variational dropout.
  • PyTorch implementation is a way more accurate and reproduces results on MNIST and the toy problem. It requires python 3.6 and pytorch 0.3.

Citation

If you found this code useful please cite our paper

@article{neklyudov2018variance,
  title={Variance Networks: When Expectation Does Not Meet Your Expectations},
  author={Neklyudov, Kirill and Molchanov, Dmitry and Ashukha, Arsenii and Vetrov, Dmitry},
  journal={7th International Conference on Learning Representations},
  year={2019}
}

About

Variance Networks: When Expectation Does Not Meet Your Expectations, ICLR 2019

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published