Learning Discrete Weights Using the Local Reparameterization Trick

This repository is the implementation of LR-nets, as reported in the paper "Learning Discrete Weights Using the Local Reparameterization Trick".

Authors: Oran Shayer, Dan Levi, Ethan Fetaya

tensorflow==1.x

numpy==1.16.x

For ImageNet runs, pre-trained full precision weights can be downloaded from: https://drive.google.com/file/d/1BiKBMCW8RoPf8SVpscQTDYUSBk7n8ELz/view?usp=sharing

If you use our code for research, please cite our paper:

@article{shayer2017learning,

title={Learning discrete weights using the local reparameterization trick},

author={Shayer, Oran and Levi, Dan and Fetaya, Ethan},

journal={arXiv preprint arXiv:1710.07739},

year={2017}

}