This repository contains an implementation of variational autoencoders with the focus of their application as a binary hasher. encoder.py: this module contians the variational encoder implemetaion and only could be used for inference. decoder.py: this module contaings the variational decoder implementaion and only could be used for inference. vae_gaussian.py: this module contains both the variational decoder and encoder, and only could be used to train joinly the decoder and encoder. search.py: this module employs the hash codes generated by encoders to perform a search in particular k nearest neighbors.