Skip to content

Here we have the basic models to build a variational autoencoder from scratch; only using numpy. Also it comes with a formal explanation and coded in a way that it's easy to modify it.

Notifications You must be signed in to change notification settings

mcbosch/VAE-Tutorial-Scratch

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 

Repository files navigation

Variational Autoencoders

This repository is aimed to practice with Variatonial Autoencoders to model complex distributions and making generative models.

This repository is based in the following paper: Tutorial on Variational Autoencoders. That paper offers a repository where we can find a code for a VAE example for the MNIST dataset. Nevertheless, in this repository we'll build a VAE class only for practicing our coding abiities and get a deeper understanding. After that, we'll compare our results with a VAE coded with packages such as caffe or PyTorch. Additionaly, we will add a more mathematical explanaiton of the Variational Bayesian theory and how is used in this models.

I hope this repository helps anyoene in their path on untherstanding Variational Bayesian models and VAEs.

Repository Structure

Variational Bayesian Models Overview

Variational Autoencoders Overview

Formal Explanation of Backpropagation in VAEs

Bibliography

About

Here we have the basic models to build a variational autoencoder from scratch; only using numpy. Also it comes with a formal explanation and coded in a way that it's easy to modify it.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages