An Implementation of Variational Autoencoders for Collaborative Filtering (Liang et al. 2018) in PyTorch.

This repo gives you an implementation of VAE for Collaborative Filtering in PyTorch. It's model is quite simple but powerful so i made a success reproducing it with PyTorch. Every data preprocessing step and code follows exactly from Authors' Repo.

I implemented MultiDAE too but didn't test it so MultiDAE is excluded from main.py but its code is available in models.py

PyTorch 0.4 & Python 3.6
Numpy
TensorboardX

python main.py --cuda for full training.

You should execute python data.py first to download necessary data and preprocess MovieLens-20M dataset.