PyTorch implementation of the Variational Autoencoder with Spatial Broadcast Decoder.
Create a conda environment with all the requirements (edit environment.yml
if you want to change the name of the environment):
conda env create -f environment.yml
Activate the environment
source activate pytorch
Generate the data
python generate.py
Train the model
python train.py
Training on GPU for 100 epochs should take about 3 hours.
Acknowledgements
Thanks to Nicholas Watters for very helpful pointers for the implementation.