Reproducibility code updated for current revisions is available in "revision1" branch. Please clone this specific branch for up-to-date reproducibility code.
This repository contains the code and instructions to reproduce the results from the VEGA paper.
Set up your own environment with Conda and install required dependencies:
pip install -r requirements.txt
Instructions:
-
Download archive with preprocessed data at: https://drive.google.com/file/d/1XPLH1jmc-C0OQgfUksgp-GxA0h2ik35g/view?usp=sharing
-
Unpack data to the data directory using this command in this directory:
tar zxvf /path/to/vega-data.tar.gz -C ./data --strip-components 1
- Run the bash script to train all models:
bash train_all_models.sh
Once done training, models are in ./trained_models/
- Use the jupyter notebooks to reproduce figures