tkda-h3 / vamo

Variational Autoencoder for Materials Optimization

vamo

Variational Autoencoder for Materials Optimization

Required Packages

• python >= 3.4.x
• numpy >= 1.11.x
• scipy >= 0.18.x
• pandas >= 1.11.x
• tensorflow >= 1.x
• keras >= 2.x
• combo development version

Install Guide

• Install COMBO with the develop version

  1. Download or clone the github repository, e.g.

     git clone https://github.com/tsudalab/combo.git git checkout develop

  2. Run setup.py install

     cd combo python setup.py install

• Install TensorFlow

  • Install the tensorflow suitable for your machine according to the install guide shown in https://www.tensorflow.org/install/

• Install Keras

  pip3 install keras

Usage

1. run encoder_Si-Ge.ipynb in order to learn the variational encoder
2. run vamo_Si-Ge.ipynb for searching the structure with the maximum templature