An OpenAI gym Environment based on a kinetic Monte Carlo (KMC) Simulation engine for a materials synthesis problem. Most rinforcement learning (RL) environments are for robotics or control, and are not heavily physics-based. Here, we aim to utilize RL for a materials optimization problem - stochastic growth from molecular beam epitaxy or pulsed laser deposition.
You must first install the KMC engine, kmcsim (available here : https://github.com/ramav87/kmcsim/tree/Rama-branch) and then you can install this environment with
pip install -e .
Stay tuned for usable notebook with Deep Q learning