Faithful Python implementation of the paper "Towards Deep Symbolic Reinforcement Learning" by Garnelo et al.
Work in progress, please feel free to contribute.
Confirmed to work with Python 3.8.5.
pip install -r requirements.txt
or
pip install -r requirements.freeze
Run with python main.py.
- Object tracking occasionally messes up, and relations get jumbled. Cause unknown.
- Agent breaks upon encountering a cross during an episode