This repository provides a research platform for the card game Hanabi, offering an RL environment compatible with OpenAI Gym and a lower-level game interface for non-RL methods. It is intended for researchers and developers experimenting with AI agents for Hanabi.

How It Works

The platform offers two interfaces: rl_env.py provides a standard OpenAI Gym-like API for reinforcement learning agents, while pyhanabi.py offers a lower-level interface suitable for methods like Monte Carlo Tree Search. This dual approach allows for flexibility in agent development and experimentation.

Quick Start & Requirements

• Install: pip install . (after cloning) or pip install git+repo_url
• Prerequisites: C++ compiler (e.g., sudo apt-get install g++), pip (e.g., sudo apt-get install python-pip), numpy for examples.
• Examples: python examples/rl_env_example.py, python examples/game_example.py

Highlighted Details

• OpenAI Gym-compatible RL environment.
• Lower-level interface for non-RL methods like MCTS.
• Research platform specifically for Hanabi AI experiments.

Maintenance & Community

This is not an officially supported Google product. No community links or contributor information are provided in the README.

Licensing & Compatibility

The README does not specify a license.

Limitations & Caveats

The project is described as a research platform and is not officially supported. No specific limitations or caveats are detailed in the README.