gym-2048

Open AI gym environment for the game 2048 and agents to play it.

Getting Started

This project contains an Open AI gym environment for the game 2048 (in directory gym-2048) and some agents and tools to learn to play it.

simple_rl_agent.py - Simple RL agent which learns the quality of moves for a particular board state
train_deep_network.py - Trains a deep neural network to play from SL data
gather_training_data.py - records training data to train neural network from in CSV form

Other training_data files are for manipulating the training data.

What things you need to install the software and how to install them

For gym environment:

gym
numpy

For agents and recording training data:

pygame
tensorflow

For running tests:

pytest

You can get set up to develop this project with:

python setup.py develop

You can run the unit tests for this project by running:

py.test

Please report issues or create pull requests if you want to contribute.

This project is licensed under the MIT License - see the LICENSE.txt file for details