AI Learns To Play Snake! Reinforcement Learning With PyTorch and Pygame

In this Python Reinforcement Learning Project I build an AI and then teach it to play Snake! I build everything using Pygame and PyTorch.

Key steps:

• Implement the Snake game.
• Implement the Agent that controls the game.
• Implement the neural network to predict the moves and train it.

After 500+ games on V1

After 1000+ games on V1

Model haven't beat its own record but it gets better averaged results every game

TODO:

• Avoid the AI to trap itself
• Don't let her get in a loop for too long
• Increase game speed to increase training speed

Dependencies

I have used Anaconda to create the python env

conda create -n pygame_env python=3.7

I used pyGame to create the game environment

pip install pygame

I used pyTorch to create the model

pip install torch torchvision

Extra deps:

pip install matplotlib ipython