Skip to content

README.md

Latest commit

 

History

History
38 lines (29 loc) · 917 Bytes

README.md

File metadata and controls

38 lines (29 loc) · 917 Bytes

Snake AI Game

A reinforcement learning agent that learns to play Snake using Deep Q-Learning inside a single Jupyter Notebook.

🚀 Features

  • Snake game built with pygame
  • DQN agent with replay memory + target training
  • Reward shaping:
    • +10 for eating food
    • -10 for dying
    • -0.1 step penalty
    • +1 closer to food, -1 farther
  • Model uses:
    • Adam optimizer
    • Bellman function( Mean square loss)

⚙️ Setup

git clone https://github.com/Surya-Raghuram/SnakeGameAI.git
cd SnakeGameAI
python -m venv venv
source venv/bin/activate   # or venv\Scripts\activate on Windows
pip install -r requirements.txt

Training

Run the notebook cells to train the agent. Trained models are saved as model.pth. (One pretrained model for 300 loops is included. ->Model)

image

The training graph of the pretrained model attached