Neural-Jump

A platformer game where AI agents learn to play through reinforcement learning.

Table of Contents

• Introduction
• Features
• Installation
• Usage
• License

Introduction

SuperMarioAI is a platformer game inspired by the classic Mario games. The project aims to implement AI agents that learn to play the game using deep reinforcement learning techniques. The AI agents navigate through levels, avoiding obstacles, collecting rewards, and improving their performance over time.

Features

• Mario-like platformer gameplay
• AI agents with neural networks
• Dynamic learning and adaptation
• Visual representation of AI decision-making
• Level design using text files

Installation

1. Clone the repository:

   git clone https://github.com/Ayfri/Neural-Jump.git
   

   Copy

2. Install the required dependencies:

   pip install -r requirements.txt
   

   Copy

3. Run the game:

   python run-ai.py
   

   Copy

Usage

1. Start the game by running the run-ai.py script.
2. Watch the AI agents learn to play the game.
3. Observe the agents' performance and decision-making process.
4. Modify the game settings and level design to experiment with different scenarios.

You can also run run-game.py to play the game manually.

License

This project is licensed under the GNU General Public License v3.0. See the LICENSE file for more information.