Skip to content

Latest commit

 

History

History
98 lines (83 loc) · 2.93 KB

README.md

File metadata and controls

98 lines (83 loc) · 2.93 KB

BrawlAI

A solution for optimizing ranked drafting sequences in the mobile game Brawl Stars by Supercell.

Python PyTorch FastAPI TypeScript TailwindCSS Next JS React PostgreSQL

✨ Features

  • 🎯 Optimized drafting sequences for competitive play
  • ⚡ Real-time analysis and suggestions
  • 📱 Responsive, mobile-friendly interface
  • 🤖 AI-powered decision making
  • 🔄 Regular updates based on meta/balance changes

🚀 Usage

Free usage at brawl-ai.com

💻 Running Locally

Prerequisites

  • Node.js (v18 or higher)
  • Python 3.8+
  • PostgreSQL (optional, for training)
  • Brawl Stars API key

Quick Start

  1. Clone the repository
git clone https://github.com/ANDI-neV/brawl-ai.git
cd brawl-ai
  1. Set up configuration, create config.ini in backend/src:
[Credentials]
api = YOUR_API_KEY
  1. Install dependencies:
# Frontend
cd frontend
npm install

# Backend
cd ../backend
pip install -r requirements.txt
  1. Start the application
# Run frontend
npm run dev

# Run backend
python src/web_server.py

🛠 Development

Training a custom model

The project includes a pre-trained model at backend/src/out/models/ai_model.pth (model will be regularly updated in relation to new updates/balance changes), but you can train your own:

  1. Set up database configuration in config.ini (and adjust URL parameters/database loading according to your setup)
[Credentials]
username = YOUR_USERNAME
password = YOUR_PASSWORD
host = YOUR_HOST_URL
database = YOUR_DATABASE_NAME
api = YOUR_API_KEY
  1. Create the required tables in the database (battles and players)
  2. Feed the database with and adjust parameters to your liking:
python src/feeding.py
  1. Configure training parameters in ai.py
  2. Run the training script:
python src/ai.py

📝 License

GPLv3

🙏 Acknowledgments

  • Brawl Stars API for providing game data
  • BrawlAPI for providing high-quality images for brawlers and maps