v1.0.0 — Initial Release

🥅 Football Goalkeeper Training Using Reinforcement Learning

What's included in this release:

• Interactive 2D football simulation built with Pygame
• Rule-based AI goalkeeper with threat estimation and shot prediction
• Reinforcement Learning (PPO) agent trained using PyTorch
• State machine behavior: IDLE → TRACKING → READY → DIVING → RECOVERING
• Real-time performance metrics (saves, goals, confidence, threat level)
• Enhanced visuals: grass textures, goal nets, ball trails, and UI overlays

How to run:

1. Install dependencies: pip install pygame torch numpy
2. Run: python main.py

Known Limitations:

• Currently 2D only (3D extension planned)
• Single attacker mode
• Training requires manual episode runs