A machine learning-based football game where you train your own AI team using reinforcement learning. Tweak the parameters, train from scratch or continue from a checkpoint, and watch two AI teams compete against each other on the pitch.
Watch two fully trained AI teams compete against each other in a live match. The interface displays real-time match statistics including the score, ball possession percentage, completed passes, goals scored, and other key metrics for both teams. This is where all the training pays off as you see your AI players make decisions, pass, and score on their own.
The landing page of the application where you can navigate to training, matches, and your saved models.
The core of the project. Train your AI football team and monitor live statistics like reward, epochs, and performance over time. You can start a model from scratch or continue training an existing one. All key parameters are adjustable including learning rate, epochs, discount rate, and entropy coefficient, giving you full control over how your team learns.