This project implements a neural network-based chess AI using TensorFlow and Keras. The model uses convolutional layers and residual blocks to predict the best chess moves and evaluate board states. It combines policy and value predictions to create a robust chess-playing AI, inspired by AlphaZero's architecture.
python deep-learning tensorflow neural-networks residual-networks chess-ai mixed-precision-training data-generators reinforcement-learning-concepts fen-parsing board-representation-in-machine-learning
-
Updated
Nov 20, 2024 - Python