Skip to content

Latest commit

 

History

History
31 lines (16 loc) · 868 Bytes

README.md

File metadata and controls

31 lines (16 loc) · 868 Bytes

NaiveGo

This is a simple gomoku program, which is a reproduction version of the AlphaZero paper.

Setting up and Running NaiveGo

If you just want to play with the provided model, run the following order in the current directory

python gui.py

And if you want to train the neural network, run the following order in the current directory

python train.py

File structure

gui.py: Interface for playing with the NavieGomoku

train.py: Interface for training neural network

gomoku_board.py: Realization of the gomoku board state

mcts.py: Performing the Monte Carlo Tree Search algorithm

alpha.py: Performing the AlphaZero Version MCTS algorithm

policy_network.py: Neural network to evaluate the current board state.

Performance

img