junxiaosong/AlphaZero_Gomoku resource

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

3625 stars · 1020 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

A Python implementation of the AlphaZero algorithm that trains a Gomoku (Five in a Row) AI purely through self-play using MCTS and a policy-value neural network. It supports PyTorch, TensorFlow, and Theano/Lasagne backends and is designed as an educational reference that can train a decent model on a single PC.

Use cases

• learn how the AlphaZero algorithm works
• train a Gomoku AI from scratch via self-play
• play against a self-learned board game AI
• study Monte Carlo Tree Search combined with neural networks
• experiment with swapping deep learning frameworks in an RL pipeline
• reproduce AlphaGo Zero-style training on a small game

When to choose

• you want a small, readable codebase to understand AlphaZero self-play training
• you need a Gomoku AI trainable on a single PC in hours
• you want to compare PyTorch, TensorFlow, and Theano implementations of the same RL algorithm

When to avoid

• you need a production-strength Go, chess, or Shogi engine
• you want a polished game application with a full GUI
• you need distributed or large-scale RL training infrastructure

Facets

learning-resource · maturity maintenance

reinforcement-learning machine-learning deep-learning game-engine reinforcement-learning artificial-intelligence machine-learning python cross-platform alphazero mcts gomoku self-play board-game pytorch tensorflow theano educational game-development

Member repositories

For agents

markdown · JSON · MCP: product_card(name="junxiaosong/AlphaZero_Gomoku")