Cryolite/kanachan
A Japanese (Riichi) Mahjong AI Framework
This project helps Japanese Riichi Mahjong enthusiasts and researchers build their own AI opponents or analysis tools. It takes raw game records from Mahjong Soul and converts them into an annotated format suitable for training AI models. The output is a trained Mahjong AI model capable of making strategic decisions during gameplay. This is ideal for those looking to create an AI that can challenge top human players or existing AI programs.
332 stars. No commits in the last 6 months.
Use this if you want to train a custom Japanese Riichi Mahjong AI using your own collected Mahjong Soul game data and computational resources.
Not ideal if you are looking for a pre-trained Mahjong AI model, pre-collected game data, or a tool to simply play against an existing AI, as this project focuses on the training framework itself.
Stars
332
Forks
43
Language
Python
License
—
Category
Last pushed
Feb 16, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/transformers/Cryolite/kanachan"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
SwanHubX/SwanLab
⚡️SwanLab - an open-source, modern-design AI training tracking and visualization tool. Supports...
mdsrqbl/omnihuman
AI model that understands text & humanoids.
stas00/ml-engineering
Machine Learning Engineering Open Book
labmlai/annotated_deep_learning_paper_implementations
🧑🏫 60+ Implementations/tutorials of deep learning papers with side-by-side notes 📝; including...
analyticalrohit/AI-ML-Cheatsheets
All Stanford Cheatsheets: Artificial Intelligence, Transformers, LLMs, Deep Learning, Machine...