AIPython★ 1
MaxwellCCC/hangzhou-mahjong
Hangzhou Mahjong AI Trainer
Framework for training and evaluating Hangzhou Mahjong agents using public-information Transformers and paired branch RL, featuring the r2600 checkpoint.
// KEY FEATURES
- Implements public‑information Transformer models for Mahjong state encoding.
- Uses paired branch reinforcement learning to improve policy convergence.
- Includes pretrained r2600 checkpoint for immediate evaluation and fine‑tuning.
New repositories every 30 minutes
REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.
Join on Telegram