KataGo¶
KataGo is a neural-network-based Go engine. It is one of the strongest open-source Go programs, using deep learning and Monte Carlo Tree Search (MCTS).
MOSAIC integrates KataGo as a built-in opponent for Go environments (OpenSpiel Go, PettingZoo Go).
Installation¶
KataGo requires a binary and model weights:
# Ubuntu / Debian
sudo apt-get install -y katago
# Or download from GitHub releases
# https://github.com/lightvector/KataGo/releases
Model weights are stored in var/models/katago/.
Service Module¶
The KataGo integration lives in:
gym_gui/services/go_ai/katago_service.pygym_gui/services/go_ai/gtp_engine.py(shared GTP base class)
KataGo communicates via the GTP (Go Text Protocol).
Usage in MOSAIC¶
Human vs KataGo: Play Go against the engine using the GUI.
Agent vs KataGo: Train or evaluate RL/LLM agents against KataGo.