Tencent/PhoenixGo
A Go AI program implementing AlphaGo Zero's deep reinforcement learning and Monte Carlo Tree Search algorithm.
Not currently ranked — collecting fresh signals.
star history
PhoenixGo is a Go-playing AI agent that replicates the AlphaGo Zero paper’s approach, using deep neural networks trained via self-play reinforcement learning. The system employs Monte Carlo Tree Search guided by neural network evaluations for move selection. It supports GPU acceleration using CUDA, cuDNN, and TensorRT, and was the 2018 World AI Go Tournament champion.
Frequently asked
- What is Tencent/PhoenixGo?
- A Go AI program implementing AlphaGo Zero's deep reinforcement learning and Monte Carlo Tree Search algorithm.
- Is PhoenixGo open source?
- Yes — Tencent/PhoenixGo is an open-source project tracked on heatdrop.
- What language is PhoenixGo written in?
- Tencent/PhoenixGo is primarily written in C++.
- How popular is PhoenixGo?
- Tencent/PhoenixGo has 2.9k stars on GitHub.
- Where can I find PhoenixGo?
- Tencent/PhoenixGo is on GitHub at https://github.com/Tencent/PhoenixGo.