Source: leela-zero Section: games Priority: optional Maintainer: Ximin Luo Build-Depends: debhelper (>= 11), cmake, libgtest-dev, libeigen3-dev, libboost-dev, libboost-program-options-dev, libboost-filesystem-dev, libopenblas-dev, opencl-headers, ocl-icd-libopencl1, ocl-icd-opencl-dev, qtbase5-dev, zlib1g-dev, clinfo , mesa-opencl-icd | opencl-icd Standards-Version: 4.1.5 Homepage: https://github.com/gcp/leela-zero Vcs-Browser: https://salsa.debian.org/infinity0/leela-zero Vcs-Git: https://salsa.debian.org/infinity0/leela-zero.git Package: leela-zero Architecture: any Depends: ${shlibs:Depends}, ${misc:Depends} Recommends: opencl-icd, clinfo Description: Go engine with no human-provided knowledge, modeled after the AlphaGo Zero paper A Go program with no human provided knowledge. Using MCTS (but without Monte Carlo playouts) and a deep residual convolutional neural network stack. . This is a fairly faithful reimplementation of the system described in the Alpha Go Zero paper "Mastering the Game of Go without Human Knowledge". For all intents and purposes, it is an open source AlphaGo Zero. . https://deepmind.com/documents/119/agz_unformatted_nature.pdf . No network weights are in this repository. If you manage to obtain the AlphaGo Zero weights, this program will be about as strong, provided you also obtain a few Tensor Processing Units. Lacking those TPUs, the author recommends a top of the line GPU - it's not exactly the same, but the result would still be an engine that is far stronger than the top humans. . Recomputing the AlphaGo Zero weights will take about 1700 years on commodity hardware. Upstream is running a public, distributed effort to repeat this work. Working together, and especially when starting on a smaller scale, it will take less than 1700 years to get a good network (which you can feed into this program, suddenly making it strong). To help with this effort, run the leelaz-autogtp binary provided in this package. The best-known network weights file is at http://zero.sjeng.org/best-network