From 806a642eb1d68ef5bd62660547073118ffac3fd8 Mon Sep 17 00:00:00 2001 From: Mike Brown Date: Fri, 16 Jun 2023 00:20:57 +0000 Subject: [PATCH] update README.md to reflect current build from source status on master (#92729) Signed-off-by: Mike Brown To avoid new contributor issues when building master a couple README.md comments will help... This change: ~~- Documents the current support restriction to apt package `g++-9` #91328 ** noting here that with the commit in https://github.com/pytorch/pytorch/pull/92911 g++-11.3 appears to build and run local tests at least as well as g++9, so this restriction may be overcome with that PR merge depending on success and CI updates.~~ (fixed now) - Documents wip status for CUDA 12 #91122 (by forwarding to support matrix per suggestion) Pull Request resolved: https://github.com/pytorch/pytorch/pull/92729 Approved by: https://github.com/kit1980 --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 4b75dbe6e65..050ad53b5d6 100644 --- a/README.md +++ b/README.md @@ -160,8 +160,8 @@ If you are installing from source, you will need: We highly recommend installing an [Anaconda](https://www.anaconda.com/distribution/#download-section) environment. You will get a high-quality BLAS library (MKL) and you get controlled dependency versions regardless of your Linux distro. -If you want to compile with CUDA support, install the following (note that CUDA is not supported on macOS) -- [NVIDIA CUDA](https://developer.nvidia.com/cuda-downloads) 11.0 or above +If you want to compile with CUDA support, [select a supported version of CUDA from our support matrix](https://pytorch.org/get-started/locally/), then install the following: +- [NVIDIA CUDA](https://developer.nvidia.com/cuda-downloads) - [NVIDIA cuDNN](https://developer.nvidia.com/cudnn) v7 or above - [Compiler](https://gist.github.com/ax3l/9489132) compatible with CUDA