From aee7486b11ba7404d8487d91a932131b90464be3 Mon Sep 17 00:00:00 2001 From: Janne Hellsten Date: Wed, 13 Oct 2021 13:24:42 +0300 Subject: [PATCH] Update README and Troubleshooting for GCC versions for #2 --- README.md | 1 + docs/troubleshooting.md | 1 + 2 files changed, 2 insertions(+) diff --git a/README.md b/README.md index fcce585..7938456 100644 --- a/README.md +++ b/README.md @@ -56,6 +56,7 @@ While new generator approaches enable new media synthesis capabilities, they may * 1–8 high-end NVIDIA GPUs with at least 12 GB of memory. We have done all testing and development using Tesla V100 and A100 GPUs. * 64-bit Python 3.8 and PyTorch 1.9.0 (or later). See https://pytorch.org for PyTorch install instructions. * CUDA toolkit 11.1 or later. (Why is a separate CUDA toolkit installation required? See [Troubleshooting](./docs/troubleshooting.md#why-is-cuda-toolkit-installation-necessary)). +- GCC 7 or later (Linux) or Visual Studio (Windows) compilers. Recommended GCC version depends on CUDA version, see for example [CUDA 11.4 system requirements](https://docs.nvidia.com/cuda/archive/11.4.1/cuda-installation-guide-linux/index.html#system-requirements). * Python libraries: see [environment.yml](./environment.yml) for exact library dependencies. You can use the following commands with Miniconda3 to create and activate your StyleGAN3 Python environment: - `conda env create -f environment.yml` - `conda activate stylegan3` diff --git a/docs/troubleshooting.md b/docs/troubleshooting.md index b7218b7..a96c143 100644 --- a/docs/troubleshooting.md +++ b/docs/troubleshooting.md @@ -18,6 +18,7 @@ Make sure you've installed everything listed on the requirements section in the - **ninja** - PyTorch uses [Ninja](https://ninja-build.org/) as its build system. - **GCC** (Linux) or **Visual Studio** (Windows) + - GCC 7.x or later is required. Earlier versions such as GCC 6.3 [are known not to work](https://github.com/NVlabs/stylegan3/issues/2). #### Why is CUDA toolkit installation necessary?