Re: pytorch and CUDA

["M. Zhou" <lumin@debian.org>]

Hi Andrius,

I was slowly pushing this forward (yes, very slow).
Long time ago, building pytorch cuda version is impossible due to
the lack of nvidia-cudnn. At that time I tried to package the upstream
binary but the license terms to be a blocker.

Now, I kept src:nvidia-cudnn a downloader package, and let
it Provides: libcudnn.so . We have enough build dependencies
for the pytorch-cuda variant, but I have not yet tried that.

Things remains to be done for enabling the pytorch-cuda variant:

1. build bin:pytorch-src from src:pytorch. It's beneficial to other
    variants like pytorch-rocm. In that case, we can simply reuse
    the same source package bin:pytorch-src, and create blank
    packages src:pytorch-cuda that Build-Depends: pytorch-src
    with CUDA enabled in d/rules.

    I have tried this the last year (search for "pytorch-src" in the
    git history), but some trivial bug blocked me and I eventually
    postponed this attempt.

2. toggle NO_CUDA=0, USE_CUDA=ON, USE_CUDNN=ON.
    We may also export CUDA_HOME=/usr to help the cmake
    find the nvcc executable.

    The expected extra build dependencies are:
    nvidia-cuda-toolkit, libcub-dev, nvidia-cudnn

Any help is appreciated.


On Wed, 2023-02-15 at 17:32 +0200, Andrius Merkys wrote:
> Hello,
> 
> I am interested in building and installing pytorch with CUDA support 
> locally for personal use. I noticed there are comments about
> possibility 
> of pytorch with CUDA in pytorch's files under debian/, but I would
> like 
> to see a branch in salsa repository which would build it with CUDA
> support.
> 
> Has anyone succeeded in building pytorch with CUDA? Would there be an
> interest of having CUDA-enabled salsa branch to build such unofficial
> Debian packages for personal use?
> 
> Best,
> Andrius
>