You are not logged in.
In the official repo, e.g. python-pytorch-cuda is always compatible with cuda(v13.1).
But in venv pytorch is installed by pip,
pytorch installed by
pip3 install torch torchvision --index-url https://download.pytorch.org/whl/cu130Is it compatible with cuda 13.1 in the official repo?
Is it better to use conda to make torch and cuda always compatible?
Or what is the best practice to install pytorch with cuda support in venv?
Last edited by jronald (2026-01-13 05:30:52)
Offline
`pip3 install torch torchvision --index-url https://download.pytorch.org/whl/cu130` will install cuda 13.0 in venv
Offline
Is it better to use conda to make torch and cuda always compatible?
Only if you use torch & cuda exclusively with conda.
Keep in mind conda is known to break archlinux functionality unless conda is blocked from interfering with system installed python .
Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.
clean chroot building not flexible enough ?
Try clean chroot manager by graysky
Offline
uv in enough for me now,
and pytorch also doesn't support conda any more,
so I'll not use conda now.
Offline