PyTorch version on my system is 1.7.1
installed via conda. The CUDA version on my system is 11.2
and driver version is 460.32.03
. But on doing import torch
, I get an OSError
as attached below
It used to work perfectly but then CUDA Toolkit got updated. Does anybody know that is it due to the version mismatch or some other error I am unaware of? I am currently running Fedora 33 with NVIDIA GeForce GTX 1050 GPU. Any help will be appreciated. Thanks!