luiswyss

I am also facing this issue. It relates to the following open issue in the PyTorch repo: [https://github.com/pytorch/pytorch/issues/123097](https://github.com/pytorch/pytorch/issues/123097). I have not yet been able to find a solution.

This can be solved analogously to https://github.com/facebookresearch/esm/issues/677. By using cudatoolkit-dev instead of cudatoolkit and by downgrading mkl to mkl=2024.0.0.

I am facing the same issue and am linking the relevant issue from the openfold repo: https://github.com/aqlaboratory/openfold/issues/427

I was able to solve the issue by modifying the environment as follows: replace ` - conda-forge::cudatoolkit==11.3.*` with ` - conda-forge::cudatoolkit-dev==11.3.*` and include: ` - conda-forge::mkl=2024.0.0` since otherwise you will...