Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix for pytorch-0.4.1 cuda-enabled build. #46562

Merged
merged 2 commits into from Sep 23, 2018

Conversation

akamaus
Copy link
Contributor

@akamaus akamaus commented Sep 12, 2018

As reported at #46032 pytorch build currently fails. The reason, I believe, is ongoing pytorch and libcaffe merge and related dependencies problems. Basically this patch fixes hard-coded dependency inside build directory.

Also openblas was updated because autograd tests failed with older version, see pytorch/pytorch#11133. Patches were merged into upstream and no longer necessary.

  • Built on platform(s)
    • [ X] NixOS

@xeji
Copy link
Contributor

xeji commented Sep 15, 2018

Maybe I don't understand the issue here. According to hydra https://hydra.nixos.org/jobset/nixpkgs/trunk/jobs-tab?filter=pytorch pytorch builds just fine.

@akamaus
Copy link
Contributor Author

akamaus commented Sep 16, 2018

@xeji. Both @jyp and me experienced problems building it ourselves. Installation process stops saying there is reference to build directory in RPATH section of one of libraries built. I suspect hydra only test cpu-only build, but most people need gpu-enabled one.

@xeji
Copy link
Contributor

xeji commented Sep 16, 2018

Thanks for explaning. The openblas update causes a mass rebuild, so please rebase this PR on the staging branch.

@akamaus
Copy link
Contributor Author

akamaus commented Sep 19, 2018

@xeji I rebased PR and checked to make sure it builds on staging with and without cuda.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

4 participants