fix: use tarball instead of git source

conda-forge / pytorch-cpu-feedstock

A conda-smithy repository for pytorch-cpu.

BSD 3-Clause "New" or "Revised" License

17 stars 43 forks source link

fix: use tarball instead of git source #230

Closed baszalmstra closed 4 months ago

baszalmstra commented 5 months ago

Checklist

[x] Used a personal fork of the feedstock to propose changes
[x] Bumped the build number (if the version is unchanged)
[ ] Reset the build number to 0 (if the version changed)
[x] Re-rendered with the latest conda-smithy (Use the phrase code>@<space/conda-forge-admin, please rerender in a comment in this PR for automated rerendering)
[x] Ensured the license file is being packaged.

This PR is an attempt to reduce the time it takes to download the source from git by downloading a tarball provided with the release instead. This tarball includes all submodules and is created for every release. The downside is that it also contains symlinks which might be a problem on windows.

conda-forge-webservices[bot] commented 5 months ago

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

I do have some suggestions for making it better though...

For recipe:

It looks like the 'libtorch' output doesn't have any tests.

baszalmstra commented 5 months ago

@hmaarrfk Based on the one macOS build pipeline that just succeeded, this seems to work!

hmaarrfk commented 5 months ago

i'm somewhat happily shocked.... using git was almost a blocker for getting this project into conda-forge.

I'll trigger the cuda builds (i need to push to your branch) just to ensure they work, after the remaining of the OSX ones passes.

baszalmstra commented 5 months ago

@hmaarrfk All the osx builds succeeded! 🎉 Would you be able to trigger the linux builds?

hmaarrfk commented 5 months ago

changing the source is a nice to have at this point. I don't think it is worth a full rebuild (and redownload) for end users.

Perhaps try to update the version of pytorch too.

https://github.com/conda-forge/pytorch-cpu-feedstock/pull/221

might offer some clues to take.