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

Support BF16 #43

Merged
merged 8 commits into from
Dec 30, 2024
Merged

Support BF16 #43

merged 8 commits into from
Dec 30, 2024

Conversation

findmyway
Copy link
Contributor

No description provided.

chengchangxu added 3 commits December 5, 2024 17:55

Verified

This commit was signed with the committer’s verified signature.
vkubiv Volodymyr Kubiv
@findmyway
Copy link
Contributor Author

findmyway commented Dec 6, 2024

Hi @pabloferz ,the CIs are now all failed. Should we update the julia version to 1.10 considering it is the new LTS?

Project.toml Outdated Show resolved Hide resolved
src/DLPack.jl Outdated Show resolved Hide resolved
@pabloferz
Copy link
Owner

pabloferz commented Dec 28, 2024

The main issue with CI is not the julia version but some changes upstream in how the DLPack API is handled in python. This requires a separate fix for supporting versioned DLManagedTensors. I'll add a separate PR for that. In the meantime I think we can merge this. As long as your jax/pytorch/numpy version is not too recent, the package should still work.

pabloferz
pabloferz previously approved these changes Dec 29, 2024
@pabloferz
Copy link
Owner

I'm unsure about the macOS PyCall failures, but I don't have a macOS machine to debug at the moment. Let me see what I can do about it.

@pabloferz pabloferz dismissed their stale review December 29, 2024 19:08

macOS - PyCall failures

@findmyway
Copy link
Contributor Author

Thanks for the review! (Also congratulations!😄)

I'm unsure about the macOS PyCall failures, but I don't have a macOS machine to debug at the moment. Let me see what I can do about it.

This might be related. JuliaGPU/CUDA.jl#2579

Also I added some restrictions to the jax version to avoid an error similar to jax-ml/jax#15624

Overall, I think we can merge this first and wait for the release of julia@v1.10.8

@pabloferz pabloferz merged commit 36d4c55 into pabloferz:main Dec 30, 2024
6 of 8 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants