KEMBAR78
Add CUDA 12.8 installation and Linux CD Docker images by tinglvv · Pull Request #145557 · pytorch/pytorch · GitHub
Skip to content

Conversation

@tinglvv
Copy link
Collaborator

@tinglvv tinglvv commented Jan 23, 2025

Add CUDA 12.8 installation script
Add Magma for CUDA 12.8
Add to Linux sbsa/x86 manywheel and libtorch Dockers

TODO: Cudnn Update to 9.7.0 once it is available

@tinglvv tinglvv requested review from a team and jeffdaily as code owners January 23, 2025 23:41
@pytorch-bot
Copy link

pytorch-bot bot commented Jan 23, 2025

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/145557

Note: Links to docs will display an error until the docs builds have been completed.

❌ 5 New Failures, 10 Cancelled Jobs

As of commit 04a77bd with merge base bdc2c2a (image):

NEW FAILURES - The following jobs have failed:

CANCELLED JOBS - The following jobs were cancelled. Please retry:

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@pytorch-bot pytorch-bot bot added the topic: not user facing topic category label Jan 23, 2025
@tinglvv
Copy link
Collaborator Author

tinglvv commented Jan 23, 2025

Error
Unable to find image 'pytorch/manylinux-builder:cuda12.8-main' locally

Seems it is not possible to do all docker changes at once. Let me break this into installation script, then magma/libtorch update.

@cpuhrsch cpuhrsch requested a review from malfet January 24, 2025 03:47
@cpuhrsch cpuhrsch added the triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module label Jan 24, 2025
@johnnynunez
Copy link
Contributor

johnnynunez commented Jan 24, 2025

Add CUDA 12.8 installation script Add Magma for CUDA 12.8 Add to Linux sbsa/x86 manywheel and libtorch Dockers

TODO: Cudnn Update to 9.7.0 once it is available

Issue: #145570

cc @atalman @malfet @ptrblck @nWEIdia

new cudnn is out and tensorrt 10.8 @tinglvv
image

ldconfig
}

function prune_128 {
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We don't ship cuDNN or cuBLAS in our binaries, so do we need to nvprune anything?

@tinglvv
Copy link
Collaborator Author

tinglvv commented Jan 27, 2025

Closing to follow up in #145567.
Manylinux docker need to be built first due to
docker: Error response from daemon: manifest for pytorch/manylinux-builder:cuda12.8-main not found: manifest unknown: manifest unknown.

@tinglvv tinglvv closed this Jan 27, 2025
pytorchmergebot pushed a commit that referenced this pull request Jan 27, 2025
Breaking #145557 into two parts.
Need to have manylinux-cuda12.8 in order to build magma.

Issue: #145570

Pull Request resolved: #145567
Approved by: https://github.com/nWEIdia, https://github.com/atalman
nWEIdia pushed a commit to nWEIdia/pytorch that referenced this pull request Jan 27, 2025
Breaking pytorch#145557 into two parts.
Need to have manylinux-cuda12.8 in order to build magma.

Issue: pytorch#145570

Pull Request resolved: pytorch#145567
Approved by: https://github.com/nWEIdia, https://github.com/atalman
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

open source topic: not user facing topic category triaged This issue has been looked at a team member, and triaged and prioritized into an appropriate module

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants