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

[BUG] Fix H100 crash/compat with Marlin #654

Open
wants to merge 6 commits into
base: main
Choose a base branch
from

Conversation

Qubitium
Copy link
Contributor

@Qubitium Qubitium commented Apr 27, 2024

PR:

Backport H100 crash fix from vllm marlin kernel

  • Fixed H100 crash for marlin

TESTS:

  • PASSED: test_q4.py
  • PASSED: test_quantization.py

Ref: vllm-project/vllm#4218

@Qubitium Qubitium changed the title [WIP] backport improved marlin kernel from vllm [WIP] use improved marlin kernel from vllm Apr 27, 2024
@Qubitium Qubitium changed the title [WIP] use improved marlin kernel from vllm Fix H100 compat with Marlin Apr 27, 2024
@Qubitium
Copy link
Contributor Author

@fxmarty Ready for merge. All relevant tests passing. I do not have H100 to run tests on but vllm already validated this piece of code on H100 so it should be good.

@Qubitium Qubitium changed the title Fix H100 compat with Marlin Fix H100 crash/compat with Marlin Apr 27, 2024
@Qubitium Qubitium marked this pull request as ready for review April 27, 2024 15:14
@Qubitium Qubitium changed the title Fix H100 crash/compat with Marlin [BUG] Fix H100 crash/compat with Marlin Apr 28, 2024
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.

None yet

1 participant