Skip to content

Conversation

@HDCharles
Copy link
Collaborator

@HDCharles HDCharles commented Nov 6, 2025

Summary

Adding divisibility check for block and group quantization to improve UX

solves vllm-project/llm-compressor#1983

Summary

Adding divisibility check for block and group quantization

Signed-off-by: HDCharles <charlesdavidhernandez@gmail.com>
Copy link
Collaborator

@brian-dellabetta brian-dellabetta left a comment

Choose a reason for hiding this comment

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

Nice, I think this is preferable to the previous closed PR. One comment:

return target_to_scheme[targets[0]]


def _validate_group_or_block_size(modules: list[tuple[str, Module]]) -> None:
Copy link
Collaborator

Choose a reason for hiding this comment

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

I think the type is incorrect for match_named_modules output

Suggested change
def _validate_group_or_block_size(modules: list[tuple[str, Module]]) -> None:
def _validate_group_or_block_size(modules: Generator[tuple[str, Module]]) -> None:

Copy link
Collaborator

@kylesayrs kylesayrs left a comment

Choose a reason for hiding this comment

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

Clarifying some things...

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.

4 participants