support AR format FP8 in vLLM#1798
Open
Zhenzhong1 wants to merge 5 commits into
Open
Conversation
Contributor
There was a problem hiding this comment.
Pull request overview
This PR aims to enable exporting block-wise FP8 quantization using the auto_round / auto_round:fp8 format (targeting vLLM compatibility), rather than forcing users to export with the standalone fp8 format.
Changes:
- Removes the format-compatibility rewrite that previously replaced
auto_roundwithfp8for block-wise FP8 configs. - Updates FP8 export to always emit
weight_block_sizefor tuplegroup_size, and additionally emitsmodules_to_not_convertalongsideignored_layersfor layers kept in high precision.
This was referenced May 11, 2026
wenhuach21
reviewed
May 12, 2026
| formats = tmp_format_name.split(",") | ||
| if isinstance(ar.group_size, tuple) and any(["auto_round" in f.lower() for f in formats]): | ||
| logger.warning( | ||
| "`auto_round` format can't be used for deploying block-wise fp8 quantization now, use `fp8` instead." |
Contributor
There was a problem hiding this comment.
Please keep the warning and change it to: auto_round:fp8 format only supports vLLM inference for now. We recommend using the FP8 format via --format fp8 instead .
wenhuach21
approved these changes
May 12, 2026
wenhuach21
reviewed
May 12, 2026
|
|
||
| if isinstance(ar.group_size, tuple) and any(["auto_round" in f.lower() for f in formats]): | ||
| logger.warning( | ||
| "auto_round:fp8 format only supports vLLM inference for now. We recommend using the FP8 format via --format fp8 instead." |
yiliu30
approved these changes
May 13, 2026
Contributor
yiliu30
left a comment
yiliu30
left a comment
There was a problem hiding this comment.
Please add some UTs; others LGTM.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Related Issues
#1536
Type of Change
New feature
Test
Output Model:
