Feat (ex/llm): custom quantizer plugin allows to modify quant model#1508
Open
Giuseppe5 wants to merge 2 commits into
Open
Feat (ex/llm): custom quantizer plugin allows to modify quant model#1508Giuseppe5 wants to merge 2 commits into
Giuseppe5 wants to merge 2 commits into
Conversation
nickfraser
reviewed
May 18, 2026
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Reason for this PR
In certain scenarios it is useful to have fine-grained control over what happens on your model once quantization has been applied.
This could be useful to share quantizers in scenarios like exporting to vLLM, where for certain quantization configuration it is mondatory to have Q/K/V quantizers shared, as well as Gate Proj/Up Proj.
Changes Made in this PR
Extend the current mechanism to define custom quantizer so that it also allows to define a custom function to modify the quant model.
If inheriting from BaseQuantizer, it is possible to only override one of the two methods while preserving the default behaviour (no-op) for the other one.
Testing Summary
Added tests