Add capability to deal with linear without bias (if input Dequant zero-offset is 0)#2
Merged
Victor-Jung merged 1 commit intomainfrom Jun 10, 2025
Merged
Conversation
The current version of DeepQuant simply assumes that all linear modules have a bias, otherwise it skips unifying the Dequant nodes. This modification enables unification of Dequant blocks even when there is no `biasDequantNode`. This implementation is incomplete as it assumes that the input Dequant zeroPoint is 0.
Collaborator
|
Thanks for this contribution! This is definitely a step in the right direction for handling linear layers without bias terms I agree that the implementation is incomplete regarding zero-point handling (the assumption that input dequant zero-point = 0) and bit-width considerations, but I think this is acceptable for now. I'm down to merge this as it improves functionality without breaking existing behavior. @Victor-Jung are you in agreement? We can always enhance the zero-point and bitwidth handling in future iterations. |
Victor-Jung
approved these changes
Jun 10, 2025
Member
Victor-Jung
left a comment
Victor-Jung
left a comment
There was a problem hiding this comment.
LGTM and okay for handling zp and bw better in later PRs
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
3 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
The current version of DeepQuant simply assumes that all linear modules have a bias, otherwise it skips unifying the Dequant nodes.
This modification enables unification of Dequant blocks even when there is no
biasDequantNode.This implementation is incomplete as it assumes that the input Dequant zeroPoint is 0.