Update int8-int4 QAT config to use Int8DynamicActivationIntxWeightConfig#3391
Merged
Merged
Conversation
andrewor14
approved these changes
Sep 30, 2025
andrewor14
left a comment
Contributor
There was a problem hiding this comment.
Thanks, cc @danielhanchen
Contributor
Author
|
@andrewor14 @danielhanchen I don't have permission to merge |
Contributor
|
@metascroy I think only issue here is backwards compatibility. Is it possible to handle that case? Or would you not recommend running with |
Contributor
|
@mmathew23 I think unsloth QAT requires torchao 0.13.0 actually. This is a new feature added very recently and uses some primitives added in 0.13.0, so I would recommend upgrading to that version instead |
abiswas-realadvice
pushed a commit
to abiswas-realadvice/unsloth
that referenced
this pull request
May 14, 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
4 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
TorchAO is deprecating AffineQuantizedTensor (AQT), which is what Int8DynamicActivationInt4WeightConfig is based on.
We migrate int8-int4 QAT to use Int8DynamicActivationIntxWeightConfig instead, which does not depend on AQT.
cc @andrewor14