Fix Tensor.to() #90
Merged
Fix Tensor.to() #90
Conversation
This file contains 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
By overriding a native tensor method, we prevent torch from blindly assigning the content of a QTensor to another using Tensor.data.
younesbelkada
approved these changes
Feb 20, 2024
SunMarc
reviewed
Feb 20, 2024
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Outdated suggestions cannot be applied.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.
There were several issues with
Tensor.to()inquantoTensorsubclasses.First, the
PackedTensorandQbitsTensordispatch forTensor.to()were not implemented.Second, when moving an entire
nn.Module,pytorchdoes not only callTensor.to()on the weight/bias, but also tries to do a shallow copy into the existing weight/bias.Adding a proper dispatch at the function level to tag the
QTensorclass as incompatible with shallow copy solves the issue.