New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Yet another tensor refactoring #193

Merged

dacorvo merged 3 commits into main from yet_another_tensor_refactoring

May 3, 2024

Collaborator

dacorvo commented May 3, 2024

What does this PR do?

This makes QTensor an abstract class, with all its main features implemented in the new QBytesTensor class.

This is more consistent with the QBitsTensor class design:

both classes implement the QTensor API (qtype, axis, dequantize()),
both classes use the same functional dispatch (for linear mainly),
each class has its own ATEN dispatch (mostly empty for QBitsTensor, extended for QBytesTensor if axis is None).

dacorvo added 3 commits

May 3, 2024 15:29


          refactor(tensor): make QTensor an abstract class

5be3954


          refactor(tensor): remove 8-bit groupwise support

ae042ab


          test(compile): still not working with pt 2.3.0

685b415

dacorvo marked this pull request as ready for review

May 3, 2024 15:32

dacorvo merged commit 6e44e96 into main

11 checks passed

dacorvo deleted the yet_another_tensor_refactoring branch

May 3, 2024 15:32

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet