Skip to content

Conversation

@sfvaroglu
Copy link
Contributor

This PR adds support for per-channel quantization for broadcast ops (add/subtract/multiply).

@sfvaroglu
Copy link
Contributor Author

@AndrewZhaoLuo @anwang2009 @mbrookhart please review

Copy link
Contributor

@anwang2009 anwang2009 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

just some nits, LGTM!

Copy link
Contributor

@AndrewZhaoLuo AndrewZhaoLuo left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Will read a bit more deeply tomorrow

@sfvaroglu sfvaroglu requested a review from AndrewZhaoLuo March 28, 2022 18:23
Copy link
Contributor

@AndrewZhaoLuo AndrewZhaoLuo left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM, please take a look at the failing tests.

@sfvaroglu sfvaroglu force-pushed the sevin/bcast_per_channel branch from e3d2b35 to 89910c6 Compare March 29, 2022 16:44
@AndrewZhaoLuo
Copy link
Contributor

@sfvaroglu what's the status of this PR?

@sfvaroglu
Copy link
Contributor Author

I'm trying to fix the failing test. I couldn't reproduce locally (no arm-cpu), so it's going slow.

@sfvaroglu sfvaroglu force-pushed the sevin/bcast_per_channel branch from 8d86cd3 to dda96a3 Compare April 4, 2022 19:59
@sfvaroglu
Copy link
Contributor Author

@AndrewZhaoLuo This PR is ready to go.

@AndrewZhaoLuo
Copy link
Contributor

LGTM, thanks for this @sfvaroglu B)

@AndrewZhaoLuo AndrewZhaoLuo merged commit 5ec061f into apache:main Apr 5, 2022
pfk-beta pushed a commit to pfk-beta/tvm that referenced this pull request Apr 11, 2022
…0718)

* Add per-channel quantization to QNN add/subtract/multiply

* Add feedback

* Add feedback - round 2

* Fix for arm test

* Add params to the test

* Try again

* Try int

* Move lhs_axis and rhs_axis

* Add as an attribute

* Add quotes
mehrdadh pushed a commit to mehrdadh/tvm that referenced this pull request Apr 11, 2022
…0718)

* Add per-channel quantization to QNN add/subtract/multiply

* Add feedback

* Add feedback - round 2

* Fix for arm test

* Add params to the test

* Try again

* Try int

* Move lhs_axis and rhs_axis

* Add as an attribute

* Add quotes
Lucien0 pushed a commit to Lucien0/tvm that referenced this pull request Apr 19, 2022
…0718)

* Add per-channel quantization to QNN add/subtract/multiply

* Add feedback

* Add feedback - round 2

* Fix for arm test

* Add params to the test

* Try again

* Try int

* Move lhs_axis and rhs_axis

* Add as an attribute

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

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants