[AdithyaReddy9]

Summary:

Allow static per-tensor and per-channel affine weight specifications. Retain the channel axis when deriving bias quantization parameters. Serialize identical per-channel convolution weight zero points as the scalar required by TOSA and reject conflicting zero points explicitly.

Reviewed By: rascani

Differential Revision: D128346146

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/24337

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 2 New Failures, 1 Cancelled Job, 10 Pending

As of commit a031e6e with merge base 186ca41 ():

NEW FAILURES - The following jobs have failed:

• trunk / test-torchao-huggingface-checkpoints (lfm2_5_1_2b, mt-l-arm64g4-16-62, executorch-ubuntu-22.04-gc... / linux-job (gh)
  RuntimeError: Command bash /__w/_temp/exec_script failed with exit code 1
• trunk / test-torchao-huggingface-checkpoints (lfm2_5_1_2b, mt-l-x86iavx512-8-64, executorch-ubuntu-22.04-... / linux-job (gh)
  RuntimeError: Command bash /__w/_temp/exec_script failed with exit code 1

CANCELLED JOB - The following job was cancelled. Please retry:

• pull / nxp-mcuxpresso-test / linux-job (gh)
  Could not load credentials from any providers

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[meta-codesync]

@AdithyaReddy9 has exported this pull request. If you are a Meta employee, you can view the originating Diff in D128346146.

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: AdithyaReddy9 / name: Adithya Reddy Nallabolu (ab974ba)

[github-actions]

This PR needs a release notes: label

If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "release notes: none"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

[AdrianLundell]

Thanks for the contribution, the change looks good but please add full E2E integration tests to ensure affine weights are supported across the full pipeline, preferably using pytest and the test pipelines used in other tests, and make sure the linter passes.

[AdithyaReddy9]

Thanks for the contribution, the change looks good but please add full E2E integration tests to ensure affine weights are supported across the full pipeline, preferably using pytest and the test pipelines used in other tests, and make sure the linter passes.

Added pytest E2E coverage through quantization, export, TOSA lowering and numerical reference comparison, including nonzero affine weight zero points and bias. All 10 local tests pass

[rascani]

Looks like the lintrunner is failing (this happens sometimes with internal diffs exported). Can you fix that?

[rascani]

Review automatically exported from Phabricator review in Meta.