NXP backend: Add QuantizeFusedConvBnBiasAtenPass call to integration tests pipeline

[StrycekSimon]

Adds call to QuantizeFusedConvBnBiasAtenPass in conversion pipeline for integration tests. The pass enables QAT support for models containing biasless convolutions in integration tests.

Test plan

Covered by NXP internal tests.

cc @robert-kalmar @JakeStevens @digantdesai

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/18904

• 📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

✅ You can merge normally! (1 Unrelated Failure)

As of commit 7e596ce with merge base 4ac044b ():

FLAKY - The following job failed but was likely due to flakiness present on trunk:

• Copilot code review / Cleanup artifacts (gh) (detected as infra flaky with no log or failing log classifier)

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

[Copilot]

Pull request overview

Adds an ATen-graph transform into the NXP integration-test quantization flow to better support QAT graphs that include biasless convolutions (by ensuring fused/introduced conv biases are quantized/dequantized appropriately before lowering).

Changes:

• Import QuantizeFusedConvBnBiasAtenPass into the NXP integration-test utilities.
• Invoke QuantizeFusedConvBnBiasAtenPass(default_zero_bias=True) after convert_pt2e(...) when use_qat is enabled.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[JakeStevens]

seems nxp tests are failing with incorrect number of nodes.

Doe sthis change reduce number of q/dq since activation and batch norm are observed togetehr with conv if fusable?

[StrycekSimon]

I will look at this in a sec. I was primarily trying to resolve our internal tests hanging on this PR.

Edit: The tests expected bias to be introduced in biasless convolutions. This changed for PTQ after I introduced the conditioning of QuantizeFusedConvBnBiasAtenPass. I changed the default_zero_bias=False and adjusted the tests.

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 1 comment.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.