Fix regnet on XNNPACK

[GregoryComer]

Summary

The regnet_y_32gf model from torchvision does not lower on XNNPACK. This is because a "-" in a generated weight name (from conv + bn fusion) causes things to blow up. I've updated the fusion logic to properly sanitize the generated weight name and added a test for regnet to validate the fix.

Here's the part of the eager mode model responsible for this:

(trunk_output): Sequential(
    (block1): AnyStage(
      (block1-0): ResBottleneckBlock(
        (proj): Conv2dNormActivation(
          (0): Conv2d(32, 232, kernel_size=(1, 1), stride=(2, 2), bias=False)
          (1): BatchNorm2d(232, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
        )

cc @digantdesai @mcr229 @cbilgin

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/14396

• 📄 Preview Python docs built from this PR

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

❌ 1 New Failure, 3 Cancelled Jobs

As of commit 5798adb with merge base cb42db2 ():

NEW FAILURE - The following job has failed:

• Build Presets / windows (windows) / build (gh)

CANCELLED JOBS - The following jobs were cancelled. Please retry:

• Test ARM Backend / test-arm / test-backend-linux (arm_tosa, models) / linux-job (gh)
  ##[error]The operation was canceled.
• Test CoreML Backend / test-coreml / test-backend-macos (coreml_static_int8, operators) / macos-job (gh)
  ##[error]The operation was canceled.
• Test XNNPACK Backend / test-xnnpack / test-backend-linux (xnnpack_static_int8_per_tensor, models) / linux-job (gh)
  ##[error]The operation was canceled.

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

[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.

[mergennachin]

This is more performant

return name.translate(str.maketrans('.-', '__'))

[digantdesai]

causes things to blow up

Can you put the error log? I am trying to figure out what's the exact issue..

[digantdesai]

Seems like we are sensitive to names, we should put assert somewhere downstream where we check not INVALID_CHARS sneaks in..