Ship the CUDA delegate in the wheel

[shoumikhin]

Closing this as a duplicate.

An earlier push failed partway through and the tooling allocated a new branch on the retry, so this pull
request and the one that superseded it were both left open with the same title and base. The live one is
the later of the two: it carries five files rather than four, and the rest of the series stacks on it.

No work is lost. Everything here plus the later fixes is in the pull request that replaced it.

[shoumikhin]

Stack from ghstack (oldest at bottom):

• -> Ship the CUDA delegate in the wheel #21643
• Ship the quantized kernels as their own library #21642
• Ship a C++ SDK in the wheel #21639
• Split the runtime out of the Python bindings extension #21610

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21643

• 📄 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, 9 Unclassified Failures

As of commit fc781ee with merge base 70594d3 ():

NEW FAILURES - The following jobs have failed:

• nightly / link-check / lint-urls (gh)
• nightly / link-check / lint-xrefs (gh)

UNCLASSIFIED FAILURES - DrCI could not classify the following jobs because the workflow did not run on the merge base. The failures may be pre-existing on trunk or introduced by this PR:

• Build Aarch64 Linux Wheels / pytorch/executorch / build-wheel-py3_14-cpu-aarch64 (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  pyarrow
• Build Aarch64 Linux Wheels / pytorch/executorch / upload / upload-wheel-py3_14-cpu-aarch64 (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  Unable to download artifact(s): Artifact not found
• Build Linux Wheels / pytorch/executorch / build-manywheel-py3_14-cpu (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  pyarrow
• Build Linux Wheels / pytorch/executorch / upload / upload-manywheel-py3_14-cpu (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  Unable to download artifact(s): Artifact not found
• Build macOS Wheels / pytorch/executorch / build-wheel-py3_14-cpu (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  pyarrow
• Build macOS Wheels / pytorch/executorch / upload / upload-wheel-py3_14-cpu (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  Unable to download artifact(s): Artifact not found
• Build Windows Wheels / pytorch/executorch / build-wheel-py3_14-cpu (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  pyarrow
• Build Windows Wheels / pytorch/executorch / upload / upload-wheel-py3_14-cpu (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  Unable to download artifact(s): Artifact not found
• Test CUDA Builds / export-model-cuda-artifact (SocialLocalMobile, Qwen3.5-35B-A3B-HQQ-INT4, quantized-int4-tile-packed) / linux-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  RuntimeError: Command bash /__w/_temp/exec_script failed with exit code 1

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