[ExecuTorch][WebGPU] squeeze_copy + unsqueeze_copy test suites (cases.py op-test framework)

[JulianCloudNTH]

Stack from ghstack (oldest at bottom):

• [ExecuTorch][WebGPU] Flatten landed-op test dirs to test/ops/test_<op>.py #20435
• [ExecuTorch][WebGPU] cat op test suite (cases.py op-test framework) #20399
• [ExecuTorch][WebGPU] Add cat op + ValueList graph support (aten.cat.default) #20398
• [ExecuTorch][WebGPU] permute_copy op test suite (cases.py op-test framework) #20397
• [ExecuTorch][WebGPU] Add permute_copy + IntList graph support (aten.permute_copy.default) #20396
• [ExecuTorch][WebGPU] slice_copy op test suite (cases.py op-test framework) #20395
• [ExecuTorch][WebGPU] Add slice_copy op (aten.slice_copy.Tensor) #20394
• -> [ExecuTorch][WebGPU] squeeze_copy + unsqueeze_copy test suites (cases.py op-test framework) #20393
• [ExecuTorch][WebGPU] Add squeeze_copy + unsqueeze_copy (flat copies) #20392
• [ExecuTorch][WebGPU] sigmoid op test suite (cases.py op-test framework) #20391
• [ExecuTorch][WebGPU] Add sigmoid op (aten.sigmoid.default) #20390
• [ExecuTorch][WebGPU] select_copy op test suite (cases.py op-test framework) #20363
• [ExecuTorch][WebGPU] Add select_copy op (aten.select_copy.int) #20362
• [ExecuTorch][WebGPU] view_copy op test suite (cases.py op-test framework) #20361
• [ExecuTorch][WebGPU] Add view_copy op (aten.view_copy.default) #20360
• [ExecuTorch][WebGPU] mul op test suite (cases.py op-test framework) #20359

Registers aten.squeeze_copy.dims and aten.unsqueeze_copy.default in the cases.py op-test framework: a _squeeze_suite of 3 configs (squeeze leading/middle/multiple size-1 dims) and a _unsqueeze_suite of 3 configs (insert dim at front/middle/last) that generate_op_tests exports via VulkanPartitioner and compares to a torch golden on Dawn. Also adds test/ops/squeeze/test_squeeze.py (SqueezeModule + CONFIGS + _op_delegated smoke test), test/ops/unsqueeze/test_unsqueeze.py (UnsqueezeModule + CONFIGS + _op_delegated smoke test), and the two partitioner-allowlist entries in tester.py.
@exported-using-ghexport

Differential Revision: D108793152

Differential Revision: D108793152

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/20393

• 📄 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:

• [ROCm] MI350 CI jobs will have longer queue times due to CI migration

❌ 1 New Failure, 1 Unrelated Failure

As of commit 69e223a with merge base 8a77e7e ():

NEW FAILURE - The following job has failed:

• pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
  RuntimeError: Command docker exec -t 20c37d3ab096e9d2c03bd965c61c3cbb6e567f83cd69d9b4f9d00f777cfd38f0 /exec failed with exit code 92

BROKEN TRUNK - The following job failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• pull / test-llama-runner-qnn-linux (fp32, qnn_16a16w, qnn) / linux-job (gh) (trunk failure)

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.

[JulianCloudNTH]

@claude review

[claude]

Claude finished @JulianCloudNTH's task in 4m 10s —— View job

---

Review: squeeze_copy + unsqueeze_copy test suites

• Read changed files
• Review cases.py suites
• Review test_squeeze.py / test_unsqueeze.py
• Review tester.py allowlist entries
• Post review feedback

This is a clean, low-risk, test-only PR that follows the established per-op pattern (view_copy, select, sigmoid). The structure mirrors the existing files well. A few notes below — nothing blocking.

Correctness ✅

The key thing I checked is the overload mismatch risk: the CONFIGS use int dims (0, 1, 2) but the allowlist (tester.py:28) and the Vulkan op registry (backends/vulkan/op_registry.py:1202) only register aten.squeeze_copy.**dims**, not .dim. This is correct — the edge dialect normalizes torch.squeeze(x, int) to squeeze_copy.dims, which is exactly what the existing test_vulkan_backend_squeeze relies on (backends/vulkan/test/test_vulkan_delegate.py:1275, also an int dim). The multi config's tuple dim (0, 2) additionally exercises the multi-dim path. The _op_delegated(edge, "squeeze_copy") assertion is self-checking here — if an int dim ever fell back to CPU, the smoke test would fail loudly. Good defensive coverage.

unsqueeze configs (front=0, mid=1, last=2 on rank-2 inputs) are all valid dim positions. ✅

Minor / nits

1. Import-block formatting (cases.py:52-61) — the new squeeze/unsqueeze imports are separated by blank lines that break the otherwise contiguous alphabetical import block (note view_copy directly above has no leading blank). lintrunner/usort will likely want these regrouped. Worth running lintrunner -a before landing to avoid a CI lint failure.

2. CONFIGS key-naming inconsistency — squeeze uses dim0/mid/multi while unsqueeze uses front/mid/last. Purely cosmetic, but aligning them (e.g. both front/mid/...) would read more uniformly.

3. Comment drift between the two _op_delegated helpers — squeeze says # op must be absorbed into the delegate, not left as a CPU-fallback node. while unsqueeze adds top-level. Same logic, just inconsistent wording.

4. Helper duplication — _det_input, _lower/_export, _delegated, _op_delegated are copy-pasted across the per-op test files. This already matches the existing convention (view_copy duplicates the same), so it's consistent with the repo and CLAUDE.md's "match existing style" guidance — flagging only in case there's appetite to hoist these into a shared ops/_smoke.py helper as the op count grows. Not needed for this PR.

5. The two suites don't reuse the existing _fn_config_suite helper (cases.py:149) because that helper hardcodes construct={"fn": fn} whereas these need construct={"dim": dim}. That's a reasonable reason to inline; no change needed.

Summary

LGTM. The only thing I'd actually do before landing is run lintrunner -a for the import grouping (item 1). The rest are cosmetic. No bugs or correctness issues found.
• branch gh/JulianCloudNTH/42/head