Add Qwen 3.6 MoE model and switch CI to Qwen3.6-35B-A3B-HQQ-INT4

[mergennachin]

Qwen 3.6 MoE shares architecture and runner with Qwen 3.5 MoE. Add a stub README pointing to the existing qwen3_5_moe example. Update CI scripts and cuda.yml to use the Qwen 3.6 prequantized checkpoint. Improve qwen3_5_moe README: add quick-start section for prequantized weights, list available prequantized checkpoints, and clean up terminology.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/18955

• 📄 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:

• [CI[B200] Smoke test encounters CUDA Unknown error for dgxb200-03 and dgxb200-04

❌ 5 New Failures, 1 Cancelled Job, 3 Unrelated Failures

As of commit 655fa02 with merge base 75ba558 ():

NEW FAILURES - The following jobs have failed:

• pull / android / build-android (gh)
  Process completed with exit code 2.
• pull / unittest-editable / macos / macos-job (gh)
  export/tests/test_target_recipes.py::TestTargetRecipes::test_mv3_model
• trunk / test-torchao-huggingface-checkpoints (lfm2_5_1_2b, linux.arm64.2xlarge, executorch-ubuntu-22.04-g... / linux-job (gh)
  RuntimeError: Command docker exec -t f506c7c3c2d63321f325511db37a0c7fa73a5be23adf4185605eab450a90f7ec /exec failed with exit code 1
• trunk / test-torchao-huggingface-checkpoints (phi_4_mini, linux.arm64.2xlarge, executorch-ubuntu-22.04-gc... / linux-job (gh)
  RuntimeError: Command docker exec -t 87c341acc03066a6b1e22fdf04ff48899497d8fd98ce44d3421f7b024783f8ae /exec failed with exit code 1
• trunk / test-torchao-huggingface-checkpoints (qwen3_4b, linux.arm64.2xlarge, executorch-ubuntu-22.04-gcc1... / linux-job (gh)
  RuntimeError: Command docker exec -t f7718632755d44392745bbd1a2ddcdf241847f74e89f4bf61bc8bf6a150991ce /exec failed with exit code 1

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

• trunk / test-models-macos-cpu (vit, xnnpack-quantization-delegation) / macos-job (gh)
  ##[error]The operation was canceled.

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

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

• pull / unittest / windows / windows-job (gh) (trunk failure)
  ##[error]The operation was canceled.
• pull / unittest-editable / windows / windows-job (gh) (trunk failure)
  ##[error]The operation was canceled.
• trunk / unittest-release / windows / windows-job (gh) (trunk failure)
  ##[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.

[Copilot]

Pull request overview

Adds Qwen 3.6 MoE documentation and switches CUDA CI/export scripts to use the Qwen3.6-35B-A3B-HQQ-INT4 prequantized checkpoint, leveraging the existing Qwen 3.5 MoE runner/export pipeline.

Changes:

• Add a stub qwen3_6_moe README that points to the qwen3_5_moe example and links the Qwen 3.6 prequantized INT4 checkpoint.
• Update CUDA workflow + CI scripts to use SocialLocalMobile/Qwen3.6-35B-A3B-HQQ-INT4.
• Improve qwen3_5_moe README with a prequantized quick-start and clearer “prequantized checkpoint” terminology.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
examples/models/qwen3_6_moe/README.md	Adds a minimal pointer README for Qwen 3.6 MoE and links prequantized weights.
examples/models/qwen3_5_moe/README.md	Adds quick-start + clarifies prequantized export terminology and checkpoints list.
.github/workflows/cuda.yml	Switches CUDA CI model matrix and runner selection to Qwen 3.6 INT4 checkpoint.
.ci/scripts/test_model_e2e.sh	Updates supported HF model ID mapping to Qwen 3.6 checkpoint.
.ci/scripts/export_model_artifact.sh	Updates supported HF model ID mapping to Qwen 3.6 checkpoint for export.

---

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

[Copilot]

In the quick-start snippet, the placeholder still says <path-to-bundle>, but the README now consistently describes --prequantized as a prequantized checkpoint directory. Update the placeholder wording to match (e.g., <path-to-prequantized-checkpoint>), to avoid confusion about what path is expected.

Suggested change

	python export.py --prequantized <path-to-bundle>
	python export.py --prequantized <path-to-prequantized-checkpoint>

[Copilot]

This comment still refers to “Qwen3.5 MoE”, but the matrix entry below is now Qwen3.6-35B-A3B-HQQ-INT4. Update the comment to reflect Qwen 3.6 (or both 3.5/3.6) so the exclusion rationale stays accurate.

Suggested change

	# Qwen3.5 MoE uses a prequantized checkpoint, only tile-packed
	# Qwen3.6 MoE uses a prequantized checkpoint, only tile-packed

[Copilot]

This exclusion-block comment still says “Qwen3.5 MoE uses a prequantized checkpoint”, but the excluded model is now Qwen3.6-35B-A3B-HQQ-INT4. Update the comment to match the new model name (or make it version-agnostic) to keep the workflow self-explanatory.

Suggested change

	# Qwen3.5 MoE uses a prequantized checkpoint, only tile-packed
	# Qwen3.6-35B-A3B-HQQ-INT4 uses a prequantized checkpoint, only tile-packed