[executorch][muse-glimmer] Add unified CUDA sample_token entry point

[Gasoonjia]

Stack from ghstack (oldest at bottom):

• [executorch][muse-glimmer] Run DFlash sampling on CUDA #22492
• [executorch][muse-glimmer] Make temperature 0 an exact argmax in the solo sampler #22631
• [executorch][muse-glimmer] Run DFlash sampling on CUDA #22630
• [executorch][muse-glimmer] Run DFlash sampling on CUDA #22678
• [executorch][muse-glimmer] Add CUDA speculative sampling orchestration #22491
• -> [executorch][muse-glimmer] Add unified CUDA sample_token entry point #22490
• [executorch][muse-glimmer] Add CUDA residual correction sampling #22489
• [executorch][muse-glimmer] Add CUDA excluded-token correction #22677
• [executorch][muse-glimmer] Add CUDA speculative acceptance #22487
• [executorch][muse-glimmer] Add CUDA categorical sampling #22486
• [executorch][muse-glimmer] Add CUDA sampling probabilities #22485
• [executorch][muse-glimmer] Add CUDA argmax sampling primitive #22676
• [executorch][muse-glimmer] Add CUDA excluded-token correction #22488
• [executorch][muse-glimmer] Add CUDA argmax sampling primitive #22484

Add the CUDA counterpart of muse_glimmer::sample_token. Preserve greedy argmax, temperature/top-k/top-p probability construction, optional probabilities-only output, and stochastic categorical sampling behind one device-resident interface.

Differential Revision: D117826288

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/22490

• 📄 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, 1 Pending, 2 Unclassified Failures

As of commit 0b45917 with merge base d0da5f8 ():

NEW FAILURE - The following job has failed:

• Lint / lintrunner (gh)
  >>> Lint for examples/models/muse-glimmer/tests/sampling_cuda_test.cpp:

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:

• MLX / test-mlx / test-mlx (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
• MLX / test-mlx-voxtral-realtime / test-mlx-voxtral-realtime (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  ##[error]fatal: Fetched in submodule path 'extension/llm/tokenizers/third-party/json', but it did not contain e6cafa573aac6ed9227f752a5371c0b3f436307d. Direct fetching of that commit failed.

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.