Skip to content

[executorch][muse] Serve MG batching on CUDA from the launcher - #24009

Merged
Gasoonjia merged 1 commit into
gh/Gasoonjia/48/origfrom
gh/Gasoonjia/49/orig
Oct 20, 2026
Merged

Gasoonjia merged 1 commit into
gh/Gasoonjia/48/origfrom
gh/Gasoonjia/49/orig

Conversation

@Gasoonjia

Copy link
Copy Markdown
Contributor

This PR was created by the merge bot to help merge the original PR into the main branch.
ghstack PR number: #23733 by @Gasoonjia
^ Please use this as the source of truth for the PR details, comments, and reviews
ghstack PR base: https://github.com/pytorch/executorch/tree/gh/Gasoonjia/49/base
ghstack PR head: https://github.com/pytorch/executorch/tree/gh/Gasoonjia/49/head
Merge bot PR base: https://github.com/pytorch/executorch/tree/gh/Gasoonjia/48/orig
Merge bot PR head: https://github.com/pytorch/executorch/tree/gh/Gasoonjia/49/orig
Differential Revision: D124871961
@diff-train-skip-merge

Pull Request resolved: #23733

`serve_batching.py --backend cuda` launches the CUDA batching worker behind the
same FastAPI app, SessionRuntime, multiplexed worker client, chat template and
streaming parser as MLX. It requires `--data-path` (the AOTI weights blob) and
rejects `--pos-embed-path`, since the CUDA artifact is text only; image
requests fail during preparation.

BATCHING_SERVING.md gains a CUDA section: export with
`export_solo.py --backend cuda --use-offgraph-kv-cache`, build with
`make muse-glimmer-batching-cuda`, launch, and how resident sessions, prefix
snapshots and the capture lane share the exported cell pool.
ghstack-source-id: 451855086
@exported-using-ghexport

Differential Revision: [D124871961](https://our.internmc.facebook.com/intern/diff/D124871961/)
@pytorch-bot

pytorch-bot Bot commented Oct 20, 2026 •

Copy link
Copy Markdown

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/24009

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

⏳ No Failures, 186 Pending

As of commit eb4148c with merge base 897d28f (image):
💚 Looks good so far! There are no failures yet. 💚

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

@meta-cla meta-cla Bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Oct 20, 2026
@Gasoonjia
Gasoonjia added this pull request to stack #24012 October 20, 2026 03:42
@Gasoonjia
Gasoonjia merged commit dfdd5db into main Oct 20, 2026
263 of 271 checks passed
@Gasoonjia
Gasoonjia deleted the gh/Gasoonjia/49/orig branch October 20, 2026 03:43

This branch was successfully deployed

1 active deployment
cadence — eb4148c9 Deployed Oct 20, 2026 by Gasoonjia via hifi-op-test / hifi4 #34141
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants