Skip to content

Qwen 3.5 MoE Metal: Use max-sized prefill example for dynamic inputs#18956

Open
manuelcandales wants to merge 1 commit intogh/manuelcandales/176/headfrom
gh/manuelcandales/177/head
Open

Qwen 3.5 MoE Metal: Use max-sized prefill example for dynamic inputs#18956
manuelcandales wants to merge 1 commit intogh/manuelcandales/176/headfrom
gh/manuelcandales/177/head

Conversation

@manuelcandales
Copy link
Copy Markdown
Contributor

With alloc_graph_input=False, ExecuTorch sets the input tensor's
numel_bound_ from the serialized example size. A small example (T=2)
prevents runtime inputs larger than 2 tokens. Use max_seq_len-1 as
the prefill example size so any prompt length is accepted at runtime.

Authored with Claude.

[ghstack-poisoned]
@pytorch-bot
Copy link
Copy Markdown

pytorch-bot bot commented Apr 16, 2026

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/18956

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:

❌ 1 New Failure, 1 Cancelled Job, 4 Unrelated Failures

As of commit 58fe35f with merge base 5707e2a (image):

NEW FAILURE - The following job has failed:

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

FLAKY - The following jobs failed but were likely due to flakiness present on trunk:

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

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

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 Apr 16, 2026
manuelcandales added a commit that referenced this pull request Apr 16, 2026
With alloc_graph_input=False, ExecuTorch sets the input tensor's
numel_bound_ from the serialized example size. A small example (T=2)
prevents runtime inputs larger than 2 tokens. Use max_seq_len-1 as
the prefill example size so any prompt length is accepted at runtime.

Authored with Claude.


ghstack-source-id: 7118150
ghstack-comment-id: 4263712315
Pull-Request: #18956
@manuelcandales manuelcandales requested review from metascroy and removed request for larryliu0820 and lucylq April 16, 2026 22:14
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.

1 participant