Skip to content

[Doc][Recipe] Use renamed Mamba prefix flag in Nemotron-3.5-Super-VL FP8 - #62451

Merged
esmeetu merged 3 commits into
vllm-project:mainfrom
namgyu-youn:recipes/nemotron-mamba-shared-prefix-flag
Oct 22, 2026
Merged

[Doc][Recipe] Use renamed Mamba prefix flag in Nemotron-3.5-Super-VL FP8#62451
esmeetu merged 3 commits into
vllm-project:mainfrom
namgyu-youn:recipes/nemotron-mamba-shared-prefix-flag

Conversation

@namgyu-youn

@namgyu-youn namgyu-youn commented Oct 19, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Overview

The FP8 variant still passes --enable-mamba-fine-grained-prefix-cache, renamed to --enable-mamba-shared-prefix-checkpoint in v0.31.0 (#57382). With the recipe's 0.31.0 floor, the FP8 command never parses.

Test Plan

node scripts/validate-source.mjs --source vllm --dir recipes   (vllm-project/recipes) → valid
FP8 renderings in the generated JSON                          → old flag 0, new flag 8
grep -- --enable-mamba-shared-prefix-checkpoint vllm/engine/arg_utils.py  @v0.31.0, v0.32.0 → present
Pull Request Checklist

  • I used vLLM's /pr-checklist skill.

  • AI assistance was used during the creation of this PR.

  • Design Fit:

  • Testing and Validation:

  • Code Quality and Style:

  • Pull Request Contents:

@namgyu-youn
[Doc][Recipe] Use renamed Mamba prefix flag in Nemotron-3.5-Super-VL FP8
0d1fdcf 
Signed-off-by: namgyu-youn <namgyu.dev@gmail.com>
@namgyu-youn
namgyu-youn requested a review from esmeetu as a code owner October 19, 2026 10:10
claude[bot]
claude Bot reviewed Oct 19, 2026
View reviewed changes

@claude claude Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Claude Code Review

This pull request is from a fork — automated review is disabled. A repository maintainer can comment @claude review to run a one-time review.

@namgyu-youn

namgyu-youn commented Oct 19, 2026

Copy link
Copy Markdown
Contributor Author

cc @esmeetu working on this module. PTAL, thanks!

@esmeetu esmeetu added the ready ONLY add when PR is ready to merge/full CI is needed label Oct 22, 2026
@esmeetu

esmeetu commented Oct 22, 2026

Copy link
Copy Markdown
Member

/ci run

@github-actions

github-actions Bot commented Oct 22, 2026

Copy link
Copy Markdown

✅ @namgyu-youn, CI is now available for this PR.

  • /ci run starts upstream CI; /amd-ci run starts AMD CI only.
  • Run commands test your branch with every commit currently on its upstream target branch. If it is behind, the bot merges the target branch into it first (pull before pushing again). Append --allow-stale to a run command to test an outdated branch as is, at your own risk.
  • /ci retry retries failed jobs in the CI build for the current PR head. If the current head has no CI build, it starts a new CI build for the current head containing only jobs that failed in the latest earlier CI build for this PR.
  • /amd-ci retry retries failed jobs in AMD CI for the current PR head. Use /amd-ci run when the current head has no AMD CI build.
  • /ci cancel cancels scheduled or running CI builds for this PR branch; /amd-ci cancel does the same for AMD CI only.

@github-actions

github-actions Bot commented Oct 22, 2026

Copy link
Copy Markdown

❌ This PR is 204 commits behind upstream main. The bot could not merge upstream main into this branch (API returned 422: expected head sha didn’t match current head ref.). No new CI build was started. Either merge or rebase onto the latest main yourself and rerun /ci run, or use /ci run --allow-stale to test this branch as is.

@namgyu-youn

namgyu-youn commented Oct 22, 2026

Copy link
Copy Markdown
Contributor Author

/ci run

@github-actions

github-actions Bot commented Oct 22, 2026

Copy link
Copy Markdown

✅ Triggered Buildkite CI #96569 for commit fc134d009020.

@esmeetu
esmeetu merged commit 2f8e0de into vllm-project:main Oct 22, 2026
11 checks passed
@namgyu-youn
namgyu-youn deleted the recipes/nemotron-mamba-shared-prefix-flag branch October 22, 2026 06:24
github-actions Bot added a commit to vllm-project/recipes that referenced this pull request Oct 22, 2026
@github-actions
Sync recipes from vllm-project/vllm@7d5335936c40, vllm-project/vllm-o…
49f6a34 
…mni@949907817ee9

vllm-project/vllm@main:recipes 0746329d2abe → 7d5335936c40 [latest.vllm]
vllm-project/vllm@0746329...7d53359

Commits:
- vllm-project/vllm@2f8e0de8f6f0 [Doc][Recipe] Use renamed Mamba prefix flag in Nemotron-3.5-Super-VL FP8 (vllm-project/vllm#62451)
- vllm-project/vllm@4643f4185314 [Doc] DeepSeek-V4.1-Flash: fix TP4 c32 max-num-seqs in the B200 AgentX guide (vllm-project/vllm#62971)

Recipes:
- deepseek-ai/DeepSeek-V4.1-Flash (modified): https://recipes.vllm.ai/deepseek-ai/DeepSeek-V4.1-Flash
- nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4 (modified): https://recipes.vllm.ai/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4

vllm-project/vllm-omni@main:recipes 6b35bca6ffff → 949907817ee9 [latest.vllm-omni]
vllm-project/vllm-omni@6b35bca...9499078

Commits:
- vllm-project/vllm-omni@34aca93dc424 [Experimental] Clarify Noisy KV step_replicas, D, and clean semantics (vllm-project/vllm-omni#9169)

Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
github-actions Bot added a commit to shantanusharma/recipes that referenced this pull request Oct 22, 2026
@github-actions
Sync recipes from vllm-project/vllm@12c01a22eb2f, vllm-project/vllm-o…
bbd3006 
…mni@717f8a3499fb

vllm-project/vllm@main:recipes 0dd9da117dc7 → 12c01a22eb2f [latest.vllm]
vllm-project/vllm@0dd9da1...12c01a2

Commits:
- vllm-project/vllm@2f8e0de8f6f0 [Doc][Recipe] Use renamed Mamba prefix flag in Nemotron-3.5-Super-VL FP8 (vllm-project/vllm#62451)

Recipes:
- nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4 (modified): https://recipes.vllm.ai/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4

vllm-project/vllm-omni@main:recipes 6b35bca6ffff → 717f8a3499fb [latest.vllm-omni]
vllm-project/vllm-omni@6b35bca...717f8a3

Commits:
- vllm-project/vllm-omni@34aca93dc424 [Experimental] Clarify Noisy KV step_replicas, D, and clean semantics (vllm-project/vllm-omni#9169)

Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@claude claude[bot] claude[bot] left review comments
@esmeetu esmeetu esmeetu approved these changes

Assignees

No one assigned

Labels

ready ONLY add when PR is ready to merge/full CI is needed

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@namgyu-youn @esmeetu