Skip to content

[vllm] fix: restore vllm patch - #7076

Merged
wuxibin89 merged 11 commits into
verl-project:mainfrom
tardis-key:main
Jul 17, 2026
Merged

[vllm] fix: restore vllm patch#7076
wuxibin89 merged 11 commits into
verl-project:mainfrom
tardis-key:main

Conversation

@tardis-key

@tardis-key tardis-key commented Jul 16, 2026

Copy link
Copy Markdown
Collaborator

What does this PR do?

enable vllm patch

fix ci error introduced by #7061

Checklist Before Starting

  • Search for similar PRs. Paste at least one query link here: ...
  • Format the PR title as [{modules}] {type}: {description} (This will be checked by the CI)
    • {modules} include fsdp, megatron, veomni, sglang, vllm, rollout, trainer, ci, training_utils, recipe, hardware, deployment, ray, worker, single_controller, misc, perf, model, algo, env, tool, ckpt, doc, data, cfg, reward, fully_async, one_step_off
    • If this PR involves multiple modules, separate them with , like [megatron, fsdp, doc]
    • {type} is in feat, fix, refactor, chore, test
    • If this PR breaks any API (CLI arguments, config, function signature, etc.), add [BREAKING] to the beginning of the title.
    • Example: [BREAKING][fsdp, megatron] feat: dynamic batching

Test

For changes that can not be tested by CI (e.g., algorithm implementation, new model support), validate by experiment(s) and show results like training curve plots, evaluation results, etc.

API and Usage Example

Demonstrate how the API changes if any, and provide usage example(s) if possible.

# Add code snippet or script demonstrating how to use this

Design & Code Changes

Demonstrate the high-level design if this PR is complex, and list the specific changes.

Checklist Before Submitting

Important

Please check all the following items before requesting a review, otherwise the reviewer might deprioritize this PR for review.

@tardis-key
tardis-key marked this pull request as draft July 16, 2026 09:43

@gemini-code-assist gemini-code-assist Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Code Review

This pull request introduces an import of npu_vllm_patch in verl/utils/vllm/__init__.py. However, the relative import syntax used (import .npu_vllm_patch) is invalid in Python and will cause a SyntaxError. It should be corrected to from . import npu_vllm_patch.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

Comment thread verl/utils/vllm/__init__.py Outdated
@tardis-key tardis-key changed the title restore vllm patch [vllm] fix: restore vllm patch Jul 16, 2026
@tardis-key
tardis-key force-pushed the main branch 3 times, most recently from 0431e8d to bee7f5d Compare July 17, 2026 03:08
@tardis-key
tardis-key marked this pull request as ready for review July 17, 2026 06:19
@tardis-key

Copy link
Copy Markdown
Collaborator Author

@Secbone please review

_FakeFSDPModel was missing config, can_generate attributes needed
by the full save_checkpoint codepath after LoRA weight saving.
Added _FakeConfig class and delegated missing attrs to
_fsdp_wrapped_module.
_FakePEFTModel.state_dict() produces 4 LoRA keys (2 direct +
2 peft-prefixed), not 2. The filter correctly keeps all 4.
@wuxibin89
wuxibin89 merged commit d718b8c into verl-project:main Jul 17, 2026
100 of 109 checks passed
@Secbone

Secbone commented Jul 17, 2026

Copy link
Copy Markdown
Contributor

LGTM 👍

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants