[fix] Remove redundant parameter passing

[qibaoyuan]

PLEASE FILL IN THE PR DESCRIPTION HERE ENSURING ALL CHECKLIST ITEMS (AT THE BOTTOM) HAVE BEEN CONSIDERED.

Purpose

Remove unnecessary parameter passing to simplify the code.

Test Plan

cli as given

Test Result

1. online

vllm-omni serve  /mnt/user/qibaoyuan/Qwen2.5-Omni-3B --omni --port 8791 \
--stage-configs-path ~/qwen2_5_omni.yaml

result:

3. offline
   3.1 text

3.2 use_audio_in_video

---

Essential Elements of an Effective PR Description Checklist

• The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
• The test plan, such as providing test command.
• [x ] The test results, such as pasting the results comparison before and after, or e2e results
• (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
• (Optional) Release notes update. If your change is user facing, please update the release notes draft.

BEFORE SUBMITTING, PLEASE READ https://github.com/vllm-project/vllm-omni/blob/main/CONTRIBUTING.md (anything written below this line will be removed by GitHub Actions)

[chatgpt-codex-connector]

💡 Codex Review

vllm-omni/vllm_omni/entrypoints/omni_stage.py

Lines 243 to 245 in 1b6bae9

	log_file=model._log_file
	in_q=model._in_q
	out_q=model._out_q

Use stage queues instead of model string in sync worker

_stage_worker now pulls _log_file, _in_q, and _out_q from model (lines 243‑245), but init_stage_worker calls it with the model name string from _start_stage_processes. On the synchronous path this immediately raises AttributeError before the worker can log readiness or process batches, so every stage process crashes and the orchestrator never runs.

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[qibaoyuan]

💡 Codex Review

vllm-omni/vllm_omni/entrypoints/omni_stage.py

Lines 243 to 245 in 1b6bae9

	log_file=model._log_file
	in_q=model._in_q
	out_q=model._out_q

Use stage queues instead of model string in sync worker
_stage_worker now pulls _log_file, _in_q, and _out_q from model (lines 243‑245), but init_stage_worker calls it with the model name string from _start_stage_processes. On the synchronous path this immediately raises AttributeError before the worker can log readiness or process batches, so every stage process crashes and the orchestrator never runs.

ℹ️ About Codex in GitHub

thx, fixed

[Gaohan123]

lgtm, thanks!