update vllm version to 0.13.0

[billyang-scale]

Pull Request Summary

What is this PR changing? Why is this change being made? Any caveats you'd like to highlight? Link any relevant documents, links, or screenshots here if applicable.

This PR update vllm version to 0.13.0

Test Plan and Usage Guide

How did you validate that your PR works correctly? How do you run or demo the code? Provide enough detail so a reviewer can reasonably reproduce the testing procedure. Paste example command line invocations if applicable.
verify this version work correcly on local devbox

(APIServer pid=1) INFO 02-09 14:31:36 [api_server.py:1351] vLLM API server version 0.13.0
(APIServer pid=1) INFO 02-09 14:31:36 [utils.py:253] non-default args: {'port': 5005, 'model': '/workspace/model_files', 'enforce_eager': True, 'served_model_name': ['Qwen/Qwen3-4B-Instruct-2507'], 'tensor_parallel_size': 4}

[billyang-scale]

The issue was that after upgrading to vLLM 0.13.0, the --attention-backend argument is now included in vLLM's built-in argument parser. The manual addition at vllm_server.py:85 was causing a conflict.

[socket-security]

Review the following changes in direct dependencies. Learn more about Socket for GitHub.

Diff	Package	Supply Chain Security	Vulnerability	Quality	Maintenance	License
	vllm@​0.11.0 ⏵ 0.13.0	+1

View full report