[Bugfix] when set offline model running error

[lengrongfu]

Purpose

Fix: #23684

Test Plan

Test Result

• HF_HUB_OFFLINE=1 python3 offline_test.py

from vllm import LLM, SamplingParams

prompts = [
    "Hello, my name is",
    "The president of the United States is",
    "The capital of France is",
    "The future of AI is",
]
sampling_params = SamplingParams(temperature=0.8, top_p=0.95)

llm = LLM(model="/tmp/Qwen1.5-7B")

outputs = llm.generate(prompts, sampling_params)

# Print the outputs.
for output in outputs:
    prompt = output.prompt
    generated_text = output.outputs[0].text
    print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")

---

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.
• 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 in the Google Doc.

[gemini-code-assist]

Code Review

This pull request addresses a crash that occurs when running vLLM in offline mode (HF_HUB_OFFLINE=1). The root cause was that creating a default EngineArgs instance for logging purposes would trigger network access. The fix cleverly avoids this by initializing the default arguments with the already-resolved model path from the provided arguments, which prevents the crash. The logic for detecting a non-default model is also correctly adjusted. This is a solid and well-targeted fix for the reported bug.