fix: honor server hub for default Nano model

[LauraGPT]

Summary

• fix funasr-server --hub ms --model fun-asr-nano so the default Nano model honors the requested hub instead of forcing Hugging Face
• apply the same hub selection to the AutoModel fallback path when vLLM is unavailable
• update server app regression tests for both vLLM and fallback loading

Fixes #3321.

Verification

• red first: the two new hub regression tests failed with hf before the fix
• python3 -m pytest -q tests/test_server_app_openai_segments.py tests/test_cli.py tests/test_realtime_ws_service.py -k "hub or server or fallback or default_fun_asr_nano"
• python3 -m py_compile funasr/bin/_server_app.py funasr/bin/server.py
• python3 -m pytest -q tests/test_server_app_openai_segments.py tests/test_docs_funasr_install_commands.py tests/test_collect_growth_metrics.py
• git diff --check

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.