[bugfix] remove vllm speech route

[linyueqian]

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

Purpose

Fix 400 Bad Request when calling /v1/audio/speech with Qwen3-TTS voices (e.g., Vivian, Ryan).

vllm's built-in /v1/audio/speech route validates the voice parameter against OpenAI's voice list (alloy, echo, shimmer, etc.), rejecting any Qwen3-TTS voice names. Because vllm-omni never removed this upstream route, vllm's handler took precedence over omni's, which has proper Qwen3-TTS validation.

The fix adds _remove_route_from_router(router, "/v1/audio/speech", {"POST"}) before registering omni's speech endpoint, consistent with how /v1/chat/completions, /health, and /v1/models are already handled.

Fixes #1041 and addresses this comment.

Test Plan

# Start server                                                                                                           
vllm-omni serve /path/to/Qwen3-TTS-12Hz-1.7B-CustomVoice \                                                               
    --stage-configs-path vllm_omni/model_executor/stage_configs/qwen3_tts.yaml \                                         
    --host 0.0.0.0 --port 8800 \                                                                                         
    --trust-remote-code --enforce-eager --omni                                                                           
                                                                                                                         
# Test with Qwen3-TTS voice (previously returned 400)                                                                    
curl -X POST http://localhost:8800/v1/audio/speech -H "Content-Type: application/json" -d '{"input": "今天天气真好", "voice": "Ryan", "instructions": "用开心的语气说"}' --output test_ryan.wav -w "\nHTTP Status: %{http_code}\n"

Test Result

test_ryan.wav

---

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.

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)

[david6666666]

LTGM