Integrate HuggingFace TTS models into vLLM-Omni with CUDA graph acceleration.
npx skills add https://github.com/vllm-project/vllm-omni --skill add-tts-model
Please help me install this Agent Skill. Skill: add-tts-model Source: https://github.com/vllm-project/vllm-omni/tree/main/.claude/skills/add-tts-model Command: npx skills add https://github.com/vllm-project/vllm-omni --skill add-tts-model
Integrate a new text-to-speech model into vLLM-Omni from HuggingFace reference implementation to enable production-ready, streaming audio serving with CUDA graph acceleration and robust deployment workflows.
Wiring a HuggingFace TTS model into vLLM-Omni for production-ready streaming audio.
High-intent search queries and answers about installing and using this skill.
To integrate a HuggingFace TTS model into vLLM-Omni, you apply stage separation patterns, configure online serving, and generate artifacts like stage configs and end2end.py to enable production-ready streaming audio.
Yes, vLLM-Omni supports CUDA graph acceleration for streaming audio. It applies this acceleration alongside async chunk streaming to deliver low-latency audio output for production-grade TTS deployments.
Two-stage and single-stage patterns dictate the architecture for HuggingFace TTS reference integration. The Skill enforces phase-specific architecture patterns and rigorous validation to correctly separate and configure each stage.
Yes, you can use async chunk streaming for low-latency text-to-speech in vLLM-Omni. The integration process configures online serving with async streaming to deliver audio chunks efficiently.
Yes, pre-commit and DCO workflows are required for production TTS deployments. The Skill enforces these workflows alongside rigorous validation to ensure robust, production-grade code delivery for vLLM-Omni.
Adding a TTS model to vLLM-Omni generates artifacts such as stage configs, end2end.py, and documentation. These artifacts support end-to-end delivery and testing for the integrated streaming audio service.
hsliuustc0106Launch and configure vLLM-Omni API servers for OpenAI-compatible models.
handsomelong922Deploy scalable LLM serving with vLLM for OpenAI-compatible endpoints.
bobmatnycAutomate pre-merge checks for linting, testing, documentation, security, and CI/CD status.
leonling-llImplement two-stage pipeline optimizations for Gluon GEMM kernels on CDNA3 and CDNA4 GPUs.
