Skip to content

vllm_omni.model_executor.models ¶

Modules:

Name	Description
`bagel`
`common`
`cosyvoice3`
`covo_audio`
`dreamzero`
`dynin_omni`
`fish_speech`
`glm_image`
`glm_tts`	GLM-TTS model support for vLLM-Omni.
`higgs_audio_v2`	vllm-omni integration for boson-ai's higgs-audio v2 (two-stage TTS).
`hunyuan_image3`
`lance`	Lance (ByteDance) — vLLM-Omni model_executor package.
`mammoth_moda2`
`mimo_audio`
`ming_flash_omni`
`minicpmo_4_5`
`moss_tts`
`moss_tts_nano`
`omnivoice`
`output_templates`
`qwen2_5_omni`
`qwen3_omni`
`qwen3_tts`
`registry`
`utils`
`voxcpm2`
`voxtral_tts`
`whisper_utils`

OmniModelRegistry `module-attribute` ¶

OmniModelRegistry = _ModelRegistry(
    {
        None: {
            model_arch: (
                _LazyRegisteredModel(
                    module_name=_resolve_module_name(
                        mod_relname
                    ),
                    class_name=cls_name,
                )
            )
            for model_arch, (mod_relname, cls_name) in (
                items()
            )
        },
        None: {
            model_arch: (
                _LazyRegisteredModel(
                    module_name=f"vllm_omni.model_executor.models.{mod_folder}.{mod_relname}",
                    class_name=cls_name,
                )
            )
            for model_arch, (
                mod_folder,
                mod_relname,
                cls_name,
            ) in (items())
        },
    }
)