Skip to content

vllm_omni.model_executor.models.moss_tts.pipeline ¶

Pipeline topology for all MOSS-TTS variants (2-stage: talker → codec).

MOSS_TTS_PIPELINE `module-attribute` ¶

MOSS_TTS_PIPELINE = PipelineConfig(
    model_type="moss_tts",
    model_arch="MossTTSDelayModel",
    stages=(
        StagePipelineConfig(
            stage_id=0,
            model_stage="moss_tts",
            execution_type=LLM_AR,
            input_sources=(),
            owns_tokenizer=True,
            engine_output_type="latent",
            async_chunk_process_next_stage_input_func=f"{_PROC}.talker2codec_async_chunk",
            sampling_constraints={"detokenize": False},
        ),
        StagePipelineConfig(
            stage_id=1,
            model_stage="moss_tts_codec",
            execution_type=LLM_GENERATION,
            input_sources=(0,),
            final_output=True,
            final_output_type="audio",
            engine_output_type="audio",
            model_arch="MossTTSCodecDecoder",
            sync_process_input_func=f"{_PROC}.talker2codec",
            sampling_constraints={"detokenize": True},
        ),
    ),
)

MOSS_TTS_REALTIME_PIPELINE `module-attribute` ¶

MOSS_TTS_REALTIME_PIPELINE = PipelineConfig(
    model_type="moss_tts_realtime",
    model_arch="MossTTSRealtime",
    stages=(
        StagePipelineConfig(
            stage_id=0,
            model_stage="moss_tts",
            execution_type=LLM_AR,
            input_sources=(),
            owns_tokenizer=True,
            engine_output_type="latent",
            async_chunk_process_next_stage_input_func=f"{_PROC}.talker2codec_async_chunk",
            sampling_constraints={"detokenize": False},
        ),
        StagePipelineConfig(
            stage_id=1,
            model_stage="moss_tts_codec",
            execution_type=LLM_GENERATION,
            input_sources=(0,),
            final_output=True,
            final_output_type="audio",
            engine_output_type="audio",
            model_arch="MossTTSCodecDecoder",
            sync_process_input_func=f"{_PROC}.talker2codec",
            sampling_constraints={"detokenize": True},
        ),
    ),
)