vllm_omni.model_executor.models.minicpmo_4_5 ¶

Modules:

Name	Description
`batched_token2wav`	Strict, state-explicit batching for MiniCPM-o 4.5 Token2wav.
`cuda_graph_wrapper`
`duplex`
`minicpmo_4_5_code2wav`	Strict batched codec-to-waveform stage for MiniCPM-o 4.5.
`minicpmo_4_5_omni`
`minicpmo_4_5_omni_llm`
`minicpmo_4_5_omni_tts`	MiniCPM-o 4.5 native autoregressive Talker.
`minicpmo_4_5_token2wav`	MiniCPM-o 4.5 Token2wav adapter over in-tree `StepAudio2Token2WavCore`.
`pipeline`	MiniCPM-o 4.5 pipeline topology (frozen).

MINICPMO45_DUPLEX_CODEC_TOKENS_PER_CHUNK `module-attribute` ¶

MINICPMO45_DUPLEX_CODEC_TOKENS_PER_CHUNK = 26

MINICPMO45_DUPLEX_TURN_END_CODEC_TOKENS = (
    MINICPMO45_DUPLEX_CODEC_TOKENS_PER_CHUNK * 4
)