FunAudioLLM / SenseVoice

Multilingual Voice Understanding Model
https://funaudiollm.github.io/
Other
2.61k stars 249 forks source link

非常厉害的模型,CPU占用确实比Paraformer模型低很多。希望能增加分角色功能和热词功能。 #116

Open wjq8811 opened 2 weeks ago

wjq8811 commented 2 weeks ago

官方的示例比较少,虽然支持实时语音识别,但是文档里没说,如果可以的话,希望能出一个麦克风输入实时转写的示例。 谢谢,辛苦了。

linexjlin commented 2 weeks ago

我写了一个不过,是基于 siliconflow 家的 OpenAI transcription API 的 https://github.com/linexjlin/realtime_transcription