search

modelscope / FunASR

A Fundamental End-to-End Speech Recognition Toolkit and Open Source SOTA Pretrained Models, Supporting Speech Recognition, Voice Activity Detection, Text Post-processing etc.
https://www.funasr.com
Other
6.46k stars 688 forks source link

paraformer-zh支持的语音格式 #1681

Closed wuchunlong-comac closed 5 months ago

wuchunlong-comac commented 5 months ago

paraformer-zh看文档是需要16000采样的单通道音频，能支持双通道么

LauraGPT commented 5 months ago

yes! code is here: https://github.com/alibaba-damo-academy/FunASR/blob/main/funasr/utils/load_utils.py#L94