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 687 forks source link

When running paraformer, an error occurs in the first step of extracting cmvn features. #1744

Closed didi222-lqq closed 2 months ago

didi222-lqq commented 4 months ago

image torch 2.2.0 cuda 11.8 torchaudio 2.2.0 torchvision 0.17.0

LauraGPT commented 2 months ago

You could solve it by google.