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
6.47k stars 688 forks source link

在C++中读取token出错 #1490

Closed OgataAsatsuki closed 4 months ago

OgataAsatsuki commented 7 months ago

https://github.com/alibaba-damo-academy/FunASR/blob/5e7eb6f160c48861cbcd39825a0cb98f98538772/runtime/onnxruntime/src/vocab.cpp#L29-L44 我是用的paraformer-zh-streaming模型,代码中在YAML文件中读取token,但是config.yaml中并没有token_list,执行函数会导致vocab为空,然后在 https://github.com/alibaba-damo-academy/FunASR/blob/5e7eb6f160c48861cbcd39825a0cb98f98538772/runtime/onnxruntime/src/vocab.cpp#L156 出错

token_list应该从token.json中读取

lyblsgo commented 7 months ago

c++ runtime now only support model version V1.x.x, please use v1.0.6