audio_codec-encodec-zh_en-general-16k-nq32ds640-pytorch模型,Huggingface和Modelscope下载的参数不同
Opened this issue · 0 comments
yellowcomets commented
Huggingface下载的model.pth md5sum是4285b02283940535f4636787dc801dda,Modelscope下载的model.pth md5sum是277ad4e9033a825d78622d691df47b82,对同一音频进行编码得到的code也不相同,不是同一模型吗?