How to undo generation of softlinks of wav files when using Speaker Verification model?

[xiaofu2730]

Things just like this

[anautsch]

Hi @Opsamcxf, please clarify what happened, so we know where to help (it must be obvious but we want to be on the right track). Which speaker verification model is used how, to result in which wav files, and which softlinks are created - and why do they need to be undone?

[xiaofu2730]

Thank you for your reply! The problem I mentioned is that speechbrain generates softlinks of relevant audio files when using ECAPA_TDNN model. I want to undo this function for its messing up my working directory. Thank you!

…

------------------ 原始邮件 ------------------ 发件人: "Andreas ***@***.***>; 发送时间: 2022年4月1日(星期五) 晚上9:48 收件人: ***@***.***>; 抄送: ***@***.***>; ***@***.***>; 主题: Re: [speechbrain/speechbrain] How to undo generation of softlinks of wav files when using Speaker Verification model? (Issue #1352) Hi @Opsamcxf, please clarify what happened, so we know where to help (it must be obvious but we want to be on the right track). Which speaker verification model is used how, to result in which wav files, and which softlinks are created - and why do they need to be undone? — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: ***@***.***>

[Spectra456]

I also have this problem, it's annoying and load audio always using softlink which is linked to first version of the file.

[TParcollet]

Hi, we took note of this issue and are starting to discuss it internally :-) Thanks !