jumon / whisper-punctuator

Zero-shot multimodal punctuation insertion and truecasing using Whisper
MIT License
96 stars 5 forks source link

Question: What if the text is already punctuated? #8

Open mirix opened 1 year ago

mirix commented 1 year ago

Hello,

If we wish to repunctuate and recapitalise some text that is already punctuated and capitalised, do we need to remove the existing puctuation and to convert everything to lower-case beforehand or will the procedure handle that?

Cheers,

Ed

frznkshn commented 12 months ago

Hello,

If we wish to repunctuate and recapitalise some text that is already punctuated and capitalised, do we need to remove the existing puctuation and to convert everything to lower-case beforehand or will the procedure handle that?

Cheers,

Ed

According to my tests, this library preserves the appearance of the text, whatever it may be, and adds punctuation to it. So, first of all, you should clean the input text and then use this library.