Using Whisper's text-timing functionality on a pre-existing transcript

[threaded]

Hello, I’m interested in applying Whisper’s text timing functionality to a preexisting transcript (one generated by a human). Is this something the model can do?