Closed Carsonthemonkey closed 1 year ago
This should be an option when we have a setting menu. For the edge case where someone is talking about public speaking or something where they will intentionally use words like um or uh as part of what they are saying.
Whisper already does this to some degree, so I don't think it is necessary
To optimize the number of tokens we are sending to GPT API, we should filter out filler words in the transcript before sending it to ChatGPT. This could also be an option for the user on the transcript. We might want to look at the npm package natural to accomplish this easily