Open DIGist opened 1 year ago
This feature could be very helpful indeed for further correction of outputted text manually
UPVOTE! This would be incredibly powerful for my use-case. I am transcribing all files in the system and sometimes it makes just NO sense at all. Knowing that the average confidence is low, would be a huge help.
Ideally it would be word level confidence, as shown via --print-colors
.
One can log the output into a file then open it via this
vim -c ":term ++hidden ++open cat < FILENAME >"
This will enable one to view it as it was in vim to view.
So for spot checking long transcriptions, it would be helpful to have the confidence score output to some file.. either as an extra column in the CSV or maybe some metadata in VTT or even it's own log or something.
Thanks! This is amazing work!