In generated transcripts, the first word of a new sentence sometimes starts with a lowercase letter instead of uppercase.
So like this: "Speaker 0: That is great. furthermore I would like to discuss XY."
(I am experimenting with modifiying the diarized.py script to add arguments for transcribing without diarization or for outputting srt with word-level timestamps, but this issue occurs with the original, unmofied diarize.py)