Audio Transcription Labelling - Aiming to not duplicate labelling process between sessions

Hello, does prodigy automatically remove any files/clips that have already been labelled if you pass the same database name:

python -m prodigy audio.transcribe speaker_transcripts label.jsonl --loader jsonl --fetch-media

Hi @bev.manz :wave:

Prodigy does not automatically remove files or clips that have already been labelled from a database. However, when you run a labelling instance multiple times with the same database name, examples that have already been annotated in the database will not be presented to you.

If there are going to be multiple annotators, Prodigy has nice customisability around whether you want labelling overlap or not, how many times an example should be labelled etc. You can read more about task routing here.