Chorus v1

A single model capable of transcribing two overlapping speakers. No separate diarisation needed. Helpful for meetings and dialogues.