https://github.com/Const-me/Whisper
I had fun with both of these. They will both do realtime transcription. Bit you will have to download the training data sets…