Use any Whisper model on the hub for long transcription using the transformers pipeline! Just add chunk_length_s parameter to the pipeline and it'll slide through the audio to create a transcription.

→ View original post on X — @reach_vb