Fix: 30s issue during inference

#1
by abar-uwc - opened

Use return_timestamps argument and remove chunk_length_s this will work for longer duration of audio.

i am running on 1 hour hindi audio, still i get some missing transcriptions,

Sign up or log in to comment