Problem with super long audio
#5
by
mrigankr - opened
Hi! Thanks for this awesome model. When I run this with a long audio about 12 mins long, it misses a large chunk and does not transcribe like 2 min chunks in between. Can you help?
Hey @mrigankr - thanks for flagging. This is indeed unexpected behaviour. What delay are you running with? The default of 480ms? If you're able to share the audio file I can try and repro on my end