Created by: lightsailpro

I tested the en8k model using the demo server. For longer audio clips e.g. 28 seconds of wave file, it seems that only last 10 seconds or so of the wave file got transcribed. The 1st 18 seconds just got discarded. Does anyone know why?