removed tokens array from transcripts

set self.model to loaded model

fixed pt should be pytube

fixed error in verbose and time transcription

fixed documentation

change handler to use whisper model and include segments

Duplicate from philschmid/openai-whisper-endpoint