Listen, Attend and Spell model for end-to-end ASR system - View it on GitHub