Deep Speech: Scaling up end-to-end speech recognition

Re-assign community