Performant and accurate speech recognition built on Pytorch - View it on GitHub