Sequence-to-sequence model with LSTM encoder/decoders and attention - View it on GitHub