jxieeducation/DIY-Data-Science

Please make Pull Requests for good resources, or create Issues for any feedback! Seq2Seq solves the traditional fixed-size input problem thatEffective Approaches to Attention-based Neural Machine Translation prevents traditional DNNs from mastering sequence based tasks such as translation and question answering. It has been shown to have state of the art performances in English-French and English-German translations and in responding to short questions. Seq2Seq was first introduced in late 2014 by 2 papers (Sequence to Sequence Learning with Neural Networks and Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation) from Google Brain and Yoshua Bengio's group. The two papers took a similar approach in machine translation, in which Seq2Seq was developed upon.