This repo is the final project of COSC-572.
We aim to build a Chatbot based the movie dialogs dataset to produce interactive conversations. We will experiment with both generative and retrieval-based models to compare the results. In addition to using existed model, the Sequence To Sequence (seq2seq) model introduced in Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation, we will try to improve it using policy gradient with reward function.