cascadeLSTMDRL

An end-to-end framework demo for goal-driven dialogue

Introduction

It is a tensorflow based implementation of cascaded LSTM network (based Deep Reinforcement Learning) to solve goal-driven dialogue.

The architecture of the model is:

Reference

[1] Yue Ma, Xiaojie Wang, Cascaded LSTMs based Deep Reinforcement Learning for Goal-driven Dialogue

@article{cascadeLSTMDRL,
  title={Cascaded LSTMs based Deep Reinforcement Learning for Goal-driven Dialogue},
  author={Yue Ma and Xiaojie Wang},
  journal={The Sixth Conference on Natural Language Processing and Chinese Computing, NLPCC 2017},
  year={2017}
}

Requirements:

• python3
• numpy
• tensorflow 0.9.0
• tqdm