How to utilize ? Please download the dataset and the GLOVE word embeddings to the local area. The specific url has been given in the paper, and prepareData.py is used to process the dataset ¢Ù.

¢Ú Adjust the parameters of the experiment in config.py.

¢Û Training with the train.py.

Experimental environment£º

Python==3.5 tensorflow-gpu>=1.3.0 numpy+mkl >=1.13.1

We will update the framework of this model in succession, and continue to carry out the next work. If there are any questions or mistakes, please kindly contact us in time. Thank you very much!