This repository contains code and data used in the following paper:

@inproceedings{lan2018toolkit,
 author = {Lan, Wuwei and Xu, Wei},
 title = {Neural Network Models for Paraphrase Identification, Semantic Textual Similarity, Natural Language Inference, and Question Answering},
 booktitle = {Proceedings of COLING 2018},
 year = {2018}
} 

The SPM_toolkit contains 4 state-of-the-art neural networks for sentence pair modeling tasks, including: Paraphrase Identification, Semantic Textual Similarity, Natural Language Inference, and Question Answering.

1. The model was implemented with Python 2.7, PyTorch 0.4.0 and Torchtext 0.1.1 .
2. Sample command to run (navigate to the model folder first): python main.py