Text-Classification-Pytorch

Description

This repository contains the implmentation of various text classification models like RNN, LSTM, Attention, CNN, etc in PyTorch deep learning framework along with a detailed documentation of each of the model.

Text Classification is one of the basic and most important task of Natural Language Processing. In this repository, I am focussing on one such text classification task and that is Sentiment Analysis. So far I have covered following six different models in this repo.

• RNN
• LSTM
• LSTM + Attention
• Self Attention
• CNN
• RCNN

Requirements

• Python==3.6.6
• PyTorch==0.4.0
• torchtext==0.2.3

Downloads and Setup

Once you clone this repo, run the main.py file to process the dataset and to train the model.

$ python main.py

References

• A Structured Self-Attentive Sentence Embedding : Paper
• Convolutional Neural Networks for Sentence Classification : Paper
• Recurrent Convolutional Neural Networks for Text Classification : Paper