PyTorch implementation for Bi-directional Attention Flow model

Question answering system based on the paper https://arxiv.org/abs/1611.01603. This is a slightly modified version where its answer selection module use bilinear function, giving slight improvement in accuracy over the original model. Part of the code are from https://github.com/jojonki/BiDAF.

In Order to run,

1. Download SQuAD dataset and GloVe embeddings

(~850 MB, this will download files to $HOME/data):

chmod +x download.sh; ./download.sh

2. Preprocess SQuAD data

python -m squad.prepro

Then place the processed data and unzipped GloVe embeddings into the data directory (by default it is ./data/squad)

3. Training

To train, run the following command.

python main.py

To test,

python main.py --test 1 --resume <PATH_TO_SAVED_PARAMS>