This repository contains source code for the systems described in:
Addressing Semantic Drift in Question Generation for Semi-Supervised Question Answering
BERT-QG, BERT-QA, BERT-QPC (Code is coming soon...)
The code requires Python 3. Some basic python dependencies are specified in "requirement.txt".
pip install -r requirements.txt
By the way, a Python 3 virtual environment could be set up and run with:
virtualenv name_of_environment -p python3
source name_of_environment/bin/activate
Setup Standford CoreNLP environment by running the following commands:
mkdir LIB/corenlp
cd LIB/corenlp; wget http://nlp.stanford.edu/software/stanford-corenlp-full-2018-10-05.zip; unzip stanford-corenlp-full-2018-10-05.zip
export CORENLP_HOME=$PWD/stanford-corenlp-full-2018-10-05
Download GloVe word embeddings:
mkdir LIB/glove
cd LIB/glove; wget http://nlp.stanford.edu/data/glove.840B.300d.zip; unzip glove.840B.300d.zip
Download ELMo:
mkdir LIB/elmo
cd LIB/elmo
wget https://s3-us-west-2.amazonaws.com/allennlp/models/elmo/2x4096_512_2048cnn_2xhighway_5.5B/elmo_2x4096_512_2048cnn_2xhighway_5.5B_weights.hdf5
wget https://s3-us-west-2.amazonaws.com/allennlp/models/elmo/2x4096_512_2048cnn_2xhighway_5.5B/elmo_2x4096_512_2048cnn_2xhighway_5.5B_options.json
Setup ELMo environment:
git clone https://github.com/allenai/bilm-tf.git
cd bilm-tf; python setup.py install
Download BERT:
cd LIB/bert
wget https://storage.googleapis.com/bert_models/2018_10_18/uncased_L-12_H-768_A-12.zip
unzip uncased_L-12_H-768_A-12.zip
Download SQuADv1.1 dataset:
mkdir LIB/squad
cd LIB/squad
wget https://rajpurkar.github.io/SQuAD-explorer/dataset/train-v1.1.json
wget https://rajpurkar.github.io/SQuAD-explorer/dataset/dev-v1.1.json
Download the article title list of SQuAD QG test set:
wget https://raw.githubusercontent.com/xinyadu/nqg/master/data/doclist-test.txt
Download QQP dataset:
python LIB/download.py
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent