SQuAD

Introduction

NLP system trained on Stanford Question Answering Dataset (SQuAD). SQuAD tests the ability of a system to not only answer reading comprehension questions, but also abstain when presented with a question that cannot be answered based on the provided paragraph.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.

PREREQUISITES

Below you can find the instructions for downloading the dataset, the pre-trained embeddings and the weights of our trained model.

Please note that the software will automatically download all of them, so you are not required to specifically follow the below procedure.

Optional Manual Procedure

First of all you have to create a new folder /data at the root level of the project.

The original dataset is available at the following link. Once you have downloaded it, you have to place the unzipped file training_set.json inside the /data/raw/ folder.

The pre-trained Glove embeddings used are available at the following link. Once you have downloaded it, you have to place the unzipped file Glove_50.txt inside the /data/raw/ folder.

The weights of the trained model are available at the following link. Once you have downloaded it, you have to place the unzipped file DRQA.h5 inside the /data/checkpoints/ folder.

Apple M1

If you are using the new Apple M1 chip please be sure to have installed hdf5 by running:

$ brew install hdf5

INSTALLING

below you can find all the scripts for installing based on your OS/processor

$ make
    > "+------------------------------------------------------+"
    > "|         OS         |  Hardware  |    Setup Command   |"
    > "+------------------------------------------------------+"
    > "|   Windows/Linux    |   - GPU    |  'make setup.CPU'  |"
    > "|   Windows/Linux    |   + GPU    |  'make setup.GPU'  |"
    > "|    Apple macOS     |    + M1    |  'make setup.M1'   |"
    > "|    Apple macOS     |    - M1    |  'make setup.CPU'  |"
    > "+------------------------------------------------------+"

for instance if you have MacOS with Intel chip you have to run:

$ make setup.CPU

or alternatively you can find all the different version of the requirements inside the /tools/requirements folder.

Running the Tests

Training

You can train the model from scratch using your custom dataset by running:

$ python src/train.py "<path_of_your_json_dataset>"

Inference

You can run the inference procedure on a specific dataset by running:

$ python src/compute_answers.py "<path_of_your_json_dataset>"

once you have done this, you can see the output generated inside the /data/predictions/answers.pred.json file.

Evaluations

You can evaluate the performances of the inference by running:

$ python src/evaluate.py "<path_of_your_json_dataset>" data/predictions/answers.pred.json

Authors

Matteo Conti - author - contimatteo
Francesco Palmisano - author - Frankgamer97
Primiano Arminio Cristino - author - primianocristino
Luciano Massaccesi - author - fruscello

License

This project is licensed under the MIT License - see the LICENSE.md file for details

contimatteo/SQuAD