Extractive-QA

2024 Applied Deep Learning Homework 1

This repository is implementation of HW1 for Applied Deep Learning course in 2024 at National Taiwan University.

Download dataset and model checkpoint

To download the datasets and checkpoints of Multiple Choice & Question Answering models, run the following command:

bash ./download.sh

Reproducing best result

To reproduce best result, run the following command:

Note: you can define your own argument paths if you have your own data.

bash ./run.sh data/context.json data/test.json pred/prediction.csv

Training

Multiple Choice

To train the multiple choice model, run the following command:

bash ./train_mc.sh

Question Answering

To train the question answering model, run the following command:

bash ./train_qa.sh

Implementation

Due to the lack of training resources, the model is trained using google colab PRO, with A100 GPU equipped.