Data

Before starting, download the needed files.

Data set

For the model training we are using the SQuAD dataset - (Stanford Question Answering Dataset).

Download SQuADv1 and place the files in the squad-v1 folder.

squad-v1\dev-v1.1.json
squad-v1\train-v1.1.json

Word embeddings

For the generation of the incorrect answers we're going to use the pretrained glove embeddings. Download glove.6B.zip and place it in the embeddings folder.

embeddings\glove.6B.300d.txt