DMN+ implementation in TensorFlow for question answering on the bAbI 10k dataset.

Structure and parameters from Dynamic Memory Networks for Visual and Textual Question Answering which is henceforth referred to as Xiong et al.

Adapted from Stanford's cs224d assignment 2 starter code and using methods from Dynamic Memory Networks in Theano for importing the Babi-10k dataset.

Original code is here: https://github.com/barronalex/Dynamic-Memory-Networks-in-TensorFlow

This code has been modified with GRU cells in answer module

Install Tensorflow with python 2.7.x (https://www.tensorflow.org/install/)

sudo apt-get install python-pip python-dev python-virtualenv

virtualenv --system-site-packages tensorflow

source ~/tensorflow/bin/activate

pip install --upgrade tensorflow

pip install --upgrade tensorflow-gpu

Run the included shell script to fetch the data

bash fetch_babi_data.sh

Process filename in data folder

python process_file.py

Create file properties.py

# path_to_glove
glove_path="/home/alex/Documents/nlp/code/data"
# glove file
glove_file=""
train_folder="en"
# dimension of word_embedidng
embed_size = 100
# output size of rnn
hidden_size = 100
# either softmax or rnn
answer_prediction="rnn"
use_glove=True

Use 'dmn_train.py' to train the DMN+ model contained in 'model.py'

python train.py --task_id 2

Once training is finished, test the model on a specified task

python test.py --task_id 2

The l2 regularization constant can be set with -l2-loss (-l). All other parameters were specified by Xiong et al and can be found in the 'Config' class in 'model.py'.