The code found here contains an RNN and an LSTM implemented both in Keras.
Performance is comparable to that of the Facebook baseline LSTM results.
This project utilizes the following additional libraries and you may need to install them on your machine before running the project.
- keras: https://keras.io/
- pandas: https://pandas.pydata.org/
- matplotlib: https://matplotlib.org/
It's advisable to have the anaconda environment on your machine.
To run the project simply open up your terminal and run your environment.
After activating the environment run the command jupyter notebook
in your terminal
if you wish to see the project in jupyter.
If you'd like to contribute please contact the author.
Inspiration taken from Keras team bAbI RNN implementation found here:
https://github.com/keras-team/keras/blob/master/examples/babi_rnn.py.