This assignment is a part of the "MIT 6.S191: Introduction to Deep Learning" lecture series. All lecture slides and videos are available on the course website.
Do have a look at the amazing course at http://introtodeeplearning.com
It is advised to use Google Colab for running this notebook.
In this assignment we'll be building an RNN model with LSTM architecture for music generation. We will train a model to learn the patterns in raw sheet music given in ABC notation and then use this model to generate new music. We've built the model using tensorflow and keras API primarily.
This notebook also used a package called "mitdeeplearning" which contains some convienence functions and will be imported along with the other dependencies.
This package can be simply installed by using the command:
pip install mitdeeplearningThe model we use follows this architecture:
