In this project we are focusing on generating music automatically using Recurrent Neural Network(RNN).
- Let's break the project program into 3 major part:
- Data preparation
- recover the notes played in each midi file
- data processing
- Training
- Prediction
- Data preparation
In the following software and hardware list, you can run the code file in this repository.
Software | OS |
---|---|
Music21, Anaconda Package Python 2.x/3.x, TensorFlow, Keras, numpy, glob | Ubuntu 16.04 or greater |
- Use
pip
to install this packages:
E.g.
pip install Music21
you’ll need to make sure you have pip
available. You can check this by running
pip --version
if you don't have pip
, i suggest this link to install it
To train the network you run RNN.py.
E.g.
python RNN.py