A project that trains a LSTM recurrent neural network over a dataset of MIDI files. More information can be found on the writeup about this project or the final report written. Warning: Some parts of this codebase are unfinished.
- Python 2.7
- Anaconda
- Numpy (http://www.numpy.org/)
- Tensorflow (https://github.com/tensorflow/tensorflow) - 0.8
- Python Midi (https://github.com/vishnubob/python-midi.git)
- Mingus (https://github.com/bspaans/python-mingus)
- Matplotlib (http://matplotlib.org/)
- Run
./install.sh
to create conda env, install dependencies and download data source activate music_rnn
to activate the conda environment- Run
python nottingham_util.py
to generate the sequences and chord mapping file todata/nottingham.pickle
- Run
python rnn.py --run_name YOUR_RUN_NAME_HERE
to start training the model. Use the grid object inrnn.py
to edit hyperparameter configurations. source deactivate
to deactivate the conda environment