A music generation ML model with a focus on Lo-Fi music (INCOMPLETE/STILL IN THE WORKS).
Using Tensorflow LSTMs and Sequential layers, I tried making a Deep Learning model that will generate Lo-Fi music given snippets of training examples.
Python 3.x Install following packages using pip:
- Music21
- Tensorflow
- Keras
- glob
- Download the folder containing Lo-Fi midi snippets (or any midi files of your choice)
- To train the model, you simply just run each code cell in model.ipynb and run the last cell to generate music
I will probably look at papers/articles related to music generation to try and improve this model, as it is still in it's preliminary stages right now.