PROJECT ASSUMPTIONS

1. Pretrained model can be use
2. Groovae 4bar config is used
3. The dataset of Groovae 4bar is downloaded using tsds_name argument in config.py
4. You can deploy and use this program everywhere as long as docker is installed
5. The output of this program is 4 (four) .mid files created by MusicVAE model

HOW TO RUN:

1. Clone this repository
2. Make sure you have docker installed
3. Build the image named pormagenta (maybe you need to wait for few minutes)

   docker build -t pormagenta . 

4. Run the image

   docker run --rm -ti -v $PWD/results:/home/results pormagenta

Prove