Music Transformer

This is the basic environment for GANs & Roses music generation model, using Magenta Music Transformer.

For the end-to-end, ready for deployment art to music pipeline, check out our deploy repository. If you are interested in geting to know more about the GANs & Roses project, you might enjoy reading our Medium article.

Prerequisites

• Docker INSTALL

Set up

Clone repository:

git clone https://github.com/GANs-N-Roses/transformer.git

Move to the project directory:

cd transformer

Build the image (this should take around 3.6 GB in disk):

docker-compose build

Run it.

docker-compose run

It will download the music transformer pretrained models (only the first time on the first run) and then start the container.

When it's done, it will print the URL to access the Jupyter Notebook from the host machine:

roses  |     To access the notebook, open this file in a browser:
roses  |         file:///root/.local/share/jupyter/runtime/nbserver-1-open.html
roses  |     Or copy and paste one of these URLs:
roses  |         http://b3ce3c5ccffe:8888/?token=c42836b292095cb6e374e09e6765e90ae3455338ba7ca842
roses  |      or http://127.0.0.1:8888/?token=c42836b292095cb6e374e09e6765e90ae3455338ba7ca842

Also, it's possible to access the container bash directly:

Check the container name using

docker ps

And then:

docker exec -it your-container-name bash

Usage

Your local src directory is synced with the Jupyter Notebook's working directory (/music), so any files in this directory will be visible within the Docker container and ready to be loaded to Jupyter. Similarly, any Notebook you save will appear in this directory.

Learn how to use the music transformer with the demo.ipynb Notebook (in src).

Also, you may add or remove dependancies by modifying requirements.txt.

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Please make sure to update tests as appropriate.

License

MIT