Collaborative Research: FW-HTF-R: Toward an Ecosystem of Artificial-intelligence-powered Music Production (TEAMuP)

We will provide a software framework that lets deep learning practitioners easily integrate their own PyTorch models into the open-source Audacity Digital Audio Workstation. This framework will let deep learning developers easily integrate new deep-models into Audacity, a free and open-source DAW that has logged over 100 million downloads since 2015. Developers will be able to upload their trained PyTorch model to HuggingFace’s Model Hub. The HuggingFace repository will be made accessible through Audacity’s user interface and will let end users load a model in a manner similar to traditional plugins. We will integrate this work into the main Audacity build, in collaboration with developers at the Muse group. In addition to this, we will develop illustrative deep learning models to be deployed on HuggingFace for use in Audacity.