Music generator generation of continuous personalized music

Abstract

Techniques are disclosed relating to automatically generate new music content. In some embodiments, a computing system receivers user input specifying a user-defined music control element. The computing system may train a machine learning model to change both composition and performance parameters based on user adjustments to the user-defined music control element. In embodiments in which composition and performance subsystems are on different devices, one device may transmit configuration information to another device, where the configuration information specifies how to adjust parameters based on user input to the user-defined music control element. Disclosed techniques may facilitate centralized learning for human-like music production while allowing individualized customization for individual users. Further, disclosed techniques may allow artists to define their own abstract music controls and make those controls available to end-users.