You can generate very interesting music simply by working with MIDI as opposed to sampled audio (slashing complexity by orders of magnitude!) and starting from a good model architecture. Daniel D. Johnson's (formerly known as Hexahedria, hired by Google Brain) model biaxial-rnn-music-composition is from 2015, requires very few resources for training or inference, and still delivers compelling, SOTA-or-close results wrt. improvising ("noodling") classical piano. Github
https://github.com/danieldjohnson/biaxial-rnn-music-composit... , you may also want to check out user kpister's recent port to Python 3.x and aesara:
https://github.com/kpister/biaxial-rnn-music-composition (Hat tip:
https://news.ycombinator.com/item?id=30328593 )