This is a rather comprehensive post in Medium by Alex Yu on state of the art methods for automatic music generation using Deep Learning. So what I like about it is that it doesn’t just introduce the MAESTRO dataset and its use, but also describes general DL in rather plain and approachable terms.
https://towardsdatascience.com/generating-music-with-artificial-intelligence-9ce3c9eef806
So just saving the link for future reference.