Musenet

MuseNet Overview

MuseNet is an innovative AI tool developed by OpenAI that is capable of generating 4-minute musical compositions with 10 different instruments, and can combine styles from country to Mozart to the Beatles. It uses a 72-layer network with 24 attention heads, with full attention over a context of 4096 tokens, which allows it to remember long-term structure in a piece. MuseNet has two modes: simple and advanced. In simple mode, users can listen to random pre-generated samples, while in advanced mode, users can interact with the model directly to create entirely new pieces.

MuseNet Highlights

• MuseNet can blend generations in novel ways, such as generating a piece in a pop style with piano, drums, bass, and guitar using the first 6 notes of a Chopin Nocturne.
• It uses composer and instrumentation tokens to give more control over the kinds of samples it generates. These tokens were prepended to each sample during training time, so the model would learn to use this information in making note predictions.
• MuseNet uses the recompute and optimized kernels of Sparse Transformer to train a 72-layer network with 24 attention heads—with full attention over a context of 4096 tokens. This long context may be one reason why it is able to remember long-term structure in a piece.