MusicRNN

MusicRNN implements Magenta's LSTM-based language models, which are useful if you want to generate a NoteSequence to continue an input one. In the demos below, we call continueSequence() with the provided input, and display the result. Each section uses a different model checkpoint (melody, drums or chord pitches) to generate the output sequence.

Melody Continuation

Drums Continuation

Chord Pitches Improv Continuation