Magenta is a research project exploring the role of machine learning in the process of creating art and music.

This is the home for our serializable NoteSequence representation along with utilities for:

• creating them from various formats (e.g., MIDI, abc, MusicXML)
• manipulating them (e.g., slicing, quantizing)
• extracting components (e.g., melodies, drum tracks, chord)
• exporting them (e.g., to MIDI or audio)
• or converting them to representations useful for model training (e.g., one-hot tensors)

To learn more see the main magenta repo and have look at the hello magenta notebook.

Installation

pip install note-seq

Notes

Importing note-seq increases the value of pretty_midi.pretty_midi.MAX_TICK to 1e10. This may change the expected behavior of pretty_midi when using it alongside of note-seq. For example, loading corrupted MIDI files with pretty_midi may result in out of memory errors.