Skip to content

NotaGen

Generate classical sheet music with a symbolic music model

Visit official site

What NotaGen is

The model is described as a symbolic music generation model that uses large language model training methods to produce classical sheet music. It goes through pre-training, fine-tuning and reinforcement learning stages, and the material lists efficient generation, varied styles and high-quality output.

What you can do with it

  • Pre-train a base model on a large-scale music dataset
  • Fine-tune for a specific classical style
  • Optimise results with CLaMP-DPO without manual labels
  • Choose among NotaGen-small, NotaGen-medium and NotaGen-large

Getting started

  • Set up a Python environment and install dependencies from the README
  • Download pre-trained weights for the small, medium or large model
  • Fine-tune the model with your own dataset
  • Apply CLaMP-DPO reinforcement learning to refine the scores

Application scenarios

  • A music creator generates classical-style scores
  • A music educator uses generated scores as teaching material
  • A researcher explores symbolic music generation

Pros & cons

✓ What we like

  • Three model sizes are listed: NotaGen-small, NotaGen-medium and NotaGen-large
  • Both local deployment and online Colab notebook use are supported

! What to watch out for

  • The documented setup installs a Python environment and dependency libraries
  • Fine-tuning requires a dataset supplied by the user

FAQ

What training stages are described?

The material names pre-training, fine-tuning and reinforcement learning with CLaMP-DPO.

Which model sizes are listed?

The feature list names NotaGen-small, NotaGen-medium and NotaGen-large.

How can it be accessed?

The material lists a Gradio demo, local deployment and an online Colab notebook.

Last reviewed: 2026-11-15

More AI music creation tools

View all →

How we review