Apply diffusion models using the new Hugging Face diffusers package to synthesize music instead of images.
