These models generate and modify music from text prompts and raw audio. Key capabilities:
- Music generation - Create original music compositions and continuations based on text prompts.
- Audio super-resolution - Increase sample rates and add high frequency detail. Improves audio fidelity.
- Audio inpainting - Fill in missing sections based on surrounding content. Can remove vocals or instruments.
Models we recommend
Generate music from a prompt or melody
Stable diffusion for real-time music generation
Four-bar drum loop generation
Text-to-audio generation with latent diffusion models
AudioSR: Versatile Audio Super-resolution at Scale
Music inpainting of melody and chords
Bach chorale generation and harmonization
Tools to train a generative model on arbitrary audio samples
Remix the music into another styles with MusicGen Chord
Emotional conditioned music generation using transformer-based model.
Controllable Text-to-Music Generation
MusicGen fine-tuned on chamber choir music
Generate music restricted to chord sequences and tempo
Generate music in stereo, restricted to chord sequences and tempo
MAGNeT: Masked Audio Generation using a Single Non-Autoregressive Transformer
MusicGen fine-tuned on an epic orchestral style