AI-powered melody generation for musicians and producers.
Created by Patricia Clover (PC) — blending code and creativity.
MelodyMind is a simple but powerful tool that generates unique melodies based on user input. Just type in a mood, genre, or theme — and let the AI handle the rest.
Built for musicians, producers, and hobbyists who want a spark of inspiration in their workflow. 🎶
- 🎯 Prompt-based melody generation (e.g., “uplifting jazz riff” or “moody lo-fi loop”)
- 🧠 Trained RNN model with custom music embeddings
- 💽 Output in
.mid(MIDI) format for use in any DAW - 🖤 Minimal interface: no fluff, just music
git clone https://github.com/PatriciaClover/MelodyMind.git
cd MelodyMindpip install -r requirements.txtpython run.py --prompt "dreamy ambient intro" --output output.midTry generating a melody with:
python run.py --prompt "nostalgic piano ballad"Check the examples/ folder for sample input/output.
MelodyMind uses an RNN trained on a dataset of genre-tagged MIDI clips. Prompts are converted into embeddings, then decoded into sequences of notes, rhythms, and velocities.
Want to train your own model? Check out the melodymind/model/ directory.
- PyTorch 🎯
- pretty_midi 🎼
- scikit-learn 🧪
Patricia Clover (PC) Musician-turned-coder passionate about AI music generation.
"Where there's music, there's magic; where there's code, there's limitless possibility."
Have ideas, questions, or want to collaborate? Feel free to open an issue or reach out via GitHub Discussions.