You do not need a technical background to understand the basics of how AI music works. It comes down to learning patterns and then generating new audio that fits a request.
Training on patterns
During training, the model is exposed to large amounts of music and learns how melody, rhythm, harmony and vocals typically fit together across styles and languages.
Generating from your prompt
When you provide a prompt and lyrics, the model generates audio that matches the requested style, producing music and singing together. Autunes runs on the Zori models — Zori3 for Indian-language vocals and Zori4 XL for high-fidelity output.