affige/genmusic_demo_list

Curated index of 100+ research demos spanning song generation (lyrics-to-vocal synthesis via diffusion and transformer architectures) and text-to-music/audio generation, organized by task type with direct links to papers and interactive implementations. Covers emerging generative approaches including flow matching, diffusion models, and transformer-based conditioning techniques—enabling researchers to benchmark across diverse synthesis paradigms from melody composition to multi-track audio control. Primarily targets academic evaluation and reproducibility by mapping arxiv papers to their corresponding web demos and GitHub repositories.

775 stars. Actively maintained with 4 commits in the last 30 days.