A handpicked collection of music clips across 38 subgenres, curated for training audio
language models (ACE-Step, MusicGen, AudioLDM, and similar) with subgenre-conditioned
LoRA / LoKr adapters.
Subgenres selected by the author for aesthetic distinctness and demo punch, covering
electronic (deep house, idm, trance, industrial, dark ambient, electropop, eurodance, edm),
hip-hop / soul / funk (jazz funk, disco, rnb), rock (psychedelic, gothic… See the full description on the dataset page:
https://huggingface.co/datasets/akshan-main/tunefine-data.
