Use https://github.com/BlinkDL/ChatRWKV/tree/main/music to run rwkv-music models.
https://huggingface.co/BlinkDL/rwkv-4-music
https://huggingface.co/BlinkDL/rwkv-5-music
MIDI model (RWKV-4 sounds better):
Tokenizer: https://github.com/briansemrau/MIDI-LLM-tokenizer
Training data: https://huggingface.co/datasets/breadlicker45/bread-midi-dataset
ABC model (RWKV-5 sounds better if you use topp=0.5 without topk):
Training data: load_dataset("sander-wood/massive_abcnotation_dataset")["train"]
https://huggingface.co/datasets/sander-wood/massive_abcnotation_dataset