• simple@lemm.eeOP
    link
    fedilink
    English
    arrow-up
    1
    ·
    1 year ago

    It looks like most of the training data was electronic music, but when the model releases people should be able to fine-tune it to different things.