• klankin@piefed.ca
    link
    fedilink
    English
    arrow-up
    1
    ·
    5 days ago

    Theres not enough training data from one actor to make a useful model, they usually distill existing models.