!! Created for Super-Tech-FD: https://civarchive.com/models/251473/super-tech-fd , won't work on other checkpoints !!

• Trained from over 120 images, taken from the show

• Hand captioned and tagged dataset, following this layout: dronev [action]+[expression], [additional tags], [portrait, half body or full body], [image quality]

• Additionally trained on examples from N and Uzi to increase variety

  -Technically able to generate them, with activation words "uzidoorman" and "goobern", but doesn't perform well