This version was trained on 120 images of collages of often wildly different styles. If you want control, consistency and predictability, this probably isn't for you.
モデル説明
A series of experimental LoRAs trained to alter and warp faces in various ways.
