Trained on 36 images created solely by prompts and bodysliders, 18 vectors, learning rate 0.004, 15 steps pr image/embedding and 1650 steps in total. Responds well to all manner of checkpoints.
Modellquelle
Quellenauszug
Trained on 36 images created solely by prompts and bodysliders, 18 vectors, learning rate 0.004, 15 steps pr image/embedding and 1650 steps in total. Responds well to all manner of checkpoints.
Quellen
1 QuelleModellartefakte
1 ArtefaktQuellenauszüge
1 AuszugSource context: 183 downloads · Type TextualInversion · Base model SD 1.5