Trained with 30 images using base "anything-v4.0-pruned-fp32" model.

Works great with models merge "0.6(linneamix_v74) + 0.4(0.6(realdosmix_) + 0.4(ddosmix_V2))"

and loras

microwaist_03Lora

beautifulDetailedEyes_v10