V1.0 was trained on 57 photos by the famous Russian photographer Stepan Kvardakov.V2.0 was trained on a larger dataset of 300 photos and 4000 stepsLora perfectly captures beautiful Slavic models, slender figures, beautiful tans and glowing skin, and the author's retouch metod.Recommended sampling method: EulerRecommended schedule type: bong tangent or beta

ArtofCK for Z-Image

Lora is trained on 300 photographer's photos at 4000 steps

Files with SHA256: 5b6d81fc6ee4db8711477d39148a75788c244bb83b532bf3c6454966240c94de