Valeria

Trained on a 26 image SDXL 1024x1024 dataset with detailed portraits, full body poses and high variety lighting. Best results at LoRa strength 0.-0.9