Compared to similar "all fours with one leg raised" LoRAs, this model focuses more on a pose with a relatively lower leg lift. This LoRA strictly affects the posture and is compatible with humans, female characters, and more. Note that since this LoRA was trained on 4:3 images, using a 4:3 aspect ratio during generation will yield better results.