This version is focused on Better anatomy, Dynamic lighting and Dynamic angle, realistic life like images and slightly biased towards NSFW.
This version is Based on DMD2 ( https://huggingface.co/tianweiy/DMD2 )
Steps: 4, 5, 6 steps seems to be best
Sampler: Euler a,
CFG scale: 1, 1.5
Face restoration: CodeFormer,
Size: 720x1260, And all SDXL Compatible Sizes.
Clip skip: 2,
ADetailer model: Required
Negative prompt: Stable_Yogis_PDXL_Negatives-neg