Custom SDXL LoRA trained for generating realistic photos and portrait concepts. Works best with SDXL base models at weight 0.8 - 1.0.
V1.safetensors