I have developed a model for ComfyUI that generates random prompts to create ultra-high-definition images. However, I frequently encounter issues with hand deformities in the generated images. If anyone can improve my workflow, I would greatly appreciate it.