A custom image-generation pipeline with a trained LoRA model that locks a character's face and body across unlimited scenes and poses. Includes curated dataset preparation and custom LoRA training on Stable Diffusion XL, driven through an automated ComfyUI workflow — plus a from-scratch AMD/ZLUDA GPU setup.
0 likes · 0 views · Aug 17, 2026
A custom image-generation pipeline with a trained LoRA model that locks a character's face and body across unlimited scenes and poses. Includes curated datas...
