Update: Please if you see any of my models somewhere else report or let me know! Thnx
Trained on base model xl 1.0 + FLUX.1-dev + zit + sd1.5
Update: Added SD 1.5 model.
Update: Pony version added! (Please read version info. before using it)
Update: Illustrious version added.
Update: Added perfection style model (Hands + feet + all in one)
Update: Added Detailed style (Hand focus + all in one)
Feel free to check hands model as well.
Please read version info (about version) And choose which version works best for you. Additionally you can combine multiple versions together as long as you manage to balance lora weight.
Feedback and ratings are much appreciated.
Other models to check:
Skin Tone (Cinematic Photography) Style XL (Human skin color)
Realistic Skin Texture style XL (Detailed Skin)
Description
zit v2.1
lower or rise lower weight if needed.
FAQ
Comments (2)
Found a great workflow for fixing anatomical errors (like extra toes) that even Flux and standard Inpainting engines were struggling with.
Instead of using a dedicated Inpaint model, I used Masked img2img on the Illustrious XL base by simply connecting Load Image to Vae Encode to Set Latent Noise Mask. This preserved the global lighting and structure while the LoRA handled the anatomy perfectly.
Workflow & Settings:
Method: Masked img2img (Full image context)
Denoising Strength: 0.35
Model: BabesIllustrious (by Stable Yogi)
LoRA: Feet XL (Strength 1.0)
Sampler: DPM++ SDE
Steps: 30 | CFG: 6.0 | Seed: 3
Positive: Realistic underside of foot from behind, 5 toes, embedding:Stable_Yogis_Anatomy_Postives_V1
Negative: blurry, deformed, more than 5 toes, embedding:Stable_Yogis_Anatomy_Negatives_V1-neg
Surprisingly, neither PAG nor Lying Sigma changed the result—the model and LoRA combo was strong enough to fix the anatomy on its own at this denoise level
在实际操作中我发现这个模型会出现镜头聚焦不到脚的特点，虽然脚的问题少了，但也变模糊了