We all know that SD3-medium sometimes outputs blurry or unfinished faces. This LoRA fixes those problems by 80-90%.

In my testing, the Euler sampler works best with this LoRA.

So, try it out and don't forget to give feedback or suggestions!

🌐 Any inquiries about my model are more than welcome. Reach out on Discord ➡️ "shoponskp" Let's keep the conversation going