Please use this LoRA responsibly and with respect towards the real life person.

Trained on SeaArt using 49 images found around the web, mostly high res images. Images used were no older then 4 years to avoid the LoRA looking too young. This issue still sometimes has an issue with the top of her head (her hair) being cut off is several issues. It's better then in version 1.0, but might still happen. Generating 4 images at a time is recommended. I have not been able to fully fix this at the moment. Strength: 1.0 (experiment if you want)Tested checkpoints: Flux1.D fp8 and Flux-devTrigger word is no required to activate the LoRA, but can of course be used to force it.