Footjob POV SDXL 1.0

Hope you enjoy this LoRA. I recommend using Base SDXL 1.0 as it was trained on that but feel free to try other SDXL versions. Better results if you use resolution 896 x 1152.

FOR the SD 1.5 Version --> (click here)

-FOR USE with keywords: woman, pov, footjob, feet, penis (though just: POV, Footjob is fine)

You can get decent results with just using: POV, footjob if you use Base SDXL with it.

-Negative prompts: text, watermark, logos, extra toes, extra feet, extra penis, blurry, fuzzy

Negative prompts aren't a necessity but can improve quality of life.

-Use Model Strenght 1.0 (if footjobs don't happen anymore go with 1.10-1.50 though avoid 1.5 since it may cause artifacts) if image is too blurry, artifact-ridden or fuzzy stick with 1.0 strength.

Also go with 40 steps or around that and use EULER as a Sampler. Tweak the above values until you find what works for you. Remember it is hard for SD to get feet right so adjust as needed.

This was trained doing 10 epochs and 20 images of POV footjobs. I may train it again upping the steps for better results though I am kinda satisfied with this one.

Feel free to share good images generated with this :)