Milena D

TLDR (Because I don't care)

Use Milena D <lora:MilenaD:1> in positiv promt.

If you like the LoRa, please feel free to share your generated images here and leave a review (or at least a rating) ;D

About

Another ???? model known from MetArt... As always: The main reason why I trained this LoRa is the fact that I wanted to see her in different costumes. Just being ????? is boring. ^^

Training

Used 35 highres only images as training data...

Usage

There are no weight adjustments required, and the basic prompting is quite simple:

Milena D <lora:MilenaD:1>

In addition, I have made the best images using the DPM++ SDE Karras sampler with a fairly high number of steps - around 50 steps is the sweet spot for me. Activating face fix depends on the situation, if the eyes are screwed up I always retry the same seed with face fix activated...

Disclaimer (Because some people are stupid af)

All people used in training pictures or shown in my sample pictures are over the age of 18. And yes it is stupid that I even have to mention it... 🤦