Premium character LoRAs
Original synthetic characters, bundles and AI tools are in my shop:
Join the community on Discord: https://discord.gg/nB79wRNtMT
Wedgie Girls
A concept LoRA for Krea 2 focused on amateur photos of women giving themselves a wedgie: thongs, panties, bodysuits and bikini bottoms pulled up hard with both hands so the fabric digs between the buttocks or into the crotch, seen from behind, bent over, kneeling or in mirror selfies. It gives an amateur photo look.
Usage
Start with the trigger word WedgieRK, then describe the garment (thong, panties, bodysuit, bikini bottom; colour, lace or cotton), how high and tight she pulls it and where it digs in, what else she wears (crop top, jeans pulled down, stockings) or that she is topless, the pose and the camera angle, the room and end with the light.
WedgieRK, Photograph of a woman with <a black lace thong pulled up high with both hands>, <seen from behind, bent over the bed>. The lighting is <soft daylight>.
Recommended weight: 1. Made for Krea 2 Turbo (8 steps, CFG 1).
Prompt Examples
"WedgieRK, Photograph of an adult woman with light skin and long brown hair. She is wearing a white crop top and pulling her thong, panties, or bikini bottom up high with both hands, causing the fabric to dig into her buttocks."
"WedgieRK, Photograph of an adult woman with light skin and dark brown hair, wearing a green crop top and black socks, standing in a bathroom. She is topless and has her white thong pulled up high, exposing her buttocks and creating a deep indentation in her flesh."
"WedgieRK, Photograph of a woman with light tan skin and long blonde hair, kneeling on a patterned rug. She is topless, with her back to the camera, exposing her bare back, buttocks, and the top of her thighs."
Training
- Base model: Krea 2
- Type: Concept LoRA
- Steps: 4000
- Rank: 32
- Optimizer: adamw, constant learning rate
- Engine: ai-toolkit
Generation settings
Trained on Krea 2, and works on Krea 2 Turbo for faster generation.
- Raw: ~30 steps, guidance 3.5
- Turbo: ~8 steps, guidance 1.0
The gallery images were generated with krea2_turbo_mxfp8.safetensors, at 8 steps, guidance 1, LoRA weight 1.



















