A submissive POV of a woman wearing a strap-on: you look up at her from below her hips, or from kneeling height in front of her. The harness, the dildo's colour and material, how she holds it, who she is, what she wears, the scene and the light are all yours to prompt. Trained on Krea 2 Raw.
Trigger: pvstrp9 strap-on POV
- Start the prompt like this:
A pvstrp9 strap-on POV photo, seen from a low angle looking up at her from below her hips, orseen from kneeling height in front of her, orseen from straight in front of her. - Then describe her, and
wearing a <harness> with a <colour, material> dildo: e.g. a white leather harness with a translucent purple jelly dildo, a tan leather strap harness with a thick pastel pink silicone dildo. - Say how she holds it: gripping the shaft and pointing it at the lens, holding it upright against her stomach, stroking it with her fingertips, holding the tip lightly.
- Want her face? Describe it and say she looks down at the camera. Want a faceless shot?
Add
framed from her chin down. - Finish with her clothes, what's behind her and a named light source.
- Portrait sizes work best for the low angle (832x1216); square works for kneeling height.
- Weight 0.8–1.0. The showcase images use 0.9.
Executives in a corner office, a dominatrix in a candlelit crypt, a burlesque star backstage, an athlete in a warehouse gym. All showcase images are made with this LoRA on a Krea 2 checkpoint. Results vary by checkpoint, so test it on yours.
If you like my work, leave a comment, like, or whatever! It'll motivate me to make more and share more. Thank you!
Description
Rank 32 on Krea 2 Raw, 750 steps (ai-toolkit, lr 1.5e-4, sigmoid timesteps), 40 crops at 512 cut at the chin, captioned with the camera angle, the harness, the dildo's colour and material, the grip, her clothes and the background. Later checkpoints were sharper but started to change prompts without the trigger, so this is the earlier one on purpose.