Before using, consider the following:
I have only been in this space for a few days, be prepared for inconsistency on V1.0.
I am no lawyer, however ethically this LORA should never be used for profit, as it was trained on real people.
This LORA is very sensitive and adjusting the weight on the same seed is your friend. Using RV1.3, I found acceptable results on 0.3-0.4
, 0.6
, and 0.9
. It also produces varying quality results at the beginning/end of the prompt.
Favorable to a 16:9 portrait orientation, such as 432x786.
Legs/toes are FUBAR'd in some cases, and require a future fix.
Lighting can be an issue.
I recommend using the Euler a sampling method.
If you'd like to help me out as I'm new to the space, or discuss, you can contact me on Discord. This is not yet a place to ask for assistance, as I'm still figuring out this space.
There will be more to add as I discover more about LORA.
Base prompt: kneeling, from behind
Kneeling locations: on bed, on floor, in chair, in office chair, on stool, in car, on couch
Prompt enhancements: ass spread, legs spread, thong, panties, no panties
Initial Release.
Go ahead and upload yours!
Your query returned no results β please try removing some filters or trying a different term.