v1.0
This is my second LoRa model. Not perfect, trained in only one pose (probably overtrained?).

Tips: scale 0.6-0.8 will be good
training model (also used to generation): abyssorangemix2_Hard
Keywords: masturbating under panties, squatting

Can be very difficult to use with other models (but with incredible prompting u can do what u want). Not tested characters. Can be used with other LoRas (but also need good prompting).

v2.0
If only you knew how many attempts it took to make even what you see now. It turned out to be much more difficult to remake this model than I thought. What we have as a result: it doesn’t work very well with character LoRA's (you need to change the weight of either the character’s LoRA or this LoRA), for me personally, it works better with a resolution of 768x768 (hires fix), sometimes an object in hand that looks like a smartphone or something may appear - this is because for the dataset. From the pros - now you can more accurately get a character by prompts, that is, the model is less overbaked. I don't know if I will ever return to this model, because it put a lot of pressure on my mentally. So sorry if it didn't live up to your expectations.

Tips: scale 0.8 (for most) - 1 will be good (for char LoRA's need lower or higher)
Training model: NAI
Used to generation model: AOM3
Keywords: masturbating under panties, squatting

If u know how to improve model, let me know