Do you know this? You have a good prompt, you want to add more ?????? elements to your creation, but the SDXL checkpoint doesn't seem to understand it.

You don't want to use a pony model, however, so as not to lose the diversity of SDXL.

This is where my LoRa can help you. It allows you to integrate ?????? elements, half-????? or ????? women without changing the overall concept or having to use a different checkpoint.

The LoRa was trained with over 600 photos. It works best with a weight of 0.8. A trigger is not required, but it is sometimes better to use the tokens "nsfw, nude".

If you like my LoRa or my other models, I would be happy if you gave me a "thumbs up". Please post your pictures directly here on the model page.