My first attempt at training the LoRA model for the SDXL network was done on a small dataset of artworks by the artist ThePit. It is advisable to set the CLIP skip parameter to 2.
Description
FAQ
Comments (4)
Hello, I really like this style. Can you make a version of SD1.5?
Hi. There are many LoRA models for 1.5 with the style of this artist that work much better than mine. To find them, search for "thepit" and apply the necessary filters. Enjoy!
OMG IT'S FINALLY HERE!
Are there any keywords I should be using?