Generates pictures of the pixie healer Challua. I used in game images to train this and they worked (for the most part)
Trigger words: challua_wf, green hair, green eyes, pointy ears, tan skin
Normal (leaf dress) outfit: leaf dress, boots, hair ornament
Christmas (holly dress) outfit: holly dress, flower dress, boots, hair ornament This is less stable
Creating portraits returns better overall quality due to the limited training set of images. The model is overtrained to produce green dresses even when prompted for other colors, so put that as a negative for a specific use case
I intend on trying to include the Halloween outfit in the next version, which will hopefully round out the LoRA capabilities.
Enjoy!