This is my nineteenth style lora so go easy with me on this one.Summary of the 2 versions. V1.0: Main version of the lora, trained on the full dataset for illustrious and noobai. Indigo: Trained for <a rel="ugc" href="https://civarchive.com/models/579632/indigo-furry-mix-xl">Indigo Furry Mix</a> on the 3 main characters. Check the version details for more info. This lora aims to replicate the style of Anhes' cell-shaded art. It's mainly used in their Lush Puppies comic (which is awesome btw check it out), but I included plenty of other stuff to flesh it out. I've been wanting to do it since I made the lora for Anhes' other style, but I held off til now as I knew it would take forever, and it did. Extremely happy with the outcome though. The main cast of the comics are all trained. I only tested the males, but the females should be there too. Obviously, characters with more panels will be better trained. They're listed below with the tags I recommend. well trained:<ul><li>lug, wolf, white fur, white body, yellow eyes, neck tuft, muscular, bara</li><li>theo, beagle, white fur, brown fur, floppy ears, yellow eyes, short stack</li><li>phan, beagle, floppy ears, white fur, brown fur, yellow eyes, muscular</li><li>bruce, rottweiler, black fur, brown fur, muscular</li><li>carlo, pit bull, grey fur, grey body, muscular</li><li>nobu, malamute, orange fur, white fur, chubby</li></ul>lesser trained but kinda might work:<ul><li>clarisse</li><li>lenora</li><li>rorie</li><li>thomas</li></ul>

Anhes Lush Puppies Style + Characters

Made by an anonymous user specifically for <a target="_blank" rel="ugc" href="https://civitai.com/models/579632/indigo-furry-mix-xl">Indigo Furry Mix XL V10.</a> They limited the dataset to just the 3 main characters, Theo, Phan and Lug. However, they made some improvements to the images and tagging. Its trained differently so its a completely different beast to the main lora. Make sure you use the linked checkpoint, and use a strength of around 0.5-0.6 depending of the complexity of the image. 0.7 is good for closeups and inpaints. Thanks again to that user for the awesome work, and a reminder to anyone else that my datasets are open to anyone hoping to make their own versions.

Files with SHA256: 90a5ad9cfceae50ef914868cdd7f5c6772f4460d858cf2494611c2a6aaf3cf01