This model is the result of several different models I merged with the intent to achieve a versatile all-in-one style checkpoint for furry, feral and anime styles.

It took me several days to create this and used many different techniques like cosine, block weight merging and utilizing training on top.

To achieve different styles and results, you have to adjust the sampling method, cfg scale and clip skip. This is because of the wide range this model is capable of and so you have to be good at prompt crafting.

Please note: The core of this model consists of furry checkpoints, hence it mainly creates anthropomorphic characters, you can also have a wide range of humans or human like characters with the according prompt.

Useful Extensions

!After Detailer (automatic inpainting for faces and hands to improve bad faces and hands)
ControlNet (let's you control image composition, pose and almost anything you could desire)
Agent Scheduler (essentially a queuing system to increase your workflow)
Ultimate SD Upscale (good for users with limited VRAM as you can upscale to increase quality)
e621 Prompt (You can copy and paste any e621 post URL to get a prompt to generate an image as shown in the showcase)

Useful LoRAs

Useful prompt embeddings

Credit and acknowledgement

Without the creators of the following models it wouldn't been possible to create this merge.

Give them the credit that they deserve ❤️

(Order is alphabetical and doesn't reflect a ranking)