Anima version is available!
Nova Furry XL
A 2d/2.5d furry checkpoint model that can have great details on any type of furs, scales and feathers, which aims to be XL-fied Peaki Furry
Rules
You cannot use the generated images for commercial use if it's not edited (or just turning it to black and white)
Advertising is always welcome
Recommended settings:
Sampler: Euler a for best performance
Steps: 20-50 steps
(Pony)
CFG Scale: 6-7
Prompt: score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_furry, BREAK, furry, anthro, detailed face eyes and fur, {Animal Type}
Negative Prompt: blurry, jpeg artifacts, username, watermark, signature, normal quality, worst quality, large head, low quality, text, error, missing fingers, extra digits, fewer digits, bad eye
(Illustrious)
CFG Scale: 3-5
Prompt: masterpiece, best quality, amazing quality, very aesthetic, high resolution, ultra-detailed, absurdres, newest, scenery, furry, anthro, {Prompt}, BREAK, depth of field, detailed fluffy fur, volumetric lighting
Negative Prompt: human, multiple tails, modern, recent, old, oldest, graphic, cartoon, text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, lowres, bad anatomy, bad hands, missing fingers, extra digits, fewer digits, cropped, very displeasing, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, signature, watermark, username, simple background, conjoined, bad ai-generated
Description
Enhancing details and posing
NoobAI EPS v1.1 + Illustrious v2.0 stable DARE applied
FAQ
Comments (27)
What's new in V10?
https://civitai.com/posts/20820508 - Nova Furry 9a
https://civitai.com/posts/20823841 - Nova Furry 10
Of course Civitai cocked up the post order, but you can identify the artist tag via the file name and make your own comparisons. Besides the style tag, it's the same seed on the same machine, using different checkpoints.
100% of my images have pretty heavy black lines around subject, I've tried negative prompt for cell shaded and dark lines around subject, but no go. any tips?
Try using no lines on prompt, outlines for negative prompt
Waos
I love this model, but can I request you train it for king cheetahs? Virtually no models have the ability to generate the king cheetah mutation. Thank you!
Requesting a LoRA on the bounty board would probably let you get your desired species from any supported models.
Big fan of the model. But my only nitpick with the model is it favor too much for females. Whenever I try to make male characters. Especially male mobians from sonic (series), it keeps making them into a girl or a femboy with eyelashes. Could you add a little more male images into the model. That would be really appreciated.
I love this model! One small issue though, it seems to have some trouble with canine penises in general, I keep getting human penises unless I use a lora specifically for that, which tends to alter style, and I feel like a furry-focused model should have no problem with this concept. Is there any prompt in particular I can use to get canine penises?
For a furry model it is shockingly terrible at sheaths. I've never had any issue with just prompting "canine penis", though.
I have a problem, it keeps making the area around vaginas human skin colored. It's most noticeable with white or gray fur, but even with black fur you can see it's pinker. Is there any way to stop this from happening?
can you give an example of a prompt you're using and having this happen on?
This has been my go-to model for generating...near anything...since I've started Stable Diffusion. Out of the box it works fantastic.
Great model, one caveat is that it doesn't work well with inpainting for some reason, any time I use it to inpaint, it comes out really blurry. Is there a workaround I'm missing?
fancy meeting you here, cutie <3
@Crody I appreciate the reply, but its specifically to do with this model, using other ones doesn't give me much trouble (i've been using the inpaint crop and stitch nodes by lquesada on comfyui for over a year now, have used them with over 50 different checkpoints at least, so I'm fairly experienced). For some reason the nova furry models perform poorly when making edits in my experience, just wanted to know if there was something specific to this model I was missing.
Hi all! If you use Swarm UI and are having the black gen issue, download the SDXL VAE from Civit and select it manually in Swarm. That should fix the issue!
can someone help me?
vae=none but in the last step, the image acquires a green tint?
any of sampler , scheduler , steps same result
Use sdxl_vae
Amazing for generating dragons! I'm actually getting better results using dpmpp_2m_sde_gpu over the recommended Euler. Looks cleaner this way.
This model is amazing not only for anthro characters, but humans too! It really well retains any style LORA I throw at it.
Another W model
Love these models, been here since 0.8!
I have noticed that in V10 e6 style artist tags are not nearly as effective as in previous versions, which does make it harder to fine-tune the style. I prefer to not use any LORAs as this model is typically powerful enough to create almost anything on it's own.
Do you have any plans for reinforcing artist styles in the future @Crody?
Expanding artist tags is not the primary goal but I'd say with each updates, variations of artist tags will expand
@Crody I think there was a misunderstanding, do you know of a good way to reinforce existing artist styles between models? I've tried dialing up the strength of the tag itself, but from V7a to V10 a lot of style tags have become diminished and no longer work or have a very weak effect.
@antiman9 Probably it has to do with merging process and I don't know how to reinforce them on v10
On the next version (v11), I'll try my best to include them as well