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
Changes
Closer looks to the IL v10
Better style adherence
NoobAI EPS v1.1 + Illustrious v2.0-stable + ChenkinNoob 0.5 DARE applied
FAQ
Comments (35)
What's the difference Nova Furry XL 18 А and 18 B
@Crody could you put it in the model description in the future so we dont have to go looking for what changed with each version?
I hope version 17 stays for onsite generation it’s the best imo
I can't get version 18 to render anthro coyotes as well as v17 does. Version 17 is awesome for anthro coyotes and v18 produces inferior images.
Not specific to just this model, but no joke, unless you are prompting a specific character/species it keeps putting that half shaved comb-over hairdo you see on like... female V from Cyberpunk and no amount of prompting gets rid of it. I just need someone who understands model training to understand this issue, and please... please train the model to also understand that that hair style is in fact HAIR so when I put it in the negative prompt it stops doing it on every single non-character specific image ever.
Otherwise, good job, thank you, and I'm not blaming you in particular just to be clear again (I have a dozen models and they ALL do it).
I dunno. Version 18 seems like a step backwards in detail. I tried the same prompt on 18a all the way down to 13.5. 18 just doesn't have the detail the others have.
13.5 and 14 is still my best to choose.
Same. 13.5, 14 and 16. I've been re-arranging my workflows and adding adetailer trying to work around it. Right now im exploring flux options.
@aecii_3d Flux as a base model or flux as a detailer?
@morafermi643 as a detailer
Hey there @Crody!
I really like your models. I am just starting using AI and I noticed taht with each version you show sample images with the difference of each model. I want to know how do you do that? Every time I try to see the same image just for a tweak it gives me a completely different image.
A good method to test small changes in a prompt, is to set the seed to be the same for all tests you do.
I would assume that is what Crody does. Having a set of specific prompts with specific seeds, so the Checkpoint's version being the only difference.
@dragonofdarkness1992 Thank you for the info, but can you tell me what a seeds?
@oathtojoy117 search in your sampler
i have problem at get my blue i search for my dragon give me all time of soft blue and not of blue like the 0000FF and give me character full black and i can't keep my color want on the model of base i am miss idea i need like help if possible i search a get my blue and same style art i have just i can reach what i search
Use graphic software to adjust colors
@refactoredx graphic software? what you use for for that?
@DrawarDragon same ??? my pinks are nearly while
@mechapineman I'm not trying to do what people advised me to do, like using SwarmUI or ComfyUI, which is more complex, but I can't post the models I made with ChatGPT just because I'm trying to learn. My specialty is correcting small flaws in images, as sometimes it's not simple, or adding something.
As I understand author no more gonan work on illsutrious checkpoints?
Didn't @Crody say about a month ago, that Nova Furry XL IL won't be discontinued on productions, in other words will be continued? It's visible somewhere in the discussions...
@FurorSaxonicus I really want to continue but I need to focus on Anima right now
I think I'll adjust the updates from monthly to bimonthly (1 in 2 months)
@Crody Oh, okay. Sorry, I may have missed that Info.
@FurorSaxonicus No worries, it's the first time ever to share this infomation
@Crody So you gonna continue workins on illsutrious models jsut because yo uwork also now on anima its gonan take little longer.I udnerstand right?
@matriksAi As long as I feel the needs to create, yes
@Crody Why Anima?
@refactoredx Lighter files, better architecture for merge base and more flexible towards prompting
@123inactace For the stability, Illustrious is still better because of how long and how well the trainings are
For the capability, model size and future training, I think Anima is far better
With time, I think Anima would be the best base model
It's like reset of the flows except we are starting it from much better knowledges compared to back when Illustrious v0.1 was released, which was almost 2 years ago
@Crody Do you know if its possible to you illustrious and nobai loras with anima checkpoints or not?
@matriksAi The system is completely different so no it's not possible
We have to recreate the LoRA by using same training image on Anima model
@Crody Thanks for info,its really sad so many good loras for illsutrious and nobbai but if they not compatible with anima its problem
@matriksAi Good news is that the Anima knows much more things than Illustrious models are
So with right prompts, I think 3/4 of the things that we needed to use LoRA to do the thing on Illustrious can be made without using anything for Anima
Much better results with IL17