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
Remove crispiness and enhance the posing
NoobAI EPS 1.1 + Illustrious v1.1 DARE applied
FAQ
Comments (34)
@Crody
So what is the major difference between illustrious 6A vs illustrious 6B? Does one have more noobAI in it or is one more compatible with Illustrious loras?
The key difference is the base coloring style
After IL v4.0, the coloring style gets similar to Pony versions
I saw lots of people still using IL v3.0 because of the prompt adherence and coloring style so I just wanted to update it with using newer base system
The result is IL v6B: extend from IL v3.0, which has more flat coloring rather than volumetric coloring
IL v6A is same as IL v5.0 but it has improved posing and reduction in noises
Both of them are good and compatible with Illustrious / NoobAI LoRAs
big thankies deadshotz for asking the all important question
Illustrious v6 model is labeled as NoobAI so it looks like crap in the Civit generator. Only v3 works for me but now that version isn't in the generator.
Changed into Illustrious
Also, please allow advanced mode for generation setting in Settings for greater experiences
Fantastic and quite easy to get the hang off.
Bruh...Do you want me to be a true furry ?
I demand this damn Nova Animal v3.0 FUCK ME!
But you hardly know them. At least offer to buy them dinner first...
curious (because i always am about this) what tagging is this model based off? danbooru? derpibooru? e621?
Probably danbooru and e621 since Illustrious is trained on danbooru and NoobAI is trained on e621
But I know that danbooru tags works much better compared to e621 tags
Amazing
Why do you use the "BREAK" keyword in nearly every prompt? What's it supposed to do, and why is it followed by what look to be innocuous keywords?
It separates prompts before and after by adding 75 tokens
This prevents prompt mixing and prompt disfunction (prompt not working)
BREAK used in case when you want to prevent color mixing: wanted to have green hair and red eyes but got red hair and red eyes instead
detail enhancement prompt also works similar to this so to prevent them from not working, I use BREAK
@Crody Keep in mind that BREAK doesn't work in ComfyUI because the creator thinks that an industry standard is way too good for his program.
@thelesbiancarwasher Someone made the BREAK function for ComfyUI
Your welcome love this model❤️
You are 1st
really like the model, basically the only one I use, however I find the fingers and eyes tend to be very inconsistant. pretty much the only way I am able to get good eyes is if I use "looking at viewer" and "close-up" in regards to fingers I have to just keep upscaling the same image and prayging one of the seeds is good
Have you tried using ControlNet? It can send skeleton information to the model dictating the position of things like fingers.
Pony v6 lo quitaron f,,,,,
What's the difference between version A and B?
It's documented in this comment thread
https://civitai.com/models/503815?modelVersionId=1626156&dialog=commentThread&commentId=763592
@Crody Thanks!
Nice
pretty damn useful used this the most
Love this model! It's actually surprisingly good at human characters too, so much so that I'm sometimes using it instead of my main anime model! Keep up the good work!
How do I properly use the "BREAK" tag?
When you approach a tag limit like 73/75, but you still want to add description, make a new line. Write BREAK . Make another new line and repeat most important part of the image you are generating or focusing in. Add additional tags of what you want after. For example:
solo fox sleeping, bed, from behind, back, (your previous tags describing character)
BREAK
solo fox sleeping, bed, from behind, back, (your new tags describing surrounding or small details)
That way you tell SD what to do on each of those tag chunks and what part is really important and it should focus on working the most, if you don't do BREAK separation SD might focus on wrong parts of the image, like making that bag of chips, slightly to the side REALLY detailed or excessive.
@smetanlol Okay, thank you for the in-depth explanation!
@RabbitEnjoyer - Worth noting that BREAK is an Automatic1111 concept. ComfyUI doesn't handle tokens the same way, so it does nothing. CivitAI is comfyUI backend, therefore BREAK does nothing for online generation.
@unseelie23 Thank you for the additional details. I don't use ComfyUI right now, but I intend on using it in the future.
Awesome!!💚🐾✨
I love it!
yup :3