Deadpan AF. She's not too interesting nor is she exciting, she functions mostly as a home security system.
Redone because if I've done Kitty then I have to do the rest and also the face has to be accurate!
I have a deep rooted OCD for that I think. I mean... humans have a dedicated part of the brain only to recognize faces. ...come on man, it can't be just me.
Recommended style tags:
animated, 2000s \(style\)Outfits:
hair pulled back, earrings, hairband, bead choker, sleeveless shirt, purple skirt, flip-flops, bracelet, anklet
jumpsuit, cape, puffy sleeves, white gloves, white boots, high heel boots -she has blue eyes in this outfit-
hair pulled back, hairband, sunglasses, over-rim eyewear, bikini, see-through sarong
I always generate with Euler a, 25 steps, CFG 5, very basic for IL and pony.
Schedule type: automatic
Recommended Resolutions: (SDXL standards) - Pony/IL too
640 x 1536 / 1536 x 640
768 x 1344 / 1344 x 768
832 x 1216 / 1216 x 832
896 x 1152 / 1152 x 896
1024 x 1024
I train everything on IL XL 0.1 so it's compatible with any IL model, but the problem with most checkpoint is that they are leaning towards anime too much and cause the chin and eyes to deform regardless how hard I train it. Legends say it should be allright with NoobAI aswell, I dunno', never tried that, it's an anime wasteland I was told...
Description
FAQ
Comments (5)
Yep, it's just you. :-p
Just look at most char lora creators, most just gave up on illustrious faces long ago...
I mean, I'm as OCD as all heck, so I can appreciate where you're coming from, but... sorry man, no. It's just you.
You'll no doubt be richly rewarded for your support when AI takes over. Or maybe Skynet will come for you first, whichever.
i know what most lora creators are like... no attention to face, outfit, tags and other details, its a fucking horror show out there. ive stopped browsing loras long ago
wanted to make a lora training guide but decided no to since ive mentioned the attention to detail is low so theres no point really.
the weak will fall!
ps: the effort i put into this is for my own amusement, im not here to amaze anyone, i just share the fruits.
@xJollyboyx Yeah, the weak have already fallen. It's long been clear which of this site's creators are the select few who actually put in the effort.
The rest are like grass -- good for background. And yeah, you can chew on it; but only if you just cannot find the food you're looking for.
It's actually remarkable (to me...) how much of the progress this field has made has been by private/opensource/etc people only in in for their private amusement.
That's a huge part of why it's such a cool field. And a refreshing break from our over-commercialized world.
@ravemry9 true, true...
the only reason i started creating loras was because i had too many characters at first i wanted. got lucky and i managed to collab with another creator, helped him with a few datasets i gathered for him, and that gave the initial push to start doing this on my own. so basically i had a very vague understanding but that was enough to get me started and the snowball effect was crazy!
@xJollyboyx That's pretty cool.
Especially that it was enough to get you started down the model creation path.
I've noticed a parallel thing in the IRL art scene (painting, some sculpting), IME the artists that have what it takes to reach the top often end up striking out on their own with vague instructions and end up carving their own niche for themselves. It makes sense it'd generalize to other forms of art like AI.
And yeah, I feel like many of the stages of my skillz in AI came thru similar snowball type of stuff, especially lucky meetings on this site where we ended up connecting and sharing methods. That's why i love this site, despite all the issues.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.