FilthyDegenerate

AI Slop 20 [SCAT]

https://civitai.com/user/xDegenerate

AI Slop 19

https://civitai.com/user/xDegenerate

AI Slop 18 [SCAT]

https://civitai.com/user/xDegenerate

AI Slop 17

https://civitai.com/user/xDegenerate

AI Slop 16

https://civitai.com/user/xDegenerate

AI Slop 15

https://civitai.com/user/xDegenerate

AI Slop 14

https://civitai.com/user/xDegenerate

AI Slop 13

https://civitai.com/user/xDegenerate

AI Slop 12

https://civitai.com/user/xDegenerate

AI Slop 11

https://civitai.com/user/xDegenerate

AI Slop 10

https://civitai.com/user/xDegenerate

AI Slop 9

https://civitai.com/user/xDegenerate

AI Slop 8

https://civitai.com/user/xDegenerate

AI Slop 7

https://civitai.com/user/xDegenerate

AI Slop 6

https://civitai.com/user/xDegenerate

AI Slop 5

https://civitai.com/user/xDegenerate

AI Slop 4

https://civitai.com/user/xDegenerate

AI Slop 3

https://civitai.com/user/xDegenerate

AI Slop 2

https://civitai.com/user/xDegenerate

AI Slop 1

https://civitai.com/user/xDegenerate

Starfire - Teen Titans [LORA][REQUEST]

Don't send private messages pls, I'm a introvert and a big schizo (if you haven't noticed by four years of bizarre animations as JohnDoe1970), I don't like interacting with people privately, if you want to say something drop a comment.

Also TAAAAAAAAAAAAAAAAAAAAAAGGGING!, this thing, this tagging thing, is so important, I used the same settings, one gave me a tangled mess of bullshit, generating body horror, the other gave me THIS, a beautiful LoRA, thing is, I like a c...

High Protein Latte [LORA][CONCEPT]

This is one is FUBAR, well at least I tried, I'm 100% sure AI can do it, is just my tagging that is shitty.

Got a new idea for next concept.

https://civitai.com/user/BootyAttackArt

Apu Spills His Tendies [LORA][CONCEPT]

AI has a easy time learning concepts more than characters, bizarre I know, but has some trouble with overlays as it often misplaces it, be it shadows, fake screenshorts or viewfinders.

Also I fucked up and didn't paid attention so a long sleeve shirt got backed in, so add sleeveless to positive and sleeves to negative and it should help a bit.

And I got the 'universal' settings for character and concepts down, I think.

2024-12-19 16:21:26 +0000 UTC View Post

Morning Paper [LORA][CONCEPT]

https://civitai.com/user/BootyAttackArt

My first custom concept, seeing if I can put on 'paper' what's on my head

Blue - Wolf's Rain [LoRA]

Getting the hang of LoRA creation, this one I tagged manually, and I included her Wolf form as well, sadly I only got 9 images of her so the end result is not that good.

https://civitai.com/user/BootyAttackArt

Gwen - Total Drama Island [LoRA]

My god this bitch gave me TROUBLE, I'm a newbie but holy shit this was hard, so I noticed every jabroni on Civitai has a 213mb LoRA, well thats for a reason, higher ranks mean the AI will learn faster, specially 'tiny' details, namely her lips, which the upper is slighly darker than the lower, I was trying to save space making sub 100mb~ LoRA's but that costs accuracy, a lot of it, plus a high batch size seens to degrade the quality of the LoRA, after all we are getting more speed out of if, ...

Raven - Teen Titans [LoRA]

https://civitai.com/user/BootyAttackArt

Another LoRA, learned THE settings, and learned how to overtrain and how it is a good thing, overtraining simply means I can let it cook for an hour then simply do a sample check and use the correct LoRA, I learned it like most things I learn, by pure accident, the LoRA was giving me some fucked up results at 1.0 and I was desperate because I 'lost' one hour, but then ...

AI Slop 166

https://civitai.com/user/BootyAttackArt

AI Slop 165

https://civitai.com/user/BootyAttackArt

The Simpsons Style [LoRA]

Third LoRA, faced a lot of issues with this one, seens they are undercooking, thus increasing the LoRA's weight is necessary, this is not a bad thing as burnt render is worse, but still I want to be able to trigger it with the default value aka 1.0

I believe it is the batch size that it is screwing me over, I want a higher value to save time, but that saving perhaps is making the LoRA weight have less impact as If I used a lower batch size (1).

But I learned even more today, speci...