Note #1: Version 2 is up and running and works well with all the models that I tested! For whatever reason I like making my favorite characters look like goths apparently lol.
What does it do you ask? This more SFW Change-A-Character is focused on transforming your waifu from that normal, adorable to girl to that hot goth gal of your dreams. There is no specific outfit it is trained on but more the idea of a goth (with it purposefully going towards black outfits and heavier make-up). Feel free to attach whatever you want in the prompt for a base outfit you'd like and how it becomes more gothic.
How to use it:
(I use Sampling Method DPM++ SDE Karras and Upscaler R-ESRGAN 4x+ Anime6B).
0.7 to 1 depending on model/lora/etc.
Main Triggers
GothicC - Just to fortify the idea of a goth. Doesn't need to be used if you don't want it as impactful.
makeup, lipstick, mascara, black nails - Another list of prompts I like using since it helps round out the look.
And then whatever else you want! You can, but don't have to, specify a type of outfit.
Just To Drop Some Prompt Words That Are Fun
gloves, pants, dress, elbow gloves, veil, jewelry, thighhighs, hat, high heels, gothic, necklace, fishnets, skirt, spikes, spiked collar, cross, torn clothes, earrings, collar, shoulder cutout, cross necklace, fingerless gloves, pantyhose, belt, choker, black nails, tattoo, fishnet pantyhose, black lips, makeup, arm tattoo, piercing, miniskirt, detached sleeves, crop top, corset, shorts, garter straps, eyeshadow, nail polish, multicolored hair, see-through - You don't have to use any of these, mostly just here for some common words that were appearing in training.
Extra, Extra!: Should be well trained and usable in most models. Images shown for example.
Picture Descriptions:
1-3: Hasaku with just this Lora.
4: MoonAnime with just this Lora.
5-6: RevAnimated with just this Lora.
7-8: Dreamshaper with just this Lora.
7: AnythingV3 combo'd with my released Rikka Lora.
7: AnythingV3 combo'd with my released Akane Lora.
7: AnythingV3 combo'd with my released Volume Lora.
7: AnythingV3 combo'd with my released Kohaku Lora.
8: AnythingV3 combo'd with my released Idol Locon.
8: AnythingV3 combo'd with my released Idol Locon.
8: AnythingV3 combo'd with my released Naughty Nurse Locon.
How was it trained?
It was trained on animefull_final with 41 images with the usual self-pruning and tagging and a wide range of different images. I like keeping my prompts simple for the first few pictures so that you can see it without a whole bunch of mumbo jumbo. Make your pictures better than mine though!
What's next for this?
I think it's fine? LMK.
What's next that isn't this?
More character Loras. (Freya from Danmachi, Mio and Ushio from Summertime Render, Junko and Kaede from Danganronpa, Manon from Street Fighter, Yuna Yunis from Legendary Hero is Dead, Jill Valentine from Resident Evil, Rupee from Nikke, Spectre Tenebria from E7, Zephia from Fire Emblem Engage, Ivy from Soul Calibur 5, Heine from Combatants Will Be Dispatched, Conqueror Lilias from E7, Evelynn from LoL (base version), Laharl-Chan from Disgaea 5, Miku from Idaten Deities, Kasumi from P5, Ikumi Mito from Food Wars, Alice Nakiri from Food Wars, Lulu from FF10, Dominique de Sade from Vanitas, and Dos from One Punch Man.) And maybe Wilhelm van Astrea, my favorite Re:Zero husbando or a Prinny from Disgaea.
V2's of some of the previous Loras. Specifically make them more diversely acceptable with other models and LORA's.
I also have some more quite NSFW Locon's. I decided to list them as well, so if people have preferences on which to fix up first let me know. Specifically, Batta-Style Brainwashing, Gasugao (EbiBlue style), Dildo Riding, Mnemosyne Style, Foot Worship, Head Infected, Living Clothes, Masturbating while Clothed (I had trouble getting masturbation with clothes on, so making a Locon), Slutify/Hookerify, Shy Gal-ify (I had no idea this was a thing but I found so many images of it when looking at other stuff I just tossed it in), Bowsette Transformation (Makes your waifu kinda look like Bowsette without changing other aspects), Fallen Angel, Pony Bondage (specifically with the capability of being "riden"), X-ray Vision (glasses that give you x-ray vision of a girl, this one has been hard), Vore (Waifu mid-way being eaten specifically), More Brainwashing (in ideation for this still), Bound Yuri (two girls tied up, breast to breast and likely kissing), Tied Till Orgasm (basically the waifu tied up with vibrators and being overwhelmed with pleasure), Beta Censorship (Automatically censors any images you create to make them BetaSafe), Robotification, Tentacles, Pillory, Latex Corruption, Combat Drone (hard to describe but basically making them a mindless sci-fi combat warrior), JAV Supervillain, and Blacked.
And, uh, I want to make a few SFW Locon's so I have a Drow, Cyberpunk, Vampire, Sailor, Courtesan, Egyptian, Naughty Schoolgirl, Witch, Samurai, Tomboy, Ninja, Maid, forest fairy.
I'm also planning to split up my Evil-ify your Waifu into two separate files, Lustful Corruption (Done) and Pure Evil.
Or just give me some random suggestion or priority opinions. That works too.
I've also added a couple of my extreme NSFW preview images to my twitter (https://twitter.com/TwoMoreTimes89). Let me know if you think I should work on them more for a submission here.
Description
FAQ
Comments (11)
Can't seem to get this one to work like the others. Tried mimicking your yaoyorozu pic but kept coming out weird
@PangolinMaster Weird. I just downloaded and retested and it seems fine on my end. What portion is coming out unexpected?
@PangolinMaster Can you test the first pictures prompt as well to see if that one is working? That should come out basically 100% the same with the same prompt.
@TwoMoreTimes89 Tried that and the issue is it keeps distorting the face for some reason. Want to me to post to the gallery?
Also seems to be changing the style to a more realistic one for some reason.
@PangolinMaster Ya, go ahead and I can see if I can triage it. I copy pasted the original myself and it's fine so I'm very curious :).
@TwoMoreTimes89 Posted! All to do with the eyes. I kept playing with the weights but they distort and go weird no matter what I do. Could it be because I'm not using hires fix?
@PangolinMaster Nah, hires fix shouldn't make THAT big of a difference. What checkpoint are you using (Scratch that, I used to model hash to find it was hassaku)? Plugging in that exact prompt still get me a decent picture (when I do 512x768, otherwise it just focuses on the corset). And the fact that it focuses on just the corset is still way different.
@PangolinMaster Wait. I figured it out and I have no idea why it looks so whacky with this. It's actually just the width x height. Change to 512x768 and it works perfectly. But I have NO idea why 512x512 turns out so bad. Maybe it needs to be trained on 512x512 images as well? I totally don't know but I'll try to learn it too lol. I've never tested heavily on non-512x768 (unless it's something that makes sense for it like laying sideways) or hassaku itself.
@PangolinMaster Scratch that. It works BETTER but to get it to be actually quite good on Hassaku it seems you need both 512x768 AND hires fix... And then it still might not be perfect perfect. I'll try to test more on Hassaku since it's relatively popular and I've only really been focusing on AnythingV3, AbyssOrange, and RevAnimated (with MeinaMix and PerfectWorld at times).
@TwoMoreTimes89 I see, I should have known. All the other lycons worked with 512x512 so I assumed this would too but I should have known seeing your test images. My bad! Also hassaku works with the others but I will try with abyss orange
@PangolinMaster I'm glad you called it out because I know why this one doesn't work Hassaku. There wasn't enough good gothic anime images of high quality and a variety of clothing so I had to grab some images of real life models wearing different outfits. Now, I'm not sure why Hassaku isn't able to process those differences, where-as the other checkpoints can, but it's good for me to note if I try this strategy again. I'll make a note about it in the page and potentially make a less robust, but compatible with Hassaku, version sometime this week. And will likely avoid real-life pictures being added to the data-set unless I'm super stretched for high quality images... Which for this one, I am unfortunately (without losing the robustness of different outfits).
Appreciate the call-out because it was some good learning for me!
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.










