Please check out my Ko-Fi page if you are interested in supporting me.
LoRA commissions open! -> Fiverr
This is a LoHa LoRA! You need the extension! Why? They do a lot better with multi-concepts.
It struggles a little with her tail, unfortunately.
I recommend 0.5-1 LoRA weight and v2.19 is better at fidelity for her original clothes.
For versatility, you can change pretty much anything on her:<lora:MajinAdnroid21v118:0.7>(majin android 21, tail, black sclera, red eyes)
For her ORIGINAL CLOTHES: <lora:MajinAdnroid21v118:0.7>(majin android 21, tail, black sclera, red eyes, white hair, long hair, hoop earrings, bracelet, black nails, tube tank top, yellow choker, white harem pants, detached sleeves)
Follow the steps in the Model Version notes to reproduce my images.
Sample images generated with AbyssOrangeMix3A1.
All negative embeds should be easy to find. and Prompts should be included in the sample images! Click the little icon in the bottom right to see it.
You might be able to get her with a penis, if you prompt for it, 3 images in the dataset had one.
Also, if you want stronger skin color, check my Colored Skin LoRA.
It was trained on 93 512x512 images scraped from the internet and 20 epochs using Anime_final_pruned.
Danbooru tags generated by Waifu Diffusion 1.4 Tagger Extension and fine tuned.
Description
Generate using a weight of 0.5-1.0 like this: <lora:MajinAdnroid21v216:0.7>.
How I archived my generations:
txt2img: DPM++ SDE Karras with 20 steps, 512x512(can go up to 768 for wide or tall), Restore faces on (using CodeFormer weight 1.0 in Settings), Hires-fix on, to 1024x1024 (double original size) with your preferred upscaler and Denoise 0.4, CFG 7. Then send to img2img!
img2img: DPM++ SED Karras with 20 steps, Restore faces on (CodeFormer weight 1), rescale to 1536x1536, CFG 7, Denoise between 0.3-0.7 (based on how much you want to improve the image), SAME PROMPT! Send to extras when happy with img2img/inpaint results.
extras: Scale to x3 (final will be 4608x4608, I suggest going lower, I did x3 just for the samples here) Upscaler 1 - FatalAnime 4x, Upscaler 2 - SwinIR 4x with 0.10 visibility, GFPGAN 0.10 visibility, CodeFormer 0.10 visibility.
If anyone is curious about my version names: v{version}.{trainingEpochs}