11th Release, 堕姫 - Daki from Kimetsu no Yaiba.
Put everything in Filter for seeing all images!
Examples Image is now active, Click on Show more for Weight & Trigger recommendation, Also click on About This Version at Right Corner for more Info.
I didn't not find a Lora on Daki & Not so much Art on it, so again i created a LORA Like MTSpider.
This Lora is trained on 114+ Images with Good quality, I sorted everything manually, Only picked the best possible. (SFW + NFSW)
AnimeFullFinalPrunedFP16 used, so he work with the majority of Anime Model, you can see in examples images. (Abyss Model + 7th Corneo's Heaven's Model.)
Well, he can do Default Skin with Black hair & Demon Skin with Multicolored Hair White & Green Hair.
About Trigger Words & Prompts: DakiV4 (For Trigger) & for prompts just refer to examples image.
Weight recommended: 0.6-1. - Sometime weight at 1 is too strong so the Best result is 0.8.
Negative Embedding used: EasyNegative, BadPrompt2.
Leave a Review & Share you're image in comments if you appreciate this Work :).
Description
If Skin is not 100% white, force it with: White Skin prompt.
If Black Hair is not 100% Black, again force it with: ((Black hair)) prompt.
You can also have Facial mark with Facial mark prompt, but it's not very precise.
蕨姫 / Warabihime Skin is also low on datasets, so color not precise.
FAQ
Comments (15)
So funny i just made one too
Ahah, yea it's happen's to me that i create a lora and the next day someone did it too ahah this technology Evolve really fast, sometime also i create a lora that people have already, but in a more better quality to my taste ahah, and btw don't be shy to post ur Lora, it's good to have diversity, some people will prefer the one you Created or Other it's depends of taste.
I wanna see yours too, please upload
great
looks awesome man thank you, it'd be epic if you tried making Douma, he's my favorite
I think it's possible to create him, Many characters (Specially Male characters) Miss on Civitai or IA Art in General.
boss，lora is not working，why
Just copy my prompts in my examples images, i think you missed something or the Realistic model you used don't work with This Lora, i was not sure if this Lora work on Realistic model, because it's a Anime Lora, but i just tested on ''Deliberate model'' and it's work atm (Not perfect, because it is optimised for Anime version) But work.
A little tips: Just Drag and Drop the image example you want in Png Info and click on Send to TXT2IMG.
It will copy all of the MetaData of the image & Gain you time.
@KXRO THKS! is working，
@Lonexd6688 Nice picture ahah.
Did you train her with images of different styles or just with screenshots from the anime?
It's MIXED But Low in Anime Screencaps - Majority Of 80-90% Of Original Art at Max Quality possible when i did create her. (So it's good for Original Art).
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.