N°56 Release, 竈門禰豆子/Nezuko Kamado (Demon Form) - Kimetsu No Yaiba - 鬼滅の刃/AnimeTV Style.
Put everything in Filter for seeing all images!
Click on Show more for Weight & Trigger recommendation, Also click on About This Version at Right Corner for more Info.
Another Character's Request for @DogMaR34 - A little more harder Than Shinobu, Tho i did get it Right, Also a Full Tryhard datasets, Same Technique as my other's Lora's, Full AnimeTV Like Style.
About this Lora:
114+ Images, All Are AnimeTV Like/Anime Screencap (1080p) & Official Artwork (4K) Quality, All Manually Picked & Tried.
All Screencap & Artwork were Upscaled with x4-AnimeSharp for Better output Quality + Manually crop by certain part.
Very precise in AnimeTV Like Style. (Working perfect on AutismMixXL & PonyV6).
Can do with & without ''Gag'' + Without Horns, Veins & Tattoo (So 3 Forms Demon/Happy+Sun).
Precise in the default outfit skin (0.8-1.0 Weight) & Ok/Good in flexibility (0.7-0.9 Weight) See Examples images.
Can do many Angles, Tho some can miss, but 80% are Ok. (Some data were missing).
Eyes is Very precise 99% Of time, I get it too ahah.
PS: Trigger words is: DNezukoXLV4, (See examples images).
Description
Recommendation for Official Default Skin: I recommend to put: Exactly like in my Examples Images for Prompts, For a Better Precision. (0.8-1.0 Weight).
For AnimeTV Style: Just need a good Model's like PonyV6 Or AutismMixXL you obtain it pretty easily. (I already optimized Prompts & Negative Prompts in Examples Images).
For Flexibility: for like clothes that don't exist put 0.6-1 Weight - (0.7-0.9 Best Result too), See my Examples images for this. (RAW Best = (0.8 Weight).
Also if you don't want something to appear, put it in Negative Prompt. (Again See Examples Images).
FAQ
Comments (9)
I'm happy I didn't get to make the dataset yet😅 lol. Good job.
Datasets was not easy to have actually ahah, especially in her 3 forms, I did retrain her 2x Time KEKW
@KXRO Yeah, they requested it to me, but I was preoccupied with my own stuff. Seems like only screencaps datasets work way better on Pony than NAI.
@IndolentCat Mhh, Actually my first try was a Little bit Mixed with some FanArt that's look like AnimeTV too (Because i didn't have all the angles i wanted), But it's fked the ''Full AnimeTV Style'' A little, So i removed them after 2x try, And now it's good ahah, in Pony you can't Mix even a little otherwise the style is Fked compared to NAI, SD 1.5 (NAI), is still a little bit better on Having Original Art+Anime Coloring at the same time.
@KXRO interestingly enough Nezuko have many common trait with one of my OC. Pony managed to get that right after 3 tries. ( i did train her like 50+ times before).
the horns vein and tattoo are clearly there ! bieautiful and flexible
Then your OC Is probably Nice too ahah, Thx np.
PS: (Un autre goat français par la ? ahah)
Hi, could you make the version for SD1.5?
Probably, when i finish all my Request yea, Tho in SD 1.5, i'm not sure if it's will be as precise, But i can try ahah
@KXRO ;)
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.