Kasumi
from
Tanetsuke Ojisan to NTR Hitozuma Sex The Animation
(種付おじさんとNTR人妻セックス The Animation)
This model was trained entirely on anime screencaps.
Preface:
My third SDXL model. ( ͡° ͜ʖ ͡°)
Base Model: Pony Diffusion V6 XL (https://civarchive.com/models/257749?modelVersionId=290640)
Prompts:-
Trigger Words:
kasumi_olis, long hair, brown hair, low ponytail, green eyes, hair bow, huge breastsOutfit 1:
blue jacket, grey dress
Optional:
black thighhighs / shoulder bagOutfit 2:
pink babydoll, covered nipples, see-through, crotchless panties, pussy, uncensored
Optional:
breasts out (replace covered nipples) / mole on breast / strap slipAccessories:
hairclip / black choker / black ribbon / wedding ringOptional:
1girl / solo / full-face blush / constricted pupils / facelessKnown Issue:
Her outfits can be hit and miss sometimes. Using Negative Prompts may help.
Weight: 1 (feel free to lower it, especially if you don't care about her outfit)
Hopefully some of you will like this model.
Description
FAQ
Comments (10)
Hajimete no Hitozuma female characters pony xl?
hi, could you tell me how you create the XL models? I know how to create the sd1.5 ones but not these. could you help me?
Hi,
I don't know how you trained your 1.5 models, but I guess you used Kohya-ss. While I did the same for my 1.5 Loras, I trained my SDXL Loras using OneTrainer on my 8GB VRAM GPU. Maybe you can try using OneTrainer too.
OneTrainer: https://github.com/Nerogar/OneTrainer
@JasonCMK to create an xl pony what should I select in one trainer?
Hi. Every time I try to make some art with this, it just comes up with bad images. I’m not changing the prompts, but what additional sauce would you say to use?
Hi,
Since I don't know how experienced you are with PonyXL, I suggest you first try generating images with the same prompts, settings and model I used in the sample images.
I didn't use additional sources. I just upscaled the images with Kohya HRFix after generating them.
@JasonCMK Thanks
@scott_toulson714
Hi,
In fact, there were some mistakes on my part.
First, the Kohya HRFix actually doesn't seem to provide many benefits in my use case.
Second, I upscaled my images twice with a scale of 1.5.
Besides, sometimes I roughly fixed irregular fingers or toes after the first upscale using Photoshop.
Error: Could not load the lora model! Reason: Could not find text_encoder_2.text.projection in the given object!
It doesn't work, it's a pity :(
Hi, I did some research and believe that your problem is related to EasyDiffusion. Could you try using the LoRA with another software?
For your reference: https://civitai.com/models/233180/kawaii-makeup-xl?dialog=commentThread&commentId=393794
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.