home models images videos 3D Models articles comics challenges updates shop

Killua Zoldyck SDXL Lora

Name: Killua Zoldyck SDXL Lora
Rating: 5 (41 reviews)
Author: l0ckd0wn

706

3.8k

Updated: Jun 30, 2025

character

anime male man hunterxhunter

Download

1 variant available

SafeTensor

217.89 MB

Verified: 3 years ago

Download (217.89 MB)

Details

Type

LoRA

Stats

466

3.6K

742

Reviews

Positive

(30)

Published

Feb 10, 2024

Base Model

Pony

Training

Steps: 20,000

Epochs: 50

Hash

AutoV2

01933DEAEE

Trigger Words

killua_zoldyck

Tensors

Recommended Resources

About this version

This version is a slight improvement over the previous version as now you have more control over the quality and style of the character. Now you can change between the anime style with "hxhanime" in the promt and a more free fan art style without "hxhanime". Also I captioned the images manually with the same syntax as the base model. ponyDiffusionV6XL.
The quality can be controlled via using "score_4_up", "score_5_up", "score_6_up", "score_7_up", "score_8_up", "score_9". Quality is based on my personal preference but 8 and 9 are also improve the likeness to the character.

default creator card background decoration

l0ckd0wn

Joined Nov 18, 2022

License:

CreativeML Open RAIL++-M

Please look at the base model for the lora. It will perform better and sometimes only with the base model. The base model is named in the version name.

With the most recent version there are some new additions.
You can use "hxhanime" in the prompt to get killua from the anime. Also I've captioned the images following the prompting of the PonyDiffusion V6XL Model. So using something like score_8_up may improve quality.

Also I've captioned them according to their rating (rating_safe, rating_questionable, rating_explicit)

I want to mention that the the current version of hxhanime results often have a low resolution. That's because I didn't have any high quality hxh screenshots of killua. If you have high quality screenshots from the anime (1024x1024 px and up) I'd be glad if you could send them to me so I can train a new model.