









Crazy dataset with super detailed images and awesome wizard and witches!
You should use the trigger word hkmagic
The lora adds some realy cool details!
Version 2 focuses on 2D anime style
Some times it gives boss battle images, add like multiple people in negative prompt or solo/1girl in normal prompt.
The LoRA weight I have been using is 0.85-1, but feel free to try and experiment!
I have used ComfyUI when generated pictures and the settings im using are:
Sampler: Euler A
Scheduler: Normal
Steps: 50
cfg: 5
Example prompts used:
"A close up 2D anime-style illustration of an evil little red wizard demon with glowing black eyes and a point ears, red skin, wearing a black, tattered magical robe with red details, conjuring a massive black blazing black fireball with both hand, smoke and black fire swirling around him, the firelight casting dramatic shadows on his face, torn cape billowing behind him, fantasy burning tavern in the background, fiery color palette of crimson, amber, and dark black, clean cel-shaded anime linework, high-detail, Japanese anime concept art style, hkmagic"
First image is with my LoRA and second is without my LoRA, same prompt and model.
Recomended vae that I use is the official sdxl_vae
1. 轉载模型僅供學習與交流分享,其版權及最終解释權归原作者。
2. 模型原作者如需認領模型,請通過官方渠道联系海藝AI工作人員進行認證。我們致力於保護每一位創作者的權益。 點擊去認領
