1.Do not use to create NSFW images of minors.
2.Do not use for criminal purposes.
3.Please create images at your own risk.
4.This model follows creativeml-openrail-m.
Denoising strength: 0.4, Hires upscaler: R-ESRGAN 4x+ Anime6B
SD1.5 model includes
https://huggingface.co/cafeai/cafe-instagram-sd-1-5-v6
https://huggingface.co/acheong08/f222
https://huggingface.co/LarryAIDraw/koji_v10
https://huggingface.co/unkounko/BalloonMix
https://huggingface.co/Kotajiro/kisaragi_mix
https://huggingface.co/naclbit/trinart_stable_diffusion_v2
https://huggingface.co/Vsukiyaki/ShiratakiMix
https://huggingface.co/dwarfbum/Hassaku
https://huggingface.co/deadman44/SD_Photoreal_Merged_Models
https://huggingface.co/scriche/AIOMonsterGirls
https://huggingface.co/Lorim/The_WonderMix
SDXL model includes
https://huggingface.co/John6666/kosa-v2-sdxl
https://civarchive.com/models/383364/CreaPrompt_Lightning_Hyper-SDXL
https://huggingface.co/TensorStack/iNiverseMix-XL-onnx
https://huggingface.co/natsusakiyomi/HimawariMixs
https://huggingface.co/Eugeoter/artiwaifu-diffusion-1.0
https://civarchive.com/models/402573/SixDragonsEatingCake
https://civarchive.com/models/128351/anime-natural-language-xl
and some self-made Low-Rank Adaptation.
Description
Adjusted to make it easier to output in natural language when the CFG scale is 5.
FAQ
Comments (4)
Interesting variance
Versatility never pays.
However, I need a 2.5D model to make LoRA.
モデルそれぞれ何が違うのでしょうか...
The new version is suitable for extension with Detection Detailer and Hires.
As a result, the output image will contain more similar compositions.
If you want to create various designs even if it is unstable, we recommend using the old version.