Covering a Diverse Range of Styles! Three Models with Distinct Personalities
Hello, it’s kojirom.
Characters with larger bust sizes can easily gain a lot of fans. The models I’m introducing today are no exception, creating women with a considerable amount of volume. However, I’ve also perceived great charm and potential beyond just that.
These models have been successful in highlighting the unique personalities of the characters. I believe they may become quite popular in the future.
Now, let’s introduce the three models that are the focus of today’s discussion.
caulkinumV2 for Anime Style – ARCT | Stable Diffusion Checkpoint | Civitai
[Author = 489]
儿童绘本插画MOMO_v1.0 – v1.0
儿童绘本插画MOMO_v1.0 – v1.0 | Stable Diffusion Checkpoint | Civitai
[Author = MarkP]
schExcelsior Mix [Legacy] – Redux-v3.0
schExcelsior Mix [Legacy] – Redux-v3.0 | Stable Diffusion Checkpoint | Civitai
[Author = DreamExplorer]
‘CaulkinumV2 for Anime Style’:
This model creates simple yet realistic-looking beautiful girl characters.
‘儿童绘本插画MOMO_v1.0’:
This model is likely well-suited for creating picture books, icons, and manga.
‘SchExcelsior Mix [Legacy]’:
It excels in generating strong, wild-looking female characters.
Make sure to use them in your creations!
caulkinumV2 for Anime Style
note: Site Summary
caulkinumV2’s glow,
Anime forms in light’s dance,
Elysium’s child.Lost recipe blends,
Hierarchical merging,
Silicon whispers.Star marks, signs of love,
Variations in series,
Joy in creation.
儿童绘本插画MOMO_v1.0
note: Site Summary
The article discusses achieving high-definition and detailed visuals using the SD1.5 model, similar to SDXL quality, with a resolution close to 2.1768. It includes information about the original LoRA origin and descriptions of the 儿童绘本插画MOMO, a model for children’s book illustrations in a cartoon and anime animation style. The process involves simple description inputs with the addition of the trigger word “illustrations” and utilizing high-definition restoration techniques. Key steps include using 20 iterations, a redrawing scale of around 0.6, and the R-ESRGAN 4x+ Anime6B enlargement algorithm. The Euler a, dpm++2M SDE karras sampling method is recommended, with a 2x magnification and a weight range of 0.6-0.95 for best results. Positive and negative keywords are provided for guidance, focusing on enhancing quality and avoiding low-resolution or undesirable elements.
schExcelsior Mix [Legacy]
note: Site Summary
A legacy version of a model mix, previously purged, has been restored for community use. Its removal was due to drama. The model works easily without much complexity or guidance. It’s a “thirsty” checkpoint, requiring users to tone down NSFW content in the negative prompts. The negative prompts are pretty standard, with ‘worst quality’ and ‘low quality:1.4’ being used alongside EasyNegative. The model mixer was Schneed, whose account is now deactivated.
All Illustrations Are Commercially Usable
Some illustrations may have incomplete details such as hands and fingers. We kindly ask for your understanding.
When using these images, please feel free to crop or modify them as needed.
We encourage you to fully express your creativity and use these illustrations with a free and imaginative approach.
The illustrations featured on this site are available for commercial use.
Configuration Parameters and More for Stable Diffusion Web UI
Positive prompt: {1 girl|wearing a swimwear}
Negative prompt: EasyNegativeV2
Steps: 80
Sampler: DPM++ 2M SDE Karras
CFG scale: 6
Seed: 1587948557
Size: 512×512
Model hash: 6553284567
Model: schexcelsiorMixLegacy_reduxV30
VAE hash: f921fb3f29
VAE: klF8Anime2_klF8Anime2VAE.pt
Denoising strength: 0.5
Clip skip: 2
Hires upscale: 1.5
Hires steps: 40
Hires upscaler: 4x-AnimeSharp
Version: v1.7.0