Sign In

detail_slider

395

Not recently loaded

Download

1 variant available

bf16 SafeTensor

ms_Realistic_Anime_silder_xiaowuhen.safetensors

BF16, good balance • 71.02 MB

Verified:

Type
LoRA
Stats

31

170

678

Generation License Fee

1 / 10 images

Reviews
Published

Oct 23, 2026

Base Model

MOSS Image

Hash
AutoV2
2A9503F9F2
ComfyUI_01321_.jpg

MOSS-IMAGE-SFT 动漫到真实滑块。

权重负:【动漫】-1==>-20,建议正常使用在-10以内。

权重正:【写实】+1==>+10,正常使用+5足够了。

参数:四步加速LoRA运行8,采样器:euler_ancestral,调度器:BETA57

--

MOSS-IMAGE-SFT: Animation to Realistic Slider.

Negative weight: [Animation] -1 == > -20, recommended to use within -10 for normal usage.

Positive weight: [Realistic] +1 ==> +10, +5 is enough for normal usage.

Parameters: Four-step acceleration LoRA run 8, Sampler: euler_ancestral, Scheduler: BETA57


Krea 2 类动漫光影氛围滑块 LoRA(Slider LoRA)

这是一个基于 Krea 2 Raw 训练的滑块 LoRA,同一构图下,平滑地把画面从「明亮通透」推到「浓郁动漫色调」。它不改人物长相、姿势和构图,只改曝光、光比、色温与阴影层次。

推荐权重 日常微调 ±0.5 ~ ±1.5;强风格 ±2.0 ~ ±2.5;更宽的范围可以根据图像的具体情况自行探索。

重要提示:如果图像不包含或只有少量光影、画风描述效果会更佳。。。

---

Krea 2 Anime Light and Shadow Tone Slider LoRA (Slider LoRA)

This is a slider LoRA trained using Krea 2 Raw. With the same composition, it smoothly transitions the image from "bright and clear" to "rich anime tones." It doesn't change the characters' faces, poses, or composition—it only alters exposure, contrast, color temperature, and shadow levels.

Recommended range: Fine-tuning for daily use ±0.5 ~ ±1.5; for strong style ±2.0 ~ ±2.5; feel free to experiment with wider ranges based on the specific image.

Important note: This effect works best if the image doesn't contain much or any light and shadow—descriptions like "anime-style" will give better results...


这是一个由于改善皮肤质感的滑块LORA。

推荐强度范围-1<----->+1.5

其他强度范围可以自行探索,目前来看。超出推荐范围可能会改变画面的整体细节的丰富程度。有点小瑕疵,先就这样吧。我也许很快会更新下一个优化版本以及一个皮肤美白的工作流。也许。

-

This is a slider LORA for improving skin texture.

Recommended intensity range: -1<----->+1.5

You can experiment with other intensity ranges on your own for now. Going beyond the recommended range might change the overall detail richness of the image. There are some minor flaws, but let's leave it as is for now. I might update with the next optimization version and a skin whitening workflow soon. Maybe.