Type | |
Stats | 727 192 |
Reviews | (108) |
Published | Aug 13, 2023 |
Base Model | |
Hash | AutoV2 51394068F1 |
欢迎使用MiaoKa_reality_mix模型 这个模型 是MiaoKa_reality的融合模型 皮肤纹理白皙,可以画出多种效果 中国风 2.5D风格都可以 兼容二次元动漫风格模型
Welcome to use MiaoKa_reality_mix model This model is MiaoKa_reality fusion model skin texture white, can draw a variety of effects Chinese style 2.5D style can be compatible with the secondary animation style model
模型内置VAE 非必要 建议不要额外添加VAE 否则部分场景会过曝
VAE is built into the model. Do not add additional VAE; otherwise, some scenarios may be overexposed
模型反馈请站内私信
建议出图参数(Suggested drawing parameters)
采样方法(Sampler):
Euler a:25~30 步
DPM++ SDE Karras: 20~ 30 步
DPM++ 2M Karras:20 ~40 步
提示词引导系数 (CFG Scale): 7
人像人物分辨率(Character resolution): 512x768, 768x768,512x1024
背景图风景图分辨率(Background and landscape resolution): 768x512, 1024x512, 1536x512
高清修复(Hires.fix): R-ESRGAN 4x , 高分迭代步数 15~30 重绘幅度 0.3~ 0.5
CLIP 终止层数(Clip Skip): 2
负面提示词(Negative):
推荐用嵌入式embeddings模型中的 EasyNegative,negative_hand-neg 或者单独使用 (worst quality, low quality:2),monochrome, zombie, (interlocked fingers:1.2)
EasyNegative in the embedded embeddings model is recommended, negative_hand-neg or alone (worst quality, low quality:2),monochrome, zombie, (interlocked fingers:1.2)
本模型中不包含任何真人素材图像源和训练代码来自互联网,模型仅用于科研兴趣交流
This model does not include any real-life material, image sources, and training code from the internet. The model is only used for scientific research interest exchange.