<h1 id="heading-113">VivCharMix</h1>Recommended DPM++ 2M Karras Hires Nearest 0.5 negative Prompts: (low quality,EasyNegative:1.2)No need for VAEAdvantages:(1) Generates good-looking faces, suitable for turning ACG characters into real-life people, preserving their looks, charm, and clothing details(2) Highly adaptive and tolerant to ACG Loras (recommended weight: 0.3~0.8, adjust in 0.05 intervals)(3) Photorealistic ( Vivid, full colors and dynamic expressions in previous version

ViVCharMix_V5 New Era

v2.9

v4.0 recommed neg words: (low quality, worst quality, lowres, grayscale, normal quality, username, outline, poorly drawn, copyright name, character doll, scribble , silent comic, doujinshi,:1.4) Outstanding ACG female figures, backgrounds, and actions are vivid and vibrant. High tolerant. The downside is that there is a problem with the feminization of male characters.

v4.0

更好的細節呈現 Better detail presentation

v3.0

Another interesting version, deeply integrated with an excellent model - kencanmix, is merged using the mbw-R-consine method (don't ask me why; it's just testing and selection). It has outstanding light and shadow effects and foreground sharpening performance, but the body shape is not stable at different angles (within an acceptable range). I think it's better and more interesting than before. 另一個有趣的版本，和一個出色的model-kencanmix深度融合，使用mbw-R-consine方式融合(別問我為什麼，就是測試和挑選)。 擁有更出色的光影和前景銳化表現，但是不同角度體型不穩定(在可接受範圍)。覺得比之前好，有趣。

v2.1R

很好的光影感、背景，調整色調，身材 更真實，所以可以有更高的容忍值增加其他LORAs 不像V1如此鮮豔，所以V1更適合展示 所以我給予V1命名-Aesthetics、V2-Appealing Great sense of light and shadow, background, color adjustment, and body shape. It's more realistic, so it can have a higher tolerance for adding other LORAs. Unlike V1, which is so vibrant, V1 is more suitable for display. Therefore, I named V1 - Aesthetics and V2 - Appealing.

v2.0

v1.0 puned

如果沒有辦法改進場景問題，這個就是完成版If there is no way to improve the scene issue, this is the final version.

Type	Checkpoint Merge
Downloads	2,584
Uploaded	Jun 13, 2023
Base Model	Other
Hash	AutoV2 A3D7311D85