2.0版本的风格更加扁平且统一,在细节处也减少了崩掉的概率

模型简介:

适用于各种卡通风格的插画生成需求。无论是简约风格还是复杂的场景设计,该模型都能保持出色的表现,生成具有高辨识度的卡通图像。通过对几何形状、鲜艳色彩和独特线条的处理,模型在表现动感、时尚或富有创意的内容时尤为突出。

——以下例图均为FLUX基础模型+此lora直出

触发词:of a vibrant digital illustration

权重0.8左右

提示词建议:Flux对文字的理解能力很强,建议提示词以小作文形式去写,可以借助市面上的AI语言模型去对自己的提示词扩充润色再丢进去,

封面提示词:

of a vibrant digital illustration,The picture is a childlike poster,the main color is
black,looks cool personality. In the middle of the image is a headshot
of a man with orange hair and beard,pink skin,some small cuts on his
forehead and chin,Band-Aids,and headphones in his ears,giving a sense of
art and design. In this person's brain,there are all kinds of emojis
and patterns,there are big orange flowers,pink rabbits,blue clouds,green
unidentified creatures,yellow sun,hearts and stars. These elements give
life,some laugh,some cry,lively and funny. Several emojis also flew out
of the head,as if born from the brain. Overall,the design of this
poster is very creative. Thoughts in the human brain are reflected in a
variety of emojis and patterns that are impressive

生图尺寸:无固定尺寸,创作者根据自己的需求设置即可,但尺寸越大效果相对来说就越清晰细节处也不那么容易产生崩坏,如果崩坏请降低权重或多抽卡

CFG:在线webui建议保持默认3.5,comfyUI建议设置为1

高清修复设置:若出现细节崩坏,则需要开启高分辨率修复,修复方式4x_NMKD-Siax_200k,步数20,重绘幅度0.3~0.5(崩坏的越严重,幅度越大)

主要特征:

模型擅长生成具有强烈动作感和节奏感的图像,尤其适用于表现动态场景和活跃的主题,具备处理明亮饱和色彩的能力,使生成的图像在视觉上更具冲击力,但也可以根据需求调整色彩风格以适应不同项目,适合广泛应用于各类卡通风格的创意项目,包括插画、品牌设计、社交媒体素材、广告宣传等领域。