Tensor.Art
在线生图

HighContrast 高对比平涂

LOCON
368

高对比度平涂画风lora 基于Animagine XL V3训练应该在绝大部分人物为主导的纯色背景中发挥不错推荐权重为0.9-0.7
提示词:high contrast 随后请使用simple background/two-tone background + 具体颜色的背景例如:high contrast,simple background,red background,1girl, solo, short hair, one eye closed, red background, black hair, brown eyes, looking at viewer, smile,

题外话:感觉训练中的sample image不能太过相信,这炉就是生成的sample image跟异形一样,但是实际用起来效果意外不错QAQ



The high-contrast, flat-color painting style Lora model, trained on Animagine XL V3, should perform well in the majority of scenarios with characters against solid-color backgrounds.R ecommended weight between 0.9 and 0.7.
The tag is "high contrast" followed by specifics like a simple background/two-tone background with certain colors. For example: "high contrast,simple background,red background,1girl, solo, short hair, one eye closed, red background, black hair, brown eyes, looking at viewer, smile,"

Side note: One should not rely too much on the sample images from training. In this case, the generated sample images resembled aliens, but the actual results turned out to be surprisingly good.

版本详情

SDXL 1.0

项目权限

    使用权限

  • 使用时无需注明出处

  • 用于模型融合

  • 分享融合模型时使用不同的许可

    商用许可

  • 生成的图片用于商业用途

  • 作为图片生成服务来商用

  • 转售模型或出售融合模型

评论

相关帖子

描述想生成的画面,回车发送