Snoot Challenge, Character approaching and lying onto POV hands - Experiment Concept LoRA

详情

模型描述

非常感谢您的评分、点赞和传播,这让我知道我做得不错,并将继续创作资源!

LoRA权重:建议在0.8至1之间,降低权重可提升对LoRA训练图像之外内容的适应性,使您的基础模型或其他LoRA的风格更明显,但可能降低角色的准确性。

Snoot挑战实验概念LoRA
使用约90+张高质量动漫图像训练,这些图像均体现该概念。

仅使用触发词时,该LoRA可能较难单独使用,需要大量额外提示才能发挥良好效果。我建议参考我下面提供的示例提示,这些提示在我的使用中能产生稳定的结果,之后可利用您的生成结果中的种子,自行调整提示进行实验。

生成结果通常包含2至4个分镜。

触发词
bothhandssnootduck

基础Snoot挑战
bothhandssnootduck, snoot challenge, pov, pov hands, 1girl, comic, looking at viewer, 3koma, heart, out of frame, hands on another's face, heart hands, silent comic, hands on another's cheeks, 1boy, meme, multiple girls, solo focus, holding, emphasis lines, facing viewer, upper body, close-up, collarbone, depth of field

Snoot挑战 aroused girl
bothhandssnootduck, snoot challenge, pov, pov hands, 1girl, comic, looking at viewer, 3koma, heart, out of frame, hands on another's face, heart hands, silent comic, hands on another's cheeks, 1boy, meme, multiple girls, solo focus, holding, emphasis lines, facing viewer, upper body, close-up, collarbone, depth of field, long hair, breast, aroused, seductive smile, heavy breathing, tongue out,

此模型生成的图像

未找到图像。