模型简介: 众所周知,SDXL原本的大模型对动漫风格人物的绘画效果不佳,不够稳定。此模型不同于其他训练大模型方法,应用模型相减,做出了既不污染原有大模型,又对动漫画风具有高亲和力的模型。填补了当前SDXL在动漫绘画领域的空白。 基于SDXL1.0训练的多种动漫风格模型。补足了原本SDXL对二次元绘画的短板。 既可以作为原始底模进行再次微调,也可以直出多种画风的图片。 v1.2提示词参考: (masterpiece), (best quality), (super fine illustration), (beautiful and detailed eyes), {very detailed light}, {painting}, {{very fine 8K CG wallpaper}}, 1girl, solo, magic circle, crystalline splashes, flying feathers, bare shoulders, smile, onsen, steam, hair flower, gem earrings (masterpiece), (best quality), (super fine illustration), (beautiful and detailed eyes), {very detailed light}, {painting}, {{very fine 8K CG wallpaper}}, 1girl, solo, lavender eyes, lavender hair, sunset, breeze, river, beautiful and delicated, wading, white shirt, wet shirt, Neg:worst, bad, worse, negativeXL_D, lips 推荐参数: DPM++ 2M Karras 20步 CFG=5或7 分辨率1024x1344!!! 开启After Detailer插件,修脸模型就行都差不多l。 去除了一部分2.5D,人物风格更接近于动漫了。 VAE:SDXL_vae_fix16 如果喜欢2.5D风格可以暂时用1.0版本。 参考提示词有更改,见示例图或下行: (masterpiece,best quality,ultra_detailed,highres,absurdres:2), (watercolor, anime screencap:1.1), 1girl, white hair, beautiful detailed eyes, lily, wedding dress, earrings, smile, looking at viewer, (worst quality, low quality, ugly:2.4), poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, mutation, mutated, extra limbs, extra legs, extra arms, disfigured, deformed, cross-eye, blurry, (bad art, bad anatomy:1.4), blurred, text, watermark, (3d:1.2), 别忘了启用AD插件! 如果有提示词不听话,画质不佳等问题,请在评论区留言,感激不尽! |