Flux美学风格Lora,效果直逼Midjourney

9/21/2024, 5:02:50 PM | 热度:78 | 阅读本文需 6 分钟

点赞

众所周知,Flux 的美学效果是不如Midjourney的,不过我们使用一些Lora后是可以达到Midjourney的美学效果的,所以今天推荐的就是: 美学_Art 艺术风格 Lora此Lora模型可生成美学艺术化风格的图片,主要增强画风和艺术效果。

模型下载地址

LiblibAI:https://www.liblib.art/modelinfo/b1ebc20bf8e3418397e1d79b14c425d8

提示词:D_Hill,art

参数推荐:

  • LORA权重:0.85-1.2

  • 采样方法:Euler

  • 调度器:simple

  • 迭代步数:10-20

工作流(文末网盘地址也可获取)

工作流就是普普通通的Flux文生图工作流,我们只需要在模型和采样器之间串联一个Lora加载器即可

image.png

Lora效果

1. 美女

Prompt: A beautiful Chinese woman in white stands behind green curtains, with her eyes closed and head tilted to the side. The yellow light shines on half of his body from above, creating an ethereal effect. She has long black hair tied back, a minimalist portrait style, and solid color background, rendered by Cinema4D, with fluorescent colors, matte texture, soft lighting, and a minimalist environment. 

提示词:一位身穿白衣的美丽中国女子站在绿色窗帘后面,闭着眼睛,头歪向一边。黄色的光线从上方照在他的半边身体上,营造出一种空灵的效果。她有长长的黑发扎在后面,极简主义的肖像风格,纯色背景,由Cinema4D渲染,具有荧光色,哑光纹理,柔和的灯光和极简主义的环境。

image.png

2. 佛

Prompt: Buddha, vector illustration in the style of James Turrell, Hsiao Ron Cheng, Alena Aenami, Yoji Shinkawa, and Wona Lifsches, with colorful gradients, minimalism, pastel colors, neon light background, creating a mesmerizing halo effect.    Hyperrealistic 3D rendering with photorealistic textures and lighting.

提示词:佛,以James Turrell, Hsiao Ron Cheng, Alena Aenami, Yoji Shinkawa和Wona Lifsches风格的矢量插图,彩色渐变,极简主义,柔和的颜色,霓虹灯的背景,创造一个迷人的光环效果。超现实的3D渲染与逼真的纹理和照明。

image.png

 3. 帐篷

Prompt: The camping tent is located on the grass in front of the rapeseed field, surrounded by camping supplies and flowers, with a lovely cartoon design style and dreamy visual effects. The soft sculpture, network camera, bright colors, bold shapes, and flower field landscape capture the moment

提示词:露营帐篷位于油菜地前的草地上,周围环绕着露营用品和鲜花,具有可爱的卡通设计风格和梦幻般的视觉效果。柔和的雕塑,网络摄像机,明亮的色彩,大胆的造型,花田景观捕捉瞬间

image.png

4. 小狗插图

Prompt:The image is a digital illustration of a white dog's head and neck. The dog appears to be a Labrador Retriever, with a white coat and a pink nose. The background is a solid blue color, and the dog's face is the only source of light in the image. The illustration is done in a realistic style, with bold lines and shading to give it a lifelike appearance. The focus of the image is on the dog, with its head tilted slightly to the side and its eyes looking upwards.

提示词:这张图片是一只白色狗的头部和颈部的数字插图。这只狗看起来是一只拉布拉多猎犬,有着白色的皮毛和粉红色的鼻子。背景是纯蓝色,狗的脸是图像中唯一的光源。插图是在一个现实主义的风格,大胆的线条和阴影给它一个逼真的外观。图片的焦点在狗身上,它的头微微向侧面倾斜,眼睛向上看。

image.png

总结

看样子模型作者应该也算采用了大量的Midjourney美学风格的图片进行训练,出图很有Midjourney的感觉,感兴趣的可以去试试哦


0

继续阅读本文相关话题

本文收录于专题

ComfyUI教程

共收录篇

查看本专题

文章目录

模型下载地址

工作流(文末网盘地址也可获取)

Lora效果

总结

AI教程交流群

扫描二维码加入群聊

与同行学习交流&资源共通