All Images Here Are Generated By AI.

All Models

>

Royal dragon beauty 御龙 SDXL1.0 Train

Royal dragon beauty 御龙 SDXL1.0 Train

Generate AI Images. Simple and Easy

AIEasyPic is the #1 AI image generator. Generate images with simple prompts, swap faces, train models to take AI photos of yourself, and more.

Model Description by Creator

SDXL1.0正式版训练,训练分辨率为1024*1024,原大小1.7g压缩为17m,使用压缩方法sv_fro,注意当前lora仅支持SDXL

AI净土交流群电报:https://t.me/+soU-Ws67aBlkY2I1

QQ群:895633778

WebUI推荐参数:

vae:sdxl-vae

采样:Euler a

步数:20-30

高分辨率修复 (Hires. fix):高清修复一定要开!!不一定开!

高清迭代步数:15-20

放大算法:R-ESRGAN 4x+

重绘幅度:0.25-0.4

放大倍数:2或者2以上

cfg:7(可自由尝试)

压缩算法简介:

sv_fro方法是一种对矩阵进行深度压缩的方法,主要用于减少矩阵的存储空间和计算复杂度。

sv_fro方法的基本思想是利用奇异值分解(Singular Value Decomposition,SVD)将矩阵分解为三个矩阵的乘积:A = U S V^T,其中U和V是正交矩阵,S是奇异值矩阵。奇异值矩阵S是一个对角矩阵,其对角线上的元素称为奇异值。

在sv_fro方法中,只保留矩阵A中奇异值较大的部分,将较小的奇异值设为0,从而实现对矩阵的深度压缩。具体步骤如下:

1. 对矩阵A进行奇异值分解,得到U、S和V。

2. 根据预设的阈值,将S中小于阈值的奇异值设为0。

3. 使用压缩后的奇异值矩阵S以及U和V重构原始矩阵A。

通过sv_fro方法进行深度压缩后,可大幅减少矩阵的存储空间,同时还能在一定程度上降低计算复杂度。这对于需要处理大规模矩阵的任务来说,具有重要的实际应用意义。

Try Full Demo
Quick Generate
Image quality might not be as good as what you can achieve with our full generation tool.
Placeholder
Choose A Style, Type What You Want In The Image And Click Generate
Placeholder
Choose A Style, Type What You Want In The Image And Click Generate
What you want in the image.
Clear
Random
* This model is available for use and will generate similar character to the ones seen in the images.
Train Your Own Model

Images Generated With This Model

Loading...

Generate AI Images. Simple and Easy

AIEasyPic is the #1 AI image generator. Generate images with simple prompts, swap faces, train models to reimagine yourself, and more.