一键部署中文文生图模型（太乙Stable Diffusion在webui）

新闻 2023-03-22

4315 0

前言

中文意境，运笔而生。

一键部署

本文基于太乙 Taiyi-Stable-Diffusion-1B-Chinese-v0.1：

https://huggingface.co/IDEA-CCNL/Taiyi-Stable-Diffusion-1B-Chinese-v0.1huggingface.co/IDEA-CCNL/Taiyi-Stable-Diffusion-1B-Chinese-v0.1

但是很多同学没有办法用上我们的模型，所以我们做了一个一键部署webui的脚本：

感谢

@稲荷 https://github.com/IDEA-CCNL/stable-diffusion-webuigithub.com/IDEA-CCNL/stable-diffusion-webui

测试环境：Linux服务器。测试环境：Linux服务器。

第一步

因为原生的webui只支持中文，所以我们需要用到封神榜团队做的webui：

git clone https://github.com/IDEA-CCNL/stable-diffusion-webui.gitcd stable-diffusion-webui

第二步

一键配置相关环境

bash webui.sh

这一步会需要一些时间，因为会安装非常多相关的python库等。（下载模型的时候，会需要一些时间，因为我们的模型有7GB）

中间会询问，是不是需要下载：

我选了1，下载。

我的输出是这样的：

################################################################
Install script for stable-diffusion + Web UI
Tested on Debian 11 (Bullseye)
################################################################

################################################################
Running on wangjunjie user
################################################################

################################################################
Repo already cloned, using it as install directory
################################################################

################################################################
Create and activate python venv
################################################################

################################################################
Launching launch.py...
################################################################
Python 3.9.12 (main, Jun 1 2022, 11:38:51)
[GCC 7.5.0]
Commit hash: fa2990b8fbf068c79a5d6a4fbef57e662e2747ef
Installing torch and torchvision
Installing gfpgan
Installing clip
Cloning Stable Diffusion into repositories/stable-diffusion...
Cloning Taming Transformers into repositories/taming-transformers...
Cloning K-diffusion into repositories/k-diffusion...
Cloning CodeFormer into repositories/CodeFormer...
Cloning BLIP into repositories/BLIP...
Installing requirements for CodeFormer
Installing requirements for Web UI
Cloning taiyi_model into repositories/Taiyi-Stable-Diffusion-1B-Chinese-v0.1...
Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple
Obtaining file:///your/path/projects/webui/stable-diffusion-webui
ERROR: file:///your/path/projects/webui/stable-diffusion-webui does not appear to be a Python project: neither setup.py nor pyproject.toml found.
WARNING: You are using pip version 22.0.4; however, version 22.3.1 is available.
You should consider upgrading via the /your/path/projects/webui/stable-diffusion-webui/venv/bin/python3 -m pip install --upgrade pip command.
Launching Web UI with arguments: --ckpt /your/path/projects/webui/stable-diffusion-webui/repositories/Taiyi-Stable-Diffusion-1B-Chinese-v0.1/Taiyi-Stable-Diffusion-1B-Chinese-v0.1.ckpt --listen --port 12345
LatentDiffusion: Running in eps-prediction mode
DiffusionWrapper has 859.52 M params.
making attention of type vanilla with 512 in_channels
Working with z of shape (1, 4, 32, 32) = 4096 dimensions.
making attention of type vanilla with 512 in_channels
Loading weights [e2e75020] from /your/path/projects/webui/stable-diffusion-webui/repositories/Taiyi-Stable-Diffusion-1B-Chinese-v0.1/Taiyi-Stable-Diffusion-1B-Chinese-v0.1.ckpt
Applying cross attention optimization (Doggettx).
Model loaded.
Loaded a total of 0 textual inversion embeddings.
Embeddings:
Traceback (most recent call last):
File "/your/path/projects/webui/stable-diffusion-webui/venv/lib/python3.9/site-packages/gradio/networking.py", line 120, in start_server
s.bind((LOCALHOST_NAME, server_port))
OSError: [Errno 98] Address already in use

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/your/path/projects/webui/stable-diffusion-webui/launch.py", line 280, in <module>
start()
File "/your/path/projects/webui/stable-diffusion-webui/launch.py", line 274, in start
webui.webui()
File "/your/path/projects/webui/stable-diffusion-webui/webui.py", line 151, in webui
app, local_url, share_url = demo.launch(
File "/your/path/projects/webui/stable-diffusion-webui/venv/lib/python3.9/site-packages/gradio/blocks.py", line 1255, in launch
server_name, server_port, local_url, app, server = networking.start_server(
File "/your/path/projects/webui/stable-diffusion-webui/venv/lib/python3.9/site-packages/gradio/networking.py", line 123, in start_server
raise OSError(
OSError: Port 12345 is in use. If a gradio.Blocks is running on the port, you can close() it or gradio.close_all().

我这属于是端口12345给占用了。。。

第三步

运行！

我把端口改到了23333。

bash webui.sh --ckpt repositories/Taiyi-Stable-Diffusion-1B-Chinese-v0.1/Taiyi-Stable-Diffusion-1B-Chinese-v0.1.ckpt --listen --port23333

这样，你的webui就会开放在服务器的 23333 端口了。我通过vscode做了一个端口转发出来。

打开之后：

中文界面

找到settings

找到User interface，下面有一个Localization (requires restart)

把选项切换到zh_CN

划到最上面点击apply settings

再划到最下面，点击

Restart Gradio and Refresh components (Custom Scripts,ui.py, js and css only)

这时候界面会重新加载，重新加载完了以后就汉化好了（要是宕机了，提示你端口被占用，你就再启动换一个端口就行了。当然，你也可以直接修改config.json）