NovelAI

GitHub - AUTOMATIC1111/stable-diffusion-webui: Stable Diffusion web UI

Required Dependencies

GitHub - AUTOMATIC1111/stable-diffusion-webui: Stable Diffusion web UI

Python 3.10.6 and Git

miniconda 换源（添加镜像）_哈！小白要成长！的博客-CSDN博客_miniconda换源

Previous PyTorch Versions | PyTorch

Anaconda的虚拟环境的包在哪里？（详细教程）_仭的博客-CSDN博客_anaconda创建的虚拟环境存储在哪

Index of /anaconda/cloud/pytorch/win-64/ | 清华大学开源软件镜像站 | Tsinghua Open Source Mirror

Git (git-scm.com)

# 新建虚拟环境，python版本是3.10.6
conda create -n novelai python==3.10.6

# 激活虚拟环境
conda activate novelai

# 查看CUDA信息
nvcc -V
nvidia-smi

# 查看conda信息，改一下conda源，下载是时候不至于掉线。
conda info

# 下载对于cuda版本的PyTorch，不要下载CPU版本的，而是GPU版本的
conda install pytorch==1.12.1 torchvision==0.13.1 torchaudio==0.12.1 cudatoolkit=11.3 -c https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/win-64/

PyTorch VERIFICATION

Start Locally | PyTorch

To ensure that PyTorch was installed correctly, we can verify the installation by running sample PyTorch code. Here we will construct a randomly initialized tensor.

From the command line, type:

python

then enter the following code:

import torch
x = torch.rand(5, 3)
print(x)

The output should be something similar to:

tensor([[0.3380, 0.3845, 0.3217],
        [0.8337, 0.9050, 0.2650],
        [0.2979, 0.7141, 0.9069],
        [0.1449, 0.1132, 0.1375],
        [0.4675, 0.3947, 0.1426]])

Additionally, to check if your GPU driver and CUDA is enabled and accessible by PyTorch, run the following commands to return whether or not the CUDA driver is enabled:

import torch
torch.cuda.is_available()

Code from this repository

# Clone 仓库到本地
git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui.git

# 在仓库根目录下，安装依赖库
python -m pip install -r .\requirements.txt

The Stable Diffusion model checkpoint

The Stable Diffusion model checkpoint, a file with .ckpt extension, needs to be downloaded and placed in the models/Stable-diffusion directory.

AI作画保姆级教程来了！逆天，太强了！_哔哩哔哩_bilibili

权重文件下载地址（提取码: o89H ）：链接

显卡显存不足，生成图是全黑的，使用下面的命令来解决。

python .\launch.py --lowvram --precision full --no-half