如果是无显卡服务器,只能用 CPU,速度极慢,不推荐。
sudo dnf update -y
sudo dnf install -y epel-release
sudo dnf groupinstall -y "Development Tools"
sudo dnf install -y \
git \
wget \
python3 \
python3-devel \
python3-pip \
libglvnd-glx \
mesa-libGLsudo dnf config-manager --add-repo http://developer.download.nvidia.com/compute/cuda/repos/rhel8/x86_64/cuda-rhel8.repo
sudo dnf install -y nvidia-driver nvidia-driver-cuda重启:
reboot验证:
nvidia-smisudo dnf install -y cuda-toolkitpip3 install virtualenv
mkdir ~/sd && cd ~/sd
python3 -m venv venv
source venv/bin/activategit clone https://github.com/AUTOMATIC1111/stable-diffusion-webui.git
cd stable-diffusion-webuipip install --upgrade pip
pip install torch torchvision --index-url https://download.pytorch.org/whl/cu118
pip install -r requirements.txtCUDA 版本根据你驱动调整(如 cu117 / cu121)
把模型放到:
models/Stable-diffusion/常用模型:
示例:
wget -O models/Stable-diffusion/v1-5.ckpt \
https://huggingface.co/runwayml/stable-diffusion-v1-5/resolve/main/v1-5-pruned-emaonly.ckptpython launch.py --listen --port 7860访问:
http://服务器IP:7860sudo dnf install -y mesa-libGLpython -c "import torch; print(torch.cuda.is_available())"启动参数:
--medvram 或 --lowvramnohup python launch.py --listen > sd.log 2>&1 &如果你告诉我:
我可以给你更精简或定制化的安装方案(包括 CPU 版、Docker 版)。