Install Automatic1111 - Apkawa/stable-diffusion-wiki-awesome GitHub Wiki

Check compatibly version

  1. a1111 changelog grep torch https://github.com/AUTOMATIC1111/stable-diffusion-webui/releases/
  2. pytorch version https://github.com/pytorch/pytorch/blob/main/RELEASE.md#release-compatibility-matrix
  3. check cuda and nvidia driver compatibly https://docs.nvidia.com/deploy/cuda-compatibility/

For 3060 i choosed:

  • a1111==1.8.0
  • pytorch==2.1.2
  • CUDA==11.8
  • nvidia==550

Install driver

  • If you have previous installation remove it first.
sudo apt-get purge nvidia*
sudo apt remove nvidia-*
sudo rm /etc/apt/sources.list.d/cuda*
sudo apt-get autoremove && sudo apt-get autoclean
sudo rm -rf /usr/local/cuda*
  • system update
sudo apt-get update
sudo apt-get upgrade

  • install other import packages sudo apt-get install g++ freeglut3-dev build-essential libx11-dev libxmu-dev libxi-dev libglu1-mesa libglu1-mesa-dev

  • first get the PPA repository driver

sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt update
  • install nvidia driver with dependencies
V=550 sudo apt install libnvidia-common-$V libnvidia-gl-$V nvidia-driver-$V

Install CUDA

Add cuda repository

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-ubuntu2204.pin
sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600
sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/3bf863cc.pub
sudo add-apt-repository "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/ /"
sudo apt-get update

Install

export CUDA_VERSION=11.8
sudo apt install cuda-$(echo $CUDA_VERSION|sed 's/\./-/g')

echo "export PATH=/usr/local/cuda-$CUDA_VERSION/bin:\$PATH" >> ~/.bashrc
echo "export LD_LIBRARY_PATH=/usr/local/cuda-$CUDA_VERSION/lib64:\$LD_LIBRARY_PATH" >> ~/.bashrc
source ~/.bashrc
sudo ldconfig

Install cuDNN

Get latest version tarball https://developer.nvidia.com/cudnn-downloads

Or get specific version from https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/linux-x86_64/

I choose 8.9.7.29_cuda11

wget -c https://developer.download.nvidia.com/compute/cudnn/redist/cudnn/linux-x86_64/cudnn-linux-x86_64-8.9.7.29_cuda11-archive.tar.xz -O cudnn.tar.xz 
tar -xvf cudnn.tar.xz
cd cudnn-*/
sudo cp -P include/cudnn.h /usr/local/cuda-11.8/include
sudo cp -P lib/libcudnn* /usr/local/cuda-11.8/lib64/
sudo chmod a+r /usr/local/cuda-11.8/lib64/libcudnn*
  • Check cuda
nvidia-smi
nvcc -V

https://gist.github.com/Apkawa/2e1cbeced22135b2a8bad9d230f4e70c

Install A1111

Pull source code

  1. git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui
  2. cd stable-diffusion-webui
  3. git checkout v1.8.0

Prepare venv

  • Init venv
    python3 -m venv ./venv

  • Enter venv source ./venv/bin/activate

  • install 'pip install accelerate`

Install pytorch and xformers

for a1111==1.8.0 support pytorch version 2.1.2

pip install torch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2 xformers --index-url https://download.pytorch.org/whl/cu118
  • check installed xformers version pip show xformers and get version
  • enable xformers, add to webui-user.sh after COMMANDLINE_ARGS=
XFORMERS_PACKAGE='xformers==0.0.23.post1+cu118'
export  COMMANDLINE_ARGS="$COMMANDLINE_ARGS  --xformers"

FAQ:

Q: nvidia-smi NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver A: Try reboot. Also, for hybrid video check https://wiki.debian.org/NVIDIA%20Optimus or https://wiki.ubuntu.com/Bumblebee manuals

Q: RuntimeError: "LayerNormKernelImpl" not implemented for 'Half'
A: Edit webui-user.sh, uncomment and add to COMMANDLINE_ARGS, like: export COMMANDLINE_ARGS="--precision full --no-half"

Q: AssertionError: Torch is not able to use GPU; add --skip-torch-cuda-test to COMMANDLINE_ARGS variable to disable this check
A: export COMMANDLINE_ARGS="--skip-torch-cuda-test"

Q: Fail for 6Gb Nvidia videocard
A: export COMMANDLINE_ARGS="--medvram"

Q: How to run as server
A: export COMMANDLINE_ARGS="--listen --enable-insecure-extension-access"

Q: Huge memory leaks
A: Install tcmalloc. for ubuntu - apt install libtcmalloc-minimal4. See https://github.com/AUTOMATIC1111/stable-diffusion-webui/discussions/6722

Q: ./webui.sh: line 292: 18311 Segmentation fault (core dumped) after load model
A: Check a1111 supported torch version, pin torch version https://pytorch.org/get-started/previous-versions/

https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Command-Line-Arguments-and-Settings