Installation on Apple Silicon (manual)

Homebrew

First, you need to install the required dependencies using Homebrew.

brew install cmake protobuf rust python git wget

Setting up Conda

Disclaimer: I wrote this guide with Conda due to environment variables saving. VENV will work for other installations but modify your environment accordingly.

# Download the conda installation file
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-arm64.sh

# Run the conda installation
`bash Miniconda3-latest-MacOSX-arm64.sh -b -p $HOME/miniconda`

# To make sure that conda is available in the $PATH.
export PATH="$HOME/miniconda/bin:$PATH"

# Initialize Conda
conda init

# (OPTIONAL) Disable conda automatic activation, since we don't need conda running all the time
conda config --set auto_activate_base false

# Create the conda environment 
conda create -n web-ui python=3.9

# Activate the conda environment
conda activate web-ui

Setting up the repository

# Clone the repository
git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui.git

# Open the cloned repository
cd stable-diffusion-webui

# Install required base dependencies
pip install -r requirements.txt

# Missing dependency
pip install gdown 

## 
## Additional work around below is required at the moment of writing.
##

# There's a bug in protobuf that causes errors when generating images.
# Read: https://github.com/protocolbuffers/protobuf/issues/10571
# Once this gets fixed, we can remove this line
pip install protobuf==3.19.4

# Remove torch and all related packages
pip uninstall torch torchvision torchaudio -y

# Install the latest nightly build of PyTorch.
# Normally, we would install the latest nightly build of PyTorch here,
# But there's currently a performance regression in the latest nightly releases.
# Therefore, we're going to use this old version which doesn't have it.
#pip install --pre torch torchvision -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html --no-deps
pip install --pre torch==1.13.0.dev20220922 torchvision -f https://download.pytorch.org/whl/nightly/cpu/torch_nightly.html --no-deps

# Activate the MPS_FALLBACK conda environment variable
conda env config vars set PYTORCH_ENABLE_MPS_FALLBACK=1

# Activate the COMMANDLINE_ARGS conda environment variable
conda env config vars set COMMANDLINE_ARGS="--skip-torch-cuda-test"

# We close down conda to activate the environment variables
conda deactivate

Setting up models

Place your downloaded model('s) in the following folder:
stable-diffusion-webui/models/Stable-diffusion/
No need to rename it or do anything else with them.

Running web-ui

# Activate conda if not active already
conda activate web-ui

# Startup the launcher
python launch.py --precision full --no-half --opt-split-attention-v1

Updating repository

# Pull the latest changes from the repo
git pull --rebase

Kinsmir/stable-diffusion-webui-m1-guide.md