1. Install homebrew if it is not installed on your Mac yet. Open terminal and execute the following command:
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
2. Install pyenv (a plugin for creating virtual environments) and install python 3.7:
brew update
brew install pyenv
pyenv install 3.7
3. Go to your repo directory and execute the following command, which tells which python version to use:
cd MY_REPO_PATH
pyenv local 3.7
4. Install wget:
brew install wget
5. In your repo folder make init.sh script executable and run it. The shell script will clone SimSwap repository
into your repository folder and will install the required models. It will require some time, please wait until
script execution will be completed.
chmod +x init.sh
./init.sh
6. Install Powershell and apply a CPU patch on SimSwap project. Wait till execution finishes.
brew install --cask powershell
cd SimSwap
pwsh ../patch_cpu.ps1
7. Verify that the patch has been applied correctly. Go to SimSwap/models/fs_model.py, find line 51 and
check that it equals the following "device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')".
8. Install onnxruntime:
brew install onnxruntime
9. Install requirements:
cd MY_REPO_PATH
pipenv install -r requirements/mac_x86_64.txt
10. Add SimSwap directory as a source root directory (change PATH_TO_SIMSWAP_FOLDER with your actual path
to SimSwap folder):
export PYTHONPATH="${PYTHONPATH}:/PATH_TO_SIMSWAP_FOLDER
pipenv install -r requirements.txt
11. Activate your virtual environment:
pipenv shell
12. Run main.py:
python3 main.py
Follow steps described in API usage.
1. Execute steps 1, 4-8 from Mac Intel x86-64 manual.
2. Install rosetta:
sudo softwareupdate --install-rosetta
3. Install ffmpeg:
brew install ffmpeg
4. Download Anaconda from https://repo.anaconda.com/archive/Anaconda3-2023.03-MacOSX-x86_64.pkg and install
it by double clicking on the pkg file. Follow the installation steps.
5. Restart the terminal window.
6. Go to the repository folder, create and activate Conda environment by executing the following commands:
conda create -n deepfake python=3.9
conda activate deepfake
7. Install pytorch, protobuf and numpy:
conda install pytorch torchvision -c pytorch-nightly
conda install -c conda-forge protobuf numpy
8. Add SimSwap directory as a source root directory to your conda path (change PATH_TO_SIMSWAP_FOLDER with
your actual path to SimSwap folder, for example "/Users/illania/deep_fake/SimSwap"):
conda develop PATH_TO_SIMSWAP_FOLDER
9. Install requirements:
pip install -r requirements/mac_arm64.txt
10. Run main.py:
python main.py
11. Configure Anaconda in PyCharm.
11.1. Open your repo folder in PyCharm.
11.2. Close the window, that asks you about which interpreter you want to use.
11.3. Click on "No interpreter" in the bottom right corner of PyCharm window.
11.4. Select "Add new interpreter" -> "Add local interpreter".
11.5. In the "Add Python interpreter window" select "Conda environment", then "Use existing environment",
then select "deepfake" from the dropdown list. It is the name of Conda environment, that you have
created on step 6.
11.6. Now right-click on main.py and select "Run 'main'".
Follow steps described in API usage.