Installation Guide

Installation Guide#

This document provides step-by-step instructions for setting up the audio2face development environment on different operating systems.

Table of Contents#

Linux Environment Setup#

Linux Prerequisites#

Before starting, ensure you have the following system requirements:

  • Ubuntu 20.04 or compatible Linux distribution

  • Internet connection for downloading packages

Linux Step 1: Install Protocol Buffers Compiler#

Download and install protoc for protocol buffer compilation:

# Create protoc directory
mkdir -p protoc
cd protoc

# Download protoc
curl -LjO https://github.com/protocolbuffers/protobuf/releases/download/v31.1/protoc-31.1-linux-x86_64.zip

# Extract and set permissions
unzip protoc-31.1-linux-x86_64.zip
rm -f protoc-31.1-linux-x86_64.zip
chmod +x bin/protoc

# Verify installation
bin/protoc --version

# Go back to the root directory
cd ..

Linux Step 2: Set Up Python#

You need Python 3.10 or higher to run this project. This document provides one method using conda for Python installation as a reference.

Install Python using Miniconda:

# Download Miniconda installer
wget -q https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh

# Install Miniconda
bash Miniconda3-latest-Linux-x86_64.sh

# Clean up installer
rm -f Miniconda3-latest-Linux-x86_64.sh

# Configure conda channels
conda config --add channels conda-forge
conda tos accept

# Create audio2face environment with Python 3.10
conda create -n audio2face python=3.10 -y

# Activate the environment
conda activate audio2face

Linux Step 3: Install PyTorch and ONNX Runtime#

The audio2face service supports both CPU and GPU inference. Choose the appropriate installation method based on your hardware configuration.

Option A: CPU Inference Environment (Default)#

For CPU-only inference, install PyTorch and ONNX Runtime with CPU support:

# Activate the environment
conda activate audio2face

# Install PyTorch with CPU support
conda install pytorch==2.4.1 torchaudio==2.4.1 cpuonly -c pytorch

# Install ONNX Runtime with CPU support
pip install onnxruntime==1.22.0

Linux Step 4: Install the Project#

Install the audio2face package:

# Ensure you're in the project root directory
cd /path/to/audio2face

# Activate conda environment
conda activate audio2face

# Install the package
pip install .

Linux Step 5: Verify Installation#

Test that everything is working correctly:

# Activate the environment
conda activate audio2face

# Check if audio2face.apis can be imported
python -c "import audio2face.apis; print('audio2face.apis imported successfully')"

# Check if the main application runs
python main.py --help

Linux Environment Activation#

To work with the audio2face project, always activate the conda environment first:

# Activate the environment
conda activate audio2face

# Your terminal prompt should now show (audio2face)
# You can now run Python scripts and use the audio2face package

Windows Environment Setup#

Windows Prerequisites#

Before starting, ensure you have the following system requirements:

  • Windows 10/11 or compatible Windows distribution

  • Internet connection for downloading packages

Windows Step 1: Install Protocol Buffers Compiler#

Download and install protoc for protocol buffer compilation:

  1. Download protoc:

  2. Extract the files:

    • Create a protoc folder in your project root directory

    • Extract the downloaded protoc-31.1-win64.zip file into the protoc folder

    • Ensure the executable file is located at: protoc\bin\protoc.exe

  3. Verify installation:

    # Open Command Prompt in your project directory
protoc\bin\protoc.exe --version

Windows Step 2: Set Up Python#

You need Python 3.10 or higher to run this project. This document provides one method using conda for Python installation as a reference.

Install Python using Miniconda:

  1. Download and Install Miniconda:

    • Visit Miniconda Installation Guide

    • Download the Windows installer from the Anaconda website

    • Follow the official installation instructions to install Miniconda

    • Important: During installation, make sure to check “Add Miniconda3 to my PATH environment variable” or add the Miniconda3/Scripts directory to the PATH environment variable manually to enable conda commands from any terminal

  2. Create and Activate Environment:

    # Create audio2face environment with Python 3.10
conda create -n audio2face python=3.10 -y

# Activate the environment
conda activate audio2face

Windows Step 3: Install PyTorch and ONNX Runtime#

Install PyTorch and ONNX Runtime with CPU support for Windows:

# Activate the environment
conda activate audio2face

# Install PyTorch with CPU support
conda install pytorch==2.4.1 torchaudio==2.4.1 cpuonly -c pytorch

# Install ONNX Runtime with CPU support
pip install onnxruntime==1.22.0

Windows Step 4: Install the Project#

Install the audio2face package:

# Ensure you're in the project root directory
cd /path/to/audio2face

# Activate conda environment
conda activate audio2face

# Temporarily add protoc to PATH for this session
set PATH=%PATH%;%CD%\protoc\bin

# Install the package
pip install .

Windows Step 5: Verify Installation#

Test that everything is working correctly:

# Activate the environment
conda activate audio2face

# Check if audio2face.apis can be imported
python -c "import audio2face.apis; print('audio2face.apis imported successfully')"

# Check if the main application runs
python main.py --help

Windows Environment Activation#

To work with the audio2face project, always activate the conda environment first:

# Activate the environment
conda activate audio2face

# Your terminal prompt should now show (audio2face)
# You can now run Python scripts and use the audio2face package
Contents