[환경구축] StyleGAN2 | Setting up StyleGAN2 TensorFlow in GoogleColab

우여곡절 많았던 StyleGAN2 TensorFlow 모델 코랩 환경에서의 세팅 정리...
StyleGAN2 ADA TensorFlow 모델도 동일하게 환경 세팅해주면 된다.

https://github.com/NVlabs/stylegan2-ada

Requirements

• Python 3.6.9
• Colab 환경이니까 Anaconda는 따로 X
• TensorFlow 1.15 is also supported on Linux
• NVIDIA drivers, CUDA 10.0 toolkit

이 정도로 준비하면 되겠다

Python Version Downgrade

%python --version

• tensorflow 1.x version 사용을 위해 python 버전을 3.6.x로 다운그레이드해줘야함

!wget https://www.python.org/ftp/python/3.6.9/Python-3.6.9.tgz
!tar xvfz Python-3.6.9.tgz
!Python-3.6.9/configure
!make
!sudo make install

• python install

Cuda Install

!apt-get --purge remove cuda nvidia* libnvidia-*
!dpkg -l | grep cuda- | awk '{print $2}' | xargs -n1 dpkg --purge
!apt-get remove cuda-*
!apt autoremove
!apt-get update

• colab에 default로 설치되어있는 cuda를 uninstall해준다

!wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/cuda-repo-ubuntu1804_10.0.130-1_amd64.deb
!sudo dpkg -i cuda-repo-ubuntu1804_10.0.130-1_amd64.deb
!sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/7fa2af80.pub
!sudo apt-get update
!wget http://developer.download.nvidia.com/compute/machine-learning/repos/ubuntu1804/x86_64/nvidia-machine-learning-repo-ubuntu1804_1.0.0-1_amd64.deb
!sudo apt install -y ./nvidia-machine-learning-repo-ubuntu1804_1.0.0-1_amd64.deb
!sudo apt-get update

• Ubuntu 18.04에서 CUDA 10.0 install

NVIDIA Driver Intsall

# Install NVIDIA driver
#!sudo apt-get install --no-install-recommends nvidia-driver-418
!sudo apt-get -y installnvidia-driver-418
# Reboot. Check that GPUs are visible using the command: nvidia-smi

# Install development and runtime libraries (~4GB)
#!sudo apt-get install --no-install-recommends \
!sudo apt-get install -y \
    cuda-10-0 \
    libcudnn7=7.6.2.24-1+cuda10.0  \
    libcudnn7-dev=7.6.2.24-1+cuda10.0


# Install TensorRT. Requires that libcudnn7 is installed above.
# !sudo apt-get install -y --no-install-recommends libnvinfer5=5.1.5-1+cuda10.0 \
!sudo apt-get install -y libnvinfer5=5.1.5-1+cuda10.0 \
    libnvinfer-dev=5.1.5-1+cuda10.0

!apt --fix-broken install

• NVIDIA driver로 설치

Pip Install

!python -m pip install pip==20.0.2

• python을 재설치했으니 pip도 다시 설치해줘야한다

TensorFlow Install

%pip install tensorflow-gpu==1.15.3

• tensorflow-gpu==1.15.3 으로 install
• tensorflow==1.15.3으로 install하면 gpu를 찾지 못 해서 cpu로 동작 됨

Install Packages

%pip install requests==2.22.0
%pip install scipy==1.3.3
%pip install Pillow==6.2.1
%pip install https://github.com/podgorskiy/dnnlib/releases/download/0.0.1/dnnlib-0.0.1-py3-none-any.whl
%pip install tqdm

• 각종 패키지들 설치

Google Drive Mount

try:
    from google.colab import drive
    drive.mount('/content/drive', force_remount=True)
    COLAB = True
    print("Note: using Google CoLab")
except:
    print("Note: not using Google CoLab")
    COLAB = False

• 구글 드라이브 마운트 해준다

Git Clone Repository

!git clone https://github.com/NVlabs/stylegan2-ada

• stylegan2 ada repository install

%cd stylegan2-ada

NVCC Test

!echo 'export PATH=/usr/local/cuda/bin:$PATH' >>~/.bashrc

• /usr/local/cuda/bin에 있는 nvcc 디렉터리를 .bashrc의 PATH에 추가

!nvcc test_nvcc.cu -o test_nvcc -run

>> CPU hello
>> GPU hello

• TensorFlow ops 컴파일에 필요한 NVCC가 제대로 실행되는지 테스트해보기

!nvidia-smi
!nvcc -V

• GPU 확인

---

References

• https://www.celantur.com/blog/setting-up-stylegan2/