Skip to content

Instantly share code, notes, and snippets.

@scionoftech

scionoftech/tensorflow_2.3.1_cuda_10.1_installation_on_Ubuntu_18.04

Last active September 9, 2021 14:00
Show Gist options
  • Save scionoftech/47d8fb458f0a27910e7861e1f01bb631 to your computer and use it in GitHub Desktop.
Save scionoftech/47d8fb458f0a27910e7861e1f01bb631 to your computer and use it in GitHub Desktop.
Download ZIP

Revisions

  1. scionoftech revised this gist Sep 9, 2021. 1 changed file with 4 additions and 0 deletions.
    4 changes: 4 additions & 0 deletions tensorflow_2.3.1_cuda_10.1_installation_on_Ubuntu_18.04
    View file
    Original file line number Diff line number Diff line change
    @@ -57,6 +57,10 @@ cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2
    # check driver with cuda
    lspci | grep -i nvidia

    # install tensorflow 2.3.1 and pytorch 1.7.1
    pip install tensorflow==2.3.1
    pip install torch==1.7.1+cu101 torchvision==0.8.2+cu101 torchaudio==0.7.2 -f https://download.pytorch.org/whl/torch_stable.html

    # Test tensorflow_2
    import tensorflow as tf
    print(tf.config.list_physical_devices('GPU'))
  2. scionoftech created this gist Sep 8, 2021.
    71 changes: 71 additions & 0 deletions tensorflow_2.3.1_cuda_10.1_installation_on_Ubuntu_18.04
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,71 @@
    # Deeplearning Environment Setup for tensorflow_2.3.1 with CUDA 10.1 and cuDNN 7.6.0

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


    # Get latest nvidia driver
    apt-cache search nvidia-driver
    sudo apt install nvidia-driver-470

    # check nvidia driver version
    nvidia-smi


    # install cuda 10.1 method-1
    # https://developer.nvidia.com/cuda-10.1-download-archive-base?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1804&target_type=runfilelocal
    wget https://developer.nvidia.com/compute/cuda/10.1/Prod/local_installers/cuda_10.1.105_418.39_linux.run
    # run below command and follow the command-line prompts
    sudo sh cuda_10.1.105_418.39_linux.run

    # install cuda 10.1 method-2
    # https://developer.nvidia.com/cuda-10.1-download-archive-base?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1804&target_type=deblocal
    wget https://developer.nvidia.com/compute/cuda/10.1/Prod/local_installers/cuda-repo-ubuntu1804-10-1-local-10.1.105-418.39_1.0-1_amd64.deb
    sudo dpkg -i cuda-repo-ubuntu1804-10-1-local-10.1.105-418.39_1.0-1_amd64.deb
    sudo apt-key add /var/cuda-repo-10-1-local-10.1.105-418.39/7fa2af80.pub
    sudo apt-get update
    sudo apt-get install cuda


    # install cuDNN(CUDA® Deep Neural Network library) 7.6.0 version (CUDA 10.1 is compatable with cuDNN 7.5.1 - 7.6.2)
    # source https://anaconda.org/anaconda/cudnn/files
    wget https://anaconda.org/anaconda/cudnn/7.6.0/download/linux-64/cudnn-7.6.0-cuda10.1_0.tar.bz2
    tar -xvf cudnn-7.6.0-cuda10.1_0.tar.bz2 -C cuda
    sudo cp cuda/include/cudnn.h /usr/local/cuda/include
    sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64
    sudo chmod a+r /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn*
    echo 'export PATH=/usr/local/cuda-10.1/bin${PATH:+:${PATH}}' >> ~/.bashrc
    echo 'export LD_LIBRARY_PATH=/usr/local/cuda-10.1/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}' >> ~/.bashrc
    echo 'export CUDA_HOME=/usr/local/cuda' >> ~/.bashrc
    source ~/.bashrc

    # test cuDNN
    cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2
    #Should see something like below:
    #define CUDNN_MAJOR 6
    #define CUDNN_MINOR 0
    #define CUDNN_PATCHLEVEL 21
    --
    #define CUDNN_VERSION (CUDNN_MAJOR * 1000 + CUDNN_MINOR * 100 + CUDNN_PATCHLEVEL)
    #include "driver_types.h"

    # check driver with cuda
    lspci | grep -i nvidia

    # Test tensorflow_2
    import tensorflow as tf
    print(tf.config.list_physical_devices('GPU'))

    # install nvidia-docker
    # first install docker-ce
    sudo apt-get update
    sudo apt-get install -y nvidia-docker2
    sudo pkill -SIGHUP dockerd

    # Test environment and to make sure everything is installed correctly
    sudo docker run --runtime=nvidia --rm nvidia/cuda:10.1-base nvidia-smi