cuda , CUPTI and cudnn, following official site:
cuda first, tensorflow second.
https://tensorflow.google.cn/install
https://tensorflow.google.cn/install/gpu
==1.=============================================================================================
# Add NVIDIA package repositories
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/cuda-repo-ubuntu1804_10.1.243-1_amd64.deb
sudo apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/7fa2af80.pub
sudo dpkg -i cuda-repo-ubuntu1804_10.1.243-1_amd64.deb
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 ./nvidia-machine-learning-repo-ubuntu1804_1.0.0-1_amd64.deb
sudo apt-get update
# Install NVIDIA driver
sudo apt-get install --no-install-recommends nvidia-driver-430
# Reboot. Check that GPUs are visible using the command: nvidia-smi
#LL:when table printed, execute this again:
sudo apt-get update
# Install development and runtime libraries (~4GB)
sudo apt-get install --no-install-recommends \
cuda-10-1 \
libcudnn7=7.6.4.38-1+cuda10.1 \
libcudnn7-dev=7.6.4.38-1+cuda10.1
# Install TensorRT. Requires that libcudnn7 is installed above.
sudo apt-get install -y --no-install-recommends libnvinfer6=6.0.1-1+cuda10.1 \
libnvinfer-dev=6.0.1-1+cuda10.1 \
libnvinfer-plugin6=6.0.1-1+cuda10.1
==2.=============================================================================================
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-10.1/extras/CUPTI/lib64
==3.=============================================================================================
vim ~/.bashrc
export PATH="/usr/local/cuda-10.1/bin:$PATH"
source ~/.bashrc
==4.=============================================================================================
[/home/openclo-host/anaconda3]>>>"enter directly"
PREFIX=/home/opencl-host/anaconda3
...
...
Do you wish the installer to initialize Anaconda3 by running conda init?[yes|no]
[no]>>>yes
==========================
$which conda
$which pip
$which python
$which ipython
==5.=============================================================================================
#LL:fanqiang
conda install pytorch torchvision cudatoolkit=10.1 -c pytorch
==6.=============================================================================================
pip install tensorflow
fail as global pypi, instead:
pip install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorflow
more detail:
https://mirrors.tuna.tsinghua.edu.cn/help/pypi/
==7.=============================================================================================
Downloads/pycharm-community-2020.1.2/bin$ ls
format.sh fsnotifier fsnotifier64 idea.properties inspect.sh libdbm64.so log.xml printenv.py pycharm64.vmoptions pycharm.png pycharm.sh pycharm.svg pycharm.vmoptions restart.py
==8.=============================================================================================
==9.=============================================================================================end.
useless:
cuda10.1Linux x86_64:
https://developer.nvidia.com/cuda-10.1-download-archive-update2?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1804&target_type=deblocal
cuda10.0Linux x86_64:
https://developer.nvidia.com/cuda-10.0-download-archive?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1804&target_type=deblocal
anaconda_2019.03:
https://repo.anaconda.com/archive/
cudnn 7.5.0 for cuda10.0:
https://developer.nvidia.com/rdp/cudnn-archive
cuDNN Library for Linux
advance==> recovery mode ==> enable networking ==>root:
sudo apt-get install -y xserver-xorg-input-all