nvidia-cuDNN-install-docs.sh

# install cuDNN 9.0 
CUDNN_FILE="/root/cudnn-9.0-linux-x64-v7.tgz"
if [ ! -f "$CUDNN_FILE" ]; then
curl -fsSL https://nvidia.obs.myhwclouds.com/cudnn-9.0-linux-x64-v7.tgz -O 
fi
tar -xzf cudnn-9.0-linux-x64-v7.tgz

# copy related files to /usr/local/cuda/lib64 or /usr/local/cuda/include
cp -a ./cuda/include/cudnn.h /usr/local/cuda/include 
cp -a ./cuda/lib64/libcudnn* /usr/local/cuda/lib64 
chmod a+r /usr/local/cuda/include/cudnn.h  /usr/local/cuda/lib64/libcudnn*
rm -fr cudnn-9.0-linux-x64-v7*

# Update your environment variables in bash session or put them in your .bashrc file
# vim ~/.bashrc
export CUDA_HOME=/usr/local/cuda
export PATH=/usr/local/cuda/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH
export LD_LIBRARY_PATH=/usr/local/cuda/lib:$LD_LIBRARY_PATH

# install pip for centos6.x+7.x|rhel6.x+7.x
yum install python*pip -y

# install pip for ubuntu14.x+16.x
apt-get install python-pip python-dev build-essential -y

# install pip for opensuse42.x
zypper install python-pip -y

pip install --upgrade pip

# install tensorflow-gpu
pip install tensorflow-gpu==1.6.0 --trusted-host mirrors.myhuaweicloud.com -i http://mirrors.myhuaweicloud.com/pypi/web/simple