Author : Eunbin Seo
Date : 2020.02.04.
- graphic driver install refer: Link
- Check driver version and select cuda version. (our driver version is 430, select cuda 10.1)
nvidia-smi
-
Download Cuda 10.0 runfile Link
-
Run the cuda 10.0 in cuda file. If not working, you make it executable. reference
-
modify bashrc file
export PATH=$PATH:/usr/local/cuda-10.0/bin
export CUDADIR=/usr/local/cuda-10.0
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-10.0/lib64
- After modifying bashrc file, execute this command to verify the cuda version.
source ~/.bashrc
sudo dpkg -i libcudnn7_7.6.4.38-1+cuda10.0_amd64.deb
- make virtual environment
python3 -m venv ./(venv name)
we named virtual environment 'tensorflow' 2. activate the virtual environment
# source (virtual environment name)/bin/activate
source tensorflow/bin/activate
check always that the venv
first, check the version.
pip3 install tensorflow-gpu==(version)
first, check the version.
pip3 install keras==(version)
refer: Link first, check the version.
pip3 install torch===1.2.0 torchvision===0.4.0 -f https://download.pytorch.org/whl/torch_stable.html
refer: Link