Then export the following environment variables, echo 'export LD_LIBRARY_PATH=/usr/lib/cuda/lib64:$LD_LIBRARY_PATH' > ~/.bashrcĮcho 'export LD_LIBRARY_PATH=/usr/lib/cuda/include:$LD_LIBRARY_PATH' > ~/.bashrcįinally, you can now install TensorFlow GPU, pip install tensorflow Set the file permissions, sudo chmod a+r /usr/lib/cuda/include/cudnn.h /usr/lib/cuda/lib64/libcudnn* Sudo cp cuda/lib64/libcudnn* /usr/lib/cuda/lib64/ Next, copy the extracted files sudo cp cuda/include/cudnn.h /usr/lib/cuda/include/
After downloading, extract the files by running tar -xvzf cudnn-10.1-linux-圆4-v7.6.5.32.tgz. Click on Download cuDNN v7.6.5 (November 5th, 2019) for CUDA 10.1, then choose cuDNN Library for Linux.
NVIDIA CUDA TOOLKIT LICENSE SOFTWARE LICENSE
After logging in and accepting the terms of cuDNN software license agreement, a list of downloadable cuDNN software will be displayed. Choose Download cuDNN, and you'll be asked to login or create an account. Then, you can install cuDNN 7.6.5 by downloading it from this link. You can check this on your own by running whereis cuda. usr/local/cuda, usr/local/cuda-10.1), instead, it is installed in /usr/lib (i.e. The difference though is that CUDA is not installed in the usual path (i.e. This means that CUDA is successfully installed on your Ubuntu 20.04. Then you will get an output similar to the following, nvcc: NVIDIA (R) Cuda compiler driverĬopyright (c) 2005-2019 NVIDIA CorporationĬuda compilation tools, release 10.1, V10.1.243 First, you can install CUDA 10.1 by sudo apt install nvidia-cuda-toolkit.
0 kommentar(er)
