Open source software supply chain lighting plan, waiting for you>>>
Problem Description:
ImportError: libcusolver.so.8.0: cannot open shared object file: No such file or directory
First, check the/usr/local/CUDA/lib64 directory to see if libcusolver. So. 8.0 exists. If not, install CUDA correctly. If it is determined that CUDA already exists, this problem is related to the ldconfig command of Linux
Solution:
First method:
On the terminal:
export LD_LIBRARY_PATH=”$LD_LIBRARY_PATH:/usr/local/cuda/lib64”
export CUDA_HOME=/usr/local/cuda
But you have to execute this command every time you want to run tensorflow
Second method:
sudo ldconfig /usr/local/cuda/lib64
Third method:
Write path in bashrc:
vim ~/.bashrc
export LD_LIBRARY_PATH=”$LD_LIBRARY_PATH:/usr/local/cuda/lib64”
export CUDA_HOME=/usr/local/cuda
source ~/.bashrc
Similar Posts:
- Linux use the man command error: Man: command not found [How to Solve]
- [Solved] Ubuntu Install caffe error: build_release/tools/caffe: error while loading shared libraries: libcudart.so.8….
- [Solved] ImportError: libcublas.so.9.0: cannot open shared object file: No such file
- Ubuntu Error: libcublas.so.8.0 [How to Solve]
- [Linux] CONDA: command not found solution
- Solve the problem of ADB: command not found on MAC
- Linux: How to configurate Environment Variables
- Super detail: command not found: the scratch solution (add the scratch environment variable to Zsh under MAC)
- macos -bash: yarn: command not found/-bash: cnpm: command not found
- Mac ADB prompts command not found