OSError: libc10_cuda.so: cannot open shared object file: No such file or directory

📚 Installation

Environment

• OS:Red Hat Enterprise Linux Server release 7.4
• Python version:Python 3.7.9
• PyTorch version:1.7.1
• CUDA/cuDNN version:10.1
• GCC version:GCC 7.3.0
• How did you try to install PyTorch Geometric and its extensions (wheel, source): pip install torch-scatter -f https://data.pyg.org/whl/torch-1.7.1+cu101.html pip install torch-sparse -f https://data.pyg.org/whl/torch-1.7.1+cu101.html pip install torch-geometric
• Any other relevant information:

Checklist

• [yes ] I followed the installation guide.
• [ yes] I cannot find my error message in the FAQ.
• [yes ] I set up CUDA correctly and can compile CUDA code via nvcc.
• [ no] I do have multiple CUDA versions on my machine.

Additional context

i try to import torch_geometric,the error is :OSError: libc10_cuda.so: cannot open shared object file: No such file or directory. detail: Traceback (most recent call last): File “test_and_visulise.py”, line 8, in <module> from models.gat_gcn import GAT_GCN File “/home/qmy/cyp/models/gat_gcn.py”, line 5, in <module> from torch_geometric.nn import GCNConv, GATConv, GINConv, global_add_pool File “/home/qmy/anaconda3/envs/deeplearning/lib/python3.7/site-packages/torch_geometric/init.py”, line 7, in <module> import torch_geometric.data File “/home/qmy/anaconda3/envs/deeplearning/lib/python3.7/site-packages/torch_geometric/data/init.py”, line 1, in <module> from .data import Data File “/home/qmy/anaconda3/envs/deeplearning/lib/python3.7/site-packages/torch_geometric/data/data.py”, line 3, in <module> from torch_geometric.typing import OptTensor, NodeType, EdgeType File “/home/qmy/anaconda3/envs/deeplearning/lib/python3.7/site-packages/torch_geometric/typing.py”, line 4, in <module> from torch_sparse import SparseTensor File “/home/qmy/anaconda3/envs/deeplearning/lib/python3.7/site-packages/torch_sparse/init.py”, line 13, in <module> library, [osp.dirname(file)]).origin) File “/home/qmy/anaconda3/envs/deeplearning/lib/python3.7/site-packages/torch/_ops.py”, line 110, in load_library ctypes.CDLL(path) File “/home/qmy/anaconda3/envs/deeplearning/lib/python3.7/ctypes/init.py”, line 364, in init self._handle = _dlopen(self._name, mode) OSError: libc10_cuda.so: cannot open shared object file: No such file or directory

i can’t find the libc10_cuda.so in my usr/local/lib/ , but i can find libcudart.so and libcudart.so.10.1.

[sauhaardac]

Weirdly my solution above stopped working again for me. I found success with the following:

conda install pytorch==1.11.0 cudatoolkit=11.3 -c pytorch

then

conda install pyg -c pyg

[sauhaardac]

For those who run into this error again, I solved it by starting in a fresh environment and first installing pyg using:

conda install -q -y pyg -c pyg

then, installing a cuda enabled version of pytorch

conda install -q -y pytorch cudatoolkit=11.3 -c pytorch

This first installs a CPU enabled version of torch in the first line and then replaces it with a GPU enabled version of torch. For whatever reason this was the only way I could resolve the error.