I am trying to run a PyTorch geometric tutorial from this webpage; specifically, tutorial #3 on Graph Classification. Before running any cells I connected to GPU by selecting Runtime --> change runtime type --> GPU, and the command !nvidia-smi returns the following info:
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 470.74 Driver Version: 460.32.03 CUDA Version: 11.2 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 Tesla K80 Off | 00000000:00:04.0 Off | 0 |
| N/A 44C P8 31W / 149W | 3MiB / 11441MiB | 0% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+
However, when I execute the following lines from the second code block:
import torch
from torch_geometric.datasets import TUDataset
I see this error message:
OSError: libcudart.so.10.2: cannot open shared object file: No such file or directory
I am using Chrome OS Version 93.0.4577.95 (Official Build) (64-bit). Interestingly, I have run this tutorial before on GPU without problems; I only started seeing this error when I returned to re-run the notebook today. I have tried a factory reset of runtime to no effect.
I can see a number of other help articles with a similar error, but the most relevant one seems to be this post where the solution was to connect to GPU runtime. I can't find other examples of this occurring in a colab tutorial. It was my impression that colab tutorials such as this one should generally run out-of-the-box, and I'm not sure what I need to install or import to fix it. I appreciate any advice you may have, thanks.
Full stack trace:
---------------------------------------------------------------------------
OSError Traceback (most recent call last)
<ipython-input-3-00fc1f8d4ac6> in <module>()
1 import torch
----> 2 from torch_geometric.datasets import TUDataset
3
4 dataset = TUDataset(root='data/TUDataset', name='MUTAG')
5
6 frames
/usr/local/lib/python3.7/dist-packages/torch_geometric/__init__.py in <module>()
5 from .debug import is_debug_enabled, debug, set_debug
6
----> 7 import torch_geometric.data
8 import torch_geometric.loader
9 import torch_geometric.transforms
/usr/local/lib/python3.7/dist-packages/torch_geometric/data/__init__.py in <module>()
----> 1 from .data import Data
2 from .hetero_data import HeteroData
3 from .temporal import TemporalData
4 from .batch import Batch
5 from .dataset import Dataset
/usr/local/lib/python3.7/dist-packages/torch_geometric/data/data.py in <module>()
1 from typing import (Optional, Dict, Any, Union, List, Iterable, Tuple,
2 NamedTuple, Callable)
----> 3 from torch_geometric.typing import OptTensor, NodeType, EdgeType
4 from torch_geometric.deprecation import deprecated
5
/usr/local/lib/python3.7/dist-packages/torch_geometric/typing.py in <module>()
2
3 from torch import Tensor
----> 4 from torch_sparse import SparseTensor
5
6 # Types for accessing data ####################################################
/usr/local/lib/python3.7/dist-packages/torch_sparse/__init__.py in <module>()
14 ]:
15 torch.ops.load_library(importlib.machinery.PathFinder().find_spec(
---> 16 f'{library}_{suffix}', [osp.dirname(__file__)]).origin)
17
18 if torch.cuda.is_available(): # pragma: no cover
/usr/local/lib/python3.7/dist-packages/torch/_ops.py in load_library(self, path)
102 # static (global) initialization code in order to register custom
103 # operators with the JIT.
--> 104 ctypes.CDLL(path)
105 self.loaded_libraries.add(path)
106
/usr/lib/python3.7/ctypes/__init__.py in __init__(self, name, mode, handle, use_errno, use_last_error)
362
363 if handle is None:
--> 364 self._handle = _dlopen(self._name, mode)
365 else:
366 self._handle = handle
OSError: libcudart.so.10.2: cannot open shared object file: No such file or directory