OSError: libcudart.so.9.2: 无法打开共享对象文件: 没有这样的文件或目录

Question

当我得到时，试图运行脚本

(infacepytorch) ryan@ryan-7570:/media/ryan/shakira/InsightFace_Pytorch$ python face_verify.py 
Traceback (most recent call last):
  File "face_verify.py", line 9, in <module>
from Learner import face_learner
  File "/media/ryan/shakira/InsightFace_Pytorch/Learner.py", line 1, in <module>
    from data.data_pipe import de_preprocess, get_train_loader, get_val_data
  File "/media/ryan/shakira/InsightFace_Pytorch/data/data_pipe.py", line 12, in <module>
import mxnet as mx
  File "/home/ryan/virtualenvs/infacepytorch/lib/python3.5/site-packages/mxnet/__init__.py", line 24, in <module>
from .context import Context, current_context, cpu, gpu, cpu_pinned
  File "/home/ryan/virtualenvs/infacepytorch/lib/python3.5/site-packages/mxnet/context.py", line 24, in <module>
from .base import classproperty, with_metaclass, _MXClassPropertyMetaClass
  File "/home/ryan/virtualenvs/infacepytorch/lib/python3.5/site-packages/mxnet/base.py", line 213, in <module>
_LIB = _load_lib()
  File "/home/ryan/virtualenvs/infacepytorch/lib/python3.5/site-packages/mxnet/base.py", line 204, in _load_lib
lib = ctypes.CDLL(lib_path[0], ctypes.RTLD_LOCAL)
  File "/usr/lib/python3.5/ctypes/__init__.py", line 347, in __init__
self._handle = _dlopen(self._name, mode)
OSError: libcudart.so.9.2: cannot open shared object file: No such file or directory

我试过了

export LD_LIBRARY_PATH=/usr/local/cuda-9.0/lib64 && sudo ldconfig

但我仍然得到同样的错误，任何建议都会很有帮助，提前谢谢。

Answer 1

正如 talonmies 所指出的，您正在运行依赖于 Cuda-9.2 的 MXNet 版本。最好安装 Cuda-9.2 并将库路径更新为指向 9.2。

如果您

无法更新 Cuda 依赖项，您还可以使用 pip 包mxnet-cu90安装 MXNet for 9.0。