tensorflow.python.framework.errors_impl.NotFoundError: libnvinfer.so.5: cannot open shared object fi

Eclipse在使用Tensorflow-CUDA10 的时候，出现错误;

参考链接：https://blog.csdn.net/qq_20373723/article/details/80065936

官方文档：https://docs.nvidia.com/deeplearning/sdk/tensorrt-install-guide/index.html#downloading

1. Download the TensorRT local repo file that matches the Ubuntu version you are using.
2. Install TensorRT from the Debian local repo package.

   $ sudo dpkg -i  
   nv-tensorrt-repo-ubuntu1604-cuda10.0-trt5.0.2.6-ga-20181009_1-1_amd64.deb
   $ sudo apt-key add /var/nv-tensorrt-repo-cuda10.0-trt5.0.2.6-ga-20181009/7fa2af80.pub
   
   $ sudo apt-get update
   $ sudo apt-get install tensorrt
   

   安装时候出现库的错误，使用sudo apt-get -f install  ,然后 sudo apt-get install tensorrt ，自动下载几Mb文件，安装成功！

3. If using Python 2.7:

   $ sudo apt-get install python-libnvinfer-dev

   The following additional packages will be installed:

     python-libnvinfer

    

   If using Python 3.x:

   $ sudo apt-get install python3-libnvinfer-dev

   The following additional packages will be installed:

     python3-libnvinfer

    

   If you plan to use TensorRT with TensorFlow:

   $ sudo apt-get install uff-converter-tf

   The graphsurgeon-tf package will also be installed with the above command.
4. Verify the installation.

   $ dpkg -l | grep TensorRT

    

   You should see something similar to the following:

   wishchin@wishchin-HP-ENVY-Laptop-13-ad1xx:~/下载$ dpkg -l | grep TensorRT
   ii  graphsurgeon-tf                                             5.0.2-1+cuda10.0                              amd64        GraphSurgeon for TensorRT package
   ii  libnvinfer-dev                                              5.0.2-1+cuda10.0                              amd64        TensorRT development libraries and headers
   ii  libnvinfer-samples                                          5.0.2-1+cuda10.0                              all          TensorRT samples and documentation
   ii  libnvinfer5                                                 5.0.2-1+cuda10.0                              amd64        TensorRT runtime libraries
   ii  python3-libnvinfer                                          5.0.2-1+cuda10.0                              amd64        Python 3 bindings for TensorRT
   ii  python3-libnvinfer-dev                                      5.0.2-1+cuda10.0                              amd64        Python 3 development package for TensorRT
   ii  tensorrt                                                    5.0.2.6-1+cuda10.0                            amd64        Meta package of TensorRT
   ii  uff-converter-tf                                            5.0.2-1+cuda10.0                              amd64        UFF converter for TensorRT package

    

App Server Installation: When setting up servers which will host TensorRT powered applications, you can simply install any of the following:

• the libnvinfer5 package (C++), or
• the python-libnvinfer package (Python 2.7), or
• the python3-libnvinfer package (Python 3.x).

Issue the following commands if you want to run an application that was built with TensorRT using the Debian package, for example:

$ sudo dpkg -i  
nv-tensorrt-repo-ubuntu1x04-cudax.x-trt5.1.x.x-ga-yyyymmdd_1-1_amd64.deb
$ sudo apt-key add /var/nv-tensorrt-repo-cudax.x-trt5.1.x.x-ga-yyyymmdd/7fa2af80.pub

$ sudo apt-get update
$ sudo apt-get install libnvinfer5