1.获取cudnn(https://developer.nvidia.com/rdp/cudnn-download,需要注册,选择ubuntu相关的)
（注:可以从笔者的gitee获取:https://gitee.com/Jello_Smith/cudnn_deb.git,由于文件太大，因此涉及的软件包libcudnn7和libcudnn7-dev都经过split切割，直接通过cat即可拼接好)
2.安装cudnn
sudo dpkg -i libcudnn7_*.deb
sudo dpkg -i libcudnn7-dev*.deb
3.获取cuda(https://developer.nvidia.com/cuda-downloads,选择ubuntu相关的)
(注:可以从笔者的gitee获取:https://gitee.com/Jello_Smith/cuda_deb.git,文件被split分割，需要使用cat拼接)
4.安装cuda (cuda依赖cudnn)
sudo dpkg -i cuda-repo-ubuntu1804*.deb
sudo apt-key add /var/cuda-repo-<version>/7fa2af80.pub
sudo apt-get update
sudo apt-get install cuda
5.安装python相关软件
sudo apt-get install python-pip python-dev
6.安装tensorflow(tensorflow的版本需要与cuda以及cudnn匹配)
pip install tensorflow-gpu==1.5.0
7.测试
>>> import tensorflow as tf
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/home/jello/.local/lib/python2.7/site-packages/tensorflow/__init__.py", line 24, in <module>
from tensorflow.python import *
File "/home/jello/.local/lib/python2.7/site-packages/tensorflow/python/__init__.py", line 47, in <module>
import numpy as np
File "/home/jello/.local/lib/python2.7/site-packages/numpy/__init__.py", line 142, in <module>
from . import core
File "/home/jello/.local/lib/python2.7/site-packages/numpy/core/__init__.py", line 59, in <module>
from . import numeric
File "/home/jello/.local/lib/python2.7/site-packages/numpy/core/numeric.py", line 3093, in <module>
from . import fromnumeric
File "/home/jello/.local/lib/python2.7/site-packages/numpy/core/fromnumeric.py", line 17, in <module>
from . import _methods
File "/home/jello/.local/lib/python2.7/site-packages/numpy/core/_methods.py", line 158, in <module>
_NDARRAY_ARRAY_FUNCTION = mu.ndarray.__array_function__
AttributeError: type object 'numpy.ndarray' has no attribute '__array_function__'