ubuntu16.04 tensorflow-gpu版本安装好后，容易的检测代码

安装好cuda和 cudnn以后，我们用pip命令安装tensorflow-gpu版本(ubuntu 16.04)：

sudo pip install tensorflow-gpu

 

如果是安装的tensorflow 1.3版本，那就要注意了，需要安装cuda 8.0和cudnn 6.0（重要），否则会报错：

xyxt@xyxt-System-Product-Name:~/Downloads/shangyixing$ python  

Python 2.7.12 (default, Nov 19 2016, 06:48:10)   

[GCC 5.4.0 20160609] on linux2  

Type"help","copyright","credits" or"license" for more information.  

>>> import tensorflow  

Traceback (most recent call last):  

File"", line 1, in  

File"/usr/local/lib/python2.7/dist-packages/tensorflow/__init__.py", line 24, in  

from tensorflow.python import *  

File"/usr/local/lib/python2.7/dist-packages/tensorflow/python/__init__.py", line 49, in  

from tensorflow.python import pywrap_tensorflow  

File"/usr/local/lib/python2.7/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 52, in  

raise ImportError(msg)  

ImportError: Traceback (most recent call last):  

File"/usr/local/lib/python2.7/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 41, in  

from tensorflow.python.pywrap_tensorflow_internal import *  

File"/usr/local/lib/python2.7/dist-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 28, in  

_pywrap_tensorflow_internal = swig_import_helper()  

File"/usr/local/lib/python2.7/dist-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 24, in swig_import_helper  

_mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname, description)  

ImportError: libcudnn.so.6: cannot open shared object file: No such file or directory

Failed to load the native TensorFlow runtime.

See https://www.tensorflow.org/install/install_sources#common_installation_problems

for some common reasons and solutions.  Include the entire stack trace  

above this error message when asking for help.

 

我们在检测tensorflow-gpu版本是否安装好，需要运行一个python代码进行测试，在python终端输入：

import tensorflow 

>>> import tensorflow as tf  

>>> matrix1 = tf.constant([[3., 3.]])   

>>> matrix2 = tf.constant([[2.],[2.]])  

>>> product = tf.matmul(matrix1, matrix2)   

>>> sess = tf.Session()

 

输出的信息，如果有你的显卡信息，则说明你的tensorflow GPU 版本安装成功了。

2017-10-10 17:44:46.428528: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.  

2017-10-10 17:44:46.428544: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.  

2017-10-10 17:44:46.428549: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.  

2017-10-10 17:44:46.428553: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.  

2017-10-10 17:44:46.428557: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.  

2017-10-10 17:44:46.547766: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:893] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero  

2017-10-10 17:44:46.548007: I tensorflow/core/common_runtime/gpu/gpu_device.cc:955] Found device 0 with properties:   

name: GeForce GTX 1070  

major: 6 minor: 1 memoryClockRate (GHz) 1.683  

pciBusID 0000:01:00.0  

Total memory: 7.92GiB  

Free memory: 7.43GiB  

2017-10-10 17:44:46.548020: I tensorflow/core/common_runtime/gpu/gpu_device.cc:976] DMA: 0   

2017-10-10 17:44:46.548024: I tensorflow/core/common_runtime/gpu/gpu_device.cc:986] 0:   Y   

2017-10-10 17:44:46.548029: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1045] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GeForce GTX 1070, pci bus id: 0000:01:00.0)

 

参考文献

[1].【报错】ImportError: libcudnn.so.6: cannot open shared object file: No such 今更新了TensorFlow 1.3，结果运行时报ImportError: libcudnn.so.6: cannot open shared object file: No such file or directory

看了更新日志发现，1.3版本需要cuDNN v6.0做支持了，之前是5.1，于是去Nvidia官网下载cuDNN v6.0对应的系统版本安装就好了。