tensorflow的Virtualenv安装方式安装

　本文介绍了如何在ubuntu上以virtualenv方式安装tensorflow。　　

安装pip和virtualenv：

1

2

3

4

5

6

# Ubuntu/Linux 64-bit

sudo apt-get install python-pip python-dev python-virtualenv

 

# Mac OS X

sudo easy_install pip

sudo pip install --upgrade virtualenv

 创建 Virtualenv 虚拟环境：

　　进入你想安装tensorflow的父目录下，然后执行下面命令建立虚拟环境：

1

virtualenv --system-site-packages tensorflow

 激活虚拟环境并安装tensorflow：

　　对于python27，则执行如下命令：

1

2

3

4

5

6

7

8

9

10

11

12

source ./tensorflow/bin/activate  # If using bash

source ./tensorflow/bin/activate.csh  # If using csh

(tensorflow)$  # Your prompt should change

 

# Ubuntu/Linux 64-bit, CPU only:

pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.6.0-cp27-none-linux_x86_64.whl

 

# Ubuntu/Linux 64-bit, GPU enabled:

pip install --upgrade https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.6.0-cp27-none-linux_x86_64.whl

 

# Mac OS X, CPU only:

pip install --upgrade https://storage.googleapis.com/tensorflow/mac/tensorflow-0.6.0-py2-none-any.whl

 　　对于python3则执行如下命令：

1

2

3

4

5

6

7

8

9

10

11

12

source ./tensorflow/bin/activate  # If using bash

source ./tensorflow/bin/activate.csh  # If using csh

(tensorflow)$  # Your prompt should change

 

# Ubuntu/Linux 64-bit, CPU only:

pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.6.0-cp34-none-linux_x86_64.whl

 

# Ubuntu/Linux 64-bit, GPU enabled:

pip install --upgrade https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.6.0-cp34-none-linux_x86_64.whl

 

# Mac OS X, CPU only:

pip3 install --upgrade https://storage.googleapis.com/tensorflow/mac/tensorflow-0.6.0-py3-none-any.whl

 测试安装：

　　在终端执行如下命令进入python shell环境：

1

python

 　　在python shell环境中测试：

1

2

3

4

5

6

7

8

9

10

>>> import tensorflow as tf

>>> hello = tf.constant('Hello, TensorFlow!')

>>> sess = tf.Session()

>>> print(sess.run(hello))

Hello, TensorFlow!

>>> a = tf.constant(10)

>>> b = tf.constant(32)

>>> print(sess.run(a + b))

42

>>>

•  如果遇到如下错误：

1

2

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

ImportError: libcudart.so.7.0: cannot open shared object file: No such file or directory

 　　那是你的CUDA安装配置不对：

　　　　安装CUDA和CUDNN可以参考 这篇文章 。

　　且添加如下两行到你的 ~/.bashrc 文件

1

2

export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/usr/local/cuda/lib64"

export CUDA_HOME=/usr/local/cuda

•  如果遇到如下错误：

1

2

3

4

5

6

7

8

9

10

11

12

13

Python 2.7.9 (default, Apr  2 2015, 15:33:21)

[GCC 4.9.2] on linux2

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

>>> import tensorflow

I tensorflow/stream_executor/dso_loader.cc:93] Couldn't open CUDA library libcublas.so.7.0. LD_LIBRARY_PATH: :/usr/local/cuda/lib64

I tensorflow/stream_executor/cuda/cuda_blas.cc:2188] Unable to load cuBLAS DSO.

I tensorflow/stream_executor/dso_loader.cc:93] Couldn't open CUDA library libcudnn.so.6.5. LD_LIBRARY_PATH: :/usr/local/cuda/lib64

I tensorflow/stream_executor/cuda/cuda_dnn.cc:1382] Unable to load cuDNN DSO

I tensorflow/stream_executor/dso_loader.cc:93] Couldn't open CUDA library libcufft.so.7.0. LD_LIBRARY_PATH: :/usr/local/cuda/lib64

I tensorflow/stream_executor/cuda/cuda_fft.cc:343] Unable to load cuFFT DSO.

I tensorflow/stream_executor/dso_loader.cc:101] successfully opened CUDA library libcuda.so locally

I tensorflow/stream_executor/dso_loader.cc:93] Couldn't open CUDA library libcurand.so.7.0. LD_LIBRARY_PATH: :/usr/local/cuda/lib64

I tensorflow/stream_executor/cuda/cuda_rng.cc:333] Unable to load cuRAND DSO.

 　　由安装报错可知，它使用的是7.0版本，故找不到，而如果你安装的是7.5版本，则可以执行如下命令添加相应链接：

1

2

3

4

sudo ln -s /usr/local/cuda/lib64/libcudart.so.7.5 /usr/local/cuda/lib64/libcudart.so.7.0

sudo ln -s libcublas.so.7.5 libcublas.so.7.0

sudo ln -s libcudnn.so.4.0.4 libcudnn.so.6.5

sudo ln -s libcufft.so libcufft.so.7.0<br>sudo ln -s libcurand.so libcurand.so.7.0