所以我在Ubuntu 16.04LTS，python 3.5.2上的anaconda下运行Tensorflow 1.1.0，按照说明单独的anaconda环境“tensorflow”。
我想升级到1.2rc，它似乎安装如下：
(tensorflow) paul@SPEED:~$ pip install --ignore-installed --no-cache-dir --upgrade https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow_gpu-1.2.0rc1-cp35-cp35m-linux_x86_64.whl
Collecting tensorflow-gpu==1.2.0rc1 from https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow_gpu-1.2.0rc1-cp35-cp35m-linux_x86_64.whl
Downloading https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow_gpu-1.2.0rc1-cp35-cp35m-linux_x86_64.whl (89.2MB)
100% |████████████████████████████████| 89.2MB 1.6MB/s
Collecting wheel>=0.26 (from tensorflow-gpu==1.2.0rc1)
Downloading wheel-0.29.0-py2.py3-none-any.whl (66kB)
100% |████████████████████████████████| 71kB 168kB/s
Collecting six>=1.10.0 (from tensorflow-gpu==1.2.0rc1)
Downloading six-1.10.0-py2.py3-none-any.whl
Collecting numpy>=1.11.0 (from tensorflow-gpu==1.2.0rc1)
Downloading numpy-1.12.1-cp35-cp35m-manylinux1_x86_64.whl (16.8MB)
100% |████████████████████████████████| 16.8MB 1.1MB/s
Collecting markdown==2.2.0 (from tensorflow-gpu==1.2.0rc1)
Downloading Markdown-2.2.0.tar.gz (236kB)
100% |████████████████████████████████| 245kB 1.1MB/s
Collecting html5lib==0.9999999 (from tensorflow-gpu==1.2.0rc1)
Downloading html5lib-0.9999999.tar.gz (889kB)
100% |████████████████████████████████| 890kB 1.1MB/s
Collecting werkzeug>=0.11.10 (from tensorflow-gpu==1.2.0rc1)
Downloading Werkzeug-0.12.2-py2.py3-none-any.whl (312kB)
100% |████████████████████████████████| 317kB 1.3MB/s
Collecting bleach==1.5.0 (from tensorflow-gpu==1.2.0rc1)
Downloading bleach-1.5.0-py2.py3-none-any.whl
Collecting protobuf>=3.2.0 (from tensorflow-gpu==1.2.0rc1)
Downloading protobuf-3.3.0-cp35-cp35m-manylinux1_x86_64.whl (5.7MB)
100% |████████████████████████████████| 5.7MB 569kB/s
Collecting setuptools (from protobuf>=3.2.0->tensorflow-gpu==1.2.0rc1)
Downloading setuptools-35.0.2-py2.py3-none-any.whl (390kB)
100% |████████████████████████████████| 399kB 835kB/s
Collecting packaging>=16.8 (from setuptools->protobuf>=3.2.0->tensorflow-gpu==1.2.0rc1)
Downloading packaging-16.8-py2.py3-none-any.whl
Collecting appdirs>=1.4.0 (from setuptools->protobuf>=3.2.0->tensorflow-gpu==1.2.0rc1)
Downloading appdirs-1.4.3-py2.py3-none-any.whl
Collecting pyparsing (from packaging>=16.8->setuptools->protobuf>=3.2.0->tensorflow-gpu==1.2.0rc1)
Downloading pyparsing-2.2.0-py2.py3-none-any.whl (56kB)
100% |████████████████████████████████| 61kB 657kB/s
Installing collected packages: wheel, six, numpy, markdown, html5lib, werkzeug, bleach, pyparsing, packaging, appdirs, setuptools, protobuf, tensorflow-gpu
Running setup.py install for markdown ... done
Running setup.py install for html5lib ... done
Successfully installed appdirs-1.4.3 bleach-1.5.0 html5lib-0.9999999 markdown-2.2.0 numpy-1.12.1 packaging-16.8 protobuf-3.3.0 pyparsing-2.2.0 setuptools-35.0.2 six-1.10.0 tensorflow-gpu-1.2.0rc1 werkzeug-0.12.2 wheel-0.29.0
但是，当我进入python验证Tensorflow的版本时，它仍然是1.1.0。
有什么想法吗？
(tensorflow) paul@SPEED:~$ python
Python 3.5.2 |Anaconda custom (64-bit)| (default, Jul 2 2016, 17:53:06)
[GCC 4.4.7 20120313 (Red Hat 4.4.7-1)] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
>>> print(tf.__version__)
1.1.0
这绝对是一个安装程序问题我已经记录了TF团队的一个错误 这有效：
conda create --name tensorflow python=3
source activate tensorflow
pip install --ignore-installed --upgrade https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow_gpu-1.2.0rc1-cp36-cp36m-linux_x86_64.whl
python
>>> import tensorflow as tf; print(tf.__version__)
1.2.0-rc1