Anaconda安装的Tensorflow缺少来自contrib框架的'audio_ops'

我正在尝试按照音频识别网络教程.

我用python 3.6创建了一个Anaconda环境,然后按照安装说明安装GPU.

我可以运行"hello world"TF示例.

当我在音频识别网络教程/示例中运行'train.py'时,我得到:

Traceback (most recent call last):
  File "train.py", line 79, in 
    import input_data
  File "/home/philglau/speech_commands/input_data.py", line 35, in 
    from tensorflow.contrib.framework.python.ops import audio_ops as contrib_audio
ImportError: cannot import name 'audio_ops'

失败的教程中的代码是:

from tensorflow.contrib.framework.python.ops import audio_ops as contrib_audio

然后我备份了那个链,直到我可以导入它的一部分:

import tensorflow.contrib.framework as test ==> works
import tensorflow.contrib.framework.python as test --> fail: 
Traceback (most recent call last):
  File "", line 1, in 
AttributeError: module 'tensorflow.contrib.framework' has no attribute 'python'

不确定我的安装在哪里出错了.

细节:

Ubuntu 16.04
Anaconda env with python 3.6
Followed the 'anaconda' instruction on the TF install page. (GPU version)

我也尝试使用python 2.7 env for anaconda但得到了相同的结果.

看起来他们正在发布版本1.4中的audio_ops模块(https://github.com/tensorflow/tensorflow/issues/11339#issuecomment-327879009).

在v1.4发布之前,一个简单的方法是安装nightly tensorflow构建

pip install tf-nightly

或者在问题评论中链接的docker镜像.