导入错误:没有名为cloud.ml的模块 [英] Import Error: no module named cloud.ml

问题描述

我正在尝试按照说明在tensorflow中使用本地预测，如

I am trying to follow the instructions to use local predictions in tensorflow as described here. Running the command gcloud ml-engine local predict --model-dir=~/PycharmProjects/nlc/export/1/ --json-instances=test.json gives me the error:

ERROR: (gcloud.ml-engine.local.predict) Cannot import google.cloud.ml. Please verify "python -c 'import google.cloud.ml'" works. Please verify the installed cloudml sdk version with: "python -c 'import google.cloud.ml as cloudml; print cloudml.__version__'".

要获取google.cloud.ml软件包，我需要安装什么?在该链接上所说的只是

What do I need to install in order to get the google.cloud.ml package? All it says on that link is

它需要在本地安装TensorFlow SDK

It requires the TensorFlow SDK be installed locally

Tensorflow在我的系统上正常工作，并且python -c 'import tensorflow as tf; print(tf.__version__)'输出1.0.1.我还确保通过运行pip install -U google-api-python-client安装python实用程序.我还需要安装什么?

Tensorflow is working properly on my system and python -c 'import tensorflow as tf; print(tf.__version__)' outputs 1.0.1. I have also made sure to install the python utilities by running pip install -U google-api-python-client. What else do I need to install?

正在运行Ubuntu的VM上尝试这样做，这样可以满足要求:

This is being attempted on a VM running Ubuntu, so that handles the requirement:

目前仅支持基于Debian的系统.

Only Debian based systems are supported at this time.

推荐答案

这是gcloud的已知错误，将在明天(2017/03/15)发行版中修复.明天某个时间，请运行

This is a known bug with gcloud that will be fixed in tomorrow's (2017/03/15) release. Sometime tomorrow, please run

gcloud components update

给您带来的不便表示歉意.

Apologies for the inconvenience.

这篇关于导入错误:没有名为cloud.ml的模块的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！