0

Now that SKFLOW has been moved within TensorFlow I am not sure how to use it. I am running Anaconda on Ubuntu server and I installed Tensorflow as suggested in this question: Tensorflow and Anaconda on Ubuntu?

My scripts are seeing Tensorflow but not SKFLOW.

Community
  • 1
  • 1
denson
  • 2,366
  • 2
  • 24
  • 25

2 Answers2

1

I don't think that skflow is in the latest, 0.7.1, tensorflow release. And therefore you won't find it if you have just conda installed tensorflow via the post you mentioned.

For example, checking the anaconda channel for tensorflow it looks like 0.7.1 was uploaded on March 3rd while the first check-in of skflow in the tensorflow code base was on March 9th. I also don't see any notice of skflow in the recent release mentions of tensorflow.

That said, when it is released with skflow as part of it, I expect you'll be able to import it like this:

import tensorflow.contrib.skflow

Community
  • 1
  • 1
Paul
  • 5,473
  • 1
  • 30
  • 37
1

I have shared this in our Gitter channel but it might useful to repost it here.

You can use pip to install TensorFlow's nightly binaries as mentioned in the README file if you don't want to wait until next official release.

Then at least for the date I posted this, you can import skflow directly from TensorFlow via from tensorflow.contrib.skflow.python import skflow and then everything else works the same as usual.

Hope this helps.

Edit: Once this PR is merged, you'll be able to import as done in examples folder in TensorFlow, e.g. from tensorflow.contrib import skflow.

Yuan Tang
  • 696
  • 4
  • 15
  • After "pip install tf-nightly", both "from tensorflow.contrib.skflow.python import skflow" and "from tensorflow.contrib import skflow" still fail. Any suggestion? The PR you mentioned seemed to be merged to the master. – jl303 Apr 06 '19 at 11:30