Debugging Tensorflow from a Windows machine

Debugging Tensorflow from a Windows machine

I have both an Imac and a Windows 10 machine. From the Tensorflow installation website, they just have the tutorials to install for a Linux or a Mac machine. I know someone already successfully installed Tensorflow in Windows 10 using docker but I just found an easier way to run a tensorflow program locally from a Windows machine and debug it from an IDE. A similar approach can be found in my previous blog (https://web.cs.dal.ca/~anh/?blog_post=debugging-tensorflow-remotely). The only drawback of this approach is you have to install tensorflow in a Mac or a Linux machine. Here are the steps:

  1. Install PyCharms in your Windows- https://www.jetbrains.com/pycharm/(They have a student license which is very cool)
  2. Make sure you installed tensorflow in your Linux or Mac machine from this tutorial (https://www.tensorflow.org/versions/r0.11/get_started/os_setup.html)
  3. In PyCharms, go to “File –> Setting –> Project –> Project Interpreter.”
  4.  Add Remote
  5. Enter the SSH credentials of your installed Linux tensorflow server.
  6. Make sure the python location is correct.

Now you can create a new python using Tensorflow and debug it from your Pycharms IDE.


anh

Leave a Reply

Your email address will not be published. Required fields are marked *