Error while training flower about models

[keveman]

Did you compile TensorFlow from source and install it as a PIP package as mentioned in the 'Getting Started' section?

NOTE Before doing anything, we first need to build TensorFlow from source, and installed as a PIP package. Please follow the instructions at Installing From Source.

Do I need to use the Master branch, or 0.7.1

[keveman]

You need HEAD of the master branch.

On Fri, Mar 11, 2016, 9:20 PM Akshay Bhat [email protected] wrote:

Do I need to use the Master branch, or 0.7.1

—
Reply to this email directly or view it on GitHub
#11 (comment).

Thanks, I will try and update.

@AKSHAYUBHAT
Could you please give detailed procedure to solve this problem?
I have encountered the same problem.
And I have installed TensorFlow from source and installed it as a PIP package (tensorflow can be imported in python environment), as @keveman has suggested.

@waterofphysics I think you need to use the HEAD or the master branch for slim tensorflow to compile.
I haven't tried it yet.

@AKSHAYUBHAT Thank you, I think use the HEAD or the master branch for slim tensorflow to compile means use master branch of TensorFlow because this repository can't be compiled alone. Actually, I have already installed TensorFlow from sources as PIP package.
When I am in python environment,

>> import tensorflow

works well.
I am wondering why the compiled codes after

$ bazel build -c opt inception/flowers_train

didn't find tensorflow pip package and get this error?

@waterofphysics You need to install latest version of TensorFlow using the github repo and not the pip package for version 0.7.1.

@AKSHAYUBHAT Thanks for your help. Yes, I have installed the latest version of TensorFlow through
Clone the TensorFlow repository and Create the pip package and install not from Pip Installation in Download and Setup instructions...

Find the solution, thank you @AKSHAYUBHAT.
I have installed from Pip Installation before I Clone the TensorFlow repository and Create the pip package and install.
So I fix this problem through sudo pip uninstall tensorflow first, and then rebuild and pip install the sources.