max test accuracy: 0

[martin-gorner]

max test accuracy: 0

[jorgemadridportillo]

I have the same error here, using Linux Mint 17.3 and Python3.4.3.

Extracting data/train-images-idx3-ubyte.gz
Extracting data/train-labels-idx1-ubyte.gz
Extracting data/t10k-images-idx3-ubyte.gz
Extracting data/t10k-labels-idx1-ubyte.gz
max test accuracy: 0

[martin-gorner]

Is your loss NaN ?
Could it be one of the problems described on this page: https://codelabs.developers.google.com/codelabs/cloud-tensorflow-mnist/#7

If you are running one of the samples without any changes, then it is something else.

[jorgemadridportillo]

No, it is not even displaying the GUI, it seems like it is not even calling the training_step function for me...

[neoscreenager]

No GUI for me too. Program ends with max test accuracy: 0

[martin-gorner]

Have you tried upgrading to Python 3.5?
Have you tried updating your matplotlib library (pip3 install --upgrade matplotlib) ?

[jorgemadridportillo]

Updating to Python 3.5 solved the problem completely!, thanks.

[martin-gorner]

I'll leave this open for a while to hear from others.
Then I'll fold it into #5

[neoscreenager]

All is working well after upgrading to python3.5 and pulling the updated code. Thanks :)

Current environment where the updated code is working for me is python 3.5, matplotlib 2.0.0 and tensorflow 1.0.0.

[val620]

I have the same problem, and I'm running the program with CentOS 7.3.1611, Python 3.5.1, matplotlib 2.0.0, tensorflow 1.0.0.

[neoscreenager]

Is your code updated? If not then please update it to latest revision and then try.

[mehransahandi]

non of updates solved the problem, I just cancelled every GUI parts and then worked well, the only solution worked for me.

[neoscreenager]

@mehransahandi What is your environment?

[kakawait]

Did you try with other graphical engine like TkAgg or Qt5Agg (depends on py-qt5)?

[val620]

All code is updated, still doesn't work.
@mehransahandi Can you elaborate?

[KortVea]

I have had the same problem: no GUI for displaying graphs, after extracting mnist data, it goes directly to max accuracy = 0, and I've tried upgrading all to the latest( Python 2.7 in virtualenv, 3.6.1 in system with tensorflow...)
I solved it by checking (macOS) /Users/you/.matplotlib/matplotlibrc and make sure the configuration something like below
backend : TkAgg

[Lobsterrr]

I have the same problem. tensorflow-1.8.0; python 3.6.3 ; matplotlib-2.2.2.
Still cannot work it out.