Tensorflow MNIST Tutorial

This is a Docker file for Martin Gorner's "Tensorflow and deep learning, without a PhD" presentation and code lab.

NOTE: Because the code lab expects a graphical user interface, the sample scripts must be modified to ignore the animation code. This Docker image ONLY patches the first script in the series: mnist_1.0_softmax.py. Additional script patching is left as an exercise for the reader. See this GitHub issue comment for more details.

Build

To build this image:

./build.sh

Usage

To run this image:

./run.sh

With the image running, you may now run the sample code. For example:

python3 mnist_1.0_softmax.py

License

MIT License