tf-mnist

MNIST in TensorFlow

This directory builds a convolutional neural net to classify the MNIST dataset using the tf.data, tf.estimator.Estimator, and tf.layers APIs.Tensorflow 1.15.0 is used in this example.TF-2.0.0 is not supported.

Setup

To build application using standalone S2I and then run the resulting image with Docker execute:

s2i build .  centos/python-36-centos7 mnist-python

To train the model, run the following:

docker run -it mnist-python

This runs the app.sh script

python official/mnist/mnist.py --data_format 'channels_last' --stop_threshold '0.96'

The model will begin training and will automatically evaluate itself on the validation data.The checkpoint folder is /tmp/mnist_model/model.ckpt.

Exporting the model

You can export the model into Tensorflow SavedModel format by using the argument --export_dir:

python mnist.py --data_format 'channels_last' --stop_threshold '0.96' --export_dir /tmp/mnist_saved_model

The SavedModel will be saved in a timestamped directory under /tmp/mnist_saved_model/ (e.g. /tmp/mnist_saved_model/1513630966/).

Getting predictions with SavedModel Use saved_model_cli to inspect and execute the SavedModel.

saved_model_cli run --dir /tmp/mnist_saved_model/TIMESTAMP --tag_set serve --signature_def classify --inputs image=/opt/app-root/src/official/mnist/examples.npy

examples.npy contains the data from example5.png and example3.png in a numpy array, in that order. The array values are normalized to values between 0 and 1.

The output should look similar to below:

Result for output key classes:
[5 3]
Result for output key probabilities:
[[  1.53558474e-07   1.95694142e-13   1.31193523e-09   5.47467265e-03
    5.85711526e-22   9.94520664e-01   3.48423509e-06   2.65365645e-17
    9.78631419e-07   3.15522470e-08]
 [  1.22413359e-04   5.87615965e-08   1.72251271e-06   9.39960718e-01
    3.30306928e-11   2.87386645e-02   2.82353517e-02   8.21146413e-18
    2.52568233e-03   4.15460236e-04]]

Experimental: Eager Execution

Eager execution (an preview feature in TensorFlow 1.5) is an imperative interface to TensorFlow. The exact same model defined in mnist.py can be trained without creating a TensorFlow graph using:

python mnist_eager.py

Credits

TensorFlow Authors