mnist_skflow.py

import tensorflow.contrib.learn as skflow
from tensorflow.examples.tutorials.mnist import input_data
from sklearn import datasets, metrics

# Loading MNIST data
mnist = input_data.read_data_sets('MNIST_data', one_hot=True)

def max_pool_2x2(tensor_in):
    return tf.nn.max_pool(tensor_in, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1],
        padding='SAME')

def conv_model(X, y):
    # reshape X to 4d tensor with 2nd and 3rd dimensions being image width and height
    # final dimension being the number of color channels
    X = tf.reshape(X, [-1, 28, 28, 1])
    # first conv layer will compute 32 features for each 5x5 patch
    with tf.variable_scope('conv_layer1'):
        h_conv1 = skflow.ops.conv2d(X, n_filters=32, filter_shape=[5, 5],
                                    bias=True, activation=tf.nn.relu)
        h_pool1 = max_pool_2x2(h_conv1)
    # second conv layer will compute 64 features for each 5x5 patch
    with tf.variable_scope('conv_layer2'):
        h_conv2 = skflow.ops.conv2d(h_pool1, n_filters=64, filter_shape=[5, 5],
                                    bias=True, activation=tf.nn.relu)
        h_pool2 = max_pool_2x2(h_conv2)
        # reshape tensor into a batch of vectors
        h_pool2_flat = tf.reshape(h_pool2, [-1, 7 * 7 * 64])
    # densely connected layer with 1024 neurons
    h_fc1 = skflow.ops.dnn(h_pool2_flat, [1024], activation=tf.nn.relu, keep_prob=0.5)
    return skflow.models.logistic_regression(h_fc1, y)

sess = tf.Session()
builder = tf.saved_model.builder.SavedModelBuilder("./model_sklearn")
sess.run(tf.global_variables_initializer())
# Training and predicting
classifier = skflow.TensorFlowEstimator(
    model_fn=conv_model, n_classes=10, batch_size=100, steps=20000,
    learning_rate=0.001)

builder.add_meta_graph_and_variables(sess, [tf.saved_model.tag_constants.SERVING])
builder.save(True)