example.py

# load training and test data from disk
import input_data
mnist = input_data.read_data_sets("MNIST_data/", one_hot=True)

# describe model
import tensorflow as tf
x = tf.placeholder(tf.float32, [None, 784])
W = tf.Variable(tf.zeros([784, 10]))
b = tf.Variable(tf.zeros([10]))
y = tf.nn.softmax(tf.matmul(x, W) + b)

# cost function
y_ = tf.placeholder(tf.float32, [None, 10])
cross_entropy = -tf.reduce_sum(y_ * tf.log(y))
train_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)

with tf.Session() as sess:
  sess.run(tf.initialize_all_variables())
  
  # training loop
  for i in range(1000):
    batch_xs, batch_ys = mnist.train.next_batch(100)
    sess.run(train_step, feed_dict={x: batch_xs, y_: batch_ys})
  
  # evaluation
  correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1))
  accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
  print(sess.run(accuracy, feed_dict={x: mnist.test.images, y_: mnist.test.labels}))