py_x = model(X, w_h, w_o)
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(py_x, Y))
train_op = tf.train.GradientDescentOptimizer(learning_rate=0.01).minimize(cost)
predict_op = tf.argmax(py_x, 1)

sess = tf.Session()
init = tf.initialize_all_variables()
sess.run(init)

for i in range(700):
    for start, end in zip(range(0, len(X_train),1 ), range(1, len(X_train),1)):
        sess.run(train_op, feed_dict={X: X_train[start:end], Y: y_train[start:end]})
    if i % 100 == 0:
        print (i, np.mean(np.argmax(y_test, axis=1) ==
                     sess.run(predict_op, feed_dict={X: x_test, Y: y_test})))