In [ ]:

    

# imports
import os
import sys
import time
import numpy as np
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data

from tf_cnnvis import *

np.random.seed(10)

%load_ext autoreload
%autoreload 2


    

In [ ]:

    

# helper method to define model
def deepnn(x):
    x_image = tf.reshape(x, [-1, 28, 28, 1])

    W_conv1 = weight_variable([5, 5, 1, 32])
    b_conv1 = bias_variable([32])
    h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)

    h_pool1 = max_pool_2x2(h_conv1)

    W_conv2 = weight_variable([5, 5, 32, 64])
    b_conv2 = bias_variable([64])
    h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)

    h_pool2 = max_pool_2x2(h_conv2)

    W_fc1 = weight_variable([7 * 7 * 64, 1024])
    b_fc1 = bias_variable([1024])

    h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*64])
    h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)

    keep_prob = tf.placeholder(tf.float32)
    h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)

    W_fc2 = weight_variable([1024, 10])
    b_fc2 = bias_variable([10])

    y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2
    return x_image, y_conv, keep_prob


def conv2d(x, W):
    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')


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


def weight_variable(shape):
    initial = tf.truncated_normal(shape, stddev=0.1)
    return tf.Variable(initial)


def bias_variable(shape):
    initial = tf.constant(0.1, shape=shape)
    return tf.Variable(initial)


    

In [ ]:

    

#load graph and data and run training

tf.reset_default_graph()

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

# defining TF model
x = tf.placeholder(tf.float32, [None, 784])
y_ = tf.placeholder(tf.float32, [None, 10])
x_image, y_conv, keep_prob = deepnn(x)

cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits = y_conv, labels = y_))
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(y_,1))
accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))

# trainning CNN
sess= tf.Session()

sess.run(tf.global_variables_initializer())
with sess.as_default():
    for i in range(1000):
        batch = mnist.train.next_batch(50)
        if i%100 == 0:
            train_accuracy = accuracy.eval(feed_dict={x:batch[0], y_: batch[1], keep_prob: 1.0})
            print("step %d, training accuracy %g"%(i, train_accuracy))
        train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5})
        
    print("test accuracy %g"%accuracy.eval(feed_dict={x: mnist.test.images, y_: mnist.test.labels, 
                                                      keep_prob: 1.0}))

    feed_dict = {x:batch[0][1:2], y_: batch[1][1:2], keep_prob: 1.0}


    

In [ ]:

    

# deconv visualization
layers = ["r", "p", "c"]
total_time = 0

start = time.time()
    # api call
is_success = deconv_visualization(sess_graph_path = sess, value_feed_dict = feed_dict, 
                                  input_tensor=x_image, layers=layers, 
                                  path_logdir=os.path.join("Log","MNISTExample"), 
                                  path_outdir=os.path.join("Output","MNISTExample"))
start = time.time() - start
print("Total Time = %f" % (start))


    

In [ ]:

    

# close the session and release variables.
sess.close()


    

In [ ]:

    




    

In [ ]:

    




    

In [ ]: