基于《21个项目玩转深度学习》课程的项目实例源码.zip

import tensorflow as tf from tensorflow.examples.tutorials.mnist import input_data # disable debugging logs tf.logging.set_verbosity(tf.logging.ERROR) # tf.logging.WARN # import data set mnist = input_data.read_data_sets("data\MNIST", one_hot=True) # set placeholders x = tf.placeholder(tf.float32, [None, 784]) y_ = tf.placeholder(tf.float32, [None, 10]) # vector to matrix x_image = tf.reshape(x, [-1, 28, 28, 1]) # generating weight def weight_variable(shape): initial = tf.truncated_normal(shape, stddev=0.1) return tf.Variable(initial) # generating bias def bias_variable(shape): initial = tf.constant(0.1, shape=shape) return tf.Variable(initial) # convolution def conv2d(x, W): return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME') # pooling def max_pool_2x2(x): return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME') # convolution layer 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) # pooling layer 1 h_pool1 = max_pool_2x2(h_conv1) # convolution layer 2 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) # pooling layer 2 h_pool2 = max_pool_2x2(h_conv2) # matrix to vector h_pool2_flat = tf.reshape(h_pool2, [-1, 7 * 7 * 64]) # full connect layer 1 W_fc1 = weight_variable([7 * 7 * 64, 1024]) b_fc1 = bias_variable([1024]) h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1) # set dropout keep_prob = tf.placeholder(tf.float32) # connect to noting, a parameter of session must be a placeholder h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob) # full connect layer 2 W_fc2 = weight_variable([1024, 10]) b_fc2 = bias_variable([10]) y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2 # set cross entropy loss function cross_entropy = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(labels=y_, logits=y_conv)) # set gradient descent optimizer train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy) # definite accuracy correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) # create session sess = tf.InteractiveSession() # initialize variables tf.global_variables_initializer().run() # 1000 times gradient descent for _ in range(20000): batch = mnist.train.next_batch(50) # print accuracy per 100 times if _ % 100 == 0: train_accuracy = accuracy.eval(feed_dict={x: batch[0], y_: batch[1], keep_prob: 1.0}) print("step %d, training accuracy %g" % (_, train_accuracy)) train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5}) # print testing accuracy print("testing accuracy %g" % accuracy.eval(feed_dict={x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}))