tf.nn.conv2d(输入描述 卷积核描述核滑动步长padding)

 tf.nn.dropout(上层输出，暂时舍弃的概率）

if train: 输出 = tf.nn.dropout(上层输出，暂时舍弃的概率）

# coding:utf-8
import tensorflow as tf
# 每个图片的像素点为28
IMAGE_SIZE = 28
# 由于是灰度图，每个图片的通道为1
NUM_CHANNELS = 1
# 第一层卷积核的大小为5
CONV1_SIZE = 5
# 第一层使用了32个卷积核
CONV1_KERNEL_NUM = 32
# 第二层卷积核大小为5
CONV2_SIZE = 5
# 第二层使用了64个卷积核
CONV2_KERNEL_NUM = 64
# 第一层神经网络有512个神经元
FC_SIZE = 512
# 第二层神经网络有10个神经元
OUTPUT_NODE = 10# 权重w的生成函数，
# shape表示生成张量的维度
# regularizer表示正则化权重
def get_weight(shape, regularizer):w = tf.Variable(tf.truncated_normal(shape, stddev=0.1))if regularizer != None: tf.add_to_collection('losses', tf.contrib.layers.l2_regularizer(regularizer)(w)) return w# 偏置b生成函数
def get_bias(shape):b = tf.Variable(tf.zeros(shape))  return b# 卷积生成函数
def conv2d(x, w):# x输入图片描述[batch, 行分辨率，列分辨率，通道数]# w卷积核描述[行分辨率，列分辨率，通道数，卷积核个数]# strides 核滑动步长[1，行步长，列步长，1]# padding 填充模式  SAME零填充return tf.nn.conv2d(x, w, strides=[1, 1, 1, 1], padding='SAME')# 最大池化计算函数
def max_pool_2x2(x):# x 输入描述[batch，行分辨率，列分辨率，通道数]# ksize 池化核描述[1，行分辨率，列分辨率，1]# strides 池化核滑动步长[1，行步长，列步长，1]# padding 填充模式return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME') # 前向传播的网络结构
def forward(x, train, regularizer):# 第一层卷积核conv1_w = get_weight([CONV1_SIZE, CONV1_SIZE, NUM_CHANNELS, CONV1_KERNEL_NUM], regularizer) # 初始化第一层偏置项bconv1_b = get_bias([CONV1_KERNEL_NUM]) # 执行卷积计算conv1 = conv2d(x, conv1_w) # 对conv1添加偏置，并通过relu激活函数进行激活输出relu1 = tf.nn.relu(tf.nn.bias_add(conv1, conv1_b)) # 最大池化，第一层最终输出pool1pool1 = max_pool_2x2(relu1) # 第二层卷积核的深度CONV1_KERNEL_NUM等于上层卷积核的个数conv2_w = get_weight([CONV2_SIZE, CONV2_SIZE, CONV1_KERNEL_NUM, CONV2_KERNEL_NUM],regularizer) conv2_b = get_bias([CONV2_KERNEL_NUM])# 第二层卷积核的输入是第一层的输出conv2 = conv2d(pool1, conv2_w) relu2 = tf.nn.relu(tf.nn.bias_add(conv2, conv2_b))# pool2是第二层的最后输出pool2 = max_pool_2x2(relu2)# 将pool2从三维张量变成二维张量# 获得pool2的数据维度，并存入list中pool_shape = pool2.get_shape().as_list() # pool_shape[0] 的值是一个batch值# pool_shape[1] 特征的长度# pool_shape[2] 特征的宽度# pool_shape[3] 特征的深度# nodes=长度*宽度*深度，所有数据点的个数nodes = pool_shape[1] * pool_shape[2] * pool_shape[3] # 将pool2表示成以batch作为行，所有特征点为列的二维形状，# 把reshaped 喂入全连接网络中reshaped = tf.reshape(pool2, [pool_shape[0], nodes]) fc1_w = get_weight([nodes, FC_SIZE], regularizer) fc1_b = get_bias([FC_SIZE]) # 把上层的输出reshaped 乘以 本层线上的权重fc1_w，再加上偏置fc1_bfc1 = tf.nn.relu(tf.matmul(reshaped, fc1_w) + fc1_b) # 如果是训练阶段，则对fc1的50%进行dropoutif train: fc1 = tf.nn.dropout(fc1, 0.5)# 通过第二层全连接网络，初始化第二层W和bfc2_w = get_weight([FC_SIZE, OUTPUT_NODE], regularizer)fc2_b = get_bias([OUTPUT_NODE])# 上层的输出fc1和本层fc2_w相乘，再加上fc2_by = tf.matmul(fc1, fc2_w) + fc2_breturn y 

#coding:utf-8
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
import mnist_lenet5_forward
import os
import numpy as npBATCH_SIZE = 100
LEARNING_RATE_BASE =  0.005
LEARNING_RATE_DECAY = 0.99
REGULARIZER = 0.0001
STEPS = 50000
MOVING_AVERAGE_DECAY = 0.99
MODEL_SAVE_PATH="./model/"
MODEL_NAME="mnist_model" def backward(mnist):x = tf.placeholder(tf.float32,[# 每轮喂入的图片数量
    BATCH_SIZE,# 行分辨率
    mnist_lenet5_forward.IMAGE_SIZE,# 列分辨率
    mnist_lenet5_forward.IMAGE_SIZE,# 输入的通道数
    mnist_lenet5_forward.NUM_CHANNELS]) y_ = tf.placeholder(tf.float32, [None, mnist_lenet5_forward.OUTPUT_NODE])# True 意味使用dropout操作y = mnist_lenet5_forward.forward(x,True, REGULARIZER) global_step = tf.Variable(0, trainable=False) ce = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=y, labels=tf.argmax(y_, 1))cem = tf.reduce_mean(ce) loss = cem + tf.add_n(tf.get_collection('losses')) learning_rate = tf.train.exponential_decay( LEARNING_RATE_BASE,global_step,mnist.train.num_examples / BATCH_SIZE, LEARNING_RATE_DECAY,staircase=True) train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss, global_step=global_step)ema = tf.train.ExponentialMovingAverage(MOVING_AVERAGE_DECAY, global_step)ema_op = ema.apply(tf.trainable_variables())with tf.control_dependencies([train_step, ema_op]): train_op = tf.no_op(name='train')saver = tf.train.Saver() with tf.Session() as sess: init_op = tf.global_variables_initializer() sess.run(init_op) ckpt = tf.train.get_checkpoint_state(MODEL_SAVE_PATH) if ckpt and ckpt.model_checkpoint_path:saver.restore(sess, ckpt.model_checkpoint_path) for i in range(STEPS):xs, ys = mnist.train.next_batch(BATCH_SIZE) reshaped_xs = np.reshape(xs,(  BATCH_SIZE,mnist_lenet5_forward.IMAGE_SIZE,mnist_lenet5_forward.IMAGE_SIZE,mnist_lenet5_forward.NUM_CHANNELS))_, loss_value, step = sess.run([train_op, loss, global_step], feed_dict={x: reshaped_xs, y_: ys}) if i % 100 == 0: print("After %d training step(s), loss on training batch is %g." % (step, loss_value))saver.save(sess, os.path.join(MODEL_SAVE_PATH, MODEL_NAME), global_step=global_step)def main():mnist = input_data.read_data_sets("./data/", one_hot=True) backward(mnist)if __name__ == '__main__':main()

#coding:utf-8
import time
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
import mnist_lenet5_forward
import mnist_lenet5_backward
import numpy as npTEST_INTERVAL_SECS = 5def test(mnist):with tf.Graph().as_default() as g: x = tf.placeholder(tf.float32,[mnist.test.num_examples,mnist_lenet5_forward.IMAGE_SIZE,mnist_lenet5_forward.IMAGE_SIZE,mnist_lenet5_forward.NUM_CHANNELS]) y_ = tf.placeholder(tf.float32, [None, mnist_lenet5_forward.OUTPUT_NODE])y = mnist_lenet5_forward.forward(x,False,None)ema = tf.train.ExponentialMovingAverage(mnist_lenet5_backward.MOVING_AVERAGE_DECAY)ema_restore = ema.variables_to_restore()saver = tf.train.Saver(ema_restore)correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) while True:with tf.Session() as sess:ckpt = tf.train.get_checkpoint_state(mnist_lenet5_backward.MODEL_SAVE_PATH)if ckpt and ckpt.model_checkpoint_path:saver.restore(sess, ckpt.model_checkpoint_path)global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1] reshaped_x = np.reshape(mnist.test.images,(mnist.test.num_examples,mnist_lenet5_forward.IMAGE_SIZE,mnist_lenet5_forward.IMAGE_SIZE,mnist_lenet5_forward.NUM_CHANNELS))accuracy_score = sess.run(accuracy, feed_dict={x:reshaped_x,y_:mnist.test.labels}) print("After %s training step(s), test accuracy = %g" % (global_step, accuracy_score))else:print('No checkpoint file found')returntime.sleep(TEST_INTERVAL_SECS) def main():mnist = input_data.read_data_sets("./data/", one_hot=True)test(mnist)if __name__ == '__main__':main()