Tensorflow框架之AlexNet

 from datetime import datetime
 import math
 import time
 import tensorflow as tf
 batch_size=32
 num_batches=100
 n_output=100
 #定义显示节点的函数
 def print_activations(t):
     print(t.op.name, ' ',t.get_shape().as_list())

 #定义inference函数:该函数接受图像作为输入，返回最后一层pool5及相关参数
 def inference(images):
     parameters=[]
     #设置第一层卷积操作
     with tf.name_scope('conv1') as scope:
         #生成权重变量
         kernel=tf.Variable(tf.truncated_normal([11,11,3,64],dtype=tf.float32,stddev=1e-1),name='weights')
         #做卷积操作
         conv=tf.nn.conv2d(images,kernel,[1,4,4,1],padding='SAME')
         #b定义偏置值初始化为0
         biases=tf.Variable(tf.constant(0.0,shape=[64],dtype=tf.float32),trainable=True,name='biases')
         bias=tf.nn.bias_add(conv,biases)
         conv1=tf.nn.relu(bias,scope)
     print_activations(conv1)
     parameters+=[kernel,biases]
     #添加LRN层与最大池化层
     lrn1=tf.nn.lrn(conv1,4,bias=1.0,alpha=0.001/9,beta=0.75,name='lrn1')
     pool1=tf.nn.max_pool(lrn1,ksize=[1,3,3,1],strides=[1,2,2,1],padding='VALID',name='pool1')
     print_activations(pool1)
     #设置第二层卷积操作

     with tf.name_scope('conv2') as scope:
         kernel=tf.Variable(tf.truncated_normal([5,5,64,192],dtype=tf.float32,stddev=1e-1),name='weights')
         conv=tf.nn.conv2d(pool1,kernel,[1,1,1,1],padding='SAME')
         biases=tf.Variable(tf.constant(0.0,dtype=tf.float32,shape=[192]),trainable=True,name='biases')
         bias=tf.nn.bias_add(conv,biases)
         conv2=tf.nn.relu(bias,name=scope)
         parameters+=[kernel,biases]
     print_activations(conv2)
     #添加LRN层与最大池化层
     lrn2=tf.nn.lrn(conv2,4,bias=1.0,alpha=0.001/9,beta=0.75,name='lrn2')
     pool2=tf.nn.max_pool(lrn2,ksize=[1,3,3,1],strides=[1,2,2,1],padding='VALID',name='pool2')
     print_activations(pool2)

     #设置第三层卷积神经网络
     with tf.name_scope('conv3') as scope:
         kernel=tf.Variable(tf.truncated_normal(shape=[3,3,192,384],stddev=1e-1,dtype=tf.float32),name='weights')
         conv=tf.nn.conv2d(pool2,kernel,[1,1,1,1],padding='SAME')
         biases=tf.Variable(tf.constant(0.0,dtype=tf.float32,shape=[384]),trainable=True,name='biases')
         bias=tf.nn.bias_add(conv,biases)
         conv3=tf.nn.relu(bias,name=scope)
         parameters+=[kernel,biases]
     print_activations(conv3)

 #设置设置第四层卷积神经网络
     with tf.name_scope('conv4') as scope:
         kernel=tf.Variable(tf.truncated_normal(shape=[3,3,384,256],stddev=1e-1,dtype=tf.float32),name='weights')
         conv=tf.nn.conv2d(conv3,kernel,[1,1,1,1],padding='SAME')
         biases=tf.Variable(tf.constant(0.0,dtype=tf.float32,shape=[256]),trainable=True,name='biases')
         bias=tf.nn.bias_add(conv,biases)
         conv4=tf.nn.relu(bias,name=scope)
         parameters+=[kernel,biases]
     print_activations(conv4)

 #设置设置第五层卷积神经网络
     with tf.name_scope('conv5') as scope:
         kernel=tf.Variable(tf.truncated_normal(shape=[3,3,256,256],stddev=1e-1,dtype=tf.float32),name='weights')
         conv=tf.nn.conv2d(conv4,kernel,[1,1,1,1],padding='SAME')
         biases=tf.Variable(tf.constant(0.0,dtype=tf.float32,shape=[256]),trainable=True,name='biases')
         bias=tf.nn.bias_add(conv,biases)
         conv5=tf.nn.relu(bias,name=scope)
         parameters+=[kernel,biases]
     print_activations(conv5)
     pool5=tf.nn.max_pool(conv5,ksize=[1,3,3,1],strides=[1,2,2,1],padding='VALID',name='pool5')
     print_activations(pool5)
     return pool5, parameters
 #设置全连接层
 def all_contact(pool5,keep_prob):
     pool_shape=pool5.get_shape().as_list()
     nodes=[-1,pool_shape[1]*pool_shape[2]*pool_shape[3]]
     densel=tf.reshape(pool5,nodes)
     with tf.name_scope('fc1'):
         w1=tf.Variable(tf.truncated_normal([9216,1024],stddev=1e-1),name='w1')
         b1=tf.Variable(tf.constant(0.0,tf.float32,shape=[1024]),trainable=True,name='b1')
         fc1=tf.nn.relu(tf.nn.bias_add(tf.matmul(densel,w1),b1))
         #设置dropout层
         fc1_drop=tf.nn.dropout(fc1,keep_prob)
         print_activations(fc1_drop)

     with tf.name_scope('fc2'):
         w2=tf.Variable(tf.truncated_normal([1024,1024],stddev=1e-1),name='w2')
         b2=tf.Variable(tf.constant(0.0,tf.float32,shape=[1024]),trainable=True,name='b1')
         fc2=tf.nn.relu(tf.nn.bias_add(tf.matmul(fc1_drop,w2),b2))
          #设置dropout层
         fc2_drop=tf.nn.dropout(fc2,keep_prob)
         print_activations(fc2_drop)

     with tf.name_scope('fc3'):
         w3=tf.Variable(tf.truncated_normal([1024,n_output],stddev=1e-1),name='w3')
         b3=tf.Variable(tf.constant(0.0,tf.float32,shape=[n_output]),trainable=True,name='b1')
         fc3=tf.nn.relu(tf.nn.bias_add(tf.matmul(fc2_drop,w3),b3))
         print_activations(fc3)
     return fc3

 def time_tensorflow_run(session,target,info_string):
     num_steps_burn_in=10
     total_duration=0.0
     total_duration_squared=0.0
     for i in range(num_batches+num_steps_burn_in):
         start_time=time.time()
         _=session.run(target)
         duration=time.time()-start_time
         if i>=num_steps_burn_in:
             if not i %10:
                 print('%s: step %d. duration=%.3f'%(datetime.now(),i-num_steps_burn_in,duration))
             total_duration+=duration
             total_duration_squared+=duration*duration
     mn=total_duration/num_batches
     vr=total_duration_squared/num_batches-mn*mn
     sd=math.sqrt(vr)
     print('%s:%s across %d steps, %.3f+/-%.3f sec / batch'%(datetime.now(),info_string,num_batches,mn,sd))

 def run_benchmark():
     with tf.Graph().as_default():
         image_size=224
         images=tf.Variable(tf.random_normal([batch_size,image_size,image_size,3],dtype=tf.float32,stddev=1e-1))
         pool5,parameters=inference(images)
         all_contact(pool5,1.0)
         init=tf.global_variables_initializer()
         sess=tf.Session()
         sess.run(init)
         time_tensorflow_run(sess,pool5,"Forward")
         objective=tf.nn.l2_loss(pool5)
         grad=tf.gradients(objective,parameters)
         time_tensorflow_run(sess,grad,"Forward-backward")
 run_benchmark()