zoukankan      html  css  js  c++  java
  • tensorBoard使用

    版本

    启动tensorboard
    
    tensorboard --logdir=F:python历程	ensorflow完整学习笔记	ensorflow_complete_review	ensorboard	est
    
    访问路径
    localhost:6006
    
    设置统计保存文件路径
    writer = tf.summary.FileWriter(current_dir + "//..//..//tensorboard//test",
                                   sess.graph)
    
    
    设置统计点位
    tf.summary.scalar('cross_entropy', cross_entropy)
    summaries = tf.summary.merge_all()
    summ = sess.run(summaries, feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.7})
    writer.add_summary(summ, global_step=i)
    
    import tensorflow as tf
    
    import sys
    import os
    
    current_dir = os.path.abspath(os.path.dirname(__file__))
    sys.path.append(current_dir + '/..')
    import mnist_data.input_data as input_data
    
    mnist = input_data.read_data_sets(current_dir + "/../mnist_data/MNIST_data/", one_hot=True)
    
    sess = tf.InteractiveSession()
    
    # 占位符
    x = tf.placeholder("float", shape=[None, 784])
    y_ = tf.placeholder("float", shape=[None, 10])
    
    
    # 权重初始化函数
    def weight_variable(shape):
        initial = tf.truncated_normal(shape=shape, stddev=0.1)
        return tf.Variable(initial)
    
    
    def bias_variable(shape):
        initial = tf.constant(0.1, shape=shape)
        return tf.Variable(initial)
    
    
    # 卷积池化函数
    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')
    
    
    # 第一层卷积
    # variable_scope
    with tf.name_scope('layer1'):
        w_conv1 = weight_variable([5, 5, 1, 32])  # ---32个5*5的卷积核心
        b_conv1 = bias_variable([32])
        x_image = tf.reshape(x, [-1, 28, 28, 1])
        h_conv1 = tf.nn.relu(conv2d(x_image, w_conv1) + b_conv1)
        h_pool1 = max_pool_2x2(h_conv1)
    
    # 第二层卷积
    with tf.name_scope('layer2'):
        w_conv2 = weight_variable([5, 5, 32, 64])  # ---64个5*5的卷积核心
        b_conv2 = bias_variable([64])
        h_conv2 = tf.nn.relu(conv2d(h_pool1, w_conv2) + b_conv2)
        h_pool2 = max_pool_2x2(h_conv2)
    
    # 全连接层
    with tf.name_scope('layer3'):
        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)
    
    # drop_out层
    with tf.name_scope('layer_drop_out'):
        keep_prob = tf.placeholder("float")
        h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
    
    # 输出层
    with tf.name_scope('layer_out'):
        w_fc2 = weight_variable([1024, 10])
        b_fc2 = bias_variable([10])
        y_conv = tf.nn.softmax(tf.matmul(h_fc1_drop, w_fc2) + b_fc2, name='the_result')
    
    # 训练和评估模型
    with tf.name_scope('Assessment'):
        cross_entropy = -tf.reduce_mean(y_ * tf.log(y_conv))
        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, 'float'))
        tf.summary.scalar('cross_entropy', cross_entropy)
        tf.summary.scalar('accuracy', accuracy)
    
    # Use `tf.global_variables_initializer` instead.
    
    print('----')
    print(current_dir + "//..//..//tensorboard//test")
    
    writer = tf.summary.FileWriter(current_dir + "//..//..//tensorboard//test",
                                   sess.graph)
    
    summaries = tf.summary.merge_all()
    saver = tf.train.Saver(max_to_keep=3)
    sess.run(tf.global_variables_initializer())
    model_file = tf.train.latest_checkpoint('./save')
    
    if model_file:
        saver.restore(sess, model_file)
    for i in range(10000):
        batch = mnist.train.next_batch(50)
        if i % 100 == 0:
            summ = sess.run(summaries, feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.7})
            writer.add_summary(summ, global_step=i)
    
        else:
            train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.7})
        if (i + 1) % 1000 == 0:
            saver.save(sess, "./save/train_epoch_" + str(i) + ".ckpt")
    
    print("test accuracy %g" % accuracy.eval(feed_dict={x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}))
    # localhost:6006
    # tensorboard --logdir=F:python历程	ensorflow完整学习笔记	ensorflow_complete_review	ensorboard	est
    
    
  • 相关阅读:
    Key-Value是用一个不可重复的key集合对应可重复的value集合
    集合是指一个对象容纳了多个对象
    Throw是一个语句,用来做抛出例外的功能
    对于try catch放在能够很好地处理例外的位置
    一个try可以跟进多个catch语句,用于处理不同情况,当一个try只能匹配一个catch
    Try中如果发现错误,即跳出try去匹配catch,那么try后面的语句就不会被执行
    出错的方法有可能是JDK,也可能是程序员写的程序,无论谁写的,抛出一定用throw
    JAVA会将所有的错误封装成为一个对象,其根本父类为Throwable
    匿名内部类是一种特殊的局部内部类,它是通过匿名类实现接口
    静态内部类定义在类中,任何方法外,用static定义
  • 原文地址:https://www.cnblogs.com/panfengde/p/11038420.html
Copyright © 2011-2022 走看看