在线性回归中添加变量显示

在 TensorBoard 中观察损失模型的参数，损失值等变量的变化。

一、实现步骤

• 1.创建事件文件

file_writer = tf.summary.FileWrite('e:/events/test',graph=sess.graph)

• 2.收集变量
  收集对于损失函数和准确率等单值变量使用 tf.summary.scalar(name=’’,tensor)，收集高维 度变量参数使用 tf.summary.histogram(name=’’,tensor)，收集输入的图片张量能显示图片使用 tf.summary.image(name=’’,tensor)，其中 name 为变量的名字，tensor 为值。使用示例如下：

tf.summary.scalar(‘error’,error) 
tf.summary.histogram('weights',weight) 
tf.summary.histogram('bias',bias)

• 3.合并变量

merged = tf.summary.merge_all()

• 4.运行合并变量

summary = sess.run(merged)

• 5.将 summary 写入事件文件

file_writer.add_summary(summary,i)

二、实例代码

import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import tensorflow as tf
def linear_regression():
    # 1.Prepare data
    X = tf.random_normal(shape=[100,1])
    y_true = tf.matmul(X,[[0.8]]) + 0.7
    # Construct weights and bias, use variables to create
    weight = tf.Variable(initial_value=tf.random_normal(shape=[1,1]))
    bias = tf.Variable(initial_value=tf.random_normal(shape=[1,1]))
    y_predict = tf.matmul(X,weight) + bias
    # 2.Construct loss function
    error = tf.reduce_mean(tf.square(y_predict-y_true))
    # 3.Optimization loss
    optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.1).minimize(error)
    # (2)Increase variable display, collect variables
    tf.summary.scalar('error',error)
    tf.summary.histogram('weights',weight)
    tf.summary.histogram('bias',bias)
    # (3)Increase variable display, merge variables
    merged = tf.summary.merge_all()
    # Initialize variables
    init = tf.global_variables_initializer()
    # Start conversation
    with tf.Session() as sess:
        # Run initialization variables
        sess.run(init)
        print('View model parameters before training: weight: %f, partial amount: %f, loss: %f'%(weight.eval(),bias.eval(),error.eval()))
        # (1)Add variable display, create text events
        file_Writer = tf.summary.FileWriter('e:/events/test',graph=sess.graph)

        # Start training
        for i in range(100):
            sess.run(optimizer)
            print('View model parameters after training %d times: weight: %f, partial amount: %f, loss: %f'%((i+1), weight.eval(), bias.eval(), error.eval()))
            # (4)Increase variable display, run merge variable
            summary = sess.run(merged)
            # (5)Write variables to event file
            file_Writer.add_summary(summary,i)

if __name__ == '__main__':
    linear_regression()

三、运行结果

四、变量可视化

• 1.打开 CMD ，输入命令：tensorboard --logdir="e:/events/test",结果如下：
  
• 2.打开浏览器，输入http://localhost:6006/,结果如下：