TensorFlow 02——简单演练

TensorFlow 的基本要素

• tf.Session() 是最重要的会话结构

• tf.Variable 是变量

• tf.placeholder 是占位符，给变量类型和size

• run初始化，有Variable就必须初始化

  init = tf.global_variables_initializer()
  sess = tf.Session()
  sess.run(init)
  

• run训练，每执行一次训练一步，要给变量

  sess.run(train_step, feed_dict={xs: x_data, ys: y_data})
  

• 输出一个tensor的值，需要用run才行

  sess.run(loss, feed_dict={xs: x_data, ys: y_data})
  

代码：

# -*- coding: utf-8 -*-
# @Software: PyCharm Tensorflow1.13.1 python3.5.6
import tensorflow.compat.v1 as tf
tf.disable_v2_behavior()
import numpy as np

def add_layer(inputs, in_size, out_size, activation_function=None):
    weights = tf.Variable(tf.random_normal([in_size,out_size]))
    biases = tf.Variable(tf.zeros([1,out_size]) + 0.1)
    wx_plus_b = tf.matmul(inputs,weights) + biases
    if activation_function is None:
        outputs = wx_plus_b
    else:
        outputs = activation_function(wx_plus_b)
    return outputs


def main():
    x_data = np.linspace(-1,1,300)[:,np.newaxis]
    noise = np.random.normal(0,0.05,x_data.shape)
    y_data = np.square(x_data) - 0.5 + noise

    xs = tf.placeholder(tf.float32, [None, 1])
    ys = tf.placeholder(tf.float32, [None, 1])

    l1 = add_layer(xs, 1, 10, activation_function=tf.nn.relu)
    preduction = add_layer(l1, 10, 1, activation_function=None)

    loss = tf.reduce_mean(tf.reduce_sum(tf.square(ys - preduction),axis=1)) # 按行求和
    train_step = tf.train.GradientDescentOptimizer(0.1).minimize(loss)
    init = tf.global_variables_initializer()
    sess = tf.Session()
    sess.run(init)

    for i in range(1001):
        sess.run(train_step, feed_dict={xs: x_data, ys: y_data})
        if i % 100 == 0:
            print("第%d步的误差为%f"%(i, sess.run(loss, feed_dict={xs: x_data, ys: y_data})))


if __name__ == '__main__':
    main()