Update Code

02. linear-regression-with-one-variable/main.py

@@ -4,14 +4,65 @@ import tensorflow as tf
 import numpy
 import matplotlib.pyplot as pyplot
 
-numpy.random.seed()
+with tf.compat.v1.Session() as sess:
 
-# 生成 x_data
-x_data = numpy.linspace(0, 100, 500)
-# 生成 y_data
-y_data = 3.1234 * x_data + 2.98 + numpy.random.randn(*x_data.shape) * 0.9
-# 画出散点图
-pyplot.scatter(x_data, y_data)
-# 画出想要学习的目标函数
-pyplot.plot(x_data, 3.1234 * x_data + 2.98, color = "red", linewidth = 1)
-# pyplot.show()
\ No newline at end of file
+    numpy.random.seed()
+
+    # 生成 x_data
+    x_data = numpy.linspace(0, 100, 500)
+    # 生成 y_data
+    y_data = 3.1234 * x_data + 2.98 + numpy.random.randn(*x_data.shape) * 0.9
+    # 画出散点图
+    pyplot.scatter(x_data, y_data)
+    # 画出想要学习的目标函数
+    pyplot.plot(x_data, 3.1234 * x_data + 2.98, color = "red", linewidth = 1)
+    # pyplot.show()
+
+    # 构建模型
+    # https://stackoverflow.com/questions/58986126/replacing-placeholder-for-tensorflow-v2
+    x = tf.keras.Input(name = "x", shape = (), dtype = tf.dtypes.float32)
+    y = tf.keras.Input(name = "y", shape = (), dtype = tf.dtypes.float32)
+    # 定义模型
+    def model(x, w, b):
+        return tf.multiply(x, w) + b
+    # 定义模型结构
+    w = tf.Variable(2.5, name = "w0")
+    b = tf.Variable(2.0, name = "b0")
+    # predict 是预测值，前向计算
+    predict = model(x, w, b)
+
+    # 训练模型
+    # 迭代次数，即训练轮数
+    train_rounds = 10
+    # 设置学习率
+    learning_rate = 0.0251
+    # 定义损失函数，本处采用均方差
+    # 交叉熵也是一个常见的 L2 损失函数
+    loss_function = tf.reduce_mean(tf.square(y - predict))
+    # 定义优化器
+    # 本处初始化梯度下降优化器
+    # https://stackoverflow.com/questions/55682718/module-tensorflow-api-v2-train-has-no-attribute-gradientdescentoptimizer
+    # https://stackoverflow.com/questions/58722591/typeerror-minimize-missing-1-required-positional-argument-var-list
+    optimizer = tf.optimizers.SGD(learning_rate).minimize(loss_function, var_list=[w, b])
+
+    # 开始训练
+    # 记录训练步数
+    step = 0
+    # 用于保存loss值的列表
+    loss_list = []  
+    # 设置迭代轮次
+    for epoch in range(train_rounds):
+        for xs, ys in zip(x_data, y_data):
+            _, loss = sess.run([optimizer, loss_function], feed_dict = {x: xs, y: ys})
+            # 将每一次的损失值放入列表
+            loss_list.append(loss)
+            step = step + 1
+            # 每训练20个样本输出一次损失值
+            display_step = 20
+            if step % display_step == 0:
+                print("Train Epoch: %02d" % (epoch + 1),"Step: %04d" % step, "loss = ", "{:.9f}".format(loss))
+        # 每一轮训练完，有一个w,b值
+        b0temp = b.eval(session = sess)
+        w0temp = w.eval(session = sess)
+        # 画出训练模型
+        pyplot.plot(x_data, w0temp * x_data + b0temp)