使用tf.Session()时未对模型进行训练

import tensorflow as tf

# 定义模型参数
W = tf.Variable([0.3], dtype=tf.float32)
b = tf.Variable([-0.3], dtype=tf.float32)

# 输入和输出占位符
x = tf.placeholder(tf.float32)
y = tf.placeholder(tf.float32)

# 模型定义
linear_model = W * x + b

# 损失函数
loss = tf.reduce_sum(tf.square(linear_model - y))

# 优化器
optimizer = tf.train.GradientDescentOptimizer(0.01)
train = optimizer.minimize(loss)

# 初始化变量
init = tf.global_variables_initializer()

# 创建会话并运行初始化
sess = tf.Session()
sess.run(init)

# 训练数据
x_train = [1, 2, 3, 4]
y_train = [0, -1, -2, -3]

# 训练循环
for epoch in range(1000):
    sess.run(train, {x: x_train, y: y_train})

# 输出训练后的参数
print(sess.run([W, b]))