如何在Tensorflow中定义具有已知形状的标量占位符

import tensorflow as tf

# 定义标量占位符（形状为 []）
scalar_placeholder = tf.placeholder(dtype=tf.float32, shape=[])

# 使用示例
with tf.Session() as sess:
    # 传入标量值（如 3.14）
    output = sess.run(scalar_placeholder, feed_dict={scalar_placeholder: 3.14})
    print(output)  # 输出: 3.14

import tensorflow.compat.v1 as tf
tf.disable_v2_behavior()  # 禁用 TF2.x 的默认行为

# 定义标量占位符
scalar_placeholder = tf.placeholder(dtype=tf.float32, shape=[])

# 使用示例
with tf.Session() as sess:
    output = sess.run(scalar_placeholder, feed_dict={scalar_placeholder: 5.0})
    print(output)  # 输出: 5.0

import tensorflow as tf

# 定义标量占位符和简单操作
a = tf.placeholder(tf.float32, shape=[], name='input_scalar')
b = a * 2

# 运行计算图
with tf.Session() as sess:
    result = sess.run(b, feed_dict={a: 10.5})
    print(result)  # 输出: 21.0