Python – tensorflow.cond()

TensorFlow 是由 Google 设计的开源Python库，用于开发机器学习模型和深度学习神经网络。

cond( ) 如果谓词 pred 为真，则返回 true_fn()，否则返回 false_fn()。

Syntax: tensorflow.cond( pred, true_fn, false_fn, name )

Parameters:

• pred: It is a scalar which determines the callable to return
• true_fn(optional): It is returned when pred is true.
• false_fn(optional): It is returned when pred is false.
• name(optional): It defines the name for the operation.

Return: It returns the result evaluated by callable.

示例 1：

Python3

# Importing the library
import tensorflow as tf
  
# Initializing the input
x = 5
y = 10
  
  
# Printing the input
print('x: ', x)
print('y: ', y)
  
# Calculating result
res = tf.cond(x < y, lambda: tf.add(x, y), lambda: tf.square(y))
  
# Printing the result
print('Result: ', res)

输出：

x:  5
y:  10
Result:  tf.Tensor(15, shape=(), dtype=int32)

示例 2：

Python3

# Importing the library
import tensorflow as tf
  
# Initializing the input
x = 5
y = 10
  
  
# Printing the input
print('x: ', x)
print('y: ', y)
  
# Calculating result
res = tf.cond(x > y, lambda: tf.add(x, y), lambda: tf.square(y))
  
# Printing the result
print('Result: ', res)

输出：

x:  5
y:  10
Result:  tf.Tensor(100, shape=(), dtype=int32)