Python – tensorflow.DeviceSpec.replace()

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。

replace()用于覆盖 DeviceSpec 对象的规范并获取新对象。

Syntax: tensorflow.DeviceSpec.replace(**kwargs)

Parameters:

• **kwargs: This method accepts all the parameters accepted by DeviceSpec constructor.

Returns: It returns a DeviceSpec object.

示例 1：

Python3

# Importing the library
import tensorflow as tf
  
# Initializing Device Specification
device_spec = tf.DeviceSpec(job ="gfg", replica = 5)
  
# Printing the DeviceSpec 
print('Device Spec: ', device_spec.to_string())
  
# Getting new DeviceSpec object
new_device_spec = device_spec.replace(job = "gfg2")
  
# Printing the result
print('New Device Spec: ', new_device_spec.to_string())

输出：

Device Spec:  /job:gfg/replica:5
New Device Spec:  /job:gfg2/replica:5

示例 2：

Python3

# Importing the library
import tensorflow as tf
  
# Initializing Device Specification
device_spec = tf.DeviceSpec(job ="gfg", replica = 5)
  
# Printing the DeviceSpec 
print('Device Spec: ', device_spec.to_string())
  
# Getting new DeviceSpec object
new_device_spec = device_spec.replace(job = "gfg2", device_type = "CPU")
  
# Printing the result
print('New Device Spec: ', new_device_spec.to_string())

输出：

Device Spec:  /job:gfg/replica:5
New Device Spec:  /job:gfg2/replica:5/device:CPU:*