Python–张量流。device spec.make_merged_spec()
原文:https://www.geesforgeks.org/python-tensorflow-device spec-make_merged_spec/
TensorFlow 是谷歌设计的开源 Python 库，用于开发机器学习模型和深度学习神经网络。
make_merged_spec 用于合并两个 DeviceSpec 的规格。
语法:张量流。device spec.make_merged_spec(dev)
参数:
- dev: 是 DeviceSpec。
返回:返回合并规格的设备规格对象。
例 1:
蟒蛇 3
# Importing the library
import tensorflow as tf
# Initializing Device Specification
device_spec = tf.DeviceSpec(job ="gfg", replica = 5)
device_spec2 = tf.DeviceSpec(device_type ="CPU")
# Printing the DeviceSpec object
print('Device Spec: ', device_spec)
print('Device Spec2: ', device_spec2)
# Getting new DeviceSpec object
new_device_spec = device_spec.make_merged_spec(device_spec2)
# Printing the result
print('New Device Spec: ', new_device_spec)
输出:
Device Spec: <tensorflow.python.framework.device_spec.DeviceSpecV2 object at 0x7fad779e4d08>
Device Spec2: <tensorflow.python.framework.device_spec.DeviceSpecV2 object at 0x7fadb9428228>
New Device Spec: <tensorflow.python.framework.device_spec.DeviceSpecV2 object at 0x7fad779e4ca8>
例 2:
蟒蛇 3
# Importing the library
import tensorflow as tf
# Initializing Device Specification
device_spec = tf.DeviceSpec(job ="gfg", replica = 5)
device_spec2 = tf.DeviceSpec(device_type ="CPU", device_index = 2)
# Printing the DeviceSpec object
print('Device Spec: ', device_spec)
print('Device Spec2: ', device_spec2)
# Getting new DeviceSpec object
new_device_spec = device_spec.make_merged_spec(device_spec2)
# Printing the result
print('New Device Spec: ', new_device_spec)
输出:
Device Spec: <tensorflow.python.framework.device_spec.DeviceSpecV2 object at 0x7fad779e4e28>
Device Spec2: <tensorflow.python.framework.device_spec.DeviceSpecV2 object at 0x7fad779e4d08>
New Device Spec: <tensorflow.python.framework.device_spec.DeviceSpecV2 object at 0x7fadb9428228>
