TensorFlow是Google設計的開源Python庫,用於開發機器學習模型和深度學習神經網絡。
parse_from_string()用於將DeviceSpec名稱解析為其組件。
用法:tensorflow.DeviceSpec.parse_from_string( spec )
參數:
- spec:它是格式為/job:/副本:/task:/device:CPU:或/job:/副本:/task:/device:GPU的字符串,所有字段均為可選。
返回值:它返回一個DeviceSpec對象。
範例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.parse_from_string("/GPU:0")
# Printing the result
print('New Device Spec:', new_device_spec.to_string())
輸出:
Device Spec: /job:gfg/replica:5 New Device Spec: /device:GPU:0
範例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.parse_from_string("replica:2 / GPU:0")
# Printing the result
print('New Device Spec:', new_device_spec.to_string())
輸出:
Device Spec: /job:gfg/replica:5 New Device Spec: /replica:2/device:GPU:0
注:本文由純淨天空篩選整理自 Python – tensorflow.DeviceSpec.parse_from_string()。非經特殊聲明,原始代碼版權歸原作者所有,本譯文未經允許或授權,請勿轉載或複製。