Class LogicalDevice

tensorflow/python/eager/context.py:240–254

Abstraction for a device initialized by the runtime. A LogicalDevice corresponds to a initialized instance on a PhysicalDevice or a remote device available in the cluster. Tensors and operations can be placed on a specific LogicalDevice by calling `tf.device()` with the `name` of the Logica

class LogicalDevice(
    collections.namedtuple("LogicalDevice", ["name", "device_type"])):
  """Abstraction for a device initialized by the runtime.

  A LogicalDevice corresponds to a initialized instance on a PhysicalDevice or a
  remote device available in the cluster. Tensors and operations can be placed
  on a specific LogicalDevice by calling `tf.device()` with the `name` of the
  LogicalDevice.

  Fields:
    name: The fully qualified name of the device. Can be used for Op or function
      placement.
    device_type: String declaring the type of device such as "CPU" or "GPU".
  """
  pass


@tf_export("config.experimental.VirtualDeviceConfiguration")