Retrieve multiple values from the computation outfeed. Device ordinal is a
Compat aliases for migration
See Migration guide for more details.
tf.raw_ops.OutfeedDequeueTupleV2( device_ordinal, dtypes, shapes, name=None )
tensor allowing dynamic outfeed.
This operation will block indefinitely until data is available. Output
corresponds to XLA tuple element
Tensor of type
An int scalar tensor, representing the TPU device to use. This should be -1 when
the Op is running on a TPU device, and >= 0 when the Op is running on the CPU
dtypes: A list of
tf.DTypes that has length
The element types of each element in
shapes: A list of shapes (each a
tf.TensorShape or list of
The shapes of each tensor in
name: A name for the operation (optional).
A list of
Tensor objects of type