XlaRecvTPUEmbeddingDeduplicationData

public final class XlaRecvTPUEmbeddingDeduplicationData

Receives deduplication data (indices and weights) from the embedding core.

The deduplication data is a Tensor with type=DT_VARIANT. The tensor itself is an XLA nested tuple containing N elements (where N is the ratio of the number of embedding to tensor cores per TPU chip). Each element of the nested tuple is a tuple of rank 1 tensors. Each tensor either contains indices (DT_UINT32) for embedding lookup on the TensorCore or weights (DT_FLOAT) to apply to the output of the embedding lookup operation.

Public Methods

Output<Object>
asOutput()
Returns the symbolic handle of a tensor.
static XlaRecvTPUEmbeddingDeduplicationData
create(Scope scope, String config)
Factory method to create a class wrapping a new XlaRecvTPUEmbeddingDeduplicationData operation.
Output<?>
output()

Inherited Methods

Public Methods

public Output<Object> asOutput ()

Returns the symbolic handle of a tensor.

Inputs to TensorFlow operations are outputs of another TensorFlow operation. This method is used to obtain a symbolic handle that represents the computation of the input.

public static XlaRecvTPUEmbeddingDeduplicationData create (Scope scope, String config)

Factory method to create a class wrapping a new XlaRecvTPUEmbeddingDeduplicationData operation.

Parameters
scope current scope
config Serialized TPUEmbeddingConfiguration proto.
Returns
  • a new instance of XlaRecvTPUEmbeddingDeduplicationData

public Output<?> output ()