Thanks for tuning in to Google I/O. View all sessions on demandWatch on demand

tensorflow:: ops:: SparseCrossHashed

#include <sparse_ops.h>

Generates sparse cross from a list of sparse and dense tensors.


The op takes two lists, one of 2D SparseTensor and one of 2D Tensor , each representing features of one feature column. It outputs a 2D SparseTensor with the batchwise crosses of these features.

For example, if the inputs are

inputs[0]: SparseTensor with shape = [2, 2]
[0, 0]: "a"
[1, 0]: "b"
[1, 1]: "c"

inputs[1]: SparseTensor with shape = [2, 1]
[0, 0]: "d"
[1, 0]: "e"

inputs[2]: Tensor [["f"], ["g"]]

then the output will be

shape = [2, 2]
[0, 0]: "a_X_d_X_f"
[1, 0]: "b_X_e_X_g"
[1, 1]: "c_X_e_X_g"

if hashed_output=true then the output will be

shape = [2, 2]
[0, 0]: FingerprintCat64(
            Fingerprint64("f"), FingerprintCat64(
                Fingerprint64("d"), Fingerprint64("a")))
[1, 0]: FingerprintCat64(
            Fingerprint64("g"), FingerprintCat64(
                Fingerprint64("e"), Fingerprint64("b")))
[1, 1]: FingerprintCat64(
            Fingerprint64("g"), FingerprintCat64(
                Fingerprint64("e"), Fingerprint64("c")))


  • scope: A Scope object
  • indices: 2-D. Indices of each input SparseTensor .
  • values: 1-D. values of each SparseTensor .
  • shapes: 1-D. Shapes of each SparseTensor .
  • dense_inputs: 2-D. Columns represented by dense Tensor .
  • num_buckets: It is used if hashed_output is true. output = hashed_valuenum_buckets if num_buckets > 0 else hashed_value.
  • strong_hash: boolean, if true, siphash with salt will be used instead of farmhash.
  • salt: Specify the salt that will be used by the siphash function.


  • Output output_indices: 2-D. Indices of the concatenated SparseTensor .
  • Output output_values: 1-D. Non-empty values of the concatenated or hashed SparseTensor .
  • Output output_shape: 1-D. Shape of the concatenated SparseTensor .

Constructors and Destructors

SparseCrossHashed (const :: tensorflow::Scope & scope, :: tensorflow::InputList indices, :: tensorflow::InputList values, :: tensorflow::InputList shapes, :: tensorflow::InputList dense_inputs, :: tensorflow::Input num_buckets, :: tensorflow::Input strong_hash, :: tensorflow::Input salt)

Public attributes


Public attributes


Operation operation


::tensorflow::Output output_indices


::tensorflow::Output output_shape


::tensorflow::Output output_values

Public functions


  const ::tensorflow::Scope & scope,
  ::tensorflow::InputList indices,
  ::tensorflow::InputList values,
  ::tensorflow::InputList shapes,
  ::tensorflow::InputList dense_inputs,
  ::tensorflow::Input num_buckets,
  ::tensorflow::Input strong_hash,
  ::tensorflow::Input salt