Python tensorflow.unsorted_segment_sum() Examples

def testGradient(self):
    num_cols = 2
    indices_flat = np.array([0, 4, 0, 8, 3, 8, 4, 7, 7, 3])
    num_segments = max(indices_flat) + 3
    for indices in indices_flat, indices_flat.reshape(5, 2):
      shape = indices.shape + (num_cols,)
      with self.test_session(use_gpu=self.use_gpu):
        tf_x, np_x = self._input(shape, dtype=tf.float64)
        s = tf.unsorted_segment_sum(data=tf_x,
                                    segment_ids=indices,
                                    num_segments=num_segments)
        jacob_t, jacob_n = tf.test.compute_gradient(
            tf_x,
            shape,
            s,
            [num_segments, num_cols],
            x_init_value=np_x.astype(np.double),
            delta=1)
      self.assertAllClose(jacob_t, jacob_n, rtol=1e-3, atol=1e-3) 

def combine(self, x):
    """Return the output from the experts.

    When one example goes to multiple experts, the outputs are summed.

    Args:
      x: a Tensor with shape [batch, num_experts, expert_capacity, depth]

    Returns:
      a `Tensor` with shape `[batch, length, depth]
    """
    depth = tf.shape(x)[-1]
    x *= tf.expand_dims(self._nonpadding, -1)
    ret = tf.unsorted_segment_sum(
        x, self._flat_indices, num_segments=self._batch * self._length)
    ret = tf.reshape(ret, [self._batch, self._length, depth])
    return ret 

def unsorted_segment_log_softmax(logits, segment_ids, num_segments):
    """Perform an unsorted segment safe log_softmax."""
    # Note: if a segment is empty, the smallest value for the score will be returned,
    # which yields the correct behavior
    max_per_segment = tf.unsorted_segment_max(data=logits,
                                              segment_ids=segment_ids,
                                              num_segments=num_segments)
    scattered_maxes = tf.gather(params=max_per_segment,
                                indices=segment_ids)
    recentered_scores = logits - scattered_maxes
    exped_recentered_scores = tf.exp(recentered_scores)

    per_segment_sums = tf.unsorted_segment_sum(exped_recentered_scores, segment_ids, num_segments)
    per_segment_normalization_consts = tf.log(per_segment_sums)

    log_probs = recentered_scores - tf.gather(params=per_segment_normalization_consts, indices=segment_ids)
    return log_probs 

def _deduplicate_indexed_slices(values, indices):
    """Sums `values` associated with any non-unique `indices`.
    Args:
      values: A `Tensor` with rank >= 1.
      indices: A one-dimensional integer `Tensor`, indexing into the first
      dimension of `values` (as in an IndexedSlices object).
    Returns:
      A tuple of (`summed_values`, `unique_indices`) where `unique_indices` is a
      de-duplicated version of `indices` and `summed_values` contains the sum of
      `values` slices associated with each unique index.
    """
    unique_indices, new_index_positions = tf.unique(indices)
    summed_values = tf.unsorted_segment_sum(
      values, new_index_positions,
      tf.shape(unique_indices)[0])
    return (summed_values, unique_indices) 

def combine(self, x):
    """Return the output from the experts.

    When one example goes to multiple experts, the outputs are summed.

    Args:
      x: a Tensor with shape [batch, num_experts, expert_capacity, depth]

    Returns:
      a `Tensor` with shape `[batch, length, depth]
    """
    depth = tf.shape(x)[-1]
    x *= tf.expand_dims(self._nonpadding, -1)
    ret = tf.unsorted_segment_sum(
        x, self._flat_indices, num_segments=self._batch * self._length)
    ret = tf.reshape(ret, [self._batch, self._length, depth])
    return ret 

def apply_attention(attn_scores, states, length, is_self=False, with_sentinel=True, reuse=False, seq2_to_seq1=None):
    attn_scores += tf.expand_dims(misc.mask_for_lengths(length, tf.shape(attn_scores)[2]), 1)
    softmax = tf.nn.softmax if seq2_to_seq1 is None else lambda x: segment.segment_softmax(x, seq2_to_seq1)
    if is_self:
        # exclude attending to state itself
        attn_scores += tf.expand_dims(tf.diag(tf.fill([tf.shape(attn_scores)[1]], -1e6)), 0)
    if with_sentinel:
        with tf.variable_scope('sentinel', reuse=reuse):
            s = tf.get_variable('score', [1, 1, 1], tf.float32, tf.zeros_initializer())
        s = tf.tile(s, [tf.shape(attn_scores)[0], tf.shape(attn_scores)[1], 1])
        attn_probs = softmax(tf.concat([s, attn_scores], 2))
        attn_probs = attn_probs[:, :, 1:]
    else:
        attn_probs = softmax(attn_scores)
    attn_states = tf.einsum('abd,adc->abc', attn_probs, states)
    if seq2_to_seq1 is not None:
        attn_states = tf.unsorted_segment_sum(attn_states, seq2_to_seq1, tf.reduce_max(seq2_to_seq1) + 1)
    return attn_scores, attn_probs, attn_states 

def combine(self, expert_out, multiply_by_gates=True):
    """Sum together the expert output, weighted by the gates.

    The slice corresponding to a particular batch element `b` is computed
    as the sum over all experts `i` of the expert output, weighted by the
    corresponding gate values.  If `multiply_by_gates` is set to False, the
    gate values are ignored.

    Args:
      expert_out: a list of `num_experts` `Tensor`s, each with shape
        `[expert_batch_size_i, <extra_output_dims>]`.
      multiply_by_gates: a boolean

    Returns:
      a `Tensor` with shape `[batch_size, <extra_output_dims>]`.
    """
    # see comments on convert_gradient_to_tensor
    stitched = common_layers.convert_gradient_to_tensor(
        tf.concat(expert_out, 0))
    if multiply_by_gates:
      stitched *= tf.expand_dims(self._nonzero_gates, 1)
    combined = tf.unsorted_segment_sum(stitched, self._batch_index,
                                       tf.shape(self._gates)[0])
    return combined 

def segment_softmax(scores, segment_ids):
    """Given scores and a partition, converts scores to probs by performing
    softmax over all rows within a partition."""

    # Subtract max
    num_segments = tf.reduce_max(segment_ids) + 1
    if len(scores.get_shape()) == 2:
        max_per_partition = tf.unsorted_segment_max(tf.reduce_max(scores, axis=1), segment_ids, num_segments)
        scores -= tf.expand_dims(tf.gather(max_per_partition, segment_ids), axis=1)
    else:
        max_per_partition = tf.unsorted_segment_max(scores, segment_ids, num_segments)
        scores -= tf.gather(max_per_partition, segment_ids)

    # Compute probs
    scores_exp = tf.exp(scores)
    if len(scores.get_shape()) == 2:
        scores_exp_sum_per_partition = tf.unsorted_segment_sum(tf.reduce_sum(scores_exp, axis=1), segment_ids,
                                                               num_segments)
        probs = scores_exp / tf.expand_dims(tf.gather(scores_exp_sum_per_partition, segment_ids), axis=1)
    else:
        scores_exp_sum_per_partition = tf.unsorted_segment_sum(scores_exp, segment_ids, num_segments)
        probs = scores_exp / tf.gather(scores_exp_sum_per_partition, segment_ids)

    return probs 

def _deduplicate_indexed_slices(values, indices):
    """Sums `values` associated with any non-unique `indices`.
    Args:
      values: A `Tensor` with rank >= 1.
      indices: A one-dimensional integer `Tensor`, indexing into the first
      dimension of `values` (as in an IndexedSlices object).
    Returns:
      A tuple of (`summed_values`, `unique_indices`) where `unique_indices` is a
      de-duplicated version of `indices` and `summed_values` contains the sum of
      `values` slices associated with each unique index.
    """
    unique_indices, new_index_positions = tf.unique(indices)
    summed_values = tf.unsorted_segment_sum(
      values, new_index_positions,
      tf.shape(unique_indices)[0])
    return (summed_values, unique_indices) 

def _deduplicate_indexed_slices(values, indices):
    """Sums `values` associated with any non-unique `indices`.
    Args:
      values: A `Tensor` with rank >= 1.
      indices: A one-dimensional integer `Tensor`, indexing into the first
      dimension of `values` (as in an IndexedSlices object).
    Returns:
      A tuple of (`summed_values`, `unique_indices`) where `unique_indices` is a
      de-duplicated version of `indices` and `summed_values` contains the sum of
      `values` slices associated with each unique index.
    """
    unique_indices, new_index_positions = tf.unique(indices)
    summed_values = tf.unsorted_segment_sum(
      values, new_index_positions,
      tf.shape(unique_indices)[0])
    return (summed_values, unique_indices) 

def Combine(self, expert_out, multiply_by_gates=True):
    """Sum together the expert output, weighted by the gates.

    The slice corresponding to a particular batch element `b` is computed
    as the sum over all experts `i` of the expert output, weighted by the
    corresponding gate values.  If `multiply_by_gates` is set to False, the
    gate values are ignored.

    Args:
      expert_out: a list of `num_experts` `Tensor`s, each with shape
        `[expert_batch_size_i, <extra_output_dims>]`.
      multiply_by_gates: a boolean

    Returns:
      a `Tensor` with shape `[batch_size, <extra_output_dims>]`.
    """
    # see comments on ConvertGradientToTensor
    stitched = ConvertGradientToTensor(tf.concat(expert_out, 0))
    if multiply_by_gates:
      stitched *= tf.expand_dims(self._nonzero_gates, 1)
    combined = tf.unsorted_segment_sum(stitched, self._batch_index,
                                       tf.shape(self._gates)[0])
    return combined 

def embedding_lookup(embedding_matrix, indices, ids, weights, size):
  """Performs a weighted embedding lookup.

  Args:
    embedding_matrix: float Tensor from which to do the lookup.
    indices: int Tensor for the output rows of the looked up vectors.
    ids: int Tensor vectors to look up in the embedding_matrix.
    weights: float Tensor weights to apply to the looked up vectors.
    size: int number of output rows. Needed since some output rows may be
        empty.

  Returns:
    Weighted embedding vectors.
  """
  embeddings = tf.nn.embedding_lookup([embedding_matrix], ids)
  # TODO(googleuser): allow skipping weights.
  broadcast_weights_shape = tf.concat([tf.shape(weights), [1]], 0)
  embeddings *= tf.reshape(weights, broadcast_weights_shape)
  embeddings = tf.unsorted_segment_sum(embeddings, indices, size)
  return embeddings 

def embedding_lookup(embedding_matrix, indices, ids, weights, size):
  """Performs a weighted embedding lookup.

  Args:
    embedding_matrix: float Tensor from which to do the lookup.
    indices: int Tensor for the output rows of the looked up vectors.
    ids: int Tensor vectors to look up in the embedding_matrix.
    weights: float Tensor weights to apply to the looked up vectors.
    size: int number of output rows. Needed since some output rows may be
        empty.

  Returns:
    Weighted embedding vectors.
  """
  embeddings = tf.nn.embedding_lookup([embedding_matrix], ids)
  # TODO(googleuser): allow skipping weights.
  broadcast_weights_shape = tf.concat([tf.shape(weights), [1]], 0)
  embeddings *= tf.reshape(weights, broadcast_weights_shape)
  embeddings = tf.unsorted_segment_sum(embeddings, indices, size)
  return embeddings 

def testValues(self):
    dtypes = [tf.float32,
              tf.float64,
              tf.int64,
              tf.int32,
              tf.complex64,
              tf.complex128]
    indices_flat = np.array([0, 4, 0, 8, 3, 8, 4, 7, 7, 3])
    num_segments = 12
    for indices in indices_flat, indices_flat.reshape(5, 2):
      shape = indices.shape + (2,)
      for dtype in dtypes:
        with self.test_session(use_gpu=self.use_gpu):
          tf_x, np_x = self._input(shape, dtype=dtype)
          np_ans = self._segmentReduce(indices,
                                       np_x,
                                       np.add,
                                       op2=None,
                                       num_out_rows=num_segments)
          s = tf.unsorted_segment_sum(data=tf_x,
                                      segment_ids=indices,
                                      num_segments=num_segments)
          tf_ans = s.eval()
        self._assertAllClose(indices, np_ans, tf_ans)
        self.assertShapeEqual(np_ans, s) 

def embedding_lookup(embedding_matrix, indices, ids, weights, size):
  """Performs a weighted embedding lookup.

  Args:
    embedding_matrix: float Tensor from which to do the lookup.
    indices: int Tensor for the output rows of the looked up vectors.
    ids: int Tensor vectors to look up in the embedding_matrix.
    weights: float Tensor weights to apply to the looked up vectors.
    size: int number of output rows. Needed since some output rows may be
        empty.

  Returns:
    Weighted embedding vectors.
  """
  embeddings = tf.nn.embedding_lookup([embedding_matrix], ids)
  # TODO(googleuser): allow skipping weights.
  broadcast_weights_shape = tf.concat([tf.shape(weights), [1]], 0)
  embeddings *= tf.reshape(weights, broadcast_weights_shape)
  embeddings = tf.unsorted_segment_sum(embeddings, indices, size)
  return embeddings 

def find_dup(a):
  """ Find the duplicated elements in 1-D a tensor.
  Args:
    a: 1-D tensor.
    
  Return:
    more_than_one_vals: duplicated value in a.
    indexes_in_a: duplicated value's index in a.
    dups_in_a: duplicated value with duplicate in a.
  """
  unique_a_vals, unique_idx = tf.unique(a)
  count_a_unique = tf.unsorted_segment_sum(tf.ones_like(a),
                                           unique_idx,
                                           tf.shape(a)[0])

  more_than_one = tf.greater(count_a_unique, 1)
  more_than_one_idx = tf.squeeze(tf.where(more_than_one))
  more_than_one_vals = tf.squeeze(tf.gather(unique_a_vals, more_than_one_idx))

  not_duplicated, _ = tf.setdiff1d(a, more_than_one_vals)
  dups_in_a, indexes_in_a = tf.setdiff1d(a, not_duplicated)

  return more_than_one_vals, indexes_in_a, dups_in_a 

def embedding_lookup(embedding_matrix, indices, ids, weights, size):
  """Performs a weighted embedding lookup.

  Args:
    embedding_matrix: float Tensor from which to do the lookup.
    indices: int Tensor for the output rows of the looked up vectors.
    ids: int Tensor vectors to look up in the embedding_matrix.
    weights: float Tensor weights to apply to the looked up vectors.
    size: int number of output rows. Needed since some output rows may be
        empty.

  Returns:
    Weighted embedding vectors.
  """
  embeddings = tf.nn.embedding_lookup([embedding_matrix], ids)
  # TODO(googleuser): allow skipping weights.
  broadcast_weights_shape = tf.concat([tf.shape(weights), [1]], 0)
  embeddings *= tf.reshape(weights, broadcast_weights_shape)
  embeddings = tf.unsorted_segment_sum(embeddings, indices, size)
  return embeddings 

def _deduplicate_indexed_slices(values, indices):
    """Sums `values` associated with any non-unique `indices`.
    Args:
      values: A `Tensor` with rank >= 1.
      indices: A one-dimensional integer `Tensor`, indexing into the first
      dimension of `values` (as in an IndexedSlices object).
    Returns:
      A tuple of (`summed_values`, `unique_indices`) where `unique_indices` is a
      de-duplicated version of `indices` and `summed_values` contains the sum of
      `values` slices associated with each unique index.
    """
    unique_indices, new_index_positions = tf.unique(indices)
    summed_values = tf.unsorted_segment_sum(
      values, new_index_positions,
      tf.shape(unique_indices)[0])
    return (summed_values, unique_indices) 

def combine(self, x):
    """Return the output from the experts.

    When one example goes to multiple experts, the outputs are summed.

    Args:
      x: a Tensor with shape [batch, num_experts, expert_capacity, depth]

    Returns:
      a `Tensor` with shape `[batch, length, depth]
    """
    depth = tf.shape(x)[-1]
    x *= tf.expand_dims(self._nonpadding, -1)
    ret = tf.unsorted_segment_sum(
        x, self._flat_indices, num_segments=self._batch * self._length)
    ret = tf.reshape(ret, [self._batch, self._length, depth])
    return ret 

def embedding_lookup(embedding_matrix, indices, ids, weights, size):
  """Performs a weighted embedding lookup.

  Args:
    embedding_matrix: float Tensor from which to do the lookup.
    indices: int Tensor for the output rows of the looked up vectors.
    ids: int Tensor vectors to look up in the embedding_matrix.
    weights: float Tensor weights to apply to the looked up vectors.
    size: int number of output rows. Needed since some output rows may be
        empty.

  Returns:
    Weighted embedding vectors.
  """
  embeddings = tf.nn.embedding_lookup([embedding_matrix], ids)
  # TODO(googleuser): allow skipping weights.
  broadcast_weights_shape = tf.concat([tf.shape(weights), [1]], 0)
  embeddings *= tf.reshape(weights, broadcast_weights_shape)
  embeddings = tf.unsorted_segment_sum(embeddings, indices, size)
  return embeddings 

def combine(self, expert_out, multiply_by_gates=True):
    """Sum together the expert output, weighted by the gates.

    The slice corresponding to a particular batch element `b` is computed
    as the sum over all experts `i` of the expert output, weighted by the
    corresponding gate values.  If `multiply_by_gates` is set to False, the
    gate values are ignored.

    Args:
      expert_out: a list of `num_experts` `Tensor`s, each with shape
        `[expert_batch_size_i, <extra_output_dims>]`.
      multiply_by_gates: a boolean

    Returns:
      a `Tensor` with shape `[batch_size, <extra_output_dims>]`.
    """
    # see comments on convert_gradient_to_tensor
    stitched = common_layers.convert_gradient_to_tensor(
        tf.concat(expert_out, 0))
    if multiply_by_gates:
      stitched *= tf.expand_dims(self._nonzero_gates, 1)
    combined = tf.unsorted_segment_sum(stitched, self._batch_index,
                                       tf.shape(self._gates)[0])
    return combined 

def _deduplicate_indexed_slices(values, indices):
    """Sums `values` associated with any non-unique `indices`.
    Args:
      values: A `Tensor` with rank >= 1.
      indices: A one-dimensional integer `Tensor`, indexing into the first
      dimension of `values` (as in an IndexedSlices object).
    Returns:
      A tuple of (`summed_values`, `unique_indices`) where `unique_indices` is a
      de-duplicated version of `indices` and `summed_values` contains the sum of
      `values` slices associated with each unique index.
    """
    unique_indices, new_index_positions = tf.unique(indices)
    summed_values = tf.unsorted_segment_sum(
      values, new_index_positions,
      tf.shape(unique_indices)[0])
    return (summed_values, unique_indices) 

def combine(self, x):
    """Return the output from the experts.

    When one example goes to multiple experts, the outputs are summed.

    Args:
      x: a Tensor with shape [batch, num_experts, expert_capacity, depth]

    Returns:
      a `Tensor` with shape `[batch, length, depth]
    """
    depth = tf.shape(x)[-1]
    x *= tf.expand_dims(self._nonpadding, -1)
    ret = tf.unsorted_segment_sum(
        x, self._flat_indices, num_segments=self._batch * self._length)
    ret = tf.reshape(ret, [self._batch, self._length, depth])
    return ret 

def _deduplicate_indexed_slices(values, indices):
    """Sums `values` associated with any non-unique `indices`.
    Args:
      values: A `Tensor` with rank >= 1.
      indices: A one-dimensional integer `Tensor`, indexing into the first
      dimension of `values` (as in an IndexedSlices object).
    Returns:
      A tuple of (`summed_values`, `unique_indices`) where `unique_indices` is a
      de-duplicated version of `indices` and `summed_values` contains the sum of
      `values` slices associated with each unique index.
    """
    unique_indices, new_index_positions = tf.unique(indices)
    summed_values = tf.unsorted_segment_sum(values,
                                            new_index_positions,
                                            tf.shape(unique_indices)[0])
    return (summed_values, unique_indices) 

def combine(self, expert_out, multiply_by_gates=True):
    """Sum together the expert output, weighted by the gates.

    The slice corresponding to a particular batch element `b` is computed
    as the sum over all experts `i` of the expert output, weighted by the
    corresponding gate values.  If `multiply_by_gates` is set to False, the
    gate values are ignored.

    Args:
      expert_out: a list of `num_experts` `Tensor`s, each with shape
        `[expert_batch_size_i, <extra_output_dims>]`.
      multiply_by_gates: a boolean

    Returns:
      a `Tensor` with shape `[batch_size, <extra_output_dims>]`.
    """
    # see comments on convert_gradient_to_tensor
    stitched = common_layers.convert_gradient_to_tensor(
        tf.concat(expert_out, 0))
    if multiply_by_gates:
      stitched *= tf.expand_dims(self._nonzero_gates, 1)
    combined = tf.unsorted_segment_sum(stitched, self._batch_index,
                                       tf.shape(self._gates)[0])
    return combined 

def embedding_lookup(embedding_matrix, indices, ids, weights, size):
  """Performs a weighted embedding lookup.

  Args:
    embedding_matrix: float Tensor from which to do the lookup.
    indices: int Tensor for the output rows of the looked up vectors.
    ids: int Tensor vectors to look up in the embedding_matrix.
    weights: float Tensor weights to apply to the looked up vectors.
    size: int number of output rows. Needed since some output rows may be
        empty.

  Returns:
    Weighted embedding vectors.
  """
  embeddings = tf.nn.embedding_lookup([embedding_matrix], ids)
  # TODO(googleuser): allow skipping weights.
  broadcast_weights_shape = tf.concat([tf.shape(weights), [1]], 0)
  embeddings *= tf.reshape(weights, broadcast_weights_shape)
  embeddings = tf.unsorted_segment_sum(embeddings, indices, size)
  return embeddings 

def Combine(self, expert_out, multiply_by_gates=True):
    """Sum together the expert output, weighted by the gates.

    The slice corresponding to a particular batch element `b` is computed
    as the sum over all experts `i` of the expert output, weighted by the
    corresponding gate values.  If `multiply_by_gates` is set to False, the
    gate values are ignored.

    Args:
      expert_out: a list of `num_experts` `Tensor`s, each with shape
        `[expert_batch_size_i, <extra_output_dims>]`.
      multiply_by_gates: a boolean

    Returns:
      a `Tensor` with shape `[batch_size, <extra_output_dims>]`.
    """
    # see comments on ConvertGradientToTensor
    stitched = ConvertGradientToTensor(tf.concat(expert_out, 0))
    if multiply_by_gates:
      stitched *= tf.expand_dims(self._nonzero_gates, 1)
    combined = tf.unsorted_segment_sum(stitched, self._batch_index,
                                       tf.shape(self._gates)[0])
    return combined 

def __init__(self, data_parallelism, model_parallelism, gates):
    """Constructs a Dispatcher.

    Args:
      data_parallelism: a Parallelism object.
      model_parallelism: a Parallelism object.
      gates: a list of 1d integer `Tensor`s, one per datashard.
        Says which expert to use for each batch element.

    Returns:
      a DistributedSingleDispatcher
    """
    gates = data_parallelism(tf.to_int32, gates)
    self._gates = gates
    self._data_parallelism = data_parallelism
    self._model_parallelism = model_parallelism

    # Compute the sizes number of examples going from each datashard to each
    # expert.
    def _PartSizes(gates):
      return tf.unsorted_segment_sum(
          tf.ones_like(gates), gates, model_parallelism.n)

    part_sizes_by_datashard = data_parallelism(_PartSizes, gates)
    self._part_sizes_by_expert = tf.unstack(
        tf.stack(part_sizes_by_datashard), num=model_parallelism.n, axis=1)

    # These indices will be used to combine the output on the datashards.
    def _StitchIndices(gates):
      return tf.dynamic_partition(
          tf.range(tf.size(gates)), gates, model_parallelism.n)

    self._stitch_indices = data_parallelism(_StitchIndices, gates) 

def gated_regression(self, last_h, regression_gate, regression_transform):
        # last_h: [v x h]
        gate_input = tf.concat([last_h, self.placeholders['initial_node_representation']], axis=-1)  # [v x 2h]
        gated_outputs = tf.nn.sigmoid(regression_gate(gate_input)) * regression_transform(last_h)  # [v x 1]

        # Sum up all nodes per-graph
        graph_representations = tf.unsorted_segment_sum(data=gated_outputs,
                                                        segment_ids=self.placeholders['graph_nodes_list'],
                                                        num_segments=self.placeholders['num_graphs'])  # [g x 1]
        output = tf.squeeze(graph_representations)  # [g]
        self.output = output
        return output

    # ----- Data preprocessing and chunking into minibatches: 

def EmbeddingLookupFeatures(params, sparse_features, allow_weights):
  """Computes embeddings for each entry of sparse features sparse_features.

  Args:
    params: list of 2D tensors containing vector embeddings
    sparse_features: 1D tensor of strings. Each entry is a string encoding of
      dist_belief.SparseFeatures, and represents a variable length list of
      feature ids, and optionally, corresponding weights values.
    allow_weights: boolean to control whether the weights returned from the
      SparseFeatures are used to multiply the embeddings.

  Returns:
    A tensor representing the combined embeddings for the sparse features.
    For each entry s in sparse_features, the function looks up the embeddings
    for each id and sums them into a single tensor weighing them by the
    weight of each id. It returns a tensor with each entry of sparse_features
    replaced by this combined embedding.
  """
  if not isinstance(params, list):
    params = [params]
  # Lookup embeddings.
  sparse_features = tf.convert_to_tensor(sparse_features)
  indices, ids, weights = gen_parser_ops.unpack_sparse_features(sparse_features)
  embeddings = tf.nn.embedding_lookup(params, ids)

  if allow_weights:
    # Multiply by weights, reshaping to allow broadcast.
    broadcast_weights_shape = tf.concat(0, [tf.shape(weights), [1]])
    embeddings *= tf.reshape(weights, broadcast_weights_shape)

  # Sum embeddings by index.
  return tf.unsorted_segment_sum(embeddings, indices, tf.size(sparse_features))