Function batch

tensorflow/contrib/labeled_tensor/python/ops/ops.py:453–497

Rebatch a tensor. See tf.batch. Args: labeled_tensors: The input tensors. batch_size: The output batch size. num_threads: See tf.batch. capacity: See tf.batch. enqueue_many: If true, the input tensors must contain a 'batch' axis as their first axis. If false, th

(labeled_tensors,
          batch_size,
          num_threads=1,
          capacity=32,
          enqueue_many=False,
          allow_smaller_final_batch=False,
          name=None)

    tc.Collection(core.LabeledTensorLike), int, int, int, bool, bool,
    tc.Optional(string_types))
def batch(labeled_tensors,
          batch_size,
          num_threads=1,
          capacity=32,
          enqueue_many=False,
          allow_smaller_final_batch=False,
          name=None):
  """Rebatch a tensor.

  See tf.batch.

  Args:
    labeled_tensors: The input tensors.
    batch_size: The output batch size.
    num_threads: See tf.batch.
    capacity: See tf.batch.
    enqueue_many: If true, the input tensors must contain a 'batch' axis as
      their first axis.
      If false, the input tensors must not contain a 'batch' axis.
      See tf.batch.
    allow_smaller_final_batch: See tf.batch.
    name: Optional op name.

  Returns:
    The rebatched tensors.
    If enqueue_many is false, the output tensors will have a new 'batch' axis
      as their first axis.

  Raises:
    ValueError: If enqueue_many is True and the first axis of the tensors
      isn't "batch".
  """

  def fn(tensors, scope):
    return input.batch(
        tensors,
        batch_size=batch_size,
        num_threads=num_threads,
        capacity=capacity,
        enqueue_many=enqueue_many,
        allow_smaller_final_batch=allow_smaller_final_batch,
        name=scope)

  return _batch_helper('lt_batch', fn, batch_size, enqueue_many,
                       labeled_tensors, allow_smaller_final_batch, name)


@tc.returns(tc.List(core.LabeledTensor))