tf.raw_ops.ExperimentalRebatchDataset

Creates a dataset that changes the batch size.

tf.raw_ops.ExperimentalRebatchDataset(
    input_dataset, num_replicas, output_types, output_shapes, use_fallback=True,
    name=None
)

Creates a dataset that changes the batch size of the dataset to current batch size // num_replicas.

Args:

  • input_dataset: A Tensor of type variant. A variant tensor representing the input dataset.
  • num_replicas: A Tensor of type int64. A scalar representing the number of replicas to distribute this batch across. As a result of this transformation the current batch size would end up being divided by this parameter.
  • output_types: A list of tf.DTypes that has length >= 1.
  • output_shapes: A list of shapes (each a tf.TensorShape or list of ints) that has length >= 1.
  • use_fallback: An optional bool. Defaults to True.
  • name: A name for the operation (optional).

Returns:

A Tensor of type variant.