tf.raw_ops.PrefetchDataset

Creates a dataset that asynchronously prefetches elements from input_dataset.

tf.raw_ops.PrefetchDataset(
    input_dataset, buffer_size, output_types, output_shapes, slack_period=0,
    legacy_autotune=True, name=None
)

Args:

  • input_dataset: A Tensor of type variant.
  • buffer_size: A Tensor of type int64. The maximum number of elements to buffer in an iterator over this dataset.
  • 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.
  • slack_period: An optional int. Defaults to 0.
  • legacy_autotune: An optional bool. Defaults to True.
  • name: A name for the operation (optional).

Returns:

A Tensor of type variant.