Lines Matching refs:dataset
31 class _PrependFromQueueAndPaddedBatchDataset(dataset_ops.Dataset):
32 """A `Dataset` that prepends a queue to another `Dataset`.
101 """A transformation that prepends a queue to a `Dataset` and batches results.
107 Below is an example of how this dataset might be used to split incoming
109 are re-enqueued back into the dataset. A more realistic example would
114 dataset = tf.data.Dataset.from_tensor_slices([2*x for x in range(10)])
115 # Make a dataset of variable-length vectors and their lengths.
116 dataset = dataset.map(lambda count: (count, tf.ones((count,))))
118 dataset = dataset.apply(
121 dataset = dataset.prefetch(1)
123 iterator = dataset.make_one_shot_iterator()
147 incoming dataset's `output_types`.
149 If provided, the structure must match that of the incoming dataset's
150 `output_types`. If not provided, the incoming dataset's `output_shapes`
157 A `Dataset` transformation function, which can be passed to
158 @{tf.data.Dataset.apply}.
161 def _apply_fn(dataset):
163 dataset,
175 attribute of the `dataset` created by
191 `dataset.output_types[1]` and `dataset.output_shapes[1]` (the non-queue
192 output types and shapes) of the `dataset` emitted by
196 An `Operation` that enqueues `components` into the dataset(s) associated
