Attention

June 2024 Status Update: Removing DataPipes and DataLoader V2

We are re-focusing the torchdata repo to be an iterative enhancement of torch.utils.data.DataLoader. We do not plan on continuing development or maintaining the [DataPipes] and [DataLoaderV2] solutions, and they will be removed from the torchdata repo. We’ll also be revisiting the DataPipes references in pytorch/pytorch. In release torchdata==0.8.0 (July 2024) they will be marked as deprecated, and in 0.9.0 (Oct 2024) they will be deleted. Existing users are advised to pin to torchdata==0.8.0 or an older version until they are able to migrate away. Subsequent releases will not include DataPipes or DataLoaderV2. Please reach out if you suggestions or comments (please use this issue for feedback)

Demultiplexer

class torchdata.datapipes.iter.Demultiplexer(datapipe: IterDataPipe, num_instances: int, classifier_fn: Callable[[_T_co], Optional[int]], drop_none: bool = False, buffer_size: int = 1000)

Splits the input DataPipe into multiple child DataPipes, using the given classification function (functional name: demux).

A list of the child DataPipes is returned from this operation.

Parameters:

• datapipe – Iterable DataPipe being filtered

• num_instances – number of instances of the DataPipe to create

• classifier_fn – a function that maps values to an integer within the range [0, num_instances - 1] or None

• drop_none – defaults to False, if True, the function will skip over elements classified as None

• buffer_size – this defines the maximum number of inputs that the buffer can hold across all child DataPipes while waiting for their values to be yielded. Defaults to 1000. Use -1 for the unlimited buffer.

Examples

>>> # xdoctest: +REQUIRES(module:torchdata)
>>> from torchdata.datapipes.iter import IterableWrapper
>>> def odd_or_even(n):
...     return n % 2
>>> source_dp = IterableWrapper(range(5))
>>> dp1, dp2 = source_dp.demux(num_instances=2, classifier_fn=odd_or_even)
>>> list(dp1)
[0, 2, 4]
>>> list(dp2)
[1, 3]
>>> # It can also filter out any element that gets `None` from the `classifier_fn`
>>> def odd_or_even_no_zero(n):
...     return n % 2 if n != 0 else None
>>> dp1, dp2 = source_dp.demux(num_instances=2, classifier_fn=odd_or_even_no_zero, drop_none=True)
>>> list(dp1)
[2, 4]
>>> list(dp2)
[1, 3]