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)

Shuffler

class torchdata.datapipes.map.Shuffler(datapipe: MapDataPipe[_T_co], *, indices: Optional[List] = None)

Shuffle the input MapDataPipe via its indices (functional name: shuffle).

When it is used with DataLoader, the methods to set up random seed are different based on num_workers.

For single-process mode (num_workers == 0), the random seed is set before the DataLoader in the main process. For multi-process mode (num_worker > 0), worker_init_fn is used to set up a random seed for each worker process.

Parameters:

• datapipe – MapDataPipe being shuffled

• indices – a list of indices of the MapDataPipe. If not provided, we assume it uses 0-based indexing

Example

>>> # xdoctest: +SKIP
>>> from torchdata.datapipes.map import SequenceWrapper
>>> dp = SequenceWrapper(range(10))
>>> shuffle_dp = dp.shuffle().set_seed(0)
>>> list(shuffle_dp)
[7, 8, 1, 5, 3, 4, 2, 0, 9, 6]
>>> list(shuffle_dp)
[6, 1, 9, 5, 2, 4, 7, 3, 8, 0]
>>> # Reset seed for Shuffler
>>> shuffle_dp = shuffle_dp.set_seed(0)
>>> list(shuffle_dp)
[7, 8, 1, 5, 3, 4, 2, 0, 9, 6]

Note

Even thought this shuffle operation takes a MapDataPipe as the input, it would return an IterDataPipe rather than a MapDataPipe, because MapDataPipe should be non-sensitive to the order of data order for the sake of random reads, but IterDataPipe depends on the order of data during data-processing.