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)

WebDataset

class torchdata.datapipes.iter.WebDataset(source_datapipe: IterDataPipe[List[Union[Dict, List]]])

Iterable DataPipe that accepts stream of (path, data) tuples, usually, representing the pathnames and files of a tar archive (functional name: webdataset). This aggregates consecutive items with the same basename into a single dictionary, using the extensions as keys (WebDataset file convention). Any text after the first “.” in the filename is used as a key/extension.

File names that do not have an extension are ignored.

Parameters:

source_datapipe – a DataPipe yielding a stream of (path, data) pairs

Returns:

a DataPipe yielding a stream of dictionaries

Examples

>>> from torchdata.datapipes.iter import FileLister, FileOpener
>>>
>>> def decode(item):
>>>     key, value = item
>>>     if key.endswith(".txt"):
>>>         return key, value.read().decode("utf-8")
>>>     if key.endswith(".bin"):
>>>         return key, value.read().decode("utf-8")
>>>
>>> datapipe1 = FileLister("test/_fakedata", "wds*.tar")
>>> datapipe2 = FileOpener(datapipe1, mode="b")
>>> dataset = datapipe2.load_from_tar().map(decode).webdataset()
>>> for obj in dataset:
>>>     print(obj)