class torchdata.datapipes.iter.RoutedDecoder(datapipe: ~Iterable[~Tuple[str, ~io.BufferedIOBase]], *handlers: ~Callable, key_fn: ~Callable = <function extension_extract_fn>)

Decodes binary streams from input DataPipe, yields pathname and decoded data in a tuple (functional name: routed_decode).

  • datapipe – Iterable datapipe that provides pathname and binary stream in tuples

  • handlers – Optional user defined decoder handlers. If None, basic and image decoder handlers will be set as default. If multiple handles are provided, the priority order follows the order of handlers (the first handler has the top priority)

  • key_fn – Function for decoder to extract key from pathname to dispatch handlers. Default is set to extract file extension from pathname


When key_fn is specified returning anything other than extension, the default handler will not work and users need to specify custom handler. Custom handler could use regex to determine the eligibility to handle data.


Access comprehensive developer documentation for PyTorch

View Docs


Get in-depth tutorials for beginners and advanced developers

View Tutorials


Find development resources and get your questions answered

View Resources