Repeater

class torchdata.datapipes.iter.Repeater(source_datapipe: IterDataPipe[T_co], times: int)

Repeatedly yield each element of source DataPipe for the specified number of times before moving onto the next element (functional name: repeat). Note that no copy is made in this DataPipe, the same element is yielded repeatedly.

If you would like to yield the whole DataPipe in order multiple times, use Cycler.

Parameters:

• source_datapipe – source DataPipe that will be iterated through

• times – the number of times an element of source_datapipe will be yielded before moving onto the next element

Example

>>> from torchdata.datapipes.iter import IterableWrapper
>>> dp = IterableWrapper(range(3))
>>> dp = dp.repeat(2)
>>> list(dp)
[0, 0, 1, 1, 2, 2]