Method from_data_loader

opacus/data_loader.py:307–363

Creates new ``DPDataLoader`` based on passed ``data_loader`` argument. Args: data_loader: Any DataLoader instance. Must not be over an ``IterableDataset`` distributed: set ``True`` if you'll be using DPDataLoader in a DDP environment generator: R

(
        cls,
        data_loader: DataLoader,
        *,
        distributed: bool = False,
        generator=None,
        batch_first: bool = True,
        rand_on_empty: bool = False,
    )

    @classmethod
    def from_data_loader(
        cls,
        data_loader: DataLoader,
        *,
        distributed: bool = False,
        generator=None,
        batch_first: bool = True,
        rand_on_empty: bool = False,
    ):
        """
        Creates new ``DPDataLoader`` based on passed ``data_loader`` argument.

        Args:
            data_loader: Any DataLoader instance. Must not be over an ``IterableDataset``
            distributed: set ``True`` if you'll be using DPDataLoader in a DDP environment
            generator: Random number generator used to sample elements. Defaults to
                generator from the original data loader.
            batch_first: Flag to indicate if the input tensor to the corresponding module
                has the first dimension representing the batch. If set to True, dimensions on
                input tensor are expected be ``[batch_size, ...]``, otherwise
                ``[K, batch_size, ...]``
            rand_on_empty: set ``True`` to return a batch containing random numbers when encountering
                empty batches rather than tensors with zero-length batch dimensions



        Returns:
            New DPDataLoader instance, with all attributes and parameters inherited
            from the original data loader, except for sampling mechanism.

        Examples:
            >>> x, y = torch.randn(64, 5), torch.randint(0, 2, (64,))
            >>> dataset = TensorDataset(x,y)
            >>> data_loader = DataLoader(dataset, batch_size=4)
            >>> dp_data_loader = DPDataLoader.from_data_loader(data_loader)
        """

        if isinstance(data_loader.dataset, IterableDataset):
            raise ValueError("Uniform sampling is not supported for IterableDataset")

        return cls(
            dataset=data_loader.dataset,
            sample_rate=1 / len(data_loader),
            num_workers=data_loader.num_workers,
            collate_fn=data_loader.collate_fn,
            pin_memory=data_loader.pin_memory,
            drop_last=data_loader.drop_last,
            timeout=data_loader.timeout,
            worker_init_fn=data_loader.worker_init_fn,
            multiprocessing_context=data_loader.multiprocessing_context,
            generator=generator if generator else data_loader.generator,
            prefetch_factor=data_loader.prefetch_factor,
            persistent_workers=data_loader.persistent_workers,
            distributed=distributed,
            batch_first=batch_first,
            rand_on_empty=rand_on_empty,
        )