Lightning-AI/pytorch-lightning · error · TypeError

Only PyTorch DataLoader are currently supported in `setup_da

Error message

Only PyTorch DataLoader are currently supported in `setup_dataloaders`.

What it means

Raised when any object passed to `fabric.setup_dataloaders()` is not a `torch.utils.data.DataLoader`. In contrast to the Lightning Trainer, Fabric only supports plain PyTorch DataLoaders — not iterables, custom loader classes, or combined objects like LightningDataModules.

Source

Thrown at src/lightning/fabric/fabric.py:1253

            raise ValueError("An optimizer should be passed only once to the `setup_optimizers` method.")

        if any(_has_meta_device_parameters_or_buffers(optimizer) for optimizer in optimizers):
            raise RuntimeError(
                "The optimizer has references to the model's meta-device parameters. Materializing them is"
                " is currently not supported. Create the optimizer after setting up the model, then call"
                " `fabric.setup_optimizers(optimizer)`."
            )

    def _validate_setup_dataloaders(self, dataloaders: Sequence[DataLoader]) -> None:
        self._validate_launched()
        if not dataloaders:
            raise ValueError("`setup_dataloaders` requires at least one dataloader as input.")

        if any(isinstance(dl, _FabricDataLoader) for dl in dataloaders):
            raise ValueError("A dataloader should be passed only once to the `setup_dataloaders` method.")

        if any(not isinstance(dl, DataLoader) for dl in dataloaders):
            raise TypeError("Only PyTorch DataLoader are currently supported in `setup_dataloaders`.")

    @staticmethod
    def _configure_callbacks(callbacks: Optional[Union[list[Any], Any]]) -> list[Any]:
        callbacks = callbacks if callbacks is not None else []
        callbacks = callbacks if isinstance(callbacks, list) else [callbacks]
        callbacks.extend(_load_external_callbacks("lightning.fabric.callbacks_factory"))
        return callbacks

View on GitHub (pinned to 9fed5c27d2)

Solutions

1. Wrap your dataset in a `torch.utils.data.DataLoader` before calling setup_dataloaders
2. If you have a custom iterable, iterate it manually inside the training loop instead of passing it to setup_dataloaders (you lose Fabric's distributed sampler handling)
3. Check `isinstance(obj, torch.utils.data.DataLoader)` before passing

Example fix

# before
loader = fabric.setup_dataloaders(my_dataset)  # not a DataLoader

# after
from torch.utils.data import DataLoader
loader = fabric.setup_dataloaders(DataLoader(my_dataset, batch_size=32))

from torch.utils.data import DataLoader
assert all(isinstance(dl, DataLoader) for dl in dls), 'only torch DataLoaders supported'

from torch.utils.data import DataLoader
def is_torch_dataloader(x) -> bool:
    return isinstance(x, DataLoader)

When it happens

Trigger: Passing a `IterableDataset`, a plain iterable, a HuggingFace `DataLoader`-like object, a `LightningDataModule`, or any custom loader class to `fabric.setup_dataloaders()`.

Common situations: Migrating from Lightning Trainer where combined data modules or arbitrary iterables were accepted; wrapping datasets with custom batch iterators; passing a dataset directly instead of a DataLoader.

Related errors

• `setup_dataloaders` requires at least one dataloader as inpu
• A dataloader should be passed only once to the `setup_datalo
• `self.log({name}, {value})` was called, but `{type(v).__name
• The dataloader {dataloader} needs to subclass `torch.utils.d
• Filter should be a dictionary, given {filter!r}

AI-assisted analysis of Lightning-AI/pytorch-lightning@9fed5c27d2 (2026-08-28). Data as JSON: /api/errors/eac3b1fe4b79d54d. Report an issue: GitHub.