MNISTDataModule¶
- class torch_uncertainty.datamodules.MNISTDataModule(root, batch_size, eval_ood=False, ood_ds='fashion', val_split=None, num_workers=1, cutout=None, test_alt=None, pin_memory=True, persistent_workers=True)[source]¶
DataModule for MNIST.
- Parameters:
root (str) – Root directory of the datasets.
eval_ood (bool) – Whether to evaluate on out-of-distribution data.
batch_size (int) – Number of samples per batch.
ood_ds (str) – Which out-of-distribution dataset to use. Defaults to
"fashion"; fashion stands for FashionMNIST and not for notMNIST.val_split (float) – Share of samples to use for validation. Defaults to
0.0.num_workers (int) – Number of workers to use for data loading. Defaults to
1.cutout (int) – Size of cutout to apply to images. Defaults to
None.test_alt (str) – Which test set to use. Defaults to
None.pin_memory (bool) – Whether to pin memory. Defaults to
True.persistent_workers (bool) – Whether to use persistent workers. Defaults to
True.kwargs – Additional arguments.