incerto.data.create_dataloaders

incerto.data.create_dataloaders#

incerto.data.create_dataloaders(train_dataset, val_dataset=None, test_dataset=None, batch_size=128, num_workers=4, pin_memory=True, shuffle_train=True)[source]#

Create standard data loaders for train/val/test.

Parameters:

  • train_dataset (Dataset) – Training dataset

  • val_dataset (Optional[Dataset]) – Validation dataset (optional)

  • test_dataset (Optional[Dataset]) – Test dataset (optional)

  • batch_size (int) – Batch size

  • num_workers (int) – Number of data loading workers

  • pin_memory (bool) – Whether to pin memory

  • shuffle_train (bool) – Whether to shuffle training data

Return type:

Tuple[DataLoader, ...]

Returns:

Tuple of DataLoaders (train, val, test) If val or test is None, returns None for that loader

Contents