DataExtractor

class lightning_pose.data.utils.DataExtractor(data_module: LightningDataModule, cond: Literal['train', 'test', 'val'] = 'train', extract_images: bool = False, remove_augmentations: bool = True)[source]

Bases: object

Helper class to extract all data from a data module.

Attributes Summary

dataset_length

Methods Summary

`__call__`()	Call self as a function.
`get_loader`()
`iterate_over_dataloader`(loader)
`verify_labeled_loader`(loader)

Attributes Documentation

dataset_length

Methods Documentation

__call__() → Tensor, {'__torchtyping__': True, 'details': ('num_examples', 'frames', 3, 'image_width', 'image_height',), 'cls_name': 'TensorType'}] | None][source]: Call self as a function.

get_loader() → DataLoader | SemiSupervisedDataLoaderDict[source]

iterate_over_dataloader(loader: DataLoader) → Tensor, {'__torchtyping__': True, 'details': ('num_examples', 'frames', 3, 'image_width', 'image_height',), 'cls_name': 'TensorType'}] | None][source]

static verify_labeled_loader(loader: DataLoader | SemiSupervisedDataLoaderDict) → DataLoader[source]

__init__(data_module: LightningDataModule, cond: Literal['train', 'test', 'val'] = 'train', extract_images: bool = False, remove_augmentations: bool = True) → None[source]