PrepareDALI

class lightning_pose.data.dali.PrepareDALI(train_stage: Literal['predict', 'train'], model_type: Literal['base', 'context'], filenames: list[str] | list[list[str]], resize_dims: list[int], dali_config: dict | DictConfig | None = None, imgaug: str | None = 'default', num_threads: int = 1)

Bases: object

All the DALI stuff in one place.

Big picture: this will initialize the pipes and dataloaders for both training and prediction.

Attributes Summary

num_iters

Methods Summary

__call__()

Returns a LightningWrapper object.

Attributes Documentation

num_iters

Methods Documentation

__call__() → LitDaliWrapper

Returns a LightningWrapper object.

__init__(train_stage: Literal['predict', 'train'], model_type: Literal['base', 'context'], filenames: list[str] | list[list[str]], resize_dims: list[int], dali_config: dict | DictConfig | None = None, imgaug: str | None = 'default', num_threads: int = 1) → None