cli.shared.types.utils package
cli.shared.types.utils.dataloader module
- class cli.shared.types.utils.dataloader.DataLoaderCliConfigType(default: ~cli.shared.types.utils.dataloader.DataLoaderConfig = <factory>, train: ~cli.shared.types.utils.dataloader.DataLoaderConfig | None = None, test: ~cli.shared.types.utils.dataloader.DataLoaderConfig | None = None)
Bases:
object
This dataclass encapsulates the data loader configurations for different scenarios. It allows for a single default configuration and optional specific configurations.
- default: DataLoaderConfig
- test: DataLoaderConfig | None = None
- train: DataLoaderConfig | None = None
- class cli.shared.types.utils.dataloader.DataLoaderConfig(num_workers: int = 1, shuffle: bool = False, drop_last: bool = False, pin_memory: bool = True, batch_size: int = 32)
Bases:
object
This dataclass is used to store the configuration for a data loader. It can be used generically for any type of data loading scenario (e.g., training, testing).
- batch_size: int = 32
- drop_last: bool = False
- num_workers: int = 1
- pin_memory: bool = True
- shuffle: bool = False
cli.shared.types.utils.dataset module
- class cli.shared.types.utils.dataset.AttacksDatasetCliConfigType(dataset_name: str, num_classes: int, train_dataset_path: str | None = None, test_dataset_path: str | None = None, download: bool | None = True, preprocessing: PreprocessConfig | None = None, dataset_part: str | None = 'test', random_sample_size: int | None = None)
Bases:
DatasetCliConfigType
This dataclass is used to store the configuration of the dataset CLI used for attacks. It extends the DatasetCliConfigType. In addition to the attributes of the DatasetCliConfigType, it has the following attributes:
- dataset_part
The part of the dataset to be used for the attack. it can be train, test or all. This is valid if the dataset paths are not provided.
- Type:
Optional[str]
- random_sample_size
The size of the random sample to be taken from the dataset.
- Type:
Optional[int]
- class cli.shared.types.utils.dataset.DatasetCliConfigType(dataset_name: str, num_classes: int, train_dataset_path: str | None = None, test_dataset_path: str | None = None, download: bool | None = True, preprocessing: PreprocessConfig | None = None)
Bases:
object
This dataclass is used to store the configuration of the dataset CLI.
- dataset_name: str
- num_classes: int
- preprocessing: PreprocessConfig | None = None
cli.shared.types.utils.device module
cli.shared.types.utils.model module
- class cli.shared.types.utils.model.ModelCliConfigType(model_name: str, num_input_channels: int, num_classes: int, pretrained: bool, weights: str | None, is_external: bool, path_configs: ModelPathConfig, norm_config: ModelNormConfig, random_seed: int | None)
Bases:
object
This dataclass is used to store the configuration of the model CLI.
- is_external: bool
- model_name: str
- norm_config: ModelNormConfig
- num_classes: int
- num_input_channels: int
- path_configs: ModelPathConfig
- pretrained: bool
cli.shared.types.utils.target module
- class cli.shared.types.utils.target.TargetCLIConfigType(targeted: bool | None = False, auto_generate_target: bool | None = False, target_images_config: TargetImageCLIConfigType | None = None, target_labels_config: TargetLabelCLIConfigType | None = None)
Bases:
object
This dataclass is used to store the configuration of the target CLI.
- target_images_config: TargetImageCLIConfigType | None = None
- target_labels_config: TargetLabelCLIConfigType | None = None
- class cli.shared.types.utils.target.TargetImageCLIConfigType(target_images_dir: str | None = None)
Bases:
object
This dataclass is used to store the configuration of the target CLI images.
- class cli.shared.types.utils.target.TargetLabelCLIConfigType(target_labels: List[int] | None = None, target_labels_path: str | None = None, target_labels_separator: str | None = ',')
Bases:
object
This dataclass is used to store the configuration of the target CLI labels.