Skip to content

nexuml_library.scenarios.vision.mnist_resnet

nexuml_library.scenarios.vision.mnist_resnet

MNIST and FashionMNIST ResNet classification scenarios.

mnist_data

mnist_data(
    download: bool = True, root: str = "mnist"
) -> DataSpec

Create a DataSpec for MNIST image classification.

Returns:

Name Type Description
DataSpec DataSpec

MNIST dataset specification with fit and test splits.

fashionmnist_data

fashionmnist_data(
    download: bool = True, root: str = "fashionmnist"
) -> DataSpec

Create a DataSpec for FashionMNIST image classification.

Returns:

Name Type Description
DataSpec DataSpec

FashionMNIST dataset specification with fit and test splits.

mnist_resnet

mnist_resnet(
    download: bool = True,
    resnet_type: str = "resnet18",
    pretrained: bool = False,
    cifar_stem: bool = True,
    lr: float = 0.001,
    batch_size: int = 64,
    max_epochs: int = 10,
) -> ScenarioSpec

MNIST image classification with ResNet backbone.

Returns:

Name Type Description
ScenarioSpec ScenarioSpec

Assembled scenario with pipeline, training, data and evaluation.

fashionmnist_resnet

fashionmnist_resnet(
    download: bool = True,
    resnet_type: str = "resnet18",
    pretrained: bool = False,
    cifar_stem: bool = True,
    lr: float = 0.001,
    batch_size: int = 64,
    max_epochs: int = 10,
) -> ScenarioSpec

FashionMNIST image classification with ResNet backbone.

Returns:

Name Type Description
ScenarioSpec ScenarioSpec

Assembled scenario with pipeline, training, data and evaluation.