cerebras.modelzoo.data.vision.classification.mixup.RandomCutmix#

class cerebras.modelzoo.data.vision.classification.mixup.RandomCutmix[source]#

Bases: torch.nn.Module

Randomly apply Cutmix to the provided batch and targets. The class implements the data augmentations as described in the paper “CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features”. :param num_classes: number of classes used for one-hot encoding. :type num_classes: int :param p: probability of the batch being transformed. Default value is 0.5. :type p: float :param alpha: hyperparameter of the Beta distribution used for cutmix.

Default value is 1.0.

Parameters

inplace (bool) – boolean to make this transform inplace. Default set to False.

Methods

forward

param batch

Float tensor of size (B, C, H, W)
__init__(num_classes, p=0.5, alpha=1.0, inplace=False)[source]#
forward(batch, target)[source]#
Parameters

  • batch (Tensor) – Float tensor of size (B, C, H, W)

  • target (Tensor) – Integer tensor of size (B, )

Returns

Randomly transformed batch.

Return type

Tensor

__call__(*args: Any, **kwargs: Any) → Any#

Call self as a function.

static __new__(cls, *args: Any, **kwargs: Any) → Any#

previous

cerebras.modelzoo.data.vision.classification.mixup

next

cerebras.modelzoo.data.vision.classification.mixup.RandomMixup