class

EfficientNetB2Weights

extendsWeightsEnum

EfficientNetB2Weights()

Pretrained weights for lucid.models.efficientnet_b2_cls.

compound-scaled B2 at 288 (9.1 M params, 1.09 GFLOPs, top-1 80.61%).

IMAGENET1K_V1WeightEntry

ImageNet-1k V1 checkpoint (top-1 80.608% / top-5 95.310%), sourced from torchvision/EfficientNet_B2_Weights.IMAGENET1K_V1.

DEFAULTWeightEntry

Alias for IMAGENET1K_V1.

Reference: Tan, Le, "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks", ICML 2019 (arXiv:1905.11946).

>>> from lucid.models import efficientnet_b2_cls
>>> model = efficientnet_b2_cls(pretrained=True).eval()