class

EfficientNetB5Weights

extendsWeightsEnum

EfficientNetB5Weights()

Pretrained weights for lucid.models.efficientnet_b5_cls.

compound-scaled B5 at 456 (30.4 M params, 10.27 GFLOPs, top-1 83.44%).

IMAGENET1K_V1WeightEntry

ImageNet-1k V1 checkpoint (top-1 83.444% / top-5 96.628%), sourced from torchvision/EfficientNet_B5_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_b5_cls
>>> model = efficientnet_b5_cls(pretrained=True).eval()