Pretrained weights for lucid.models.vgg_19_cls.
VGG-E 19-layer config (143.7 M params, 19.6 GFLOPs, top-1 72.38%).
Attributes
IMAGENET1K_V1WeightEntryImageNet-1k V1 checkpoint (top-1 72.376% / top-5 90.876%),
sourced from
torchvision/VGG19_Weights.IMAGENET1K_V1.DEFAULTWeightEntryAlias for
IMAGENET1K_V1.Notes
Reference: Simonyan, Zisserman, "Very Deep Convolutional Networks for Large-Scale Image Recognition", ICLR 2015 (arXiv:1409.1556).
Examples
>>> from lucid.models import vgg_19_cls
>>> model = vgg_19_cls(pretrained=True).eval()