Pretrained weights for lucid.models.cvt_w24_cls.
wide CvT-w24 IN-22k→IN-1k 384 finetune (Wu et al.; 277 M params, top-1 87.7%).
Attributes
IN22K_FT_IN1K_384WeightEntryImageNet-22k pretrain + ImageNet-1k 384x384 finetune (top-1
87.7%), sourced from
transformers/microsoft/cvt-w24-384-22k.DEFAULTWeightEntryAlias for
IN22K_FT_IN1K_384.Notes
Reference: Wu, Xiao, Codella, Liu, Dai, Yuan, Zhang, "CvT: Introducing Convolutions to Vision Transformers", ICCV 2021 (arXiv:2103.15808).
Examples
>>> from lucid.models import cvt_w24_cls
>>> model = cvt_w24_cls(pretrained=True).eval()