ViT-Large checkpoint containing pre-training weights on FMoW-RGB dataset. For more information [github] [arXiv]
-