FPC64 ViT-H/16 representation weights at 256 pixels.
Converted from facebook/vjepa2-vith-fpc64-256 the same way as the
ViT-L tag. Reproduces its source to 5.3e-5 relative; the residual
grows with depth, and this network is eight blocks deeper.
FPC64 ViT-H/16 representation weights at 256 pixels.
Converted from facebook/vjepa2-vith-fpc64-256 the same way as the
ViT-L tag. Reproduces its source to 5.3e-5 relative; the residual
grows with depth, and this network is eight blocks deeper.