Perception Encoder
Collection
OpenCLIP (PE Core image + text) and timm PE Core, Spatial, Lang (ViT only) weights. NOTE: These weights do not work with original modeling code.
•
19 items
•
Updated
•
2