Now we only upload two models for creating demos for image and video classification.

More models and code can be found in our github repo: UniFormer.