Config.json on huggingface is not provided

[Saintghetto17]

Thank you for contribution to Video Analysis architectures!
Is there are any recipes, how to create a config.json same as in VideoMaeV1? I am eager to use your model in my project, but there is no wrap for that and I am not sure how to make it by myself. Could somehow explain, how to make it?

Best regards

[yinanhe]

Hi, we have converted the original VideoMAEv2 checkpoint to the huggingface version. You can get the config file of the model you want by visiting the our huggingface collection.
https://huggingface.co/collections/OpenGVLab/videomae-v2-678631493ab2f0c4642d842d

[Saintghetto17]

wow, thanks for that convenience!