Does this model have CLS token?

by risakri001 - opened Jun 23

Jun 23

Hello,
The video embeddings for this model are coming shape of torch.Size([1, 10240, 1024]). So, does it have a CLS token to get global embeddings, or do we need to take the mean of all the tokens to get the global embedding of the entire video?
Thanks.

qubvel-hf

Jun 24

Hey @risakri001 , there is no cls token in V-JEPA 2 model

risakri001

Jun 24

Okay, Thanks for answering.

risakri001 changed discussion status to closed Jun 24

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment