checkpoints and Streamo-Instruct-465K

[zhuojiewu]

Hi,

Thank you for your excellent work on Streamo. I found the paper and benchmark very insightful for streaming video understanding research.

I would like to kindly ask:

1. Are there plans to release the pretrained model checkpoints?
2. Will the original training videos used to construct Streamo-Instruct-465K be made available, or is there a list of accessible source datasets with preprocessing details?

Thank you very much for your time and for considering this request.