add video modality and video demo task

[artemorloff]

trafficstars

add support for video modality: video_llava only for now as long as there is no unique inference code style

also add fake_video dataset

will be working on another models and may be (if possible) unification of the code

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you all sign our Contributor License Agreement before we can accept your contribution.
1 out of 3 committers have signed the CLA.

:white_check_mark: artemorloff
:x: EgorSevriugov
:x: Gscraid
_{You have signed the CLA already but the status is still pending? Let us recheck it.}