facebook/

PLM-Video-Auto

[📃 Tech Report] [📂 Github] Sythetic video captions and MCQs used in PLM, please refer to the paper, Section 3, for more details. The sythetic annotations covers: YT-1B, Ego4d with captions, YT-1B with MCQAs and Ego4d with QAs.

	Dataset Structure




	
	YT-1B Captions (yt1b_cap)

Data fields are :

video_id: a string feature, unique identifier for the YouTube videoid. scene_id: a string feature, unique identifier for the scene_id.… See the full description on the dataset page: https://huggingface.co/datasets/facebook/PLM-Video-Auto.