Dataset Viewer (First 5GB)
Auto-converted to Parquet Duplicate
The dataset viewer is not available for this split.
Rows from parquet row groups are too big to be read: 286.86 MiB (max=286.10 MiB)
Error code:   TooBigContentError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning: empty or missing yaml metadata in repo card (https://huggingface.co/docs/hub/datasets-cards)

Open Sora plan collected 40,258 high-quality, watermark-free videos from open-source websites under the CC0 license. About 60% of the videos are in landscape format, with a total duration of approximately 274 hours, 5 minutes, and 13 seconds.

The dataset is divided into three main sources:

Mixkit:

  • Videos: 1,234

  • Total duration: 6h 19m 32s

  • Total frames: 570,815

  • Resolution and aspect ratio distributions (less than 1% not listed).

Pexels:

  • Videos: 7,408

  • Total duration: 48h 49m 24s

  • Total frames: 5,038,641

  • Resolution and aspect ratio distributions (less than 1% not listed).

Pixabay:

  • Videos: 31,616

  • Total duration: 218h 56m 17s

  • Total frames: 23,508,970

  • Resolution and aspect ratio distributions (less than 1% not listed).

Downloads last month
62