Download README.md from TornadoLabs/interactionbench: direct link, hf CLI and curl.
- Browser
- Download file 4.73 kB
-
https://huggingface.co/datasets/TornadoLabs/interactionbench/resolve/main/README.md
- Command line
-
hf download hf://datasets/TornadoLabs/interactionbench/README.md
-
curl -L -o README.md https://huggingface.co/datasets/TornadoLabs/interactionbench/resolve/main/README.md
license: other
license_name: academic-source-terms
license_link: https://huggingface.co/datasets/InteractionBench/InteractionBench
task_categories:
- video-classification
- text-to-video
- video-text-to-text
tags:
- video
- youtube
- tornadolabs
pretty_name: InteractionBench — YouTube source videos
configs:
- config_name: videos
default: true
data_files: metadata.jsonl
- config_name: annotations
data_files: captions.parquet
InteractionBench — YouTube source videos
812 source IDs with their source links and availability. 404 videos (43.0 GB) with 281 subtitle files, downloaded at source quality and hosted for direct download.
Coverage (2026-11-16): 404 complete YouTube source videos are hosted here. They come from a prepared selection of 812 YouTube IDs read from
InteractionBench/InteractionBench(revision86a5830976f1); the selection stopped because all matching upstream metadata files were read. Of the prepared IDs: 404 hosted, 3 unavailable at download time, 0 failed downloads, 11 not added because of the 44 GB cap, 0 in progress and 394 not attempted. Thevideosconfiguration lists every prepared ID, hosted or not:video_urlis filled only for the 404 hosted videos, andavailabilitygives the state of each row when this card was generated. The annotations cover all listed IDs.
Public release: capped at 44 GB (44,000,000,000 bytes), including media and indexes. Unit: complete YouTube source videos. Upstream clip intervals, where present, remain in the annotations; the videos have not been cut to those intervals. The index lists every prepared ID; only part of them is hosted.
What's inside
metadata.jsonl— 812 rows, one per prepared source ID:video_id,youtube_url,availability,in_prepared_selection,video_url(direct HTTPS download, CDN-backed; empty when the video is not hosted),subtitle_url,size_bytes,last_attempt_at- 404 of these videos (43.0 GB) are hosted on Hugging Face storage; the other rows keep their source link only
captions.parquet— 1,060 academic annotations (source:annotations.csv)
Quick start
from datasets import load_dataset
# videos (default config): one row per source ID; video_url is set for hosted videos only
ds = load_dataset("TornadoLabs/interactionbench", "videos", split="train")
print(ds[0])
# keep the hosted videos, then download one
import requests
hosted = ds.filter(lambda r: r["video_url"] is not None)
r = requests.get(hosted[0]["video_url"])
open(f"{hosted[0]['video_id']}.mp4", "wb").write(r.content)
# annotations for all listed source videos
ann = load_dataset("TornadoLabs/interactionbench", "annotations", split="train")
How this dataset was built
Each file is the complete YouTube source video (best available quality), fetched through the Tornado API into a temporary location, measured, then copied into the public storage only while the release stayed under 44 GB. Clip intervals from the upstream annotations are kept as annotations and are not applied to the files. Videos that were unavailable, failed or did not fit are counted in the coverage note above.
Provenance & citation
- Original dataset: https://huggingface.co/datasets/InteractionBench/InteractionBench
- Paper: See the upstream dataset card
- No explicit license is declared upstream for the annotations; rights to the third-party videos are separate.
The video files remain the property of their respective YouTube uploaders. The annotations belong to the original dataset authors — cite their paper if you use them. This re-hosting is provided for reproducible research; if you are a rights holder and want content removed, open a discussion on this repo.
Request a larger or custom delivery
This release is capped at 44 GB. Additional videos from the prepared selection or from the complete upstream ID list can be collected on request: request a delivery quote from Tornado API. What can be delivered depends on each video's current availability on YouTube and on the applicable rights: recovery of every listed source is not guaranteed, and no commercial rights to the source videos are granted. The upstream terms continue to apply.

Collected and delivered with Tornado API