Datasets:

Lanxingxuan
/

Export_RL_20260911

Modalities:

Video

Size:

1K - 10K

Libraries:

Datasets

Dataset card Data Studio Files Files and versions

xet

Community

You need to agree to share your Hugging Face information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Export RL

This dataset requires manual access approval and contains 12,000 multiple-choice QA records in three JSONL files. Each source dataset contributes 4,000 QA sampled from 500 video paths, with 8 QA per path. The sampling seed is 20260911. VideoMME and LongVideoBench use Easy/Medium QA; LVBench has no difficulty filter.

Annotations are in Export_RL/annotation/. The 1,479 shared video files are in Export_RL/video/. Each JSONL row contains only Video, Question, Options, and Answer. Video is a filename; resolve it relative to Export_RL/video/ after downloading the repository. Options maps choice letters to their text, including E for five-choice questions.

from pathlib import Path
video_path = Path(repo_root) / "Export_RL" / "video" / row["Video"]

Downloads last month: 10

Total file size:

259 GB