metadata
dataset_info:
features:
- name: question_id
dtype: string
- name: video_name
dtype: string
- name: url
dtype: string
- name: videoID
dtype: string
- name: question
dtype: string
- name: options
sequence: string
- name: answer
dtype: string
- name: type
dtype: string
- name: hints
dtype: string
splits:
- name: test
num_bytes: 445365
num_examples: 1100
- name: cot
num_bytes: 74976
num_examples: 100
download_size: 987728
dataset_size: 520341
configs:
- config_name: default
data_files:
- split: test
path: data/test-*
- split: cot
path: data/cot-*
The dataset for VideoAds benchmarks:
For video and subtitle files for the dataset, remember to unzip the videoads.zip file under the dataset cache_dir and modify the yaml config under lmms-eval
Note this dataset can be only used for academic/research application.
If you consider this dataset useful, please cite,
@article{zhang2025videoads,
title={VideoAds for Fast-Paced Video Understanding: Where Opensource Foundation Models Beat GPT-4o \& Gemini-1.5 Pro},
author={Zhang, Zheyuan and Dou, Monica and Peng, Linkai and Pan, Hongyi and Bagci, Ulas and Gong, Boqing},
journal={arXiv preprint arXiv:2504.09282},
year={2025}
}