datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
TaskMeAnything-v1-videoqa-2024
Dataset Card for TaskMeAnything-v1-videoqa-2024
TaskMeAnything-v1-videoqa-2024 benchmark dataset
🌐 Website | 📑 Paper | 🤗 Huggingface | 💻 Interface
If you like our project, please give us a star ⭐ on GitHub for latest update.
TaskMeAnything-v1-2024-Videoqa
TaskMeAnything-v1-videoqa-2024 is a benchmark for reflecting the current progress of MLMs by automatically finding tasks that SOTA MLMs struggle with using the TaskMeAnything Top-K queries.
This… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/TaskMeAnything-v1-videoqa-2024.TaskMeAnything-v1-videoqa-random
Dataset Card for TaskMeAnything-v1-videoqa-random
TaskMeAnything-v1-videoqa-random dataset
🌐 Website | 📑 Paper | 🤗 Huggingface | 💻 Interface
If you like our project, please give us a star ⭐ on GitHub for latest update.
TaskMeAnything-v1-Random
TaskMeAnything-v1-videoqa-random is a dataset which randomly sampled questions from TaskMeAnything-v1, including 2,700 VideoQA questions. The dataset contains 9 splits, while each splits contains 300 questions… See the full description on the dataset page: https://huggingface.co/datasets/weikaih/TaskMeAnything-v1-videoqa-random.Video-T3-QATextual Temporal Understanding Dataset
Temporal Reasoning Transfer from Text to Video, ICLR 2025
Project Page: https://video-t3.github.io/
In each json file, we provide LLaVA-style text QA samples, using the synthesization method described in our paper.
For example:
[
{
"from": "human",
"value": "Based on the following captions describing keyframes of a video, answer the next question.\n\nCaptions:\nThe image displays a circular emblem with a metallic appearance, conveying a… See the full description on the dataset page: https://huggingface.co/datasets/MMInstruction/Video-T3-QA.Adaption-video-qa-diverse-topics
This dataset is a remastered version of this dataset prepared using Adaption's Adaptive Data platform.
video_qa_diverse_topics
This dataset contains question-answer pairs derived from a diverse collection of video clips covering topics such as biology, history, sports, and astronomy. Each entry includes a specific question about the video content and a corresponding factual answer, alongside metadata like duration, tags, and object lists. The samples demonstrate a focus on… See the full description on the dataset page: https://huggingface.co/datasets/Reubencf/Adaption-video-qa-diverse-topics.CapRL-Video-QA-20K
CapRL-Video-QA-20K.jsonl Video Path Setup
Each value is a relative path under the Hugging Face dataset root of lmms-lab/LLaVA-Video-178K.
Example:
"videos": ["0_30_s_youtube_v0_1/videos/liwei_youtube_videos/videos/youtube_video_2024/ytb_khSwLQOthHQ.mp4"]
Required Video Data
Download the original videos from Hugging Face:
Dataset: https://huggingface.co/datasets/lmms-lab/LLaVA-Video-178K
Required subdirectories for this 20k subset:
0_30_s_youtube_v0_1… See the full description on the dataset page: https://huggingface.co/datasets/internlm/CapRL-Video-QA-20K.nlp-qa-audio-video
NLP QA Audio Video Data Notes
Dataset summary
This repository contains a preparation pipeline and a small metadata sample for NLP QA work with Audio Video inputs. It does not claim to be a complete benchmark release; the loader documents how source data is normalized and validated.
Included material
dataset.py — loading, cleaning, and split preparation code.
dataset_infos.json — schema and split metadata.
metadata_sample.jsonl — small… See the full description on the dataset page: https://huggingface.co/datasets/leond-u0114/nlp-qa-audio-video.Urban_Dynamics_VideoQA_datasetvideo-frame-qa-v5video-redbox-qa-dataset
video-redbox-qa-dataset
A dataset of videos with corresponding queries and answers.
Dataset Structure
Each entry contains:
A video file
A query/question about the video
An answer to the query
Usage
from datasets import load_dataset
dataset = load_dataset("famousdetectiveadrianmonk/video-redbox-qa-dataset")
Access an example
example = dataset["train"][0]
video_path = example["video"]
query = example["query"]
answer = example["answer"]… See the full description on the dataset page: https://huggingface.co/datasets/famousdetectiveadrianmonk/video-redbox-qa-dataset.videoqa-samples-videosnlp-qa-audio-video
NLP QA Audio Video Data Notes
Dataset summary
This data card accompanies a lightweight NLP QA loader for Audio Video metadata. It is meant for pipeline inspection, source adaptation, and reproducible split preparation.
Included material
dataset.py — loading, cleaning, and split preparation code.
dataset_infos.json — schema and split metadata.
metadata_sample.jsonl — small, human-readable records for checking the schema.
README.md — data card and… See the full description on the dataset page: https://huggingface.co/datasets/ShwetaMsm3867/nlp-qa-audio-video.nlp-qa-audio-video
NLP QA Audio Video Data Notes
Dataset summary
This repository contains a preparation pipeline and a small metadata sample for NLP QA work with Audio Video inputs. It does not claim to be a complete benchmark release; the loader documents how source data is normalized and validated.
Included material
dataset.py — loading, cleaning, and split preparation code.
dataset_infos.json — schema and split metadata.
metadata_sample.jsonl — small… See the full description on the dataset page: https://huggingface.co/datasets/lucasjlpt/nlp-qa-audio-video.dataset_127491213_nlp_qa_video_text
dataset_127491213_nlp_qa_video_text.py
Dataset Summary
A nlp qa dataset with video text modality, stored in webdataset format.
Preprocessing & Augmentation
Preprocessing: standard
Augmentation: none
Splits & Sampling
Split strategy: kfold 5
Sampling: random
Quality & Labeling
Quality filtering: moderate
Labeling: self training
Files
dataset_127491213_nlp_qa_video_text.py — main artifact of this… See the full description on the dataset page: https://huggingface.co/datasets/franciszekkaminski/dataset_127491213_nlp_qa_video_text.nlp-qa-video-text
NLP QA Video Text Data Notes
Dataset summary
This repository contains a preparation pipeline and a small metadata sample for NLP QA work with Video Text inputs. It does not claim to be a complete benchmark release; the loader documents how source data is normalized and validated.
Included material
preprocess.py — loading, cleaning, and split preparation code.
dataset_infos.json — schema and split metadata.
metadata_sample.jsonl — small… See the full description on the dataset page: https://huggingface.co/datasets/Justriyarao/nlp-qa-video-text.Japanese_Video-QA
Japanese Video-QA
Overview
Japanese Video-QA is a video question-answering benchmark focused on Japanese cultural content, designed to evaluate multimodal large language models (MLLMs) on Japanese-specific videos.
428 YouTube videos → 800 QA pairs
6 domains, 100 sub-domains covering Japanese culture
Questions generated by Gemini 2.5 Flash and manually verified
Evaluated with LLM-as-a-Judge (GPT-4o), scoring 1–3
Authors: 峯 悠大, 新立 拓也, 髙橋 和也 (NABLAS Inc.)… See the full description on the dataset page: https://huggingface.co/datasets/nablasinc/Japanese_Video-QA.Long-Video-Tuning-QAsdataset_129939201_nlp_qa_audio_video
dataset_129939201_nlp_qa_audio_video.py
Dataset Summary
A nlp qa dataset with audio video modality, stored in webdataset format.
Preprocessing & Augmentation
Preprocessing: domain specific
Augmentation: none
Splits & Sampling
Split strategy: random 90 10
Sampling: random
Quality & Labeling
Quality filtering: moderate
Labeling: semi auto
Files
dataset_129939201_nlp_qa_audio_video.py — main artifact… See the full description on the dataset page: https://huggingface.co/datasets/KABIR-BHAT/dataset_129939201_nlp_qa_audio_video.video_dataset_qagemini-video-qa-pairsqwen2.5-video-qa-pairsvideoqa-mvp-datahound_train_300k_qa_videoViet-Tiktok-VideoQA
Example by GIF 1:
Long description:
Video mở đầu với hình ảnh lá cờ đỏ sao vàng tung bay trên cột cờ cao, phía dưới là một quảng trường rộng lớn với nhiều người đi lại. Cảnh quay tiếp tục chuyển sang một buổi hoàng hôn rực rỡ với mặt trời đỏ cam đang lặn trên mặt nước, cùng lúc đó có nhiều người đang chèo thuyền kayak trên sông. Sau đó, ống kính hướng đến một tòa tháp cổ kính nhiều tầng ẩn mình giữa những tán cây xanh. Tiếp theo là hình ảnh một chàng trai trẻ đang đi bộ trên một… See the full description on the dataset page: https://huggingface.co/datasets/5CD-AI/Viet-Tiktok-VideoQA.NTLBG-VideoQAIt can be directly used for VidEntro-MLLM fine-tuning and main experiments (especially the LongVideo-Reason multiple-choice reasoning task).
captionsVideoQA