datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
DocVQA
Large-scale Multi-modality Models Evaluation Suite
Accelerating the development of large-scale multi-modality models (LMMs) with lmms-eval
🏠 Homepage | 📚 Documentation | 🤗 Huggingface Datasets
This Dataset
This is a formatted version of DocVQA. It is used in our lmms-eval pipeline to allow for one-click evaluations of large multi-modality models.
@article{mathew2020docvqa,
title={DocVQA: A Dataset for VQA on Document Images. CoRR abs/2007.00398 (2020)}… See the full description on the dataset page: https://huggingface.co/datasets/lmms-lab-encoder/DocVQA.fixtures_docvqaThis dataset includes 2 document images of the DocVQA dataset.
They are used for testing the LayoutLMv2FeatureExtractor + LayoutLMv2Processor inside the HuggingFace Transformers library.
More specifically, they are used in tests/test_feature_extraction_layoutlmv2.py and tests/test_processor_layoutlmv2.py.
docvqa-single-page-questions
Dataset Card for DocVQA Dataset
Dataset Summary
DocVQA dataset is a document dataset introduced in Mathew et al. (2021) consisting of 50,000 questions defined on 12,000+ document images.
Please visit the challenge page (https://rrc.cvc.uab.es/?ch=17) and paper (https://arxiv.org/abs/2007.00398) for further information.
Usage
This dataset can be used with current releases of Hugging Face datasets library.
Here is an example using a custom collator to bundle… See the full description on the dataset page: https://huggingface.co/datasets/pixparse/docvqa-single-page-questions.MP-DocVQAdocvqa_test_subsampled_beirBEIR version of vidore/docvqa_test_subsampled.
VisRAG-Ret-Test-MP-DocVQA
Dataset Description
This is a VQA dataset based on Industrial Documents from MP-DocVQA dataset from MP-DocVQA.
Load the dataset
from datasets import load_dataset
import csv
def load_beir_qrels(qrels_file):
qrels = {}
with open(qrels_file) as f:
tsvreader = csv.DictReader(f, delimiter="\t")
for row in tsvreader:
qid = row["query-id"]
pid = row["corpus-id"]
rel = int(row["score"])
if qid in qrels:… See the full description on the dataset page: https://huggingface.co/datasets/openbmb/VisRAG-Ret-Test-MP-DocVQA.DocVQA-2026
DocVQA 2026 | ICDAR2026 Competition on Multimodal Reasoning over Documents in Multiple Domains
Building upon previous DocVQA benchmarks, this evaluation dataset introduces challenging reasoning questions over a diverse collection of documents spanning eight domains, including business reports, scientific papers, slides, posters, maps, comics, infographics, and engineering drawings.
By expanding coverage to new document domains and… See the full description on the dataset page: https://huggingface.co/datasets/VLR-CVC/DocVQA-2026.docvqa_test_subsampled_beirBEIR version of vidore/docvqa_test_subsampled.
docvqa_test_subsampled
Dataset Description
This is the test set taken from the DocVQA dataset. It includes collected images from the UCSF Industry Documents Library. Questions and answers were manually annotated.
Example of data (see viewer)
Data Curation
To ensure homogeneity across our benchmarked datasets, we subsampled the original test set to 500 pairs and renamed the different columns.
Load the dataset
from datasets import load_dataset
ds =… See the full description on the dataset page: https://huggingface.co/datasets/vidore/docvqa_test_subsampled.MMEB-train-DocVQA-images
MMEB-train — DocVQA / Train images
Backup copy of the DocVQA Train image split used in MMEB-train (VLM2Vec) training.
Original folder structure is preserved: files live under DocVQA/Train/.
Files: 78,926 JPG images
Size: ~15 GB
docvqa_1200_examplesdoc-vqa
Dataset description
The doc-vqa Dataset integrates images from the Infographic_vqa dataset sourced from HuggingFaceM4 The Cauldron
dataset, as well as images from the dataset AFTDB (Arxiv Figure Table Database) curated by cmarkea.
This dataset consists of pairs of images and corresponding text, with each image linked to an average of five questions and answers available in both English and French.
These questions and answers were generated using Gemini 1.5 Pro, thereby… See the full description on the dataset page: https://huggingface.co/datasets/cmarkea/doc-vqa.docvqa_trainDocVQA_Processed_DatasetBToks-visrag_indomain_MP-DocVQA
BToks VisRAG MP-DocVQA
This dataset repository contains Lance-format converted data used by the open-source reproduction code for Bottleneck Tokens for Unified Multimodal Retrieval (arXiv:2604.11095).
Source
Converted from openbmb/VisRAG-Ret-Train-In-domain-data.
Subset/view: MP-DocVQA. This repository does not change upstream ownership, licensing, citation requirements, or usage restrictions.
Format
The data is stored as Lance tables for the… See the full description on the dataset page: https://huggingface.co/datasets/siyrus/BToks-visrag_indomain_MP-DocVQA.DocVQA
Large-scale Multi-modality Models Evaluation Suite
Accelerating the development of large-scale multi-modality models (LMMs) with lmms-eval
🏠 Homepage | 📚 Documentation | 🤗 Huggingface Datasets
This Dataset
This is a formatted version of DocVQA. It is used in our lmms-eval pipeline to allow for one-click evaluations of large multi-modality models.
@article{mathew2020docvqa,
title={DocVQA: A Dataset for VQA on Document Images. CoRR abs/2007.00398 (2020)}… See the full description on the dataset page: https://huggingface.co/datasets/jdchandana/DocVQA.DocVQA
Large-scale Multi-modality Models Evaluation Suite
Accelerating the development of large-scale multi-modality models (LMMs) with lmms-eval
🏠 Homepage | 📚 Documentation | 🤗 Huggingface Datasets
This Dataset
This is a formatted version of DocVQA. It is used in our lmms-eval pipeline to allow for one-click evaluations of large multi-modality models.
@article{mathew2020docvqa,
title={DocVQA: A Dataset for VQA on Document Images. CoRR abs/2007.00398 (2020)}… See the full description on the dataset page: https://huggingface.co/datasets/jnamlee/DocVQA.DocVQA
Large-scale Multi-modality Models Evaluation Suite
Accelerating the development of large-scale multi-modality models (LMMs) with lmms-eval
🏠 Homepage | 📚 Documentation | 🤗 Huggingface Datasets
This Dataset
This is a formatted version of DocVQA. It is used in our lmms-eval pipeline to allow for one-click evaluations of large multi-modality models.
@article{mathew2020docvqa,
title={DocVQA: A Dataset for VQA on Document Images. CoRR abs/2007.00398 (2020)}… See the full description on the dataset page: https://huggingface.co/datasets/EliMC/DocVQA.DocVQAdocvqa_beirThis is a copy of https://huggingface.co/datasets/jinaai/docvqa reformatted into the BEIR format. For any further information like license, please refer to the original dataset.
Disclaimer
This dataset may contain publicly available images or text data. All data is provided for research and educational purposes only. If you are the rights holder of any content and have concerns regarding intellectual property or copyright, please contact us at "support-data (at) jina.ai" for… See the full description on the dataset page: https://huggingface.co/datasets/jinaai/docvqa_beir.docvqa-lance
DocVQA (Lance Format)
A Lance-formatted version of DocVQA, a benchmark for visual question answering over document images such as industry and government scans, multi-page reports, forms, and receipts, redistributed via lmms-lab/DocVQA (DocVQA config). Each row carries the page image as inline JPEG bytes, the question and reference answer span(s), the original DocVQA question-type tags, UCSF Industry Documents Library provenance, and paired CLIP embeddings for the image and the… See the full description on the dataset page: https://huggingface.co/datasets/lance-format/docvqa-lance.docvqadocvqa
Creation
This dataset is build upon the corresponding dataset from the ViDoRe Benchmark. For more information regarding the filtering please read our paper or this discussion on github.
Disclaimer
This dataset may contain publicly available images or text data. All data is provided for research and educational purposes only. If you are the rights holder of any content and have concerns regarding intellectual property or copyright, please contact us at "support-data… See the full description on the dataset page: https://huggingface.co/datasets/jinaai/docvqa.docvqaMP-DocVQA
MP-DocVQA
MP-DocVQA is one of the 11 retrieval benchmarks used in RetrievalRouter: Joint Modality and
Architecture Selection for Document Retrieval (EMNLP 2026). Each record pairs a rendered page
image, a query, and the page's extracted text, supporting both text-based and multimodal
retrieval evaluation.
📄 Paper: https://arxiv.org/pdf/2608.25625
💻 Code: https://github.com/emrekuruu/retrieval-router
🤗 Collection: https://huggingface.co/collections/emrekuruu/retrieval-router… See the full description on the dataset page: https://huggingface.co/datasets/emrekuruu/MP-DocVQA.docvqa_disco
DocVQA_mini Dataset
A mini subset of the DocVQA dataset with 500 randomly selected question-answer pairs for document visual question answering evaluation.
Dataset Details
Total Samples: 500 QA pairs
Source: DocVQA validation set
Task: Document Visual Question Answering
Image Format: PNG (extracted from parquet-embedded images)
Features
Each sample contains:
image: Document image
question: Question about the document
answers: List of valid… See the full description on the dataset page: https://huggingface.co/datasets/kenza-ily/docvqa_disco.dutch-legal-docvqa
Dutch Legal Document VQA Dataset
Nederlandse Document VQA dataset gebaseerd op ~50.000 publieke rechtspraak uitspraken.
Statistieken
Train: 267,818 samples
Validation: 29,758 samples
Unieke documenten: ~49,596
Gebruik
from datasets import load_dataset
dataset = load_dataset("Rickkosse/dutch-legal-docvqa")
sample = dataset['train'][0]
print(f"Vraag: {sample['question']}")
print(f"Antwoord: {sample['answers']}")
sample['image'].show()
Bron… See the full description on the dataset page: https://huggingface.co/datasets/Rickkosse/dutch-legal-docvqa.DocVQATR-DocVQA-Synth
TR-DocVQA-Synth: A Large-Scale Synthetic Turkish Document Visual Question Answering Dataset
Dataset Summary
TR-DocVQA-Synth is a large-scale synthetic Turkish Document Visual Question Answering dataset designed for training and evaluating multimodal models on Turkish business documents. The dataset contains 15,000 document images and 235,000 question-answer pairs generated from structured ground-truth records.
The dataset focuses on realistic Turkish document… See the full description on the dataset page: https://huggingface.co/datasets/Ethosoft/TR-DocVQA-Synth.docvqa_1200_examples_donut