Datasets:

racineai
/

VDR_MEGA_2

license: apache-2.0
task_categories:
  - question-answering
  - visual-question-answering
  - text-retrieval
language:
  - en
  - fr
  - de
  - it
  - es
  - ar
tags:
  - multimodal
  - technical-documents
  - RAG
  - DSE
  - merged-datasets

VDR_MEGA_2

Dataset Summary

VDR_MEGA_2 is a high-quality multimodal dataset created through the merge of multiple domain-specific datasets with enhanced data processing techniques. This dataset represents our most refined approach to multimodal data generation, incorporating filtering algorithms and improved AI-assisted content generation to deliver superior quality for RAG, DSE, question answering, document search, and vision-language model training tasks.

Source Datasets

This merged dataset combines the the following datasets:

Dataset (split)	Domain	Language(s)
`racineai/VDR_Military (filtered)`	Military	EN, FR
`racineai/VDR_Energy (filtered)`	Energy/Regulation	EN, FR
`racineai/VDR_Quantum (filtered)`	Quantum	EN, FR
`racineai/VDR_ibm-research_REAL-MM-RAG (train)`	Technical/Research	EN
`racineai/VDR_Cooking_Recipes (filtered)`	Culinary Arts	Multiple
`racineai/VDR_Geotechnie (filtered)`	Geotechnie	EN, FR
`racineai/VDR_Nuclear (filtered)`	Nuclear/Regulation	EN, FR, DE, IT, ES
`racineai/VDR_2_vdr-visRAG-colpali (filtered)`	Various	EN, FR, DE, IT, ES
`racineai/VDR_Renewable_Regulation (filtered)`	Energy/Regulation	Multiple
`racineai/VDR_Quantum_Circuit_Papers (filtered)`	Quantum Computing	EN
`racineai/VDR_Hydrogen (filtered)`	Hydrogen/Regulation	EN, FR
`racineai/VDR_History_Geography (filtered)`	Education/History	Multiple
`racineai/VDR_Energy_Arabic (train)`	Energy	Arabic
`racineai/VDR_CATIE-AQ_XMRec (train)`	Various	FR

Data Fields

Each entry contains:

id (string): Unique identifier
query (string): High-quality technical/domain-specific question
image (PIL.Image): High-resolution visual rendering of source document page
language (string): Detected language of the image (queries sometimes differ on purpose)

Dataset Curators

Léo Appourchaux
Paul Lemaistre
Yumeng Ye
Mattéo KHAN
André-Louis Rochet