---
license: apache-2.0
language:
- en
- fr
- es
- it
- de
tags:
- synthetic
size_categories:
- 100K<n<1M
---


# WIP - there might be issues with the negatives


# OGC - Organized, Grouped, Cleaned

> **Intended for image/text to vector (DSE)**

## Dataset Composition

The dataset merges, shuffles, and formats data from:

- [vidore/colpali_train_set](https://huggingface.co/datasets/vidore/colpali_train_set)
- [openbmb/VisRAG-Ret-Train-Synthetic-data](https://huggingface.co/datasets/openbmb/VisRAG-Ret-Train-Synthetic-data)
- [llamaindex/vdr-multilingual-train](https://huggingface.co/datasets/llamaindex/vdr-multilingual-train)

## Dataset Statistics

| Metric | Value |
|--------|-------|
| Total rows | 700,000+ |
| Rows with negatives | ≈ 33% |
| Rows without queries (image negatives only) | ≈ 25% |

## Language Distribution

| Language| Ratio |
|--------|-------|
| English | ≈ 52%  |
| French  | ≈ 12%  |
| Spanish | ≈ 12%  |
| Italian | ≈ 12%  |
| German  | ≈ 12%  |


## Creators

Dataset curated by:
- **Paul Lemaistre** 
- **Léo Appourchaux**