OpenDoc-Pdf-Preview

OpenDoc-Pdf-Preview is a compact visual preview dataset containing 6,000 high-resolution document images extracted from PDFs. This dataset is designed for Image-to-Text tasks such as document OCR pretraining, layout understanding, and multimodal document analysis.

Dataset Summary

• Modality: Image-to-Text
• Content Type: PDF-based document previews
• Number of Samples: 6,000
• Language: English
• Format: Parquet
• Split: train only
• Size: 606 MB
• License: Apache 2.0

Each entry consists of:

• A preview image of a PDF page
• A placeholder column named pdf (currently appears empty or reserved for future metadata)

Use Cases

• Pretraining OCR or Document Layout models
• PDF snapshot-based search indexing
• Few-shot document vision evaluation
• Visual prompt tuning for vision-language models (VLMs)

How to Use

from datasets import load_dataset

dataset = load_dataset("prithivMLmods/OpenDoc-Pdf-Preview", split="train")

Notes

• The column pdf may be extended in future versions with associated metadata or textual content.
• Each sample preview is rendered from the original PDF file, representing various real-world layouts and formats.