This dataset contains outputs from davanstrien/model-paper-pages-scale processed with NuExtract3, a 4B vision-language model for document understanding.

Processing Details

Source Dataset: davanstrien/model-paper-pages-scale Model: numind/NuExtract3 Mode: structured-extraction Number of Samples: 1,767 Processing Time: 48.3 min Processing Date: 2026-06-01 14:40 UTC

Configuration

Image Column: image… See the full description on the dataset page: https://huggingface.co/datasets/davanstrien/model-paper-pages-extracted.