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

Processing Details

Source Dataset: davanstrien/dataset-paper-pages-scale Model: numind/NuExtract3 Mode: structured-extraction Number of Samples: 1,719 Processing Time: 12.7 min Processing Date: 2026-06-01 11:09 UTC

Configuration

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