README.md

metadata

dataset_info:
  config_name: funsd
  features:
    - name: id
      dtype: string
    - name: tokens
      sequence: string
    - name: bboxes
      sequence:
        sequence: int64
    - name: ner_tags
      sequence:
        class_label:
          names:
            '0': O
            '1': B-HEADER
            '2': I-HEADER
            '3': B-QUESTION
            '4': I-QUESTION
            '5': B-ANSWER
            '6': I-ANSWER
    - name: image
      dtype: image
  splits:
    - name: train
      num_bytes: 27288633
      num_examples: 149
    - name: test
      num_bytes: 9931720
      num_examples: 50
  download_size: 35837449
  dataset_size: 37220353
configs:
  - config_name: funsd
    data_files:
      - split: train
        path: funsd/train-*
      - split: test
        path: funsd/test-*
    default: true

Neural Metrics · Noisy scanned forms with key-value ground truth.

FUNSD is a small, deliberately grubby set of scanned forms annotated with entities and the links between them. It is the classic sanity check for form understanding.

We use it for: key-value pair extraction - entity linking on forms - checking that a model degrades gracefully on genuinely bad scans.

Attribution

This is an unmodified fork of nielsr/funsd-layoutlmv3, created by the Qwen team. All weights, files and behaviour are identical to upstream — we rehost it so our experiments stay reproducible and version-pinned. The original license and all credit remain with the Qwen team. If you want the canonical dataset, please use the original.

---

Original dataset card from nielsr/funsd-layoutlmv3 (click to expand)