README.md

metadata

language:
  - da
license: cc-by-4.0
size_categories:
  - 1K<n<10K
task_categories:
  - image-to-text
pretty_name: Danish Handwritten Text Recognition
tags:
  - OCR
  - HTR
dataset_info:
  features:
    - name: image
      dtype: image
    - name: authority
      dtype: string
    - name: docId
      dtype: int64
    - name: docTitle
      dtype: string
    - name: scriptType
      dtype: string
    - name: sequence
      dtype: int16
    - name: alto
      dtype: string
    - name: page
      dtype: string
  splits:
    - name: train
      num_bytes: 4694374015.813
      num_examples: 3899
  download_size: 4323086467
  dataset_size: 4694374015.813
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*