README.md

metadata

license: cc-by-sa-4.0
dataset_info:
  features:
    - name: title
      dtype: string
    - name: context
      dtype: string
    - name: question
      dtype: string
    - name: answer
      dtype: string
  splits:
    - name: train
      num_bytes: 10783899
      num_examples: 3780
  download_size: 1378446
  dataset_size: 10783899
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
task_categories:
  - question-answering
language:
  - da
size_categories:
  - 1K<n<10K

Dataset Card for DaQA

Dataset Description

• Repository: https://github.com/alexandrainst/daqa
• Point of Contact: Rasmus Larsen

Dataset Summary

This dataset contains questions and answers generated by GPT-4o for a random selection of articles from a Danish Wikipedia dump.