---
dataset_info:
  features:
  - name: anchor
    dtype: string
  - name: positive
    dtype: string
  - name: negative
    dtype: string
  - name: negative_index_pos
    dtype: int64
  splits:
  - name: train
    num_bytes: 25553993
    num_examples: 30280
  download_size: 17534937
  dataset_size: 25553993
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
language:
- da
---

This is a processed version of https://huggingface.co/datasets/DDSC/da-wikipedia-queries-gemma

The dataset was created using this script:
https://github.com/Dansk-Data-Science-Community/embedding_model/blob/main/create_processed_data.py