Datasets:
mteb
/

Modalities:
Text
Formats:
arrow
Languages:
German
ArXiv:
Libraries:
Datasets
License:
rasdani
update README
1661468
|
raw
history blame
775 Bytes
metadata
configs:
  - config_name: default
    data_files:
      - split: corpus
        path: corpus/data-00000-of-00001.arrow
      - split: queries
        path: queries/data-00000-of-00001.arrow
license: cc-by-4.0
language:
  - de
source_datasets:
  - deepset/germanquad

This dataset is derived from the GermanQuAD dataset.

This dataset takes the testset and represents it as a corpus in the BEIR information retrieval benchmark format. The corresponding qrels can be found here.

Full credit for the original dataset goes to the authors of the GermanQuAD dataset.