README.md

metadata

language:
  - 'no'
  - nb
  - nn
  - da
  - sv
  - is
  - en
size_categories:
  - 10B<n<100B
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*.json
      - split: validation
        path: data/validation-*.json
  - config_name: bad
    data_files:
      - split: train
        path: data/train-bad-*.json
      - split: validation
        path: data/validation-bad-*.json
  - config_name: medium
    data_files:
      - split: train
        path: data/train-medium-*.json
      - split: validation
        path: data/validation-medium-*.json
  - config_name: good
    data_files:
      - split: train
        path: data/train-good-*.json
      - split: validation
        path: data/validation-good-*.json