README.md

metadata

language:
  - az
license: cc-by-sa-3.0
size_categories:
  - 100K<n<1M
task_categories:
  - text-generation
pretty_name: Azerbaijani Wikipedia
dataset_info:
  features:
    - name: id
      dtype: int64
    - name: text
      dtype: string
    - name: title
      dtype: string
  splits:
    - name: train
      num_bytes: 360206818
      num_examples: 129433
  download_size: 204669649
  dataset_size: 360206818
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

You can find the scripts here

AzWiki is a snapshot of Azerbaijani Wikipedia processed specifically for the text generation task. More details regarding the cleaning and processing steps will be released.

Please go to the GitHub repository for all discussions and PRs.