README.md

metadata

dataset_info:
  features:
    - name: english
      dtype: string
    - name: darija_Arab
      dtype: string
    - name: darija_Latn
      dtype: string
    - name: dataset_source
      dtype: string
    - name: id
      dtype: string
    - name: role
      dtype: string
    - name: darija_tokens
      dtype: int64
    - name: subtopic
      dtype: string
    - name: topic
      dtype: string
    - name: annotator_dialect
      dtype: string
  splits:
    - name: train
      num_bytes: 400022028.89170796
      num_examples: 354112
    - name: test
      num_bytes: 143518
      num_examples: 850
  download_size: 213959197
  dataset_size: 400165546.89170796
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*