README.md

metadata

language:
  - rcf
dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: transcription_french
      dtype: string
    - name: transcription_creole
      dtype: string
    - name: prediction_small
      dtype: string
    - name: wer_small
      dtype: float64
    - name: prediction_large
      dtype: string
    - name: wer_large
      dtype: float64
    - name: bleu
      struct:
        - name: bleu
          dtype: float64
        - name: brevity_penalty
          dtype: float64
        - name: length_ratio
          dtype: float64
        - name: precisions
          sequence: float64
        - name: reference_length
          dtype: int64
        - name: translation_length
          dtype: int64
  splits:
    - name: train
      num_bytes: 216434499.595
      num_examples: 2809
    - name: validation
      num_bytes: 826183
      num_examples: 4
  download_size: 142881491
  dataset_size: 217260682.595
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*