---
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
  splits:
  - name: train
    num_bytes: 115513345.41
    num_examples: 1502
  - name: validation
    num_bytes: 81299189.507
    num_examples: 1311
  download_size: 142776148
  dataset_size: 196812534.917
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---