---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: transkription
    dtype: string
  - name: source
    dtype: string
  - name: predicted_transkription
    dtype: string
  splits:
  - name: train
    num_bytes: 34219570298.0
    num_examples: 626517
  download_size: 34111224093
  dataset_size: 34219570298.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---