---
dataset_info:
  features:
  - name: text
    dtype: string
  - name: audio_clean
    dtype:
      audio:
        sampling_rate: 16000
  - name: audio_noisy
    dtype:
      audio:
        sampling_rate: 16000
  - name: duration
    dtype: float64
  splits:
  - name: train
    num_bytes: 3477686388
    num_examples: 4585
  download_size: 3238505242
  dataset_size: 3477686388
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---