---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: logits
    sequence: float32
  splits:
  - name: train
    num_bytes: 609219552.4155288
    num_examples: 19003
  download_size: 569570436
  dataset_size: 609219552.4155288
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---