---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: speakers
    sequence: string
  - name: timestamps_start
    sequence: float64
  - name: timestamps_end
    sequence: float64
  splits:
  - name: train
    num_bytes: 4896683464.084507
    num_examples: 1680
  download_size: 4103566637
  dataset_size: 4896683464.084507
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---