README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: string
    - name: path
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: transcription
      dtype: string
    - name: duration
      dtype: float32
    - name: language
      dtype: string
    - name: original_speaker_id
      dtype: int64
    - name: session_id
      dtype: int64
    - name: topic
      dtype: string
  splits:
    - name: train
      num_bytes: 963574796.7957073
      num_examples: 9373
    - name: test
      num_bytes: 94867191.1518061
      num_examples: 1175
    - name: validation
      num_bytes: 101196885.98946902
      num_examples: 1071
  download_size: 1208913346
  dataset_size: 1159638873.9369824
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: validation
        path: data/validation-*