arda-argmax's picture
Upload dataset
f32df61 verified
|
raw
history blame
720 Bytes
metadata
dataset_info:
  config_name: id
  features:
    - name: path
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: sentence
      dtype: string
    - name: condition_on_prev
      sequence: int64
    - name: whisper_transcript
      dtype: string
  splits:
    - name: train
      num_bytes: 886887591.415
      num_examples: 1035
    - name: validation
      num_bytes: 445769434
      num_examples: 505
    - name: test
      num_bytes: 479298192
      num_examples: 542
  download_size: 1688264749
  dataset_size: 1811955217.415
configs:
  - config_name: id
    data_files:
      - split: train
        path: id/train-*
      - split: validation
        path: id/validation-*
      - split: test
        path: id/test-*