k_whisper_dataset / README.md
k-seungri's picture
Upload dataset
34a905b verified
metadata
dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: transcripts
      dtype: string
  splits:
    - name: train
      num_bytes: 7859062.290076337
      num_examples: 104
    - name: test
      num_bytes: 972595.9236641221
      num_examples: 14
    - name: valid
      num_bytes: 1338620.786259542
      num_examples: 13
  download_size: 8246363
  dataset_size: 10170279
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: valid
        path: data/valid-*