westbrook's picture
Upload dataset
4fc6f20 verified
metadata
dataset_info:
  features:
    - name: segment_id
      dtype: string
    - name: speaker
      dtype: string
    - name: text
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: begin_time
      dtype: float32
    - name: end_time
      dtype: float32
    - name: audio_id
      dtype: string
    - name: title
      dtype: string
    - name: url
      dtype: string
    - name: source
      dtype:
        class_label:
          names:
            '0': audiobook
            '1': podcast
            '2': youtube
    - name: category
      dtype:
        class_label:
          names:
            '0': People  and  Blogs
            '1': Business
            '2': Nonprofits  and  Activism
            '3': Crime
            '4': History
            '5': Pets  and  Animals
            '6': News and Politics
            '7': Travel and Events
            '8': Kids and Family
            '9': Leisure
            '10': N/A
            '11': Comedy
            '12': News  and  Politics
            '13': Sports
            '14': Arts
            '15': Science  and  Technology
            '16': Autos  and  Vehicles
            '17': Science and Technology
            '18': People and Blogs
            '19': Music
            '20': Society and Culture
            '21': Education
            '22': Howto  and  Style
            '23': Film  and  Animation
            '24': Gaming
            '25': Entertainment
            '26': Travel  and  Events
            '27': Health and Fitness
            '28': audiobook
    - name: original_full_path
      dtype: string
    - name: utterance_pitch_mean
      dtype: float32
    - name: utterance_pitch_std
      dtype: float32
    - name: snr
      dtype: float64
    - name: c50
      dtype: float64
    - name: speaking_rate
      dtype: float64
    - name: phonemes
      dtype: string
    - name: stoi
      dtype: float64
    - name: si-sdr
      dtype: float64
    - name: pesq
      dtype: float64
  splits:
    - name: train
      num_bytes: 1056699
      num_examples: 10
  download_size: 1059252
  dataset_size: 1056699
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*