dimex100_light / README.md
carlosdanielhernandezmena's picture
Convert dataset to Parquet
265ff40 verified
|
raw
history blame
648 Bytes
metadata
license: cc-by-nc-nd-4.0
dataset_info:
  config_name: dimex100_light
  features:
    - name: audio_id
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: speaker_id
      dtype: string
    - name: utterance_type
      dtype: string
    - name: gender
      dtype: string
    - name: duration
      dtype: float32
    - name: normalized_text
      dtype: string
  splits:
    - name: train
      num_bytes: 495948632.083
      num_examples: 5999
  download_size: 459945384
  dataset_size: 495948632.083
configs:
  - config_name: dimex100_light
    data_files:
      - split: train
        path: dimex100_light/train-*
    default: true