README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: text
      dtype: string
    - name: duration
      dtype: float32
  splits:
    - name: train
      num_bytes: 338726085.76
      num_examples: 3336
    - name: test
      num_bytes: 130852089
      num_examples: 444
    - name: dev
      num_bytes: 21530480
      num_examples: 156
  download_size: 498877304
  dataset_size: 491108654.76
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: dev
        path: data/dev-*
task_categories:
  - automatic-speech-recognition
language:
  - ar

train hours: 3.098
test hours: 1.135
dev hours: 0.187