README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 24000
    - name: sentence
      dtype: string
  splits:
    - name: train
      num_bytes: 23554422884.469326
      num_examples: 637736
    - name: test
      num_bytes: 5844567487.548676
      num_examples: 159435
  download_size: 26651689028
  dataset_size: 29398990372.018
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
language:
  - th
task_categories:
  - automatic-speech-recognition
license: cc-by-4.0