README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: sentence
      dtype: string
  splits:
    - name: train
      num_bytes: 6706332150.844
      num_examples: 17308
    - name: test
      num_bytes: 756830343.164
      num_examples: 1924
  download_size: 7310372816
  dataset_size: 7463162494.007999
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
license: apache-2.0
task_categories:
  - automatic-speech-recognition
language:
  - ja
tags:
  - japanese
  - proccessed
size_categories:
  - 1K<n<10K