---
license: apache-2.0
dataset_info:
  config_name: zh-TW
  features:
  - name: path
    dtype: string
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: sentence
    dtype: string
  - name: condition_on_prev
    sequence: int64
  - name: whisper_transcript
    dtype: string
  splits:
  - name: train
    num_bytes: 691009749.0
    num_examples: 766
  - name: validation
    num_bytes: 501044472.0
    num_examples: 554
  - name: test
    num_bytes: 552470080.0
    num_examples: 617
  download_size: 1560691983
  dataset_size: 1744524301.0
configs:
- config_name: zh-TW
  data_files:
  - split: train
    path: zh-TW/train-*
  - split: validation
    path: zh-TW/validation-*
  - split: test
    path: zh-TW/test-*
---