---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: answer
    dtype: string
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  splits:
  - name: train
    num_bytes: 345978043.0
    num_examples: 2000
  - name: validation
    num_bytes: 82306957.0
    num_examples: 500
  - name: test
    num_bytes: 542140907.0
    num_examples: 3000
  download_size: 975519022
  dataset_size: 970425907.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---