---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: input
    dtype: string
  - name: audio_id
    dtype: string
  - name: dataset
    dtype: string
  - name: task
    dtype: string
  - name: output
    dtype: string
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: question
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 18989646282.934
    num_examples: 29999
  - name: test
    num_bytes: 2848131008.24
    num_examples: 4505
  download_size: 21183565716
  dataset_size: 21837777291.173996
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---