---
dataset_info:
  features:
  - name: context
    dtype:
      audio:
        sampling_rate: 16000
  - name: instruction
    dtype: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 3425832242.0
    num_examples: 20000
  - name: test
    num_bytes: 373560758.0
    num_examples: 2000
  download_size: 3715675082
  dataset_size: 3799393000.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---