---
dataset_info:
  features:
  - name: file
    dtype: string
  - name: audio
    dtype: audio
  - name: label
    dtype: string
  - name: instruction
    dtype: string
  splits:
  - name: train
    num_bytes: 11845511188.0
    num_examples: 27002
  - name: validation
    num_bytes: 1304884322.0
    num_examples: 3001
  download_size: 12631265875
  dataset_size: 13150395510.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---