metadata
dataset_info:
features:
- name: file
dtype: string
- name: audio
dtype: audio
- name: instruction
dtype: string
- name: label
dtype: string
splits:
- name: train
num_bytes: 266185539
num_examples: 2000
- name: validation
num_bytes: 65551361
num_examples: 500
download_size: 316150982
dataset_size: 331736900
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*