metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: dataset
dtype: string
- name: text
dtype: string
splits:
- name: train
- name: evaluation
- name: test
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: evaluation
path: data/evaluation-*
- split: test
path: data/test-*