metadata
dataset_info:
features:
- name: filename
dtype: string
- name: audio
dtype: audio
- name: transcription
dtype: string
configs:
- config_name: default
data_files:
- split: train
path: metadata.csv