metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: language
struct:
- name: language
dtype: string
splits:
- name: train
num_bytes: 32443542061.635834
num_examples: 79786
download_size: 31344423962
dataset_size: 32443542061.635834
configs:
- config_name: default
data_files:
- split: train
path: data/train-*