metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 43453418573.792
num_examples: 45512
download_size: 25072136318
dataset_size: 43453418573.792
configs:
- config_name: default
data_files:
- split: train
path: data/train-*