metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 20626690917.075
num_examples: 22833
download_size: 14408578835
dataset_size: 20626690917.075
configs:
- config_name: default
data_files:
- split: train
path: data/train-*