metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 2762876771.8632507
num_examples: 33117
- name: test
num_bytes: 682890832.6977491
num_examples: 8280
download_size: 2796541093
dataset_size: 3445767604.561
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*