metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 2581392022.0813613
num_examples: 31932
- name: test
num_bytes: 664600888.8786387
num_examples: 7984
download_size: 2664828216
dataset_size: 3245992910.96
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*