metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
- name: locale
dtype: string
splits:
- name: train
num_bytes: 4025909295.625
num_examples: 18347
- name: test
num_bytes: 573716568.5
num_examples: 10700
- name: validation
num_bytes: 210877557.784
num_examples: 1304
download_size: 5366410927
dataset_size: 4810503421.909
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*