metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: language
dtype: string
splits:
- name: train
num_bytes: 4562162875
num_examples: 158735
- name: validation
num_bytes: 571384179
num_examples: 19842
- name: test
num_bytes: 573444491
num_examples: 19845
download_size: 5578463246
dataset_size: 5706991545
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*