metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 1369464434.6269412
num_examples: 2022
- name: test
num_bytes: 419044632.0470588
num_examples: 579
- name: valid
num_bytes: 212957816.4
num_examples: 289
download_size: 1330402435
dataset_size: 2001466883.0740001
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*