metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 188521682.69274625
num_examples: 2386
- name: test
num_bytes: 25075262.3871941
num_examples: 299
- name: valid
num_bytes: 24638983.83405967
num_examples: 298
download_size: 229984891
dataset_size: 238235928.91400003
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*