metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
splits:
- name: train
num_bytes: 4700514636.20434
num_examples: 13231
- name: test
num_bytes: 261475443.6871174
num_examples: 736
- name: valid
num_bytes: 261119632.81254256
num_examples: 735
download_size: 5127814508
dataset_size: 5223109712.704
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*