metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 4673179027.482265
num_examples: 5146
- name: test
num_bytes: 593177419.5930359
num_examples: 644
- name: valid
num_bytes: 585191953.8886989
num_examples: 643
download_size: 2283219574
dataset_size: 5851548400.964
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*