metadata
dataset_info:
features:
- name: client_id
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 4815147523.5
num_examples: 41604
- name: val
num_bytes: 1439996858.125
num_examples: 10559
download_size: 5660284094
dataset_size: 6255144381.625
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*