metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: speaker
dtype: string
- name: duration
dtype: float32
splits:
- name: test
num_bytes: 572971043
num_examples: 5032
- name: validation
num_bytes: 37659237.06560636
num_examples: 495
- name: train
num_bytes: 1244637895.41061
num_examples: 9260
download_size: 14725798914
dataset_size: 1855268175.476216
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*