metadata
dataset_info:
features:
- name: id
dtype: string
- name: audio
dtype: audio
- name: audio_language
dtype: string
- name: text
dtype: string
- name: prompt
dtype: string
- name: duration
dtype: float64
- name: speaker_id
dtype: string
splits:
- name: train
num_bytes: 37299338467.827
num_examples: 261657
- name: test
num_bytes: 1290149510.445
num_examples: 9265
- name: dev_test
num_bytes: 1288643290.621
num_examples: 9263
download_size: 40101372264
dataset_size: 39878131268.893005
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: dev_test
path: data/dev_test-*