metadata
dataset_info:
features:
- name: file_id
dtype: string
- name: speaker_id
dtype: string
- name: text
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: duration
dtype: float32
splits:
- name: train
num_bytes: 6201639134.875
num_examples: 60417
- name: validation
num_bytes: 773233790.5
num_examples: 7676
- name: test
num_bytes: 773416270.625
num_examples: 7715
download_size: 7661635292
dataset_size: 7748289196
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*