metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 24000
- name: speaker_id
dtype: string
- name: text
dtype: string
- name: accent
dtype: string
- name: duration
dtype: float64
- name: age
dtype: string
- name: gender
dtype: string
- name: mel
sequence:
sequence:
sequence: float32
splits:
- name: test
num_bytes: 151365166
num_examples: 525
- name: val
num_bytes: 143174827
num_examples: 508
- name: train
num_bytes: 102673518514.14
num_examples: 369122
download_size: 104929399506
dataset_size: 102968058507.14
configs:
- config_name: default
data_files:
- split: test
path: data/test-*
- split: val
path: data/val-*
- split: train
path: data/train-*