metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: genres
sequence: string
- name: artist
dtype: string
- name: vocal
dtype:
audio:
sampling_rate: 16000
- name: phonemes
dtype: string
splits:
- name: train
num_bytes: 19526012462.256
num_examples: 1708
download_size: 18730750730
dataset_size: 19526012462.256
configs:
- config_name: default
data_files:
- split: train
path: data/train-*