metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: song_id
dtype: int64
- name: genre_id
dtype: int64
- name: genre
dtype: string
- name: vocals
dtype: audio
- name: others
dtype: audio
splits:
- name: train
num_bytes: 795140764
num_examples: 100
- name: other
num_bytes: 365764632.76
num_examples: 46
- name: clean
num_bytes: 429375873.24
num_examples: 54
download_size: 1501535930
dataset_size: 1590281270
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: other
path: data/other-*
- split: clean
path: data/clean-*