metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: song_id
dtype: string
- name: structure
dtype: string
- name: start_time
dtype: float64
- name: end_time
dtype: float64
- name: tempos
dtype: string
- name: keys
dtype: string
- name: instruments
dtype: string
- name: genres
dtype: string
- name: moods
dtype: string
- name: metadata
dtype: string
splits:
- name: train
num_bytes: 797925040
num_examples: 831
- name: eval
num_bytes: 15363261
num_examples: 16
download_size: 801646719
dataset_size: 813288301
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*