metadata
dataset_info:
features:
- name: audio
dtype: audio
- 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: 26462307
num_examples: 10
- name: eval
num_bytes: 84679315
num_examples: 32
download_size: 108368116
dataset_size: 111141622
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*