metadata
dataset_info:
features:
- name: song_id
dtype: int32
- name: audio
dtype:
audio:
sampling_rate: 44100
- name: arousal_annotations
sequence: float32
- name: valence_annotations
sequence: float32
splits:
- name: train
num_bytes: 14317361526
num_examples: 1802
download_size: 16088247994
dataset_size: 14317361526
configs:
- config_name: default
data_files:
- split: train
path: data/train-*