metadata
dataset_info:
features:
- name: track_id
dtype: string
- name: tags
sequence: string
- name: texts
sequence: string
- name: audio
struct:
- name: array
sequence: float32
- name: sampling_rate
dtype: int64
splits:
- name: train
num_bytes: 2648618684
num_examples: 3000
- name: test
num_bytes: 264861630
num_examples: 300
download_size: 2915723150
dataset_size: 2913480314
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*