metadata
dataset_info:
features:
- name: track_id
dtype: string
- name: song_id
dtype: string
- name: artist_id
dtype: string
- name: artist_mbid
dtype: string
- name: title
dtype: string
- name: artist_name
dtype: string
- name: release
dtype: string
- name: year
dtype: int64
- name: tag_list
sequence: string
- name: pseudo_caption
dtype: string
- name: path
dtype: string
splits:
- name: train
num_bytes: 273872354
num_examples: 740317
download_size: 153775424
dataset_size: 273872354
configs:
- config_name: default
data_files:
- split: train
path: data/train-*