metadata
dataset_info:
features:
- name: track_id
dtype: string
- name: pseudo_caption
dtype: string
- name: tag_list
sequence: string
- name: tag_mid
sequence: string
- name: audioset_tag
sequence: string
- name: audioset_mid
sequence: string
- name: path
dtype: string
splits:
- name: balanced_train
num_bytes: 1518062
num_examples: 5688
- name: unbalanced_train
num_bytes: 115369887
num_examples: 399369
download_size: 33427982
dataset_size: 116887949
configs:
- config_name: default
data_files:
- split: balanced_train
path: data/balanced_train-*
- split: unbalanced_train
path: data/unbalanced_train-*