metadata
dataset_info:
features:
- name: text
dtype: string
- name: image
dtype: image
- name: audio
dtype: audio
- name: tag
dtype: string
- name: text_id
dtype: string
splits:
- name: train
num_examples: 2403
From https://github.com/audioset/ontology More Information needed