metadata
dataset_info:
features:
- name: image
dtype: image
- name: image_url
dtype: string
- name: embedding
list: float64
- name: metadata_url
dtype: string
- name: original_height
dtype: int64
- name: original_width
dtype: int64
- name: mime_type
dtype: string
- name: caption_attribution_description
dtype: string
- name: wit_features
struct:
- name: attribution_passes_lang_id
list: bool
- name: caption_alt_text_description
list: string
- name: caption_reference_description
list: string
- name: caption_title_and_reference_description
list: string
- name: context_page_description
list: string
- name: context_section_description
list: string
- name: hierarchical_section_title
list: string
- name: is_main_image
list: bool
- name: language
list: string
- name: page_changed_recently
list: bool
- name: page_title
list: string
- name: page_url
list: string
- name: section_title
list: string
splits:
- name: train
num_bytes: 422361
num_examples: 10
download_size: 418390
dataset_size: 422361
configs:
- config_name: default
data_files:
- split: train
path: data/train-*