metadata
dataset_info:
features:
- name: news_link
dtype: string
- name: head_lines
dtype: string
- name: text
dtype: string
- name: tags
dtype: string
- name: image_caption
dtype: string
- name: label
dtype: string
splits:
- name: train
num_bytes: 1099945587.8719704
num_examples: 220289
- name: validation
num_bytes: 25329912.91509812
num_examples: 4993
- name: test
num_bytes: 74994678.44964029
num_examples: 14987
download_size: 458599988
dataset_size: 1200270179.2367089
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*