metadata
dataset_info:
features:
- name: query
dtype: string
- name: answer
dtype: string
- name: page_contains_table
dtype: bool
- name: page_contains_figure
dtype: bool
- name: page_is_title_or_toc
dtype: bool
- name: quality
dtype: int32
- name: model_name
dtype: string
- name: document_filename
dtype: string
- name: document_url
dtype: string
- name: page_language
dtype: string
- name: page_number
dtype: int32
- name: page_image
dtype: image
splits:
- name: train
num_bytes: 4892152941.46
num_examples: 17035
download_size: 1944661464
dataset_size: 4892152941.46
configs:
- config_name: default
data_files:
- split: train
path: data/train-*