metadata
dataset_info:
features:
- name: questionId
dtype: int64
- name: question
dtype: string
- name: question_types
dtype: string
- name: image
dtype: string
- name: docId
dtype: int64
- name: ucs_document_id
dtype: string
- name: ucs_document_page_no
dtype: int64
- name: answers
dtype: string
splits:
- name: train
num_bytes: 3941631
num_examples: 8
download_size: 1449484
dataset_size: 3941631
configs:
- config_name: default
data_files:
- split: train
path: data/train-*