metadata
dataset_info:
features:
- name: questionId
dtype: int32
- name: question
dtype: string
- name: question_types
list: string
- name: image
dtype: image
- name: docId
dtype: int32
- name: ucsf_document_id
dtype: string
- name: ucsf_document_page_no
dtype: int32
- name: answers
list: string
splits:
- name: train
num_bytes: 142431357.4
num_examples: 378
- name: test
num_bytes: 13684308.6
num_examples: 42
download_size: 116069595
dataset_size: 156115666
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*