metadata
dataset_info:
features:
- name: image
dtype: image
- name: frame_id
dtype: string
- name: video_id
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: question_type
dtype: string
splits:
- name: train
num_bytes: 251267471
num_examples: 360
- name: validation
num_bytes: 31407904
num_examples: 45
- name: test
num_bytes: 31408966
num_examples: 45
download_size: 314085572
dataset_size: 314084341
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*