metadata
dataset_info:
features:
- name: context
list:
list: binary
- name: answer_key
dtype: int64
- name: caption_choices
list: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 619930689
num_examples: 1024
download_size: 619859437
dataset_size: 619930689
configs:
- config_name: default
data_files:
- split: train
path: data/train-*