metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
struct:
- name: gpt_output
dtype: string
- name: entity
dtype: string
- name: id
dtype: int64
splits:
- name: train
num_bytes: 11113586972.057
num_examples: 83807
- name: test
num_bytes: 1150585888.5
num_examples: 9310
download_size: 10766578094
dataset_size: 12264172860.557
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*