metadata
dataset_info:
features:
- name: image_id
dtype: int64
- name: coco_url
dtype: string
- name: cogvlm2_caption
dtype: string
- name: Word_Count
dtype: int64
- name: Char_Count
dtype: int64
- name: num_tokens
dtype: int64
- name: label
dtype: int64
splits:
- name: train
num_bytes: 12210806
num_examples: 12501
download_size: 7185114
dataset_size: 12210806
configs:
- config_name: default
data_files:
- split: train
path: data/train-*