metadata
dataset_info:
features:
- name: image_id
dtype: int64
- name: split
dtype: string
- name: sentences
list:
- name: raw
dtype: string
- name: sent
dtype: string
- name: sent_id
dtype: int64
- name: tokens
list: string
- name: file_name
dtype: string
- name: category_id
dtype: int64
- name: ann_id
dtype: int64
- name: sent_ids
list: int64
- name: ref_id
dtype: int64
- name: raw_anns
dtype: string
- name: raw_image_info
dtype: string
- name: raw_sentences
dtype: string
- name: image_path
dtype: string
- name: bbox
list: float64
- name: captions
list: string
- name: global_image_id
dtype: string
- name: anns_id
dtype: string
- name: sentences_ko
list:
- name: ko
dtype: string
- name: sent
dtype: string
splits:
- name: train
num_bytes: 95472933
num_examples: 42226
download_size: 41038453
dataset_size: 95472933
configs:
- config_name: default
data_files:
- split: train
path: data/train-*