README.md

---
dataset_info:
  features:
  - name: sent_ids
    list: int64
  - name: file_name
    dtype: string
  - name: ann_id
    dtype: int64
  - name: ref_id
    dtype: int64
  - 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: category_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: 88095373
    num_examples: 42404
  download_size: 34425698
  dataset_size: 88095373
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---