README.md

---
dataset_info:
  features:
  - name: image_id
    dtype: string
  - name: level
    dtype: string
  - name: question
    dtype: string
  - name: question_id
    dtype: string
  - name: image_classes
    sequence: string
  - name: flickr_original_url
    dtype: string
  - name: flickr_300k_url
    dtype: string
  - name: image_height
    dtype: int64
  - name: image_width
    dtype: int64
  - name: set_name
    dtype: string
  - name: image_name
    dtype: string
  - name: image_path
    dtype: string
  - name: image
    dtype: image
  - name: caption_id
    sequence: string
  - name: caption_str
    sequence: string
  - name: reference_strs
    sequence: string
  - name: description
    dtype: string
  splits:
  - name: train
    num_bytes: 122279587.4
    num_examples: 1200
  download_size: 120760597
  dataset_size: 122279587.4
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---