---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: pos_caption
    dtype: string
  - name: neg_caption
    dtype: string
  - name: pos_count
    dtype: int64
  - name: neg_count
    dtype: int64
  - name: pos_points
    sequence:
      sequence: float64
  - name: neg_points
    sequence:
      sequence: float64
  - name: annotated_pos_count
    dtype: int64
  - name: annotated_neg_count
    dtype: int64
  - name: image_name
    dtype: string
  - name: category
    dtype: string
  - name: video_id
    dtype: string
  - name: type
    dtype: string
  splits:
  - name: FOO
    num_bytes: 3103917428.0
    num_examples: 32200
  - name: HOU
    num_bytes: 4607723746.0
    num_examples: 33250
  - name: FUN
    num_bytes: 6789751584.0
    num_examples: 32200
  - name: 'OFF'
    num_bytes: 7965510202.0
    num_examples: 30100
  - name: OTR
    num_bytes: 8902109740.0
    num_examples: 28000
  download_size: 15724972893
  dataset_size: 31369012700.0
configs:
- config_name: default
  data_files:
  - split: FOO
    path: data/FOO-*
  - split: HOU
    path: data/HOU-*
  - split: FUN
    path: data/FUN-*
  - split: 'OFF'
    path: data/OFF-*
  - split: OTR
    path: data/OTR-*
---