---
dataset_info:
  features:
  - name: scene_id
    dtype: string
  - name: frame_id
    dtype: string
  - name: qa_id
    dtype: string
  - name: scene_description
    dtype: string
  - name: category
    dtype: string
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: context
    dtype: string
  - name: tag
    list: int32
  - name: cluster
    dtype: string
  - name: layer
    dtype: string
  - name: con_up
    dtype: string
  - name: con_down
    dtype: string
  - name: images
    struct:
    - name: cam_front
      dtype: image
    - name: cam_front_left
      dtype: image
    - name: cam_front_right
      dtype: image
    - name: cam_back
      dtype: image
    - name: cam_back_left
      dtype: image
    - name: cam_back_right
      dtype: image
  - name: key_objects
    dtype: string
  - name: num_key_objects
    dtype: int32
  - name: available_cameras
    list: string
  splits:
  - name: train
    num_bytes: 5408179.0
    num_examples: 10
  download_size: 555344
  dataset_size: 5408179.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---