---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: topdown
    dtype: image
  - name: up
    dtype: image
  - name: right
    dtype: image
  - name: down
    dtype: image
  - name: left
    dtype: image
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 185084667.0
    num_examples: 416
  download_size: 185095532
  dataset_size: 185084667.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---