---
dataset_info:
  features:
  - name: image_1
    dtype: image
  - name: image_2
    dtype: image
  - name: count
    dtype: int32
  - name: common_points
    sequence:
      sequence: float32
  - name: H
    dtype: int32
  - name: W
    dtype: int32
  - name: ratio_h
    dtype: float32
  - name: r
    sequence: int32
  - name: question
    dtype: string
  splits:
  - name: train
    num_bytes: 2183003842.837
    num_examples: 33489
  download_size: 1124236603
  dataset_size: 2183003842.837
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---