---
dataset_info:
  features:
  - name: messages
    dtype: string
  - name: image_path
    dtype: string
  - name: points
    sequence:
      sequence: float32
      length: 2
  - name: counts
    dtype: int32
  splits:
  - name: train
    num_bytes: 235089822
    num_examples: 19026
  - name: test
    num_bytes: 56350292
    num_examples: 7202
  download_size: 136631283
  dataset_size: 291440114
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---