---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: facial_landmarks
    struct:
    - name: lefteye
      list: int64
    - name: righteye
      list: int64
    - name: nose
      list: int64
    - name: leftmouth
      list: int64
    - name: rightmouth
      list: int64
  splits:
  - name: train
    num_bytes: 1148586180
    num_examples: 162770
  - name: validation
    num_bytes: 142152915
    num_examples: 19867
  - name: test
    num_bytes: 141252929
    num_examples: 19962
  download_size: 1430124101
  dataset_size: 1431992024
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---