---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: keypoints
    sequence:
      sequence:
        sequence: float64
  - name: kp_normalized
    sequence:
      sequence:
        sequence: float64
  - name: ground_truth
    sequence: string
  splits:
  - name: train
    num_bytes: 23075325831.5
    num_examples: 33100
  - name: test
    num_bytes: 822691406.68
    num_examples: 3285
  download_size: 21626670217
  dataset_size: 23898017238.18
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---