---
dataset_info:
  features:
  - name: split
    dtype: large_string
  - name: label
    dtype: int64
  - name: video_rel
    dtype: large_string
  - name: frame_idx
    dtype: int64
  - name: image
    dtype: image
  - name: face_detected
    dtype: bool
  - name: box
    dtype: large_string
  - name: subject_ids
    dtype: large_string
  splits:
  - name: train
    num_bytes: 3823649411
    num_examples: 389159
  - name: validation
    num_bytes: 643658387
    num_examples: 68661
  - name: test
    num_bytes: 405424881
    num_examples: 39851
  download_size: 4773685758
  dataset_size: 4872732679
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---