---
license: mit
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
dataset_info:
  features:
  - name: image
    dtype: image
  - name: label
    dtype:
      class_label:
        names:
          '0': breathing
          '1': digging
          '2': seafloor_resting
          '3': surface_resting
          '4': swimming
  - name: label_name
    dtype: string
  - name: source
    dtype: string
  - name: frame_number
    dtype: int64
  splits:
  - name: train
    num_bytes: 103289331760.485
    num_examples: 318567
  download_size: 145861531253
  dataset_size: 103289331760.485
---