---
dataset_info:
  features:
  - name: image
    sequence:
      sequence:
        sequence: float32
  - name: labels
    sequence: float32
  splits:
  - name: train
    num_bytes: 46536731872
    num_examples: 26231
  download_size: 5154471321
  dataset_size: 46536731872
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---