---
dataset_info:
  features:
  - name: pixel_values
    sequence:
      sequence:
        sequence: float32
  - name: label
    dtype: int64
  splits:
  - name: train
    num_bytes: 14337352920
    num_examples: 23705
  download_size: 3114082212
  dataset_size: 14337352920
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---