---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: label
    dtype: int32
  splits:
  - name: train
    num_bytes: 154806589362.012
    num_examples: 1281167
  - name: val
    num_bytes: 6724025350.0
    num_examples: 50000
  download_size: 153224015481
  dataset_size: 161530614712.012
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: val
    path: data/val-*
---