---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: idx
    dtype: int64
  - name: word
    dtype: string
  splits:
  - name: train
    num_bytes: 206639400.0
    num_examples: 100000
  - name: test
    num_bytes: 19683520.0
    num_examples: 10000
  download_size: 173346403
  dataset_size: 226322920.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---