---
dataset_info:
  features:
  - name: jpeg_bytes
    sequence: uint16
  - name: caption
    sequence: uint16
  splits:
  - name: train
    num_bytes: 112562842
    num_examples: 100001
  download_size: 70758865
  dataset_size: 112562842
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---