---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: caption
    dtype: string
  - name: filename
    dtype: string
  - name: image_id
    dtype: int64
  splits:
  - name: train
    num_bytes: 257381003
    num_examples: 5000
  download_size: 257038293
  dataset_size: 257381003
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---