---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: height
    dtype: int64
  - name: width
    dtype: int64
  - name: caption_caption
    dtype: string
  splits:
  - name: train
    num_bytes: 135063994102.0
    num_examples: 1745731
  download_size: 134800766095
  dataset_size: 135063994102.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---