---
dataset_info:
  features:
  - name: text
    dtype: string
  - name: summary
    dtype: string
  - name: image
    dtype: image
  - name: source_dataset
    dtype: string
  - name: original_split
    dtype: string
  - name: original_index
    dtype: int64
  splits:
  - name: train
    num_bytes: 1627301068
    num_examples: 20000
  download_size: 1598194547
  dataset_size: 1627301068
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---