---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: caption
    dtype: string
  - name: image_name
    dtype: string
  - name: source_folder
    dtype: string
  - name: responses
    dtype: string
  splits:
  - name: train
    num_bytes: 266115424.125
    num_examples: 4999
  download_size: 260908171
  dataset_size: 266115424.125
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---