---
dataset_info:
  features:
  - name: image_id
    dtype: string
  - name: output
    dtype: string
  - name: instruction
    dtype: string
  splits:
  - name: train
    num_bytes: 267872675
    num_examples: 548825
  download_size: 33975147
  dataset_size: 267872675
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---