---
dataset_info:
  features:
  - name: image_index
    dtype: int64
  - name: qwen3-vl-4b-instruct
    dtype: string
  - name: gpt-5
    dtype: string
  - name: gpt-5-mini
    dtype: string
  - name: gemini-3-pro-preview
    dtype: string
  - name: gemini-3-flash-preview
    dtype: string
  - name: aya-vision-8b
    dtype: string
  - name: image
    dtype: image
  - name: verbovision_caption
    dtype: string
  splits:
  - name: train
    num_bytes: 240631757
    num_examples: 100
  download_size: 240353364
  dataset_size: 240631757
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---