---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: question
    dtype: string
  - name: ground
    dtype: string
  - name: prediction
    dtype: string
  - name: similarity
    dtype: float64
  splits:
  - name: train
    num_bytes: 1446454474.5
    num_examples: 2460
  download_size: 635457102
  dataset_size: 1446454474.5
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---