---
dataset_info:
  features:
  - name: chosen
    dtype: string
  - name: img_path
    dtype: string
  - name: prompt
    dtype: string
  - name: rejected
    dtype: string
  - name: question
    dtype: string
  - name: image
    dtype: image
  - name: generated_context
    dtype: string
  - name: right_context
    dtype: string
  - name: wrong_context
    dtype: string
  splits:
  - name: train
    num_bytes: 535717682.625
    num_examples: 9387
  download_size: 512700009
  dataset_size: 535717682.625
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---