---
dataset_info:
  features:
  - name: data_source
    dtype: string
  - name: prompt
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: images
    list: image
  - name: reward_model
    struct:
    - name: ground_truth
      dtype: string
    - name: style
      dtype: string
  - name: extra_info
    struct:
    - name: gt_answer
      dtype: string
    - name: id
      dtype: string
    - name: mc_options
      list: string
    - name: original_question
      dtype: string
  splits:
  - name: train
    num_bytes: 83426778.0
    num_examples: 238
  download_size: 83312269
  dataset_size: 83426778.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---