README.md

---
dataset_info:
  features:
  - name: question
    dtype: string
  - name: options
    struct:
    - name: A
      dtype: string
    - name: B
      dtype: string
    - name: C
      dtype: string
    - name: D
      dtype: string
  - name: answer
    dtype: string
  - name: reasoning
    dtype: string
  - name: source_split
    dtype: string
  - name: source_idx
    dtype: int64
  - name: messages
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
  splits:
  - name: train
    num_bytes: 32356038
    num_examples: 20665
  download_size: 26572278
  dataset_size: 32356038
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---