README.md

---
dataset_info:
  features:
  - name: data_source
    dtype: large_string
  - name: prompt
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: ability
    dtype: large_string
  - name: reward_model
    struct:
    - name: ground_truth
      dtype: string
    - name: style
      dtype: string
  - name: extra_info
    struct:
    - name: contest_id
      dtype: string
    - name: idx
      dtype: string
    - name: question_source
      dtype: string
    - name: source
      dtype: string
    - name: tags
      dtype: string
  splits:
  - name: train
    num_bytes: 3724032203
    num_examples: 434888
  download_size: 717841308
  dataset_size: 3724032203
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---