---
dataset_info:
  features:
  - name: prompt_messages
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: answers
    sequence: string
  - name: gt
    dtype: string
  - name: rewards
    sequence: float64
  splits:
  - name: train
    num_bytes: 189878
    num_examples: 34
  download_size: 75068
  dataset_size: 189878
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---