---
dataset_info:
  features:
  - name: problem
    dtype: string
  - name: reasoning_trace
    dtype: string
  - name: reference_answer
    dtype: string
  - name: finish_thinking
    dtype: bool
  - name: reward
    dtype: float64
  - name: llm_score
    dtype: int64
  - name: llm_response
    dtype: string
  splits:
  - name: train
    num_bytes: 135377955
    num_examples: 3560
  download_size: 45095270
  dataset_size: 135377955
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---