README.md

---
dataset_info:
  features:
  - name: data_source
    dtype: string
  - name: prompt
    list:
    - name: role
      dtype: string
    - name: content
      dtype: string
  - name: reward_model
    struct:
    - name: style
      dtype: string
    - name: ground_truth
      dtype: string
  - name: problem_idx
    dtype: int64
  splits:
  - name: dsr
    num_bytes: 14805760
    num_examples: 40308
  - name: dapo_en
    num_bytes: 5649260
    num_examples: 14116
  - name: aime24
    num_bytes: 13321
    num_examples: 30
  - name: amc23
    num_bytes: 15438
    num_examples: 40
  - name: math500
    num_bytes: 159412
    num_examples: 500
  - name: aime25
    num_bytes: 17869
    num_examples: 30
  - name: minerva
    num_bytes: 149829
    num_examples: 272
  - name: olympiad
    num_bytes: 296681
    num_examples: 675
  - name: gpqa
    num_bytes: 116745
    num_examples: 198
  download_size: 20500591
  dataset_size: 21224315
configs:
- config_name: default
  data_files:
  - split: dsr
    path: data/dsr-*
  - split: dapo_en
    path: data/dapo_en-*
  - split: aime24
    path: data/aime24-*
  - split: amc23
    path: data/amc23-*
  - split: math500
    path: data/math500-*
  - split: aime25
    path: data/aime25-*
  - split: minerva
    path: data/minerva-*
  - split: olympiad
    path: data/olympiad-*
  - split: gpqa
    path: data/gpqa-*
---