---
dataset_info:
  features:
  - name: year
    dtype: int64
  - name: day
    dtype: string
  - name: problem_name
    dtype: string
  - name: problem_id
    dtype: string
  - name: target_subtask
    dtype: string
  - name: instruction_seed
    dtype: string
  - name: generation
    dtype: string
  - name: uuid
    dtype: string
  - name: metadata
    struct:
    - name: completion_tokens
      dtype: int64
    - name: prompt_tokens
      dtype: int64
    - name: prompt_tokens_details
      dtype: 'null'
    - name: total_tokens
      dtype: int64
  - name: finish_reason
    dtype: string
  - name: code
    dtype: string
  - name: code_compiles
    dtype: bool
  - name: target_subtask_score
    dtype: float64
  - name: target_subtask_status
    dtype: string
  - name: all_subtasks_points
    dtype: float64
  - name: all_subtasks_results
    list:
    - name: points
      dtype: int64
    - name: problem
      dtype: string
    - name: score
      dtype: float64
    - name: score_precision
      dtype: int64
    - name: status
      dtype: string
    - name: subtask
      dtype: string
    - name: test_results
      list:
      - name: feedback
        dtype: string
      - name: score
        dtype: float64
      - name: status
        dtype: string
      - name: test_name
        dtype: string
    - name: weighted_score
      dtype: float64
  - name: messages
    list:
    - name: content
      dtype: string
    - name: role
      dtype: string
  - name: reasoning
    dtype: string
  - name: deepseek_solution
    dtype: string
  - name: __original_row_idx
    dtype: int64
  - name: source
    dtype: string
  - name: final_reasoning_trace
    dtype: string
  - name: conversations
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  splits:
  - name: train
    num_bytes: 11725715624
    num_examples: 31600
  download_size: 3641169503
  dataset_size: 11725715624
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---