README.md

---
dataset_info:
  features:
  - name: env
    dtype: string
  - name: prompt
    dtype: string
  - name: extra
    struct:
    - name: game_data
      struct:
      - name: answer
        dtype: string
      - name: difficulty
        dtype: int64
      - name: gpt_response
        dtype: string
      - name: metadata
        struct:
        - name: dyck_sequence
          dtype: string
        - name: error_indices
          list: int64
        - name: n_errors
          dtype: int64
        - name: n_types
          dtype: int64
        - name: seed
          dtype: int64
        - name: thoughts
          list: string
        - name: total_length
          dtype: int64
        - name: trace_id
          dtype: string
      - name: question
        dtype: string
    - name: metadata
      struct:
      - name: dyck_sequence
        dtype: string
      - name: error_indices
        list: int64
      - name: n_errors
        dtype: int64
      - name: n_types
        dtype: int64
      - name: seed
        dtype: int64
      - name: thoughts
        list: string
      - name: total_length
        dtype: int64
      - name: trace_id
        dtype: string
    - name: seed
      dtype: int64
    - name: task_id
      dtype: int64
    - name: task_type
      dtype: string
  splits:
  - name: train
    num_bytes: 19265546
    num_examples: 4138
  download_size: 14319816
  dataset_size: 19265546
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---