README.md

---
dataset_info:
  features:
  - name: task_name
    dtype: string
  - name: score
    dtype: float64
  - name: success
    dtype: bool
  - name: time_taken
    dtype: float64
  - name: extra
    struct:
    - name: challenge
      struct:
      - name: env
        dtype: string
      - name: extra
        struct:
        - name: base_url
          dtype: string
        - name: dataset_index
          dtype: int64
        - name: ground_truth
          dtype: string
        - name: inputs
          dtype: string
        - name: model
          dtype: string
        - name: seed
          dtype: int64
        - name: task_id
          dtype: int64
        - name: transformed_code
          dtype: string
      - name: prompt
        dtype: string
      - name: timestamp
        dtype: float64
    - name: conversation
      list:
      - name: content
        dtype: string
      - name: role
        dtype: string
    - name: dataset_index
      dtype: int64
    - name: seed
      dtype: int64
    - name: task_id
      dtype: int64
    - name: task_type
      dtype: string
    - name: test_result
      dtype: string
    - name: usage
      struct:
      - name: completion_tokens
        dtype: int64
      - name: completion_tokens_details
        dtype: 'null'
      - name: prompt_tokens
        dtype: int64
      - name: prompt_tokens_details
        dtype: 'null'
      - name: total_tokens
        dtype: int64
  splits:
  - name: train
    num_bytes: 38961161
    num_examples: 47
  download_size: 2287022
  dataset_size: 38961161
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---