README.md

---
dataset_info:
  features:
  - name: question
    dtype: large_string
  - name: answer
    dtype: large_string
  - name: messages
    large_list:
    - name: role
      dtype: large_string
    - name: content
      dtype: large_string
    - name: tool_calls
      large_list:
      - name: id
        dtype: large_string
      - name: type
        dtype: large_string
      - name: function
        struct:
        - name: name
          dtype: large_string
        - name: arguments
          dtype: large_string
    - name: reasoning_content
      dtype: large_string
    - name: tool_call_id
      dtype: large_string
    - name: name
      dtype: large_string
  - name: model
    dtype: large_string
  - name: source
    dtype: large_string
  - name: license
    dtype: large_string
  - name: topic
    dtype: large_string
  - name: tokens_len
    dtype: int64
  - name: success_rate
    dtype: float64
  - name: tools_run_time
    large_list: float64
  - name: tool
    large_list:
    - name: type
      dtype: large_string
    - name: function
      struct:
      - name: name
        dtype: large_string
      - name: description
        dtype: large_string
      - name: parameters
        struct:
        - name: type
          dtype: large_string
        - name: properties
          struct:
          - name: code
            struct:
            - name: type
              dtype: large_string
            - name: description
              dtype: large_string
        - name: required
          large_list: large_string
  - name: libraries
    large_list: large_string
  - name: transform
    dtype: large_string
  - name: gen_kwarg
    struct:
    - name: temperature
      dtype: float64
    - name: top_p
      dtype: float64
    - name: top_k
      dtype: int64
    - name: presence_penalty
      dtype: int64
    - name: repetition_penalty
      dtype: int64
  splits:
  - name: train
    num_bytes: 411241355
    num_examples: 6241
  download_size: 403060880
  dataset_size: 411241355
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---