README.md

---
dataset_info:
  features:
  - name: id
    dtype: string
  - name: goal
    dtype: string
  - name: start_url
    dtype: string
  - name: backend
    dtype: string
  - name: model
    dtype: string
  - name: success
    dtype: bool
  - name: num_steps
    dtype: int64
  - name: num_messages
    dtype: int64
  - name: tools
    dtype: string
  - name: messages
    list:
    - name: content
      dtype: string
    - name: name
      dtype: string
    - name: role
      dtype: string
    - name: thinking
      dtype: string
    - name: tool_call_id
      dtype: string
    - name: tool_calls
      dtype: string
  splits:
  - name: train
    num_bytes: 709236
    num_examples: 4
  download_size: 719589
  dataset_size: 709236
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---