---
dataset_info:
  features:
  - name: conversation_id
    dtype: int64
  - name: index
    dtype: int64
  - name: sub_index
    dtype: int64
  - name: question
    dtype: string
  - name: tools
    dtype: string
  - name: improved_system_prompt
    dtype: string
  - name: good_answer
    dtype: string
  - name: bad_answer
    dtype: string
  splits:
  - name: train
    num_bytes: 1286319
    num_examples: 64
  download_size: 209112
  dataset_size: 1286319
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---