README.md

---
dataset_info:
  features:
  - name: model
    dtype: string
  - name: provider
    dtype: string
  - name: prompt_tokens
    dtype: int64
  - name: response_tokens
    dtype: int64
  - name: run_id
    dtype: string
  - name: stage
    dtype: string
  - name: messages
    dtype: string
  - name: source
    dtype: string
  - name: id
    dtype: string
  - name: question
    dtype: string
  - name: system_prompt
    dtype: string
  - name: user_prompt
    dtype: string
  - name: response
    dtype: string
  - name: reasoning_trace
    dtype: string
  - name: answer
    dtype: string
  - name: elapsed_s
    dtype: float64
  - name: timestamp
    dtype: string
  splits:
  - name: train
    num_bytes: 130267945
    num_examples: 4979
  download_size: 52468943
  dataset_size: 130267945
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---