---
dataset_info:
  features:
  - name: instruction_seed
    dtype: string
  - name: Rationale
    dtype: string
  - name: options
    dtype: string
  - name: correct
    dtype: string
  - name: response_seed
    dtype: string
  - name: linear_formula
    dtype: string
  - name: category
    dtype: string
  - name: response
    dtype: string
  - name: __original_row_idx
    dtype: int64
  - name: conversations
    list:
    - name: from
      dtype: string
    - name: value
      dtype: string
  splits:
  - name: train
    num_bytes: 99106802
    num_examples: 29837
  download_size: 45378046
  dataset_size: 99106802
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---