---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: response_1
    dtype: string
  - name: response_2
    dtype: string
  - name: preferred_response
    dtype: string
  splits:
  - name: train
    num_bytes: 16989
    num_examples: 10
  download_size: 14081
  dataset_size: 16989
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---