---
dataset_info:
  features:
  - name: instruction
    dtype: string
  - name: original_response
    dtype: string
  - name: explanation
    dtype: string
  - name: new_response
    dtype: string
  splits:
  - name: train
    num_bytes: 3421103
    num_examples: 546
  download_size: 1686318
  dataset_size: 3421103
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---