---
dataset_info:
  features:
  - name: reference_response
    dtype: string
  - name: target
    sequence: string
  - name: ctx
    dtype: string
  splits:
  - name: train
    num_bytes: 85254210
    num_examples: 8000
  download_size: 33211950
  dataset_size: 85254210
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---