---
dataset_info:
  features:
  - name: natural_context
    dtype: string
  - name: unnatural_context
    dtype: string
  - name: question
    dtype: string
  - name: keywords
    sequence: string
  splits:
  - name: train
    num_bytes: 35001
    num_examples: 100
  download_size: 30642
  dataset_size: 35001
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---