---
dataset_info:
  features:
  - name: prompt
    dtype: string
  - name: target
    dtype: string
  - name: meta_path
    list: int64
  - name: meta_rules
    list: int64
  - name: index
    dtype: int64
  splits:
  - name: train
    num_bytes: 75700000
    num_examples: 100000
  download_size: 15115022
  dataset_size: 75700000
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---