---
dataset_info:
  features:
  - name: question_text
    dtype: string
  - name: metadata
    struct:
    - name: date
      dtype: int64
    - name: dump
      dtype: string
    - name: file_path
      dtype: string
    - name: language
      dtype: string
    - name: language_score
      dtype: float64
    - name: token_count
      dtype: int64
    - name: url
      dtype: string
  - name: prompt
    dtype: string
  - name: orig_answer_texts
    dtype: int64
  splits:
  - name: train
    num_bytes: 1545615750
    num_examples: 200000
  download_size: 675180660
  dataset_size: 1545615750
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---