README.md

---
dataset_info:
  features:
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: metadata
    struct:
    - name: corruption_level
      dtype: float64
    - name: difficulty
      struct:
      - name: corruption_level
        sequence: float64
      - name: num_words
        sequence: int64
      - name: word_len
        sequence: int64
      - name: word_length
        sequence: int64
      - name: words
        sequence: int64
    - name: direction
      dtype: string
    - name: num_words
      dtype: int64
    - name: original_words
      sequence: string
    - name: scrambled_words
      sequence: string
    - name: sorted_words
      sequence: string
    - name: source_dataset
      dtype: string
    - name: source_index
      dtype: int64
    - name: transformed_words
      sequence: string
    - name: word
      dtype: string
    - name: word_len
      dtype: int64
    - name: word_length
      dtype: int64
  splits:
  - name: train
    num_bytes: 14365237
    num_examples: 20000
  download_size: 4943454
  dataset_size: 14365237
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---