---
dataset_info:
  features:
  - name: variants
    struct:
    - name: original
      dtype: string
  - name: metadata
    struct:
    - name: base
      dtype: string
    - name: context
      dtype: string
    - name: copy_index
      dtype: int64
    - name: degenerate
      dtype: string
    - name: duplicate
      dtype: string
    - name: is_duplicate_copy
      dtype: bool
    - name: is_duplicate_source
      dtype: bool
    - name: original_index
      dtype: int64
    - name: variety
      dtype: string
  splits:
  - name: train
    num_bytes: 13380553144
    num_examples: 4400000
  download_size: 8234808275
  dataset_size: 13380553144
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---