---
dataset_info:
  features:
  - name: metadata
    struct:
    - name: doc_id
      dtype: string
    - name: split
      dtype: int64
    - name: title
      dtype: string
    - name: token_count
      dtype: int64
    - name: url
      dtype: string
  - name: passage
    dtype: string
  - name: id
    dtype: int64
  splits:
  - name: train
    num_bytes: 1377054502
    num_examples: 1639166
  download_size: 605204760
  dataset_size: 1377054502
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---