---
dataset_info:
  config_name: corpus
  features:
  - name: doc_id
    dtype: string
  - name: passage
    dtype: string
  - name: metadata
    struct:
    - name: last_modified_datetime
      dtype: timestamp[us]
  splits:
  - name: train
    num_bytes: 3472453
    num_examples: 3523
  download_size: 2002668
  dataset_size: 3472453
configs:
- config_name: corpus
  data_files:
  - split: train
    path: corpus/train-*
---