---
dataset_info:
  features:
  - name: query_id
    dtype: string
  - name: query
    dtype: string
  - name: positive_passages
    list:
    - name: docid
      dtype: string
    - name: score
      dtype: string
    - name: text
      dtype: string
    - name: title
      dtype: string
  - name: negative_passages
    list:
    - name: docid
      dtype: string
    - name: score
      dtype: string
    - name: text
      dtype: string
    - name: title
      dtype: string
  splits:
  - name: train
    num_bytes: 55813324199
    num_examples: 1000000
  download_size: 24006659292
  dataset_size: 55813324199
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---