---
dataset_info:
  features:
  - name: query
    dtype: string
  - name: positives
    sequence: string
  - name: negatives
    sequence: string
  splits:
  - name: train
    num_bytes: 850663698
    num_examples: 50000
  download_size: 494705376
  dataset_size: 850663698
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---