README.md

---
dataset_info:
  features:
  - name: anchor
    dtype: string
  - name: positive
    dtype: string
  - name: negative
    dtype: string
  splits:
  - name: train
    num_bytes: 5507686591
    num_examples: 19368084
  - name: validation
    num_bytes: 611965082
    num_examples: 2152009
  - name: test
    num_bytes: 10435603
    num_examples: 33079
  download_size: 4254203099
  dataset_size: 6130087276
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---