---
dataset_info:
  features:
  - name: query
    dtype: string
  - name: positive
    dtype: string
  - name: negative
    dtype: string
  - name: label
    dtype: float64
  - name: task_name
    dtype: string
  splits:
  - name: train
    num_bytes: 10372831990
    num_examples: 5933988
  download_size: 6205845494
  dataset_size: 10372831990
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---