---
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: 980928129
    num_examples: 4167825
  download_size: 383499294
  dataset_size: 980928129
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---