README.md

---
dataset_info:
  features:
  - name: task_type
    dtype: string
  - name: task_desc
    dtype: string
  - name: model
    dtype: string
  - name: q
    dtype: string
  - name: pos
    dtype: string
  - name: neg
    dtype: string
  - name: scores
    struct:
    - name: neg
      dtype: float64
    - name: pos
      dtype: float64
  splits:
  - name: train
    num_bytes: 748267399
    num_examples: 499252
  download_size: 401472176
  dataset_size: 748267399
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---