---
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: paper_id
    dtype: string
  - name: text_a
    dtype: string
  - name: text_b
    dtype: string
  splits:
  - name: train
    num_bytes: 684026152
    num_examples: 300000
  - name: valid
    num_bytes: 3419088
    num_examples: 1500
  download_size: 373064009
  dataset_size: 687445240
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
---