---
dataset_info:
  features:
  - name: sentence
    dtype: string
  - name: aid
    dtype: string
  - name: sem_embed
    list: int8
  - name: rhet_embed
    list: int8
  - name: matches
    list:
    - name: m
      dtype: string
    - name: s
      dtype: float32
  - name: best_match
    dtype: string
  - name: best_scores
    dtype: float32
  splits:
  - name: train
    num_bytes: 8008918529
    num_examples: 11656366
  download_size: 13147177081
  dataset_size: 8008918529
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---