---
dataset_info:
  features:
  - name: premise
    dtype: string
  - name: hypothesis
    dtype: string
  - name: labels
    dtype: int64
  splits:
  - name: train
    num_bytes: 686675406
    num_examples: 2065271
  download_size: 369531660
  dataset_size: 686675406
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---