---
dataset_info:
  features:
  - name: premise
    dtype: string
  - name: hypothesis
    dtype: string
  - name: label
    dtype: int64
  splits:
  - name: train
    num_bytes: 73953022
    num_examples: 392702
  - name: test_matched
    num_bytes: 1868275
    num_examples: 10000
  - name: test_mismatched
    num_bytes: 1999384
    num_examples: 10000
  download_size: 52382918
  dataset_size: 77820681
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test_matched
    path: data/test_matched-*
  - split: test_mismatched
    path: data/test_mismatched-*
---