---
dataset_info:
  features:
  - name: premise
    dtype: string
  - name: hypothesis
    dtype: string
  - name: label
    dtype: int64
  - name: source
    dtype: string
  splits:
  - name: train
    num_bytes: 30134573
    num_examples: 105000
  download_size: 18361633
  dataset_size: 30134573
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---