---
dataset_info:
  features:
  - name: premise
    dtype: string
  - name: hypothesis
    dtype: string
  - name: label
    dtype:
      class_label:
        names:
          '0': entailment
          '1': neutral
          '2': contradiction
  - name: explanation
    dtype: string
  splits:
  - name: train
    num_bytes: 592228
    num_examples: 1816
  - name: validation
    num_bytes: 504564
    num_examples: 2490
  - name: test
    num_bytes: 1016528
    num_examples: 5010
  download_size: 832809
  dataset_size: 2113320
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---