---
dataset_info:
  features:
  - name: premise
    dtype: string
  - name: hypothesis
    dtype: string
  - name: label
    dtype:
      class_label:
        names:
          '0': entailment
          '1': neutral
          '2': contradiction
  - name: idx
    dtype: int32
  splits:
  - name: train
    num_bytes: 74619646
    num_examples: 392702
  - name: validation
    num_bytes: 192549.19326105766
    num_examples: 1000
  - name: test
    num_bytes: 3590464.8067389424
    num_examples: 18647
  download_size: 54787760
  dataset_size: 78402660.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---