---
dataset_info:
  features:
  - name: premise
    dtype: string
  - name: hypothesis
    dtype: string
  - name: label
    dtype:
      class_label:
        names:
          '0': entailment
          '1': neutral
          '2': contradiction
  splits:
  - name: train
    num_bytes: 9251968
    num_examples: 50000
  download_size: 6315822
  dataset_size: 9251968
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---