---
dataset_info:
  features:
  - name: promptID
    dtype: int32
  - name: pairID
    dtype: string
  - name: premise
    dtype: string
  - name: premise_binary_parse
    dtype: string
  - name: premise_parse
    dtype: string
  - name: hypothesis
    dtype: string
  - name: hypothesis_binary_parse
    dtype: string
  - name: hypothesis_parse
    dtype: string
  - name: genre
    dtype: string
  - name: label
    dtype:
      class_label:
        names:
          '0': entailment
          '1': neutral
          '2': contradiction
  - name: group
    dtype: int64
  splits:
  - name: train
    num_bytes: 21051684.075966
    num_examples: 20000
  - name: validation_matched
    num_bytes: 413343.9429444728
    num_examples: 400
  download_size: 8810362
  dataset_size: 21465028.018910475
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation_matched
    path: data/validation_matched-*
---