---
dataset_info:
  features:
  - name: premise
    dtype: string
  - name: hypothesis
    dtype: string
  - name: label
    dtype:
      class_label:
        names:
          '0': entailment
          '1': neutral
          '2': contradiction
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  splits:
  - name: train
    num_bytes: 3203992544
    num_examples: 5890530
  - name: test
    num_bytes: 37311457
    num_examples: 75150
  - name: validation
    num_bytes: 18395844
    num_examples: 37350
  download_size: 1979319378
  dataset_size: 3259699845
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: validation
    path: data/validation-*
---