---
dataset_info:
  features:
  - name: text
    dtype: string
  - name: hypothesis
    dtype: string
  - name: labels
    dtype:
      class_label:
        names:
          '0': entailment
          '1': not_entailment
  - name: task_name
    dtype: string
  - name: label_text
    dtype: string
  splits:
  - name: train
    num_bytes: 315013288.0
    num_examples: 1018733
  download_size: 206032209
  dataset_size: 315013288.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---
# Dataset Card for "dataset_train_nli"

Dataset for training a universal classifier. Additional information and training code available here: https://github.com/MoritzLaurer/zeroshot-classifier