Create README.md

README.md

@@ -0,0 +1,25 @@
+---
+language:
+- en
+
+pipeline_tag: text-classification
+---
+
+# Span NLI BERT (base)
+
+This is a **BERT-base** model ([`bert-base-uncased`][2]) fine-tuned on the [**ContractNLI**][3] dataset (non-disclosure agreements) with the **Span NLI BERT** model architecture, 
+from [*ContractNLI: A Dataset for Document-level Natural Language Inference for Contracts* (Koreeda and Manning, 2021)][1].
+
+For a hypothesis, the **Span NLI BERT** model predicts NLI labels and identifies evidence for documents as premises. 
+Spans of documents should be pre-annotated; evidence is always full sentences or items in an enumerated list in the document. 
+
+For details of the architecture and usage of the relevant training/testing scripts, check out the paper and their [Github repo][4]. 
+This model is fine-tuned according to the hyperparameters in `data/conf_base.yml` in their repo, 
+which differs from their hyperparameters that produced the best dev scores as noted in the Appendix of the paper. 
+
+ArXiv: <https://arxiv.org/abs/2110.01799>
+
+[1]: https://aclanthology.org/2021.findings-emnlp.164/
+[2]: https://huggingface.co/bert-base-uncased
+[3]: https://stanfordnlp.github.io/contract-nli/
+[4]: https://github.com/stanfordnlp/contract-nli-bert