Update README.md

README.md

@@ -1,3 +1,39 @@
----
-license: cc-by-nc-4.0
----
+---
+license: cc-by-nc-4.0
+language:
+- en
+---
+
+## Introduction
+
+The model is used to evaluate the quality of a candidate patent claim compared to the gold claim.
+
+## Example Usage
+
+```python
+from transformers import AutoModel, AutoTokenizer
+import torch 
+
+device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+tokenizer = AutoTokenizer.from_pretrained("lj408/PatClaimEval-Quality", trust_remote_code=True)
+model = AutoModel.from_pretrained("lj408/PatClaimEval-Quality", trust_remote_code=True).to(device)
+gold_claim = "1. A computer-implemented method comprising: identifying a primary code segment; ..."
+candidate_claim = "1. A computer-implemented method for managing logger source code segments in a source code development platform, ..."
+res = model.score_pair(gold_claim, candidate_claim , tokenizer, device)
+print(res)
+```
+
+Note that this evaluation method can serve as a reference, but it may not be accurate in all cases. Users should rely on evaluations by patent professionals for more precise results.
+
+## Citation
+
+If you use this model or code, please cite our paper:
+
+```
+@article{jiang2025towards,
+  title={Towards Better Evaluation for Generated Patent Claims},
+  author={Jiang, Lekang and Scherz, Pascal A and Goetz, Stephan},
+  journal={arXiv preprint arXiv:2505.11095},
+  year={2025}
+}
+```