| --- |
| base_model: |
| - Qwen/Qwen3-4B |
| language: |
| - en |
| library_name: sentence-transformers |
| license: cc-by-nc-4.0 |
| pipeline_tag: text-ranking |
| tags: |
| - finance |
| - legal |
| - code |
| - stem |
| - medical |
| --- |
| |
| <img src="https://i.imgur.com/oxvhvQu.png"/> |
|
|
| # Releasing zeroentropy/zerank-1 |
|
|
| This model is the `zerank-1` reranker as introduced in the paper [zELO: ELO-inspired Training Method for Rerankers and Embedding Models](https://huggingface.co/papers/2509.12541). |
|
|
| **Code**: [https://github.com/zeroentropy-ai/zbench](https://github.com/zeroentropy-ai/zbench) |
|
|
| In search engines, [rerankers are crucial](https://www.zeroentropy.dev/blog/what-is-a-reranker-and-do_i_need_one) for improving the accuracy of your retrieval system. |
|
|
| However, SOTA rerankers are closed-source and proprietary. At ZeroEntropy, we've trained a SOTA reranker outperforming closed-source competitors, and we're launching our model here on HuggingFace. |
|
|
| This reranker [outperforms proprietary rerankers](https://huggingface.co/zeroentropy/zerank-1#evaluations) such as `cohere-rerank-v3.5` and `Salesforce/LlamaRank-v1` across a wide variety of domains, including finance, legal, code, STEM, medical, and conversational data. |
|
|
| At ZeroEntropy we've developed an innovative multi-stage pipeline that models query-document relevance scores as adjusted [Elo ratings](https://en.wikipedia.org/wiki/Elo_rating_system). More details are available in our paper: [zELO: ELO-inspired Training Method for Rerankers and Embedding Models](https://huggingface.co/papers/2509.12541). |
|
|
| Since we're a small company, this model is only released under a non-commercial license. If you'd like a commercial license, please contact us at founders@zeroentropy.dev and we'll get you a license ASAP. |
|
|
| For this model's smaller twin, see [zerank-1-small](https://huggingface.co/zeroentropy/zerank-1-small), which we've fully open-sourced under an Apache 2.0 License. |
|
|
| ## How to Use |
|
|
| ```python |
| from sentence_transformers import CrossEncoder |
| |
| model = CrossEncoder("zeroentropy/zerank-1", trust_remote_code=True) |
| |
| query_documents = [ |
| ("What is 2+2?", "4"), |
| ("What is 2+2?", "The answer is definitely 1 million"), |
| ] |
| |
| scores = model.predict(query_documents) |
| |
| print(scores) |
| ``` |
|
|
| The model can also be inferenced using ZeroEntropy's [/models/rerank](https://docs.zeroentropy.dev/api-reference/models/rerank) endpoint. |
|
|
| ## Evaluations |
|
|
| NDCG@10 scores between `zerank-1` and competing closed-source proprietary rerankers. Since we are evaluating rerankers, OpenAI's `text-embedding-3-small` is used as an initial retriever for the Top 100 candidate documents. |
|
|
| | Task | Embedding | cohere-rerank-v3.5 | Salesforce/Llama-rank-v1 | zerank-1-small | **zerank-1** | |
| |---|---|---|---|---|---| |
| | Code | 0.678 | 0.724 | 0.694 | 0.730 | **0.754** | |
| | Conversational | 0.250 | 0.571 | 0.484 | 0.556 | **0.596** | |
| | Finance | 0.839 | 0.824 | 0.828 | 0.861 | **0.894** | |
| | Legal | 0.703 | 0.804 | 0.767 | 0.817 | **0.821** | |
| | Medical | 0.619 | 0.750 | 0.719 | 0.773 | **0.796** | |
| | STEM | 0.401 | 0.510 | 0.595 | 0.680 | **0.694** | |
|
|
| Comparing BM25 and Hybrid Search without and with zerank-1: |
|
|
| <img src="https://cdn-uploads.huggingface.co/production/uploads/67776f9dcd9c9435499eafc8/2GPVHFrI39FspnSNklhsM.png" alt="Description" width="400"/> <img src="https://cdn-uploads.huggingface.co/production/uploads/67776f9dcd9c9435499eafc8/dwYo2D7hoL8QiE8u3yqr9.png" alt="Description" width="400"/> |
