用ELO思想训练检索模型，无需标注数据即达顶尖性能
zELO: ELO-inspired Training Method for Rerankers and Embedding Models
- 基于泰隆模型将排序任务建模为ELO评分系统，实现无监督训练
- 在金融、法律、代码等多领域超越闭源模型，NDCG@10与召回率均领先
- 零样本泛化能力强，适用于跨领域及私有数据集
我们提出一种名为 zELO 的新型训练方法，其核心思想是将排序任务视为与 Thurstone 模型统计等价的问题。基于此方法，我们利用无监督数据训练了一系列先进的开源重排模型：zerank-1 和 zerank-1-small。这些模型在金融、法律、代码和 STEM 等多个领域均取得最高检索成绩，显著优于闭源商用重排器，在 NDCG@10 与 Recall 指标上均表现更优。模型还展现出强大泛化能力，在跨领域及私有客户数据集上保持零样本高性能。训练数据包含 112,000 个查询，每个查询对应 100 个文档，全程端到端训练仅耗时少于 10,000 H100 小时。
原文摘要 · Abstract (English)
We introduce a novel training methodology named zELO, which optimizes retrieval performance via the analysis that ranking tasks are statically equivalent to a Thurstone model. Based on the zELO method, we use unsupervised data in order train a suite of state-of-the-art open-weight reranker models: zerank-1 and zerank-1-small. These models achieve the highest retrieval scores in multiple domains, including finance, legal, code, and STEM, outperforming closed-source proprietary rerankers on both NDCG@10 and Recall. These models also demonstrate great versatility, maintaining their 0-shot performance on out-of-domain and private customer datasets. The training data included 112,000 queries and 100 documents per query, and was trained end-to-end from unannotated queries and documents in less than 10,000 H100-hours.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。