jhu-clsp/rank1-14b
TEXT GENERATIONConcurrency Cost:1Model Size:14.8BQuant:FP8Ctx Length:32kLicense:mitArchitecture:Transformer Open Weights Warm

The jhu-clsp/rank1-14b is a 14.8 billion parameter reasoning reranker model, based on the Qwen2.5-14B architecture, developed by jhu-clsp. It is specifically designed for information retrieval tasks, leveraging test-time compute to generate explicit reasoning chains before making relevance judgments. This model excels at breaking down complex relevance decisions into logical steps, improving performance on nuanced retrieval tasks. Its primary strength lies in its unique 'think before judging' mechanism for enhanced relevance scoring.

Loading preview...