declare-lab/trustalign_qwen2.5_7b
TEXT GENERATIONConcurrency Cost:1Model Size:7.6BQuant:FP8Ctx Length:32kLicense:cc-by-4.0Architecture:Transformer0.0K Open Weights Cold

The declare-lab/trustalign_qwen2.5_7b is a 7.6 billion parameter Qwen2.5 model developed by declare-lab, specifically trained with Trust-Align. This model demonstrates improved trustworthiness by providing answers grounded in provided documents and refusing to answer when no supporting documents are available. It excels in tasks requiring factual accuracy and citation from given contexts, making it suitable for applications demanding high reliability in information retrieval.

Loading preview...