agentrl/ReSearch-Qwen-32B-Instruct
TEXT GENERATIONConcurrency Cost:2Model Size:32.8BQuant:FP8Ctx Length:32kPublished:Mar 27, 2025License:mitArchitecture:Transformer0.0K Open Weights Cold

agentrl/ReSearch-Qwen-32B-Instruct is a 32.8 billion parameter instruction-tuned language model developed by agentrl, based on the Qwen2.5 architecture. This model is specifically trained using the ReSearch framework, which enables LLMs to reason with search operations via reinforcement learning without supervised reasoning data. It integrates search capabilities directly into the reasoning chain, making it particularly effective for complex question answering and knowledge-intensive tasks.

Loading preview...