BeastyZ/Qwen2.5-3B-ConvSearch-R1-QReCC
TEXT GENERATIONConcurrent Unit Cost:1Model Size:3.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:May 21, 2025License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold
BeastyZ/Qwen2.5-3B-ConvSearch-R1-QReCC is a 3.1 billion parameter language model based on the Qwen2.5-3B-Instruct architecture. It is fine-tuned using the QReCC dataset with the ConvSearch-R1 training method, specializing in conversational search tasks. This model is optimized for retrieving relevant information within a conversational context, making it suitable for applications requiring nuanced understanding of multi-turn queries.
Loading preview...
Model Overview
BeastyZ/Qwen2.5-3B-ConvSearch-R1-QReCC is a specialized language model built upon the Qwen2.5-3B-Instruct base architecture. With 3.1 billion parameters and a 32,768 token context length, this model has been specifically fine-tuned for conversational search applications.
Key Capabilities
- Conversational Search: Optimized for understanding and responding to queries within a multi-turn conversational context.
- QReCC Dataset Training: Leverages the QReCC dataset, which focuses on conversational question answering, to enhance its ability to handle complex, evolving search intents.
- ConvSearch-R1 Method: Utilizes the ConvSearch-R1 training methodology, indicating a focus on robust retrieval and response generation in conversational settings. The underlying code for this method is available here, and further details can be found in the associated paper here.
Good For
- Chatbots and Virtual Assistants: Ideal for systems that need to maintain context and provide relevant search results over extended conversations.
- Information Retrieval: Enhancing search engines or knowledge bases with conversational capabilities.
- Contextual Question Answering: Applications requiring the model to understand and answer questions based on prior turns in a dialogue.