tobil/qmd-query-expansion-1.7B
The tobil/qmd-query-expansion-1.7B model is a 1.7 billion parameter Qwen3-based language model developed by tobil. It is specifically fine-tuned for query expansion tasks. This model excels at rephrasing and expanding search queries to improve retrieval results. It is designed for integration into search and information retrieval systems.
Loading preview...
Overview
tobil/qmd-query-expansion-1.7B is a specialized language model built upon the Qwen3-1.7B architecture. It has been fine-tuned using a production Supervised Fine-Tuning (SFT) pipeline specifically for the task of query expansion. This model's primary function is to take an initial search query and generate expanded or rephrased versions to enhance the effectiveness of search engines and information retrieval systems.
Key Capabilities
- Query Expansion: Generates alternative or more comprehensive versions of user queries.
- Qwen3 Chat Template: Utilizes the Qwen3 chat template for input formatting, expecting queries in the format
/no_think Expand this search query: <query>. - SFT-only Training: The model checkpoint represents an SFT-only trained version, indicating a focus on direct instruction following for query expansion without additional reinforcement learning from human feedback (GRPO).
Good For
- Improving Search Relevance: Enhancing the recall and precision of search results by providing richer query inputs.
- Information Retrieval Systems: Integrating into search backends to automatically expand user queries.
- Natural Language Understanding (NLU) Pipelines: As a component for pre-processing user inputs in NLU applications where query variation is beneficial.