Overview
mmoralesf/mariana-qwen3-14b is a specialized 14 billion parameter language model, fine-tuned from the Qwen3-14B base model. Its primary purpose is to act as a virtual sales assistant for the automotive industry, specifically tailored for used car dealerships in Mexico. The model is trained to handle conversations in Mexican Spanish and integrates tool-calling functionalities.
Key Capabilities
- Automotive Sales Specialization: Optimized for dialogues related to buying and selling used cars.
- Mexican Spanish: Proficient in the nuances of Mexican Spanish for natural interactions.
- Tool Calling: Supports 11 custom tools using the Hermes format, enabling advanced functionalities.
- Domain-Specific Tasks: Capable of vehicle inventory search, providing recommendations, calculating financing options, and sharing business information (hours, locations, policies).
- Lead Management: Designed to assist with lead capture and follow-up processes.
Training Details
The model was fine-tuned using the LoRA method (r=16, alpha=32) on a synthetic dataset of automotive sales conversations. This targeted training ensures its high performance within the specified domain.
Good For
- Automated customer service for used car dealerships in Mexico.
- Virtual sales assistants requiring specialized automotive knowledge.
- Applications needing tool-calling capabilities for sales-related tasks.