NutriMinds-3B-Merged

🍏 NutriMinds-3B-Merged

Model Description

NutriMinds-3B is a highly specialized clinical AI engine developed by Shrisudarsanah at SRM University. It is designed to act as an expert clinical nutritionist, analyzing specific food items and ingredient labels against 52 distinct medical conditions (such as Type 2 Diabetes, Celiac Disease, and Chronic Kidney Disease).

The model has been fine-tuned to output evaluations in a strict, predictable JSON schema, making it ideal for integration into automated healthcare applications and recipe-generation pipelines.

Model Details

• Base Model: Qwen/Qwen2.5-3B-Instruct
• Architecture: 16-bit Merged LoRA Weights (Native Hugging Face format)
• Training Method: SFT (Supervised Fine-Tuning) via Unsloth
• Primary Language(s): English
• License: Apache 2.0

Intended Use

This model is intended to be used as a backend inference engine for the NutriMinds AI application. It accepts a patient's medical profile and a food label/item, and returns a JSON object detailing the compatibility score, risk alerts, and clinical rationale.

Note on Medical Advice: This model is an academic tool built for analytical demonstration. It should not replace professional medical diagnosis or treatment.

Prompt Format

The model requires the standard ChatML template and expects the following strict system prompt to trigger the JSON generation behavior:

You are NutriMinds AI, an expert clinical nutritionist and medical analyst.
Analyze the user's primary medical condition and food item.
Return a strict JSON response evaluating food suitability according to medical guidelines.