Arogya AI - Full Model Overview

Arogya AI is an 8 billion parameter language model, built upon Meta-Llama-3-8B and fine-tuned by Rafael and contributors, specifically for health data analysis and prediction in the Maluku Tenggara Regency, Indonesia. Unlike its LoRA adapter counterpart, this is a full merged model (~16 GB) that is ready-to-use without requiring a separate base model download, making it compatible with platforms like Ollama.

Key Capabilities

• Specialized Health Analysis: Trained on over 10,000 real health records from Maluku Tenggara, covering 169 health indicators from 2020-2029.
• Multi-Disease Support: Focuses on 7 primary diseases: Dengue Fever (DBD), Acute Respiratory Infection (ISPA), Malaria, Diarrhea, Tuberculosis (TB), Stunting, and Pneumonia.
• Geographic Specificity: Optimized for 9 sub-districts within Maluku Tenggara.
• Applications: Capable of disease prediction, trend analysis, intervention recommendations, resource allocation, and risk assessment.
• Performance: Achieves 87.3% accuracy and an F1 score of 0.86 on classification tasks, significantly outperforming the base Llama 3 8B on this specialized domain.

Good For

• Public Health Planning: Assisting in program planning and resource allocation for health initiatives in Maluku Tenggara.
• Epidemiological Research: Analyzing health data trends and assessing disease risks within the specified region.
• Developers: Deploying a standalone, specialized health AI model, particularly for Ollama environments.

Limitations

It's crucial to note the model's geographic specificity to Maluku Tenggara and its limited disease coverage. It is not intended as a substitute for professional medical advice or for individual patient diagnosis.