emylton/arogya-ai-full
The emylton/arogya-ai-full model is an 8 billion parameter Llama 3-based language model developed by Rafael and contributors, specifically fine-tuned for health data analysis and disease prediction in the Maluku Tenggara Regency, Indonesia. It excels at analyzing health data for 7 key diseases across 9 sub-districts, leveraging over 10,000 real health records. This full merged model is ready-to-use, supporting applications like disease prediction, trend analysis, and intervention recommendations within its specialized geographic and disease scope.
Loading preview...
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.