ayuag/yukt-med
Yukt-Med is a 2.7 billion parameter language model developed by ayuag, fine-tuned from Microsoft's Phi-2 base model. Specialized for the medical and healthcare domain, it is trained on over 86,000 curated medical interactions. This compact model provides concise, accurate, and non-diagnostic healthcare information, optimized for efficiency and runnable on commodity hardware with 4-bit GGUF quantization.
Loading preview...
What is Yukt-Med?
Yukt-Med is a lightweight, specialized language model (2.7B parameters) developed by ayuag, fine-tuned from Microsoft's Phi-2. It is designed specifically for the medical and healthcare domain, providing concise and accurate information based on extensive medical datasets. A key differentiator is its ultra-efficiency, with a 4-bit GGUF quantized version capable of running on devices with as little as 4GB RAM (CPU/Mobile), making it highly accessible for offline and resource-constrained environments.
Key Capabilities & Features
- Specialized Medical Knowledge: Trained on over 86,800 curated examples from datasets like ChatDoctor, MedQuad, and drug databases, focusing on conversational medical advice, Q&A, and symptom-disease mapping.
- Efficiency: Optimized for performance on commodity hardware through 4-bit GGUF quantization.
- Instruction-Following: Responds accurately to instructions using a specific prompt template.
- Production-Ready: Available in both Standard Safetensors and Compact GGUF formats.
When to Use Yukt-Med
This model is ideal for:
- Rapid medical information retrieval: Quickly access healthcare-related facts.
- Symptom analysis support: Assist in understanding potential symptoms (non-diagnostic).
- Educational purposes: Great for learning about medical conditions, drugs, and general health information.
Important Note: Yukt-Med is intended for informational purposes only and not for medical diagnosis.